=== 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: fdee58b7a5e240d23ead32de5f87cde65b0b6616 ################# 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 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/01 job id: 42887413 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/02 job id: 42887415 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/03 job id: 42887420 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/04 job id: 42887424 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/05 job id: 42887429 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/06 job id: 42887435 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/07 job id: 42887437 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/08 job id: 42887438 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/09 job id: 42887440 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/10 job id: 42887443 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/11 job id: 42887445 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/12 job id: 42887447 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/13 job id: 42887449 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/14 job id: 42887451 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/15 job id: 42887453 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/16 job id: 42887455 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/17 job id: 42887457 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/18 job id: 42887459 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/19 job id: 42887461 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/20 job id: 42887464 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/21 job id: 42887466 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/22 job id: 42887468 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/23 job id: 42887470 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/24 job id: 42887472 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/25 job id: 42887475 --- 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/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/26 job id: 42887477 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 0.023 0.038 135.926 135.927 farming_run 1 2.0 135.365 135.366 135.884 135.887 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.456488E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.026 116.710 116.711 qs_energies 1 2.0 0.000 0.000 116.433 116.436 mp2_main 1 3.0 0.000 0.001 114.168 114.171 mp2_gpw_main 1 4.0 0.019 0.026 113.044 113.047 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.835 94.330 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.410 55.906 mp2_eri_3c_integrate_gpw 272 7.0 0.155 0.172 41.721 47.153 get_2c_integrals 1 6.0 0.000 0.000 37.379 38.253 integrate_v_rspace 273 8.0 0.435 0.450 25.095 30.227 pw_transfer 6555 10.6 0.373 0.395 27.399 28.379 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 26.031 26.898 grid_integrate_task_list 273 9.0 20.917 26.531 20.917 26.531 fft_wrap_pw1pw2_100 2178 12.4 1.226 1.445 23.591 24.464 compute_2c_integrals 1 7.0 0.002 0.003 19.858 19.859 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.978 19.518 mp2_eri_2c_integrate_gpw 1 9.0 2.366 2.436 18.975 19.517 rpa_ri_compute_en 1 5.0 0.004 0.027 19.100 19.244 cp_fm_cholesky_decompose 12 8.2 17.613 18.503 17.613 18.503 cholesky_decomp 1 7.0 0.000 0.000 16.367 17.237 fft3d_s 5443 13.4 16.149 16.732 16.172 16.752 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.574 10.855 15.574 calculate_wavefunction 272 8.0 5.428 5.589 12.544 13.131 rpa_num_int 1 6.0 0.000 0.000 10.683 10.693 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.640 10.649 calc_mat_Q 8 8.0 0.000 0.000 9.354 9.455 contract_S_to_Q 8 9.0 0.000 0.000 8.774 8.874 calc_potential_gpw 544 9.5 0.006 0.006 8.302 8.698 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.285 8.634 potential_pw2rs 545 10.0 0.107 0.110 7.702 8.504 parallel_gemm_fm 14 9.1 0.000 0.000 8.358 8.442 parallel_gemm_fm_cosma 14 10.1 8.358 8.442 8.358 8.442 create_integ_mat 1 6.0 0.014 0.028 7.800 7.810 collocate_single_gaussian 272 10.0 0.040 0.043 7.501 7.790 array2fm 1 7.0 0.000 0.000 6.746 7.246 pw_scatter_s 2720 13.7 4.430 4.608 4.430 4.608 pw_gather_s 2722 13.2 3.856 4.211 3.856 4.211 array2fm_buffer_send 1 8.0 3.004 3.332 3.004 3.332 pw_poisson_solve 545 10.5 1.134 1.178 2.209 2.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.043792, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.038 397.318 397.319 farming_run 1 2.0 396.391 396.414 397.273 397.278 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.226334E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 210.576 210.576 qs_energies 1 2.0 0.000 0.000 210.326 210.333 scf_env_do_scf 1 3.0 0.000 0.000 107.004 107.004 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.094 106.101 rebuild_ks_matrix 4 6.0 0.000 0.000 106.092 106.100 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 106.092 106.100 hfx_ks_matrix 4 8.0 0.001 0.001 105.721 105.724 integrate_four_center 4 9.0 0.143 0.454 105.720 105.723 mp2_main 1 3.0 0.003 0.012 103.037 103.044 mp2_gpw_main 1 4.0 0.054 0.142 102.078 102.090 integrate_four_center_main 4 10.0 0.104 0.509 96.482 99.828 integrate_four_center_bin 265 11.0 96.379 99.821 96.379 99.821 init_scf_loop 1 4.0 0.000 0.000 92.716 92.716 mp2_ri_gpw_compute_in 1 5.0 0.071 0.083 74.931 75.984 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.511 55.567 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.801 47.349 integrate_v_rspace 95 8.0 0.396 0.572 28.148 33.499 pw_transfer 2240 10.6 0.145 0.168 29.903 30.338 ao_to_mo_and_store_B_mult_1 91 7.0 11.020 29.732 11.020 29.732 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.918 29.360 grid_integrate_task_list 95 9.0 23.436 28.987 23.436 28.987 mp2_ri_gpw_compute_en 1 5.0 0.070 0.122 26.914 28.573 fft_wrap_pw1pw2_100 730 12.4 1.266 1.428 26.641 27.036 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.824 1.879 25.169 25.180 get_2c_integrals 1 6.0 0.006 0.018 20.274 20.363 compute_2c_integrals 1 7.0 0.010 0.027 19.246 19.263 compute_2c_integrals_loop_lm 1 8.0 0.008 0.020 18.860 19.132 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.853 18.852 19.131 fft3d_s 1823 13.4 18.424 18.778 18.437 18.791 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.285 14.285 calculate_wavefunction 91 8.0 2.039 2.072 9.761 9.987 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.582 8.873 9.445 potential_pw2rs 186 10.0 0.034 0.035 8.587 9.092 local_gemm 172 8.0 8.316 8.874 8.316 8.874 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.515 8.085 8.680 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.234 8.503 calc_potential_gpw 182 9.5 0.002 0.002 7.960 8.185 collocate_single_gaussian 91 10.0 0.018 0.042 7.870 8.073 mp_sendrecv_dm3 2068 8.0 6.126 6.700 6.126 6.700 mp_sync 38 10.4 3.934 6.685 3.934 6.685 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.429 6.341 6.429 pw_gather_s 912 13.2 4.898 5.355 4.898 5.355 pw_scatter_s 910 13.7 3.956 4.282 3.956 4.282 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.073332, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 450.285568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.019 0.032 53.966 53.967 qs_mol_dyn_low 1 2.0 0.003 0.003 53.706 53.715 qs_forces 11 3.9 0.002 0.003 53.635 53.637 qs_energies 11 4.9 0.001 0.004 52.117 52.131 scf_env_do_scf 11 5.9 0.000 0.001 45.931 45.931 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.652 43.653 dbcsr_multiply_generic 2286 12.5 0.095 0.098 34.475 35.078 qs_scf_new_mos 108 7.5 0.000 0.001 33.070 33.387 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.069 33.387 ot_scf_mini 108 9.5 0.002 0.002 31.391 31.589 multiply_cannon 2286 13.5 0.183 0.191 26.267 28.231 multiply_cannon_loop 2286 14.5 1.506 1.575 25.474 27.397 velocity_verlet 10 3.0 0.001 0.001 26.304 26.308 ot_mini 108 10.5 0.001 0.001 20.111 20.364 qs_ot_get_derivative 108 11.5 0.001 0.001 17.012 17.228 mp_waitall_1 267858 16.1 9.341 16.373 9.341 16.373 multiply_cannon_metrocomm3 54864 15.5 0.067 0.074 6.120 13.967 multiply_cannon_multrec 54864 15.5 4.182 6.470 7.549 10.984 rebuild_ks_matrix 119 8.3 0.000 0.000 8.281 8.435 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.022 8.280 8.435 multiply_cannon_sync_h2d 54864 15.5 5.833 7.954 5.833 7.954 mp_sum_l 7207 12.9 5.949 7.712 5.949 7.712 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.309 7.451 qs_ot_get_p 119 10.4 0.002 0.005 6.640 6.983 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.970 6.102 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.607 6.051 init_scf_run 11 5.9 0.000 0.001 4.915 4.915 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.915 4.915 sum_up_and_integrate 119 10.3 0.012 0.014 4.803 4.809 integrate_v_rspace 119 11.3 0.002 0.002 4.791 4.798 dbcsr_mm_accdrv_process 76910 16.1 1.189 1.833 3.291 4.548 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.382 4.490 calculate_rho_elec 119 8.7 0.011 0.016 4.381 4.490 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.709 3.821 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.934 3.693 rs_pw_transfer 974 11.9 0.011 0.013 3.154 3.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.961 3.151 apply_single 119 13.6 0.000 0.000 2.961 3.151 calculate_dm_sparse 119 9.5 0.000 0.000 3.007 3.135 ot_diis_step 108 11.5 0.006 0.006 2.783 2.784 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.698 2.700 calculate_first_density_matrix 1 7.0 0.001 0.004 2.691 2.695 jit_kernel_multiply 13 15.8 2.039 2.627 2.039 2.627 density_rs2pw 119 9.7 0.004 0.004 2.476 2.565 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.435 2.506 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.435 2.436 cp_fm_redistribute_end 50 14.0 2.204 2.407 2.209 2.408 cp_fm_diag_elpa_base 50 14.0 0.197 2.366 0.198 2.375 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.345 2.347 acc_transpose_blocks 54864 15.5 0.228 0.243 1.833 2.287 init_scf_loop 11 6.9 0.000 0.002 2.257 2.257 potential_pw2rs 119 12.3 0.004 0.004 2.150 2.177 mp_sum_d 4125 12.0 1.472 2.174 1.472 2.174 wfi_extrapolate 11 7.9 0.001 0.001 2.160 2.160 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.061 2.105 grid_integrate_task_list 119 12.3 2.011 2.101 2.011 2.101 pw_transfer 1439 11.6 0.051 0.057 1.939 2.033 make_m2s 4572 13.5 0.054 0.056 1.952 2.016 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.864 1.961 make_images 4572 14.5 0.133 0.139 1.868 1.930 fft3d_ps 1201 14.6 0.365 0.473 1.645 1.737 mp_alltoall_d11v 2130 13.8 1.445 1.591 1.445 1.591 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.500 1.534 fft_wrap_pw1pw2_140 487 13.2 0.078 0.092 1.437 1.532 mp_waitany 12084 13.8 1.299 1.461 1.299 1.461 grid_collocate_task_list 119 9.7 1.286 1.366 1.286 1.366 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.848 1.242 acc_transpose_blocks_kernels 54864 16.5 0.250 0.378 0.850 1.153 prepare_preconditioner 11 7.9 0.000 0.000 1.102 1.131 make_preconditioner 11 8.9 0.000 0.001 1.102 1.131 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.043 1.080 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.967000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.545455, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.545728E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1032117. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.033 0.080 41.161 41.163 qs_mol_dyn_low 1 2.0 0.003 0.006 40.843 40.851 qs_forces 11 3.9 0.002 0.002 40.777 40.779 qs_energies 11 4.9 0.006 0.008 39.010 39.015 scf_env_do_scf 11 5.9 0.001 0.002 33.568 33.569 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.816 30.818 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.321 22.716 qs_scf_new_mos 108 7.5 0.001 0.001 21.188 21.428 qs_scf_loop_do_ot 108 8.5 0.001 0.002 21.188 21.427 ot_scf_mini 108 9.5 0.003 0.003 20.245 20.414 velocity_verlet 10 3.0 0.001 0.002 19.613 19.614 multiply_cannon 2286 13.5 0.209 0.219 16.707 18.336 multiply_cannon_loop 2286 14.5 0.905 0.980 15.412 17.047 ot_mini 108 10.5 0.001 0.001 12.542 12.777 mp_waitall_1 217478 16.2 6.816 11.947 6.816 11.947 qs_ot_get_derivative 108 11.5 0.001 0.001 10.020 10.192 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.352 9.961 multiply_cannon_multrec 27432 15.5 1.952 4.517 5.862 8.873 rebuild_ks_matrix 119 8.3 0.000 0.000 7.551 7.696 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.551 7.696 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.652 6.783 dbcsr_mm_accdrv_process 47894 16.0 3.008 5.177 3.841 5.680 qs_ot_get_p 119 10.4 0.001 0.002 4.691 4.920 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.900 4.739 sum_up_and_integrate 119 10.3 0.024 0.027 4.542 4.549 integrate_v_rspace 119 11.3 0.002 0.002 4.517 4.526 mp_sum_l 7207 12.9 2.301 4.224 2.301 4.224 init_scf_run 11 5.9 0.000 0.001 4.153 4.153 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.152 4.153 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.077 4.153 apply_single 119 13.6 0.000 0.000 3.077 4.153 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.976 4.018 calculate_rho_elec 119 8.7 0.021 0.024 3.975 4.018 make_m2s 4572 13.5 0.053 0.055 2.877 3.140 rs_pw_transfer 974 11.9 0.009 0.010 3.060 3.128 make_images 4572 14.5 0.199 0.234 2.787 3.048 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.999 3.020 multiply_cannon_sync_h2d 27432 15.5 2.143 2.793 2.143 2.793 init_scf_loop 11 6.9 0.001 0.003 2.714 2.718 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.530 2.531 calculate_first_density_matrix 1 7.0 0.001 0.003 2.472 2.475 ot_diis_step 108 11.5 0.011 0.011 2.465 2.466 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.345 2.437 density_rs2pw 119 9.7 0.004 0.004 2.324 2.397 potential_pw2rs 119 12.3 0.006 0.006 2.250 2.263 calculate_dm_sparse 119 9.5 0.000 0.000 2.120 2.200 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.099 2.101 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.057 2.059 cp_fm_redistribute_end 50 14.0 1.680 2.015 1.686 2.018 pw_transfer 1439 11.6 0.063 0.066 1.970 2.001 cp_fm_diag_elpa_base 50 14.0 0.321 1.946 0.329 1.977 grid_integrate_task_list 119 12.3 1.833 1.957 1.833 1.957 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.880 1.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.799 1.840 jit_kernel_multiply 9 16.1 0.781 1.764 0.781 1.764 prepare_preconditioner 11 7.9 0.000 0.000 1.726 1.751 make_preconditioner 11 8.9 0.000 0.002 1.726 1.751 make_images_data 4572 15.5 0.045 0.052 1.236 1.728 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.630 1.685 fft3d_ps 1201 14.6 0.498 0.554 1.589 1.620 wfi_extrapolate 11 7.9 0.001 0.001 1.620 1.620 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.093 1.604 acc_transpose_blocks 27432 15.5 0.107 0.111 1.260 1.578 mp_alltoall_d11v 2130 13.8 1.446 1.548 1.446 1.548 mp_allgather_i34 2286 14.5 0.735 1.521 0.735 1.521 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.307 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.311 1.320 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.272 1.313 grid_collocate_task_list 119 9.7 1.242 1.302 1.242 1.302 mp_sum_d 4125 12.0 0.680 1.140 0.680 1.140 make_images_sizes 4572 15.5 0.005 0.005 0.826 1.108 mp_alltoall_i44 4572 16.5 0.821 1.103 0.821 1.103 acc_transpose_blocks_kernels 27432 16.5 0.179 0.268 0.733 0.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.960 0.973 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.970 0.973 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.863 0.894 dbcsr_complete_redistribute 329 12.2 0.082 0.106 0.793 0.862 mp_alltoall_z22v 1201 16.6 0.788 0.861 0.788 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.163000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.000000, yerr=1.858641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.203136E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.058 33.521 33.523 qs_mol_dyn_low 1 2.0 0.003 0.004 33.193 33.201 qs_forces 11 3.9 0.001 0.002 33.133 33.135 qs_energies 11 4.9 0.009 0.037 31.513 31.521 scf_env_do_scf 11 5.9 0.001 0.002 26.479 26.479 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.725 23.726 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.962 17.075 velocity_verlet 10 3.0 0.001 0.001 15.992 15.994 qs_scf_new_mos 108 7.5 0.001 0.001 15.400 15.422 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.399 15.421 ot_scf_mini 108 9.5 0.002 0.003 14.656 14.670 multiply_cannon 2286 13.5 0.197 0.203 13.278 14.123 multiply_cannon_loop 2286 14.5 0.641 0.667 12.385 13.330 ot_mini 108 10.5 0.001 0.001 9.110 9.123 qs_ot_get_derivative 108 11.5 0.001 0.001 7.556 7.575 multiply_cannon_multrec 18288 15.5 1.914 2.946 6.726 7.102 rebuild_ks_matrix 119 8.3 0.000 0.000 6.552 6.566 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.551 6.565 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.799 5.813 dbcsr_mm_accdrv_process 38222 16.0 4.105 5.504 4.729 5.640 mp_waitall_1 169478 16.3 3.333 4.540 3.333 4.540 sum_up_and_integrate 119 10.3 0.030 0.031 4.133 4.138 integrate_v_rspace 119 11.3 0.002 0.003 4.103 4.111 init_scf_run 11 5.9 0.000 0.001 3.697 3.699 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.697 3.699 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.445 3.462 calculate_rho_elec 119 8.7 0.030 0.031 3.444 3.461 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.813 3.438 qs_ot_get_p 119 10.4 0.001 0.002 3.285 3.312 init_scf_loop 11 6.9 0.001 0.003 2.731 2.731 rs_pw_transfer 974 11.9 0.009 0.010 2.540 2.643 calculate_first_density_matrix 1 7.0 0.001 0.004 2.380 2.384 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.441 2.378 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.063 2.370 apply_single 119 13.6 0.000 0.000 2.062 2.369 make_m2s 4572 13.5 0.045 0.046 2.027 2.177 density_rs2pw 119 9.7 0.004 0.004 2.018 2.125 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.098 2.108 make_images 4572 14.5 0.190 0.201 1.941 2.090 potential_pw2rs 119 12.3 0.007 0.008 1.914 1.919 prepare_preconditioner 11 7.9 0.000 0.000 1.901 1.903 make_preconditioner 11 8.9 0.000 0.001 1.901 1.903 calculate_dm_sparse 119 9.5 0.000 0.000 1.889 1.901 grid_integrate_task_list 119 12.3 1.790 1.898 1.790 1.898 jit_kernel_multiply 10 16.3 0.573 1.874 0.573 1.874 pw_transfer 1439 11.6 0.063 0.067 1.847 1.859 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.752 1.832 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.812 1.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.787 1.788 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.757 1.770 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.684 1.694 multiply_cannon_sync_h2d 18288 15.5 1.386 1.656 1.386 1.656 mp_sum_l 7207 12.9 1.278 1.628 1.278 1.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.605 1.617 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.535 1.536 ot_diis_step 108 11.5 0.011 0.011 1.525 1.525 cp_fm_redistribute_end 50 14.0 1.137 1.511 1.138 1.512 cp_fm_diag_elpa_base 50 14.0 0.357 1.446 0.371 1.483 fft3d_ps 1201 14.6 0.508 0.526 1.443 1.454 acc_transpose_blocks 18288 15.5 0.074 0.075 1.345 1.370 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.309 1.324 grid_collocate_task_list 119 9.7 1.205 1.320 1.205 1.320 wfi_extrapolate 11 7.9 0.001 0.001 1.258 1.258 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.251 make_images_data 4572 15.5 0.046 0.050 0.885 1.059 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.011 1.014 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.936 0.962 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.755 0.952 acc_transpose_blocks_kernels 18288 16.5 0.208 0.217 0.909 0.924 mp_alltoall_d11v 2130 13.8 0.782 0.920 0.782 0.920 cp_fm_cholesky_invert 11 10.9 0.833 0.837 0.833 0.837 mp_alltoall_z22v 1201 16.6 0.753 0.828 0.753 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.826 0.827 mp_allgather_i34 2286 14.5 0.353 0.755 0.353 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.667 0.728 jit_kernel_transpose 5 15.6 0.701 0.713 0.701 0.713 mp_waitany 9880 13.7 0.546 0.687 0.546 0.687 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.523000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.727273, yerr=2.631657 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 546.152448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 37.603 37.604 qs_mol_dyn_low 1 2.0 0.003 0.004 37.347 37.355 qs_forces 11 3.9 0.002 0.002 37.287 37.288 qs_energies 11 4.9 0.001 0.001 35.541 35.547 scf_env_do_scf 11 5.9 0.000 0.001 30.034 30.035 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.255 26.256 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.068 19.248 velocity_verlet 10 3.0 0.001 0.001 18.949 18.951 qs_scf_new_mos 108 7.5 0.001 0.001 17.381 17.443 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.380 17.443 ot_scf_mini 108 9.5 0.002 0.003 16.415 16.470 multiply_cannon 2286 13.5 0.233 0.262 14.827 15.412 multiply_cannon_loop 2286 14.5 0.944 0.975 13.731 14.108 ot_mini 108 10.5 0.001 0.001 10.077 10.137 multiply_cannon_multrec 27432 15.5 2.346 3.008 8.640 8.990 qs_ot_get_derivative 108 11.5 0.001 0.001 8.221 8.277 dbcsr_mm_accdrv_process 47916 15.9 5.370 6.823 6.201 7.461 rebuild_ks_matrix 119 8.3 0.000 0.000 6.898 6.948 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.897 6.947 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.142 6.186 init_scf_run 11 5.9 0.000 0.001 4.142 4.142 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.142 4.142 sum_up_and_integrate 119 10.3 0.035 0.037 4.116 4.124 integrate_v_rspace 119 11.3 0.002 0.003 4.081 4.089 init_scf_loop 11 6.9 0.000 0.000 3.747 3.747 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.697 3.732 calculate_rho_elec 119 8.7 0.040 0.046 3.696 3.732 qs_ot_get_p 119 10.4 0.001 0.001 3.619 3.691 mp_waitall_1 145218 16.4 2.934 3.495 2.934 3.495 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.893 3.362 prepare_preconditioner 11 7.9 0.000 0.000 2.854 2.863 make_preconditioner 11 8.9 0.000 0.000 2.854 2.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.437 2.776 make_m2s 4572 13.5 0.055 0.057 2.599 2.702 calculate_first_density_matrix 1 7.0 0.000 0.000 2.617 2.618 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.145 2.608 apply_single 119 13.6 0.000 0.000 2.145 2.608 rs_pw_transfer 974 11.9 0.008 0.009 2.502 2.605 make_images 4572 14.5 0.271 0.331 2.489 2.588 density_rs2pw 119 9.7 0.004 0.004 2.151 2.244 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.160 2.192 calculate_dm_sparse 119 9.5 0.000 0.000 2.123 2.182 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.113 2.125 pw_transfer 1439 11.6 0.063 0.067 1.986 2.018 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.896 1.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.918 1.919 grid_integrate_task_list 119 12.3 1.835 1.911 1.835 1.911 potential_pw2rs 119 12.3 0.008 0.009 1.863 1.869 mp_sum_l 7207 12.9 1.217 1.823 1.217 1.823 ot_diis_step 108 11.5 0.012 0.012 1.815 1.815 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.762 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.705 1.717 fft3d_ps 1201 14.6 0.538 0.590 1.578 1.604 acc_transpose_blocks 27432 15.5 0.109 0.111 1.550 1.578 jit_kernel_multiply 8 15.7 0.771 1.553 0.771 1.553 fft_wrap_pw1pw2_140 487 13.2 0.083 0.092 1.483 1.521 wfi_extrapolate 11 7.9 0.001 0.001 1.467 1.467 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.414 1.416 cp_fm_redistribute_end 50 14.0 0.922 1.371 0.924 1.373 cp_fm_diag_elpa_base 50 14.0 0.424 1.293 0.444 1.334 grid_collocate_task_list 119 9.7 1.219 1.334 1.219 1.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.290 1.300 dbcsr_complete_redistribute 329 12.2 0.119 0.147 0.972 1.257 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.871 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.177 1.195 make_images_data 4572 15.5 0.045 0.049 1.056 1.160 cp_fm_upper_to_full 72 13.5 0.829 1.151 0.829 1.151 mp_alltoall_d11v 2130 13.8 0.958 1.098 0.958 1.098 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.912 1.086 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.079 1.080 multiply_cannon_sync_h2d 27432 15.5 0.988 1.032 0.988 1.032 acc_transpose_blocks_kernels 27432 16.5 0.265 0.272 0.932 0.957 mp_alltoall_z22v 1201 16.6 0.923 0.952 0.923 0.952 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.670 0.948 cp_fm_cholesky_invert 11 10.9 0.913 0.917 0.913 0.917 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.854 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.227 0.833 mp_alltoall_i22 627 13.8 0.465 0.763 0.465 0.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.604000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.454545, yerr=3.576785 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 613.044224E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.031 29.429 29.430 qs_mol_dyn_low 1 2.0 0.003 0.003 29.241 29.249 qs_forces 11 3.9 0.001 0.002 29.175 29.176 qs_energies 11 4.9 0.001 0.001 27.438 27.442 scf_env_do_scf 11 5.9 0.000 0.001 22.565 22.565 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.738 19.739 velocity_verlet 10 3.0 0.001 0.001 15.260 15.263 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.751 12.831 qs_scf_new_mos 108 7.5 0.001 0.001 11.782 11.816 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.782 11.815 ot_scf_mini 108 9.5 0.002 0.002 11.075 11.104 multiply_cannon 2286 13.5 0.231 0.241 9.819 10.291 multiply_cannon_loop 2286 14.5 0.334 0.347 8.842 9.045 ot_mini 108 10.5 0.001 0.001 6.133 6.163 rebuild_ks_matrix 119 8.3 0.000 0.000 6.084 6.108 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.083 6.108 multiply_cannon_multrec 9144 15.5 1.658 1.902 5.823 6.071 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.429 5.451 qs_ot_get_derivative 108 11.5 0.001 0.001 4.788 4.818 dbcsr_mm_accdrv_process 12550 15.8 3.188 4.084 4.062 4.141 sum_up_and_integrate 119 10.3 0.037 0.041 3.783 3.787 integrate_v_rspace 119 11.3 0.002 0.003 3.745 3.750 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.400 3.409 calculate_rho_elec 119 8.7 0.059 0.061 3.400 3.409 init_scf_run 11 5.9 0.000 0.001 3.401 3.402 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.401 3.401 qs_ot_get_p 119 10.4 0.001 0.001 3.068 3.106 init_scf_loop 11 6.9 0.000 0.000 2.799 2.800 mp_waitall_1 121218 16.5 2.230 2.745 2.230 2.745 calculate_first_density_matrix 1 7.0 0.000 0.000 2.162 2.162 make_m2s 4572 13.5 0.035 0.036 2.023 2.148 make_images 4572 14.5 0.267 0.302 1.932 2.056 rs_pw_transfer 974 11.9 0.008 0.008 1.960 2.051 prepare_preconditioner 11 7.9 0.000 0.000 2.026 2.031 make_preconditioner 11 8.9 0.000 0.000 2.026 2.031 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.998 2.000 density_rs2pw 119 9.7 0.003 0.004 1.848 1.934 grid_integrate_task_list 119 12.3 1.842 1.930 1.842 1.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.898 1.928 pw_transfer 1439 11.6 0.063 0.066 1.875 1.884 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.784 1.794 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.779 1.780 calculate_dm_sparse 119 9.5 0.000 0.000 1.740 1.758 jit_kernel_multiply 10 15.7 0.837 1.745 0.837 1.745 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.697 1.728 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.649 1.650 potential_pw2rs 119 12.3 0.010 0.010 1.542 1.550 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.471 1.473 fft3d_ps 1201 14.6 0.543 0.556 1.452 1.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.450 1.460 cp_fm_redistribute_end 50 14.0 0.729 1.441 0.730 1.442 cp_fm_diag_elpa_base 50 14.0 0.663 1.361 0.708 1.417 fft_wrap_pw1pw2_140 487 13.2 0.082 0.087 1.402 1.413 grid_collocate_task_list 119 9.7 1.276 1.359 1.276 1.359 ot_diis_step 108 11.5 0.012 0.013 1.328 1.328 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.255 1.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.200 1.237 apply_single 119 13.6 0.000 0.000 1.200 1.236 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.225 1.226 wfi_extrapolate 11 7.9 0.001 0.001 1.192 1.192 make_images_data 4572 15.5 0.039 0.042 0.974 1.181 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.136 1.140 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.878 1.108 cp_fm_cholesky_invert 11 10.9 1.059 1.063 1.059 1.063 acc_transpose_blocks 9144 15.5 0.038 0.039 0.962 0.973 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.922 mp_alltoall_d11v 2130 13.8 0.835 0.916 0.835 0.916 mp_alltoall_z22v 1201 16.6 0.782 0.817 0.782 0.817 mp_allgather_i34 2286 14.5 0.309 0.807 0.309 0.807 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.781 0.789 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 multiply_cannon_sync_h2d 9144 15.5 0.717 0.782 0.717 0.782 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.341 0.765 acc_transpose_blocks_kernels 9144 16.5 0.116 0.118 0.723 0.734 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.680 0.730 dbcsr_complete_redistribute 329 12.2 0.163 0.173 0.650 0.684 mp_sum_l 7207 12.9 0.483 0.649 0.483 0.649 jit_kernel_transpose 5 15.6 0.607 0.619 0.607 0.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.430000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=579.272727, yerr=5.010732 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 734.466048E+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_ISendRecv 4998 189067. MP_Wait 8898 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 0.017 0.036 42.552 42.552 qs_mol_dyn_low 1 2.0 0.003 0.004 42.328 42.335 qs_forces 11 3.9 0.002 0.002 42.267 42.269 qs_energies 11 4.9 0.001 0.001 40.270 40.274 scf_env_do_scf 11 5.9 0.000 0.001 34.328 34.329 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.085 26.087 velocity_verlet 10 3.0 0.001 0.001 23.898 23.903 dbcsr_multiply_generic 2286 12.5 0.100 0.101 18.130 18.334 qs_scf_new_mos 108 7.5 0.001 0.001 16.810 16.907 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.810 16.907 ot_scf_mini 108 9.5 0.002 0.002 15.692 15.794 multiply_cannon 2286 13.5 0.309 0.316 13.922 14.907 multiply_cannon_loop 2286 14.5 0.345 0.350 12.596 13.533 ot_mini 108 10.5 0.001 0.001 9.175 9.298 multiply_cannon_multrec 9144 15.5 3.404 4.825 8.527 8.627 init_scf_loop 11 6.9 0.000 0.000 8.212 8.216 rebuild_ks_matrix 119 8.3 0.000 0.000 7.123 7.275 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.122 7.274 prepare_preconditioner 11 7.9 0.000 0.000 7.241 7.256 make_preconditioner 11 8.9 0.000 0.000 7.241 7.256 qs_ot_get_derivative 108 11.5 0.001 0.001 7.064 7.165 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.730 7.131 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.436 6.574 dbcsr_mm_accdrv_process 12550 15.8 4.131 5.602 4.999 6.343 cp_fm_upper_to_full 72 14.2 3.218 4.675 3.218 4.675 mp_waitall_1 97218 16.6 3.009 4.056 3.009 4.056 init_scf_run 11 5.9 0.000 0.001 3.909 3.909 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.908 3.909 sum_up_and_integrate 119 10.3 0.064 0.066 3.886 3.892 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.856 3.861 calculate_rho_elec 119 8.7 0.118 0.121 3.855 3.860 qs_ot_get_p 119 10.4 0.001 0.001 3.689 3.832 integrate_v_rspace 119 11.3 0.003 0.003 3.821 3.829 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.589 2.998 dbcsr_complete_redistribute 329 12.2 0.287 0.291 2.120 2.984 make_m2s 4572 13.5 0.038 0.038 2.506 2.715 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.772 2.622 make_images 4572 14.5 0.351 0.384 2.384 2.592 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.228 2.495 apply_single 119 13.6 0.000 0.000 2.228 2.495 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.510 2.459 mp_alltoall_i22 627 13.8 1.561 2.438 1.561 2.438 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.393 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.506 2.349 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.298 2.299 calculate_dm_sparse 119 9.5 0.000 0.000 2.188 2.205 pw_transfer 1439 11.6 0.066 0.067 2.181 2.189 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.153 2.154 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.085 2.093 grid_integrate_task_list 119 12.3 2.052 2.084 2.052 2.084 ot_diis_step 108 11.5 0.014 0.014 2.078 2.079 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.004 2.005 density_rs2pw 119 9.7 0.004 0.004 1.915 1.937 mp_sum_l 7207 12.9 1.161 1.913 1.161 1.913 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.811 1.863 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.800 1.821 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.752 1.754 fft3d_ps 1201 14.6 0.568 0.578 1.720 1.726 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.706 1.715 cp_fm_cholesky_invert 11 10.9 1.671 1.674 1.671 1.674 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.658 1.658 cp_fm_diag_elpa_base 50 14.0 1.509 1.563 1.656 1.656 rs_pw_transfer 974 11.9 0.009 0.009 1.537 1.576 grid_collocate_task_list 119 9.7 1.484 1.516 1.484 1.516 hybrid_alltoall_any 4725 16.4 0.088 0.148 1.226 1.491 make_images_data 4572 15.5 0.043 0.046 1.214 1.474 wfi_extrapolate 11 7.9 0.001 0.001 1.456 1.457 potential_pw2rs 119 12.3 0.014 0.015 1.348 1.352 mp_alltoall_d11v 2130 13.8 1.252 1.272 1.252 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.229 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.186 1.209 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.091 1.104 acc_transpose_blocks 9144 15.5 0.037 0.038 1.003 1.070 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.052 multiply_cannon_sync_h2d 9144 15.5 1.046 1.048 1.046 1.048 mp_alltoall_z22v 1201 16.6 1.017 1.039 1.017 1.039 jit_kernel_multiply 6 15.7 0.841 1.015 0.841 1.015 qs_create_task_list 11 7.9 0.001 0.001 0.943 0.954 generate_qs_task_list 11 8.9 0.375 0.394 0.942 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.900 0.914 copy_dbcsr_to_fm 153 11.3 0.002 0.003 0.825 0.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.552000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=693.454545, yerr=9.829118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.181312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 913225. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.141 83.700 83.701 qs_mol_dyn_low 1 2.0 0.004 0.012 83.283 83.295 qs_forces 11 3.9 0.005 0.026 83.160 83.171 qs_energies 11 4.9 0.003 0.018 80.290 80.317 scf_env_do_scf 11 5.9 0.001 0.002 71.272 71.275 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.485 65.487 dbcsr_multiply_generic 2055 12.4 0.107 0.111 52.510 52.788 qs_scf_new_mos 99 7.5 0.000 0.001 48.532 48.649 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.532 48.649 ot_scf_mini 99 9.5 0.002 0.003 46.121 46.207 multiply_cannon 2055 13.4 0.178 0.187 42.781 43.824 multiply_cannon_loop 2055 14.4 1.564 1.606 41.795 42.880 velocity_verlet 10 3.0 0.003 0.019 42.421 42.422 ot_mini 99 10.5 0.001 0.003 27.957 28.041 qs_ot_get_derivative 99 11.5 0.001 0.009 21.051 21.165 multiply_cannon_multrec 49320 15.4 12.041 12.789 17.094 17.871 rebuild_ks_matrix 110 8.3 0.000 0.001 14.467 14.614 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.019 14.466 14.613 mp_waitall_1 241148 16.1 12.057 12.919 12.057 12.919 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.702 12.822 multiply_cannon_sync_h2d 49320 15.4 9.917 10.454 9.917 10.454 qs_ot_get_p 110 10.4 0.001 0.001 9.574 9.664 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.089 8.477 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.154 7.701 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.151 7.683 apply_single 110 13.6 0.000 0.001 7.151 7.683 sum_up_and_integrate 110 10.3 0.037 0.047 7.061 7.078 integrate_v_rspace 110 11.3 0.003 0.015 7.024 7.051 init_scf_run 11 5.9 0.000 0.001 6.955 6.955 scf_env_initial_rho_setup 11 6.9 0.000 0.003 6.954 6.955 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.839 6.881 ot_diis_step 99 11.5 0.006 0.008 6.570 6.571 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.229 6.374 calculate_rho_elec 110 8.6 0.020 0.024 6.228 6.373 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.120 6.160 mp_sum_l 6514 12.8 5.378 6.128 5.378 6.128 init_scf_loop 11 6.9 0.001 0.003 5.756 5.766 dbcsr_mm_accdrv_process 87628 16.1 2.101 2.188 4.932 5.194 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.078 5.079 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.583 4.584 cp_fm_redistribute_end 48 14.0 3.986 4.555 3.989 4.556 cp_fm_diag_elpa_base 48 14.0 0.561 4.495 0.565 4.519 make_m2s 4110 13.4 0.061 0.066 3.972 4.073 rs_pw_transfer 902 11.9 0.011 0.013 3.818 4.001 wfi_extrapolate 11 7.9 0.001 0.003 3.993 3.993 make_images 4110 14.4 0.177 0.190 3.875 3.979 calculate_dm_sparse 110 9.5 0.000 0.001 3.778 3.905 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.779 3.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.635 3.640 prepare_preconditioner 11 7.9 0.000 0.000 3.501 3.522 make_preconditioner 11 8.9 0.000 0.002 3.501 3.522 density_rs2pw 110 9.6 0.004 0.004 3.312 3.480 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.422 3.476 grid_integrate_task_list 110 12.3 3.255 3.392 3.255 3.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.310 3.344 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.286 3.329 pw_transfer 1331 11.6 0.054 0.065 2.904 2.974 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.817 2.890 calculate_first_density_matrix 1 7.0 0.000 0.003 2.870 2.882 potential_pw2rs 110 12.3 0.006 0.007 2.658 2.698 jit_kernel_multiply 13 15.9 2.547 2.568 2.547 2.568 mp_alltoall_d11v 2046 13.8 2.109 2.447 2.109 2.447 fft_wrap_pw1pw2_140 451 13.1 0.169 0.189 2.349 2.426 fft3d_ps 1111 14.6 0.753 0.840 2.354 2.417 acc_transpose_blocks 49320 15.4 0.221 0.230 2.278 2.351 mp_sum_d 3879 11.9 1.691 2.251 1.691 2.251 grid_collocate_task_list 110 9.6 2.086 2.202 2.086 2.202 mp_waitany 14300 13.8 1.849 2.104 1.849 2.104 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.908 1.933 make_images_data 4110 15.4 0.043 0.046 1.800 1.925 cp_fm_cholesky_invert 11 10.9 1.886 1.890 1.886 1.890 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.550 1.828 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.642 1.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.701000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.984810 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 587.874304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3531366. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.064 0.136 70.697 70.698 qs_mol_dyn_low 1 2.0 0.003 0.004 70.187 70.198 qs_forces 11 3.9 0.002 0.002 70.121 70.122 qs_energies 11 4.9 0.015 0.045 66.702 66.707 scf_env_do_scf 11 5.9 0.001 0.002 57.106 57.110 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 49.243 49.245 dbcsr_multiply_generic 2055 12.4 0.115 0.122 38.605 38.818 velocity_verlet 10 3.0 0.001 0.001 36.464 36.465 qs_scf_new_mos 99 7.5 0.001 0.001 33.202 33.335 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.202 33.335 multiply_cannon 2055 13.4 0.224 0.246 31.645 32.693 ot_scf_mini 99 9.5 0.003 0.004 31.529 31.643 multiply_cannon_loop 2055 14.4 0.931 0.952 30.290 31.222 ot_mini 99 10.5 0.001 0.001 18.624 18.758 multiply_cannon_multrec 24660 15.4 7.615 9.386 14.068 15.704 rebuild_ks_matrix 110 8.3 0.000 0.001 13.819 13.958 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 13.818 13.958 qs_ot_get_derivative 99 11.5 0.001 0.001 12.721 12.838 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.089 12.221 mp_waitall_1 186928 16.3 8.244 10.729 8.244 10.729 multiply_cannon_sync_h2d 24660 15.4 7.024 8.176 7.024 8.176 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.344 7.843 init_scf_loop 11 6.9 0.001 0.003 7.825 7.829 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.587 7.289 apply_single 110 13.6 0.000 0.001 6.587 7.289 init_scf_run 11 5.9 0.000 0.001 7.077 7.083 scf_env_initial_rho_setup 11 6.9 0.001 0.003 7.077 7.083 sum_up_and_integrate 110 10.3 0.053 0.060 6.671 6.709 integrate_v_rspace 110 11.3 0.002 0.002 6.618 6.657 dbcsr_mm_accdrv_process 52282 16.1 4.872 5.723 6.295 6.566 qs_ot_get_p 110 10.4 0.001 0.001 6.247 6.400 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.894 5.905 calculate_rho_elec 110 8.6 0.039 0.047 5.893 5.904 ot_diis_step 99 11.5 0.010 0.010 5.834 5.834 prepare_preconditioner 11 7.9 0.000 0.000 5.777 5.793 make_preconditioner 11 8.9 0.001 0.002 5.777 5.793 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.805 5.521 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.359 5.518 make_m2s 4110 13.4 0.058 0.063 4.378 4.832 make_images 4110 14.4 0.396 0.442 4.267 4.717 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.229 4.256 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.748 3.748 pw_transfer 1331 11.6 0.066 0.071 3.483 3.643 wfi_extrapolate 11 7.9 0.001 0.001 3.554 3.554 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.376 3.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.463 3.465 calculate_first_density_matrix 1 7.0 0.001 0.008 3.434 3.443 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.327 3.391 density_rs2pw 110 9.6 0.004 0.004 3.165 3.373 grid_integrate_task_list 110 12.3 3.128 3.318 3.128 3.318 rs_pw_transfer 902 11.9 0.012 0.014 3.041 3.296 calculate_dm_sparse 110 9.5 0.001 0.001 3.256 3.288 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.109 3.158 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.150 3.151 cp_fm_redistribute_end 48 14.0 2.341 3.118 2.343 3.118 cp_fm_diag_elpa_base 48 14.0 0.740 3.002 0.771 3.089 fft_wrap_pw1pw2_140 451 13.1 0.200 0.217 2.804 2.961 make_images_data 4110 15.4 0.047 0.050 2.452 2.902 fft3d_ps 1111 14.6 1.075 1.267 2.734 2.881 hybrid_alltoall_any 4261 16.3 0.102 0.439 2.163 2.831 cp_fm_cholesky_invert 11 10.9 2.716 2.723 2.716 2.723 mp_sum_l 6514 12.8 1.979 2.669 1.979 2.669 jit_kernel_multiply 12 16.3 1.072 2.624 1.072 2.624 potential_pw2rs 110 12.3 0.008 0.008 2.523 2.574 grid_collocate_task_list 110 9.6 2.120 2.299 2.120 2.299 mp_alltoall_d11v 2046 13.8 1.888 2.255 1.888 2.255 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.991 2.013 qs_energies_init_hamiltonians 11 5.9 0.004 0.032 1.945 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.905 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.819 1.831 mp_allgather_i34 2055 14.4 0.680 1.780 0.680 1.780 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.784 1.779 acc_transpose_blocks 24660 15.4 0.108 0.113 1.677 1.696 mp_irecv_dv 57340 16.2 0.660 1.564 0.660 1.564 mp_alltoall_z22v 1111 16.6 1.390 1.486 1.390 1.486 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.475 dbcsr_complete_redistribute 325 12.2 0.239 0.293 1.195 1.465 mp_waitany 10164 13.8 1.184 1.425 1.184 1.425 cp_fm_cholesky_decompose 22 10.9 1.406 1.414 1.406 1.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.698000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.090909, yerr=7.216831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 657.670144E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.098 60.591 60.593 qs_mol_dyn_low 1 2.0 0.003 0.003 60.173 60.196 qs_forces 11 3.9 0.002 0.002 60.102 60.103 qs_energies 11 4.9 0.003 0.017 56.968 56.981 scf_env_do_scf 11 5.9 0.001 0.008 48.945 48.945 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.936 39.936 velocity_verlet 10 3.0 0.002 0.012 32.859 32.861 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.875 29.162 qs_scf_new_mos 99 7.5 0.001 0.001 25.433 25.540 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.432 25.539 ot_scf_mini 99 9.5 0.002 0.003 24.194 24.316 multiply_cannon 2055 13.4 0.212 0.222 22.310 23.650 multiply_cannon_loop 2055 14.4 0.620 0.635 21.053 22.083 ot_mini 99 10.5 0.001 0.001 13.985 14.111 rebuild_ks_matrix 110 8.3 0.000 0.000 12.157 12.299 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.156 12.299 multiply_cannon_multrec 16440 15.4 3.889 5.049 9.679 10.896 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.749 10.878 mp_waitall_1 146766 16.3 7.391 10.362 7.391 10.362 qs_ot_get_derivative 99 11.5 0.001 0.001 9.481 9.600 init_scf_loop 11 6.9 0.001 0.006 8.973 8.974 prepare_preconditioner 11 7.9 0.000 0.000 7.202 7.217 make_preconditioner 11 8.9 0.000 0.003 7.202 7.217 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.380 7.177 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.523 6.893 sum_up_and_integrate 110 10.3 0.061 0.061 6.439 6.453 integrate_v_rspace 110 11.3 0.002 0.003 6.378 6.393 dbcsr_mm_accdrv_process 34862 16.1 4.606 5.234 5.646 5.807 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.108 5.632 apply_single 110 13.6 0.000 0.000 5.108 5.631 init_scf_run 11 5.9 0.000 0.001 5.609 5.609 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.609 5.609 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.546 5.558 calculate_rho_elec 110 8.6 0.058 0.058 5.545 5.557 qs_ot_get_p 110 10.4 0.001 0.001 5.376 5.522 make_m2s 4110 13.4 0.050 0.052 4.283 4.653 make_images 4110 14.4 0.390 0.505 4.166 4.535 ot_diis_step 99 11.5 0.011 0.011 4.472 4.472 multiply_cannon_sync_h2d 16440 15.4 3.661 4.323 3.661 4.323 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.209 3.946 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.695 3.701 grid_integrate_task_list 110 12.3 3.171 3.413 3.171 3.413 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.349 3.350 pw_transfer 1331 11.6 0.065 0.072 3.183 3.193 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.077 3.090 density_rs2pw 110 9.6 0.004 0.004 2.849 3.053 wfi_extrapolate 11 7.9 0.001 0.001 3.002 3.002 cp_fm_cholesky_invert 11 10.9 2.932 2.941 2.932 2.941 rs_pw_transfer 902 11.9 0.010 0.011 2.762 2.937 make_images_data 4110 15.4 0.044 0.047 2.454 2.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.892 2.893 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.770 2.773 hybrid_alltoall_any 4261 16.3 0.105 0.380 2.180 2.755 cp_fm_redistribute_end 48 14.0 1.722 2.743 1.724 2.744 cp_fm_diag_elpa_base 48 14.0 0.956 2.597 1.013 2.705 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.542 2.609 fft_wrap_pw1pw2_140 451 13.1 0.209 0.212 2.589 2.603 calculate_dm_sparse 110 9.5 0.001 0.001 2.533 2.566 calculate_first_density_matrix 1 7.0 0.000 0.003 2.521 2.522 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.438 2.504 multiply_cannon_metrocomm4 14385 15.4 0.043 0.047 0.877 2.498 fft3d_ps 1111 14.6 1.063 1.072 2.394 2.410 mp_sum_l 6514 12.8 1.692 2.378 1.692 2.378 mp_irecv_dv 48980 15.7 0.809 2.369 0.809 2.369 grid_collocate_task_list 110 9.6 2.176 2.343 2.176 2.343 potential_pw2rs 110 12.3 0.010 0.011 2.281 2.300 mp_alltoall_d11v 2046 13.8 1.789 2.124 1.789 2.124 dbcsr_complete_redistribute 325 12.2 0.323 0.354 1.504 1.974 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.954 1.955 cp_fm_upper_to_full 70 13.6 1.386 1.846 1.386 1.846 cp_fm_cholesky_decompose 22 10.9 1.723 1.742 1.723 1.742 mp_allgather_i34 2055 14.4 0.581 1.678 0.581 1.678 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.610 1.624 jit_kernel_multiply 8 16.6 0.653 1.522 0.653 1.522 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.480 1.493 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.998 1.454 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.359 1.448 mp_waitany 17072 13.8 1.134 1.342 1.134 1.342 acc_transpose_blocks 16440 15.4 0.071 0.073 1.303 1.319 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.286 1.296 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.593000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.636364, yerr=7.737426 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 737.681408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.180 67.105 67.106 qs_mol_dyn_low 1 2.0 0.005 0.017 66.645 66.658 qs_forces 11 3.9 0.008 0.050 66.529 66.535 qs_energies 11 4.9 0.011 0.054 63.074 63.092 scf_env_do_scf 11 5.9 0.001 0.001 54.242 54.245 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.829 41.830 velocity_verlet 10 3.0 0.003 0.017 38.239 38.243 dbcsr_multiply_generic 2055 12.4 0.116 0.118 30.176 30.380 qs_scf_new_mos 99 7.5 0.001 0.001 27.032 27.128 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.032 27.127 ot_scf_mini 99 9.5 0.003 0.003 25.363 25.454 multiply_cannon 2055 13.4 0.246 0.263 22.781 23.701 multiply_cannon_loop 2055 14.4 0.891 0.912 21.073 21.717 ot_mini 99 10.5 0.001 0.002 14.535 14.641 multiply_cannon_multrec 24660 15.4 4.176 6.894 12.601 13.764 init_scf_loop 11 6.9 0.001 0.003 12.367 12.368 rebuild_ks_matrix 110 8.3 0.000 0.000 12.200 12.316 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.199 12.316 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.803 10.903 prepare_preconditioner 11 7.9 0.000 0.000 10.586 10.601 make_preconditioner 11 8.9 0.000 0.001 10.586 10.601 qs_ot_get_derivative 99 11.5 0.001 0.002 10.310 10.400 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.771 10.263 dbcsr_mm_accdrv_process 52304 16.0 6.940 8.351 8.280 9.236 mp_waitall_1 126806 16.4 4.801 7.273 4.801 7.273 sum_up_and_integrate 110 10.3 0.068 0.072 6.518 6.533 integrate_v_rspace 110 11.3 0.003 0.016 6.450 6.463 make_m2s 4110 13.4 0.060 0.061 5.642 6.222 make_images 4110 14.4 0.572 0.690 5.499 6.074 qs_ot_get_p 110 10.4 0.001 0.001 5.732 5.876 init_scf_run 11 5.9 0.000 0.001 5.873 5.875 scf_env_initial_rho_setup 11 6.9 0.002 0.018 5.873 5.875 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.838 5.847 calculate_rho_elec 110 8.6 0.077 0.081 5.837 5.847 cp_fm_upper_to_full 70 13.8 3.375 4.787 3.375 4.787 ot_diis_step 99 11.5 0.011 0.013 4.183 4.183 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.080 4.148 apply_single 110 13.6 0.000 0.000 4.080 4.147 dbcsr_complete_redistribute 325 12.2 0.414 0.466 2.789 3.950 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.755 3.769 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.443 3.497 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.467 3.471 pw_transfer 1331 11.6 0.064 0.072 3.435 3.464 make_images_data 4110 15.4 0.046 0.050 2.831 3.427 grid_integrate_task_list 110 12.3 3.274 3.421 3.274 3.421 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.221 3.363 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.329 3.362 hybrid_alltoall_any 4261 16.3 0.119 0.454 2.471 3.353 multiply_cannon_sync_h2d 24660 15.4 3.181 3.305 3.181 3.305 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.251 3.252 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.165 3.213 density_rs2pw 110 9.6 0.004 0.004 2.962 3.178 wfi_extrapolate 11 7.9 0.004 0.029 3.158 3.158 cp_fm_cholesky_invert 11 10.9 3.086 3.094 3.086 3.094 calculate_dm_sparse 110 9.5 0.001 0.001 3.003 3.040 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.938 2.940 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.806 2.930 mp_alltoall_i22 605 13.7 1.693 2.906 1.693 2.906 rs_pw_transfer 902 11.9 0.010 0.011 2.657 2.891 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.769 2.802 fft3d_ps 1111 14.6 1.069 1.107 2.634 2.662 calculate_first_density_matrix 1 7.0 0.003 0.020 2.618 2.625 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.613 2.616 cp_fm_redistribute_end 48 14.0 1.301 2.578 1.303 2.580 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.520 2.556 cp_fm_diag_elpa_base 48 14.0 1.194 2.434 1.273 2.547 grid_collocate_task_list 110 9.6 2.219 2.424 2.219 2.424 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.306 2.308 potential_pw2rs 110 12.3 0.012 0.013 2.267 2.288 mp_alltoall_d11v 2046 13.8 1.996 2.284 1.996 2.284 cp_fm_cholesky_decompose 22 10.9 1.844 1.892 1.844 1.892 jit_kernel_multiply 10 15.7 1.011 1.817 1.011 1.817 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.754 1.792 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.658 acc_transpose_blocks 24660 15.4 0.102 0.104 1.595 1.639 mp_sum_l 6514 12.8 1.115 1.636 1.115 1.636 mp_allgather_i34 2055 14.4 0.806 1.609 0.806 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.580 1.598 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.856 1.549 mp_irecv_dv 62702 16.1 0.758 1.469 0.758 1.469 mp_alltoall_z22v 1111 16.6 1.348 1.429 1.348 1.429 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.252 1.351 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.106000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.000000, yerr=11.337789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 832.811008E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 58.730 58.731 qs_mol_dyn_low 1 2.0 0.003 0.004 57.827 58.416 qs_forces 11 3.9 0.002 0.003 57.719 57.721 qs_energies 11 4.9 0.002 0.002 54.076 54.080 scf_env_do_scf 11 5.9 0.000 0.001 45.212 45.212 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 36.913 36.913 velocity_verlet 10 3.0 0.001 0.001 31.822 31.826 dbcsr_multiply_generic 2055 12.4 0.106 0.107 23.633 23.766 qs_scf_new_mos 99 7.5 0.001 0.001 22.623 22.684 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.622 22.684 ot_scf_mini 99 9.5 0.002 0.002 21.357 21.393 multiply_cannon 2055 13.4 0.251 0.272 17.726 19.071 multiply_cannon_loop 2055 14.4 0.325 0.338 16.245 16.475 rebuild_ks_matrix 110 8.3 0.000 0.000 11.554 11.600 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.553 11.600 ot_mini 99 10.5 0.001 0.001 10.958 10.993 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.302 10.349 multiply_cannon_multrec 8220 15.4 3.230 4.608 7.509 8.512 init_scf_loop 11 6.9 0.000 0.000 8.241 8.242 mp_waitall_1 106626 16.5 6.558 8.240 6.558 8.240 qs_ot_get_derivative 99 11.5 0.001 0.001 7.047 7.081 prepare_preconditioner 11 7.9 0.000 0.000 6.590 6.594 make_preconditioner 11 8.9 0.000 0.000 6.590 6.594 qs_ot_get_p 110 10.4 0.001 0.001 6.397 6.438 sum_up_and_integrate 110 10.3 0.080 0.081 6.296 6.308 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.164 6.245 integrate_v_rspace 110 11.3 0.003 0.003 6.215 6.227 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.724 5.739 calculate_rho_elec 110 8.6 0.115 0.115 5.724 5.738 init_scf_run 11 5.9 0.000 0.001 5.688 5.688 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.688 5.688 dbcsr_mm_accdrv_process 17442 15.9 2.739 3.778 4.148 5.065 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.801 4.806 make_m2s 4110 13.4 0.039 0.040 4.477 4.792 make_images 4110 14.4 0.635 0.693 4.346 4.660 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.053 4.496 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.468 4.468 ot_diis_step 99 11.5 0.012 0.012 3.883 3.884 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.839 apply_single 110 13.6 0.000 0.000 3.776 3.839 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.834 3.836 cp_fm_redistribute_end 48 14.0 0.970 3.779 0.977 3.780 cp_fm_diag_elpa_base 48 14.0 2.612 3.559 2.791 3.728 grid_integrate_task_list 110 12.3 3.355 3.517 3.355 3.517 cp_fm_cholesky_invert 11 10.9 3.324 3.329 3.324 3.329 pw_transfer 1331 11.6 0.065 0.070 3.283 3.298 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.176 3.189 multiply_cannon_sync_h2d 8220 15.4 2.919 3.059 2.919 3.059 make_images_data 4110 15.4 0.039 0.043 2.459 2.947 density_rs2pw 110 9.6 0.004 0.004 2.794 2.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.885 2.886 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.377 2.840 wfi_extrapolate 11 7.9 0.001 0.001 2.821 2.821 calculate_first_density_matrix 1 7.0 0.000 0.000 2.773 2.775 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 2.664 2.666 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.643 2.655 calculate_dm_sparse 110 9.5 0.001 0.001 2.505 2.536 rs_pw_transfer 902 11.9 0.010 0.010 2.403 2.525 grid_collocate_task_list 110 9.6 2.327 2.500 2.327 2.500 fft3d_ps 1111 14.6 1.112 1.132 2.435 2.455 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.403 2.414 mp_alltoall_d11v 2046 13.8 1.742 2.101 1.742 2.101 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.050 2.066 potential_pw2rs 110 12.3 0.015 0.015 2.048 2.053 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.990 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.964 1.983 cp_fm_cholesky_decompose 22 10.9 1.805 1.819 1.805 1.819 mp_allgather_i34 2055 14.4 0.663 1.785 0.663 1.785 dbcsr_complete_redistribute 325 12.2 0.581 0.644 1.671 1.766 jit_kernel_multiply 10 15.6 1.101 1.760 1.101 1.760 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.509 1.628 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.447 1.459 qs_create_task_list 11 7.9 0.001 0.002 1.230 1.329 generate_qs_task_list 11 8.9 0.380 0.448 1.229 1.326 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.261 1.285 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.915 1.267 mp_waitany 9240 13.8 1.113 1.236 1.113 1.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.731000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.818182, yerr=15.236429 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 1.348239E+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_ISendRecv 4620 360267. MP_Wait 7524 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 0.020 0.040 88.138 88.138 qs_mol_dyn_low 1 2.0 0.003 0.003 87.809 87.820 qs_forces 11 3.9 0.002 0.002 87.741 87.742 qs_energies 11 4.9 0.001 0.001 83.663 83.665 scf_env_do_scf 11 5.9 0.001 0.001 73.558 73.558 velocity_verlet 10 3.0 0.001 0.001 56.313 56.319 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.132 44.134 init_scf_loop 11 6.9 0.000 0.000 29.339 29.341 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.922 29.001 qs_scf_new_mos 99 7.5 0.001 0.001 27.452 27.508 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.451 27.507 prepare_preconditioner 11 7.9 0.000 0.000 27.372 27.377 make_preconditioner 11 8.9 0.000 0.000 27.372 27.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.576 26.843 ot_scf_mini 99 9.5 0.002 0.002 25.641 25.675 multiply_cannon 2055 13.4 0.351 0.371 21.545 22.394 multiply_cannon_loop 2055 14.4 0.342 0.344 19.678 19.939 cp_fm_upper_to_full 70 14.2 12.875 18.384 12.875 18.384 ot_mini 99 10.5 0.001 0.001 13.898 13.930 rebuild_ks_matrix 110 8.3 0.001 0.001 13.317 13.359 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.316 13.359 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.110 12.149 dbcsr_complete_redistribute 325 12.2 1.026 1.039 7.488 10.709 multiply_cannon_multrec 8220 15.4 4.369 4.530 9.522 9.615 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.401 9.612 qs_ot_get_derivative 99 11.5 0.001 0.001 9.340 9.374 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.780 8.954 mp_waitall_1 87304 16.6 7.969 8.810 7.969 8.810 mp_alltoall_i22 605 13.7 5.434 8.663 5.434 8.663 sum_up_and_integrate 110 10.3 0.151 0.153 6.815 6.831 integrate_v_rspace 110 11.3 0.003 0.004 6.664 6.681 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.629 6.668 calculate_rho_elec 110 8.6 0.227 0.227 6.628 6.667 qs_ot_get_p 110 10.4 0.001 0.001 6.529 6.576 make_m2s 4110 13.4 0.043 0.044 5.455 5.967 cp_fm_cholesky_invert 11 10.9 5.932 5.937 5.932 5.937 init_scf_run 11 5.9 0.000 0.001 5.830 5.830 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.829 5.830 make_images 4110 14.4 0.879 0.925 5.266 5.777 dbcsr_mm_accdrv_process 11614 15.7 3.202 3.634 5.012 5.247 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.613 5.096 apply_single 110 13.6 0.000 0.000 4.613 5.096 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.692 4.969 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.814 4.821 ot_diis_step 99 11.5 0.015 0.016 4.528 4.529 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.353 4.354 multiply_cannon_sync_h2d 8220 15.4 3.956 3.961 3.956 3.961 pw_transfer 1331 11.6 0.074 0.074 3.832 3.840 grid_integrate_task_list 110 12.3 3.682 3.737 3.682 3.737 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.716 3.724 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.961 3.710 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.698 3.698 cp_fm_diag_elpa_base 48 14.0 3.159 3.358 3.696 3.696 make_images_data 4110 15.4 0.042 0.044 2.958 3.650 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.640 3.641 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.997 3.442 wfi_extrapolate 11 7.9 0.001 0.001 3.341 3.341 calculate_dm_sparse 110 9.5 0.001 0.001 3.210 3.231 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.158 3.166 density_rs2pw 110 9.6 0.004 0.004 3.101 3.128 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.977 2.981 fft3d_ps 1111 14.6 1.270 1.275 2.924 2.935 grid_collocate_task_list 110 9.6 2.646 2.700 2.646 2.700 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.519 2.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.419 2.439 calculate_first_density_matrix 1 7.0 0.000 0.000 2.379 2.380 rs_pw_transfer 902 11.9 0.010 0.010 2.300 2.345 mp_alltoall_d11v 2046 13.8 2.225 2.278 2.225 2.278 cp_fm_cholesky_decompose 22 10.9 2.251 2.267 2.251 2.267 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.172 2.244 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.174 potential_pw2rs 110 12.3 0.021 0.021 2.169 2.172 qs_create_task_list 11 7.9 0.001 0.002 1.902 1.950 generate_qs_task_list 11 8.9 0.742 0.795 1.900 1.949 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.891 1.917 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.812 1.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.138000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1212.727273, yerr=56.954672 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 627.404800E+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 4087 56729. MP_Allreduce 11258 785. MP_Sync 168 MP_Alltoall 2210 3580211. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 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 0.041 0.176 204.510 204.511 qs_mol_dyn_low 1 2.0 0.005 0.016 203.914 203.926 qs_forces 11 3.9 0.007 0.017 203.756 203.759 qs_energies 11 4.9 0.005 0.033 198.006 198.024 scf_env_do_scf 11 5.9 0.001 0.001 180.917 180.922 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 159.642 159.644 dbcsr_multiply_generic 2485 12.5 0.176 0.181 123.352 124.243 velocity_verlet 10 3.0 0.004 0.026 122.501 122.507 qs_scf_new_mos 116 7.6 0.001 0.001 121.337 121.613 qs_scf_loop_do_ot 116 8.6 0.001 0.001 121.336 121.613 ot_scf_mini 116 9.6 0.003 0.004 114.767 115.042 multiply_cannon 2485 13.5 0.235 0.245 100.196 101.659 multiply_cannon_loop 2485 14.5 2.127 2.195 97.991 99.345 ot_mini 116 10.6 0.001 0.002 65.294 65.554 multiply_cannon_multrec 59640 15.5 32.937 34.981 41.494 43.898 qs_ot_get_derivative 116 11.6 0.001 0.002 40.665 40.929 rebuild_ks_matrix 127 8.3 0.001 0.001 32.963 33.233 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.021 32.962 33.233 mp_waitall_1 288940 16.2 28.212 31.105 28.212 31.105 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.618 29.862 multiply_cannon_sync_h2d 59640 15.5 27.102 28.832 27.102 28.832 qs_ot_get_p 127 10.4 0.001 0.001 27.737 27.998 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.075 24.981 apply_single 127 13.6 0.001 0.001 24.075 24.981 ot_diis_step 116 11.6 0.008 0.011 24.245 24.246 init_scf_loop 11 6.9 0.001 0.003 21.191 21.192 qs_ot_p2m_diag 82 11.4 0.077 0.094 21.000 21.112 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.699 18.877 cp_dbcsr_syevd 82 12.4 0.004 0.005 18.497 18.499 multiply_cannon_metrocomm3 59640 15.5 0.115 0.120 15.579 18.047 prepare_preconditioner 11 7.9 0.000 0.000 16.602 16.643 make_preconditioner 11 8.9 0.000 0.002 16.602 16.643 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.796 15.972 cp_fm_diag_elpa 82 13.4 0.000 0.001 15.551 15.557 cp_fm_redistribute_end 82 14.4 12.247 15.465 12.260 15.468 cp_fm_diag_elpa_base 82 14.4 3.161 15.227 3.195 15.346 make_m2s 4970 13.5 0.103 0.110 13.712 14.059 sum_up_and_integrate 127 10.3 0.089 0.106 13.921 13.940 make_images 4970 14.5 0.399 0.420 13.531 13.884 integrate_v_rspace 127 11.3 0.003 0.006 13.832 13.853 init_scf_run 11 5.9 0.000 0.001 12.712 12.713 scf_env_initial_rho_setup 11 6.9 0.000 0.002 12.712 12.713 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.586 12.693 calculate_rho_elec 127 8.7 0.044 0.063 12.586 12.692 cp_fm_cholesky_invert 11 10.9 9.572 9.581 9.572 9.581 mp_sum_l 7804 13.0 8.299 9.412 8.299 9.412 wfi_extrapolate 11 7.9 0.003 0.018 9.238 9.238 dbcsr_mm_accdrv_process 123452 16.2 3.312 3.449 8.123 8.668 calculate_dm_sparse 127 9.5 0.001 0.001 8.485 8.586 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.993 8.137 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.809 7.919 make_images_data 4970 15.5 0.067 0.072 6.645 7.505 multiply_cannon_metrocomm1 59640 15.5 0.084 0.087 5.663 7.483 grid_integrate_task_list 127 12.3 7.033 7.451 7.033 7.451 density_rs2pw 127 9.7 0.006 0.007 6.464 7.069 pw_transfer 1535 11.6 0.074 0.104 6.793 7.069 hybrid_alltoall_any 5155 16.4 0.291 2.245 5.821 6.904 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 6.591 6.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.787 6.795 rs_pw_transfer 1038 11.9 0.016 0.018 5.861 6.425 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.911 5.990 fft_wrap_pw1pw2_140 519 13.2 0.441 0.504 5.552 5.744 fft3d_ps 1281 14.7 2.078 2.556 5.392 5.576 mp_alltoall_d11v 2401 14.1 4.127 5.376 4.127 5.376 grid_collocate_task_list 127 9.7 4.685 5.077 4.685 5.077 cp_fm_cholesky_decompose 22 10.9 4.807 4.821 4.807 4.821 mp_sum_d 4443 12.2 3.891 4.636 3.891 4.636 potential_pw2rs 127 12.3 0.009 0.010 4.544 4.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.511000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.272727, yerr=6.716207 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.270912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 6445848. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.069 195.560 195.561 qs_mol_dyn_low 1 2.0 0.003 0.004 195.101 195.115 qs_forces 11 3.9 0.003 0.003 194.950 194.952 qs_energies 11 4.9 0.002 0.006 188.184 188.196 scf_env_do_scf 11 5.9 0.001 0.003 171.257 171.266 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 137.110 137.113 velocity_verlet 10 3.0 0.001 0.002 122.544 122.547 dbcsr_multiply_generic 2507 12.6 0.188 0.195 98.802 100.176 qs_scf_new_mos 117 7.6 0.001 0.001 97.653 98.203 qs_scf_loop_do_ot 117 8.6 0.001 0.002 97.652 98.202 ot_scf_mini 117 9.6 0.004 0.005 92.820 93.466 multiply_cannon 2507 13.6 0.480 0.529 77.799 82.765 multiply_cannon_loop 2507 14.6 1.256 1.294 74.409 77.017 ot_mini 117 10.6 0.001 0.001 51.074 51.632 mp_waitall_1 226760 16.4 26.211 39.758 26.211 39.758 multiply_cannon_multrec 30084 15.6 21.707 25.535 31.535 35.778 init_scf_loop 11 6.9 0.001 0.003 34.051 34.055 rebuild_ks_matrix 128 8.3 0.001 0.001 33.106 33.595 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.105 33.594 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.840 30.280 prepare_preconditioner 11 7.9 0.000 0.000 29.623 29.678 make_preconditioner 11 8.9 0.000 0.002 29.623 29.678 qs_ot_get_derivative 117 11.6 0.001 0.002 28.896 29.530 multiply_cannon_metrocomm3 30084 15.6 0.093 0.099 16.262 28.924 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.252 28.778 qs_ot_get_p 128 10.4 0.001 0.002 23.458 24.071 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.427 23.708 apply_single 128 13.6 0.001 0.001 22.426 23.708 ot_diis_step 117 11.6 0.014 0.015 22.005 22.007 multiply_cannon_sync_h2d 30084 15.6 18.957 21.237 18.957 21.237 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.373 18.420 cp_fm_cholesky_invert 11 10.9 17.276 17.290 17.276 17.290 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.151 17.153 make_m2s 5014 13.6 0.090 0.095 14.772 16.272 make_images 5014 14.6 1.148 1.341 14.560 16.061 sum_up_and_integrate 128 10.3 0.117 0.135 14.677 14.702 integrate_v_rspace 128 11.3 0.003 0.003 14.560 14.589 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.683 13.696 cp_fm_redistribute_end 83 14.4 7.944 13.590 7.956 13.590 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.510 13.536 calculate_rho_elec 128 8.7 0.088 0.105 13.509 13.536 cp_fm_diag_elpa_base 83 14.4 5.397 13.119 5.609 13.470 init_scf_run 11 5.9 0.000 0.001 12.040 12.042 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.040 12.042 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.388 11.894 multiply_cannon_metrocomm4 27577 15.6 0.097 0.111 3.840 10.818 make_images_data 5014 15.6 0.065 0.072 9.022 10.779 mp_irecv_dv 69486 16.3 3.644 10.428 3.644 10.428 hybrid_alltoall_any 5200 16.5 0.343 1.508 7.958 10.404 dbcsr_mm_accdrv_process 62242 16.2 4.648 5.518 9.295 9.908 wfi_extrapolate 11 7.9 0.001 0.001 8.560 8.560 pw_transfer 1547 11.6 0.086 0.102 8.100 8.165 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.875 7.929 grid_integrate_task_list 128 12.3 7.174 7.624 7.174 7.624 density_rs2pw 128 9.7 0.006 0.006 7.104 7.479 cp_fm_cholesky_decompose 22 10.9 7.376 7.475 7.376 7.475 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.344 7.288 fft_wrap_pw1pw2_140 523 13.2 0.467 0.512 6.771 6.818 calculate_dm_sparse 128 9.5 0.001 0.001 6.614 6.782 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.429 6.437 rs_pw_transfer 1046 11.9 0.014 0.016 6.040 6.411 mp_sum_l 7870 13.0 4.469 6.353 4.469 6.353 fft3d_ps 1291 14.7 2.780 2.923 6.250 6.282 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.488 5.638 mp_alltoall_d11v 2415 14.1 4.927 5.610 4.927 5.610 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.428 5.495 mp_allgather_i34 2507 14.6 1.957 5.311 1.957 5.311 grid_collocate_task_list 128 9.7 4.910 5.264 4.910 5.264 potential_pw2rs 128 12.3 0.015 0.017 5.110 5.132 dbcsr_complete_redistribute 395 12.7 0.769 0.848 3.396 4.277 mp_sum_d 4455 12.2 2.729 4.089 2.729 4.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.561000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.909091, yerr=5.567022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 946.167808E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 180.785 180.787 qs_mol_dyn_low 1 2.0 0.003 0.004 180.343 180.357 qs_forces 11 3.9 0.003 0.003 180.240 180.246 qs_energies 11 4.9 0.001 0.002 173.666 173.679 scf_env_do_scf 11 5.9 0.001 0.001 157.306 157.309 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.349 121.350 velocity_verlet 10 3.0 0.001 0.001 115.031 115.033 qs_scf_new_mos 117 7.6 0.001 0.001 84.212 84.593 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.211 84.592 dbcsr_multiply_generic 2507 12.6 0.183 0.188 82.773 83.905 ot_scf_mini 117 9.6 0.003 0.004 80.046 80.463 multiply_cannon 2507 13.6 0.497 0.522 62.420 66.947 multiply_cannon_loop 2507 14.6 0.860 0.885 59.181 61.814 ot_mini 117 10.6 0.001 0.001 43.150 43.587 init_scf_loop 11 6.9 0.000 0.000 35.841 35.842 mp_waitall_1 178456 16.5 26.538 35.598 26.538 35.598 prepare_preconditioner 11 7.9 0.000 0.000 31.696 31.748 make_preconditioner 11 8.9 0.000 0.000 31.696 31.748 rebuild_ks_matrix 128 8.3 0.001 0.001 30.587 31.056 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.040 30.587 31.055 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.313 30.735 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.619 28.047 multiply_cannon_metrocomm3 20056 15.6 0.061 0.066 15.834 25.331 multiply_cannon_multrec 20056 15.6 13.191 16.385 21.840 25.016 qs_ot_get_derivative 117 11.6 0.001 0.002 23.194 23.622 qs_ot_get_p 128 10.4 0.001 0.001 22.100 22.646 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.102 21.124 apply_single 128 13.6 0.001 0.001 20.101 21.123 ot_diis_step 117 11.6 0.018 0.018 19.852 19.852 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.729 17.747 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.672 16.673 make_m2s 5014 13.6 0.082 0.087 15.174 16.209 multiply_cannon_sync_h2d 20056 15.6 14.301 16.041 14.301 16.041 make_images 5014 14.6 1.172 1.263 14.940 15.973 cp_fm_cholesky_invert 11 10.9 14.935 14.945 14.935 14.945 sum_up_and_integrate 128 10.3 0.132 0.145 14.477 14.502 integrate_v_rspace 128 11.3 0.003 0.004 14.344 14.374 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.315 13.378 calculate_rho_elec 128 8.7 0.131 0.146 13.314 13.377 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.308 13.312 cp_fm_redistribute_end 83 14.4 5.000 13.222 5.022 13.226 cp_fm_diag_elpa_base 83 14.4 7.757 12.614 8.175 13.098 init_scf_run 11 5.9 0.000 0.001 11.328 11.329 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.328 11.328 make_images_data 5014 15.6 0.061 0.068 9.400 10.837 hybrid_alltoall_any 5200 16.5 0.431 1.969 8.053 9.846 multiply_cannon_metrocomm4 17549 15.6 0.059 0.068 3.487 9.491 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.188 9.488 mp_irecv_dv 50230 16.2 3.367 9.246 3.367 9.246 dbcsr_mm_accdrv_process 41502 16.2 4.475 5.264 8.115 8.288 pw_transfer 1547 11.6 0.085 0.104 8.002 8.119 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.779 7.901 grid_integrate_task_list 128 12.3 7.394 7.883 7.394 7.883 cp_fm_cholesky_decompose 22 10.9 7.596 7.639 7.596 7.639 wfi_extrapolate 11 7.9 0.001 0.001 7.546 7.546 cp_fm_upper_to_full 105 14.5 5.727 7.413 5.727 7.413 density_rs2pw 128 9.7 0.005 0.006 6.742 7.054 fft_wrap_pw1pw2_140 523 13.2 0.476 0.534 6.748 6.871 dbcsr_complete_redistribute 395 12.7 1.237 1.291 4.875 6.650 calculate_dm_sparse 128 9.5 0.001 0.001 6.044 6.140 fft3d_ps 1291 14.7 2.691 2.886 6.050 6.124 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.721 5.727 rs_pw_transfer 1046 11.9 0.013 0.014 5.365 5.649 grid_collocate_task_list 128 9.7 5.071 5.470 5.071 5.470 mp_alltoall_d11v 2415 14.1 4.453 5.422 4.453 5.422 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.597 5.374 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.565 5.267 mp_allgather_i34 2507 14.6 1.763 5.044 1.763 5.044 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.685 4.806 potential_pw2rs 128 12.3 0.020 0.022 4.770 4.805 mp_sum_l 7870 13.0 3.369 4.758 3.369 4.758 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.097 4.134 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.364 4.064 mp_alltoall_i22 716 14.1 1.976 3.860 1.976 3.860 qs_energies_init_hamiltonians 11 5.9 0.020 0.029 3.731 3.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.787000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=898.090909, yerr=8.016512 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.159025E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56895. MP_Allreduce 11306 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 83 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.084 0.141 191.689 191.690 qs_mol_dyn_low 1 2.0 0.003 0.003 191.148 191.161 qs_forces 11 3.9 0.003 0.003 191.038 191.043 qs_energies 11 4.9 0.006 0.040 183.950 183.984 scf_env_do_scf 11 5.9 0.001 0.001 166.864 166.875 velocity_verlet 10 3.0 0.001 0.001 126.924 126.926 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.185 119.186 qs_scf_new_mos 118 7.6 0.001 0.001 82.861 83.152 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.860 83.151 dbcsr_multiply_generic 2527 12.6 0.192 0.197 80.545 81.135 ot_scf_mini 118 9.6 0.003 0.004 78.315 78.605 multiply_cannon 2527 13.6 0.570 0.615 55.317 58.467 multiply_cannon_loop 2527 14.6 1.198 1.225 51.571 53.018 init_scf_loop 11 6.9 0.001 0.003 47.550 47.551 prepare_preconditioner 11 7.9 0.000 0.000 43.424 43.446 make_preconditioner 11 8.9 0.000 0.001 43.424 43.446 ot_mini 118 10.6 0.001 0.001 43.022 43.339 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.864 41.966 multiply_cannon_multrec 30324 15.6 14.107 19.607 26.102 31.055 rebuild_ks_matrix 129 8.3 0.001 0.001 29.533 29.764 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.533 29.763 mp_waitall_1 154994 16.5 18.500 28.637 18.500 28.637 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.714 26.925 qs_ot_get_derivative 118 11.6 0.001 0.002 23.123 23.422 make_m2s 5054 13.6 0.097 0.103 20.839 21.940 qs_ot_get_p 129 10.4 0.001 0.001 21.314 21.655 make_images 5054 14.6 1.952 2.275 20.528 21.628 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.278 19.849 apply_single 129 13.6 0.001 0.001 19.277 19.848 ot_diis_step 118 11.6 0.018 0.019 19.771 19.772 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.080 17.129 cp_fm_cholesky_invert 11 10.9 16.757 16.767 16.757 16.767 cp_fm_upper_to_full 105 14.7 11.280 16.733 11.280 16.733 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.772 15.773 multiply_cannon_metrocomm3 30324 15.6 0.048 0.050 6.684 15.483 sum_up_and_integrate 129 10.3 0.142 0.154 14.429 14.466 integrate_v_rspace 129 11.3 0.003 0.004 14.287 14.327 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.456 13.479 calculate_rho_elec 129 8.7 0.176 0.192 13.455 13.479 make_images_data 5054 15.6 0.064 0.067 11.341 13.148 dbcsr_complete_redistribute 395 12.7 1.505 1.624 9.266 13.127 multiply_cannon_sync_h2d 30324 15.6 11.842 12.775 11.842 12.775 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.533 12.536 cp_fm_redistribute_end 83 14.4 2.137 12.443 2.156 12.448 cp_fm_diag_elpa_base 83 14.4 9.650 11.829 10.253 12.312 dbcsr_mm_accdrv_process 62756 16.2 7.403 8.537 11.566 12.071 hybrid_alltoall_any 5240 16.5 0.528 2.218 10.105 11.968 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.830 11.686 init_scf_run 11 5.9 0.000 0.001 11.036 11.038 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.036 11.038 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.536 10.256 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.480 9.692 mp_alltoall_i22 718 14.1 5.772 9.556 5.772 9.556 pw_transfer 1559 11.6 0.085 0.101 8.122 8.201 grid_integrate_task_list 129 12.3 7.535 8.004 7.535 8.004 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.895 7.977 cp_fm_cholesky_decompose 22 10.9 7.762 7.849 7.762 7.849 wfi_extrapolate 11 7.9 0.001 0.001 7.645 7.645 multiply_cannon_metrocomm4 25270 15.6 0.075 0.087 2.827 7.480 mp_irecv_dv 76685 16.2 2.681 7.204 2.681 7.204 fft_wrap_pw1pw2_140 527 13.2 0.484 0.496 6.924 7.015 density_rs2pw 129 9.7 0.005 0.006 6.496 6.803 calculate_dm_sparse 129 9.5 0.001 0.001 6.250 6.320 fft3d_ps 1301 14.7 2.816 2.886 6.118 6.184 mp_alltoall_d11v 2423 14.1 5.118 6.096 5.118 6.096 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.676 5.726 grid_collocate_task_list 129 9.7 5.242 5.661 5.242 5.661 rs_pw_transfer 1054 12.0 0.013 0.014 4.956 5.308 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.502 4.626 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.528 4.625 potential_pw2rs 129 12.3 0.022 0.023 4.535 4.564 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.484 4.485 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.274 4.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.690000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.818182, yerr=27.372243 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.523978E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848418. MP_ISendRecv 7740 122880. MP_Wait 20114 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 0.020 0.040 179.167 179.167 qs_mol_dyn_low 1 2.0 0.003 0.004 178.677 178.703 qs_forces 11 3.9 0.003 0.003 178.575 178.577 qs_energies 11 4.9 0.001 0.002 171.245 171.253 scf_env_do_scf 11 5.9 0.001 0.001 154.014 154.021 velocity_verlet 10 3.0 0.001 0.001 117.347 117.351 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.591 116.592 qs_scf_new_mos 118 7.6 0.001 0.001 80.303 80.444 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.302 80.444 dbcsr_multiply_generic 2527 12.6 0.185 0.191 76.025 76.766 ot_scf_mini 118 9.6 0.004 0.004 75.790 75.910 multiply_cannon 2527 13.6 0.589 0.611 55.419 59.779 multiply_cannon_loop 2527 14.6 0.453 0.468 50.408 51.962 ot_mini 118 10.6 0.001 0.001 40.991 41.138 init_scf_loop 11 6.9 0.000 0.000 37.265 37.266 mp_waitall_1 130650 16.6 27.129 35.158 27.129 35.158 prepare_preconditioner 11 7.9 0.000 0.000 33.348 33.377 make_preconditioner 11 8.9 0.000 0.000 33.348 33.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.195 31.468 rebuild_ks_matrix 129 8.3 0.001 0.001 28.942 29.122 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.941 29.121 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.435 26.602 qs_ot_get_p 129 10.4 0.001 0.001 21.585 21.723 multiply_cannon_multrec 10108 15.6 10.502 14.320 17.796 20.983 qs_ot_get_derivative 118 11.6 0.002 0.002 20.577 20.703 ot_diis_step 118 11.6 0.020 0.021 20.339 20.340 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.782 20.264 apply_single 129 13.6 0.001 0.001 19.781 20.263 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 12.632 19.977 make_m2s 5054 13.6 0.068 0.073 16.729 19.298 cp_fm_cholesky_invert 11 10.9 19.287 19.294 19.287 19.294 make_images 5054 14.6 2.326 2.708 16.419 18.993 qs_ot_p2m_diag 83 11.4 0.495 0.501 17.533 17.548 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.341 16.343 sum_up_and_integrate 129 10.3 0.184 0.194 14.340 14.384 integrate_v_rspace 129 11.3 0.004 0.004 14.156 14.209 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.534 13.586 calculate_rho_elec 129 8.7 0.260 0.271 13.533 13.586 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.135 13.137 cp_fm_diag_elpa_base 83 14.4 12.873 12.961 13.128 13.128 make_images_data 5054 15.6 0.053 0.061 10.209 12.911 hybrid_alltoall_any 5240 16.5 0.800 3.691 9.931 12.517 multiply_cannon_sync_h2d 10108 15.6 11.698 12.182 11.698 12.182 init_scf_run 11 5.9 0.000 0.001 10.439 10.439 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.439 10.439 cp_fm_cholesky_decompose 22 10.9 8.383 8.473 8.383 8.473 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.284 8.374 mp_allgather_i34 2527 14.6 3.203 8.317 3.203 8.317 grid_integrate_task_list 129 12.3 7.800 8.202 7.800 8.202 pw_transfer 1559 11.6 0.084 0.093 8.019 8.041 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.797 7.822 dbcsr_mm_accdrv_process 20926 16.1 2.625 3.400 6.926 7.571 multiply_cannon_metrocomm1 10108 15.6 0.029 0.029 4.570 7.544 wfi_extrapolate 11 7.9 0.001 0.001 7.431 7.431 fft_wrap_pw1pw2_140 527 13.2 0.504 0.526 6.750 6.777 density_rs2pw 129 9.7 0.005 0.006 6.233 6.530 calculate_dm_sparse 129 9.5 0.001 0.001 6.169 6.272 fft3d_ps 1301 14.7 2.744 2.813 5.948 5.998 grid_collocate_task_list 129 9.7 5.520 5.831 5.520 5.831 dbcsr_complete_redistribute 395 12.7 2.112 2.171 5.331 5.764 mp_alltoall_d11v 2423 14.1 4.943 5.559 4.943 5.559 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.311 5.319 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.205 5.206 rs_pw_transfer 1054 12.0 0.013 0.013 4.364 4.698 multiply_cannon_metrocomm4 7581 15.6 0.023 0.027 1.870 4.414 mp_irecv_dv 29086 15.9 1.835 4.341 1.835 4.341 potential_pw2rs 129 12.3 0.026 0.027 4.310 4.329 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.194 4.227 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.817 3.863 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.552 3.840 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.477 3.817 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.636 3.753 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.643 3.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=179.167000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1412.454545, yerr=61.005758 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410026932224 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 1957874282496 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 11444723269632 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019194212352 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019194212352 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 92.796614E+12 0.0% 0.0% 100.0% flops max/rank 11.606417E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705505824 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.126718E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342276E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108963904 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58535. MP_Allreduce 11110 1512. MP_Sync 86 MP_Alltoall 1700 36954327. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.047 303.113 303.114 qs_mol_dyn_low 1 2.0 0.003 0.003 302.544 302.559 qs_forces 11 3.9 0.003 0.005 302.439 302.446 qs_energies 11 4.9 0.002 0.002 293.720 293.733 scf_env_do_scf 11 5.9 0.001 0.001 271.051 271.059 velocity_verlet 10 3.0 0.001 0.001 218.254 218.262 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.674 141.677 init_scf_loop 11 6.9 0.000 0.000 129.106 129.113 prepare_preconditioner 11 7.9 0.000 0.000 124.233 124.268 make_preconditioner 11 8.9 0.000 0.000 124.233 124.268 make_full_inverse_cholesky 11 9.9 0.000 0.000 99.630 121.324 qs_scf_new_mos 116 7.6 0.001 0.001 98.576 98.695 qs_scf_loop_do_ot 116 8.6 0.001 0.001 98.575 98.694 ot_scf_mini 116 9.6 0.004 0.004 93.700 93.770 dbcsr_multiply_generic 2485 12.5 0.213 0.220 86.556 87.190 cp_fm_upper_to_full 104 14.8 53.377 76.929 53.377 76.929 multiply_cannon 2485 13.5 0.729 0.773 61.015 61.641 multiply_cannon_loop 2485 14.5 0.467 0.482 57.161 58.642 ot_mini 116 10.6 0.001 0.001 46.707 46.801 dbcsr_complete_redistribute 393 12.7 3.969 4.007 30.515 43.984 copy_fm_to_dbcsr 208 11.6 0.001 0.002 27.000 40.432 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.557 37.838 cp_fm_cholesky_invert 11 10.9 36.090 36.096 36.090 36.096 mp_alltoall_i22 712 14.1 22.348 35.963 22.348 35.963 mp_waitall_1 104546 16.7 31.381 35.405 31.381 35.405 rebuild_ks_matrix 127 8.3 0.001 0.001 34.381 34.483 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.380 34.482 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.012 32.103 qs_ot_get_p 127 10.4 0.001 0.001 30.982 31.114 qs_ot_p2m_diag 82 11.4 0.868 0.873 26.461 26.492 qs_ot_get_derivative 116 11.6 0.002 0.002 25.704 25.771 cp_dbcsr_syevd 82 12.4 0.005 0.006 24.677 24.680 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 21.051 22.314 make_m2s 4970 13.5 0.076 0.079 20.498 21.537 cp_fm_diag_elpa 82 13.4 0.000 0.000 21.244 21.245 cp_fm_diag_elpa_base 82 14.4 16.834 18.503 21.239 21.240 make_images 4970 14.5 3.755 3.902 20.020 21.063 ot_diis_step 116 11.6 0.022 0.022 20.953 20.954 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.196 20.452 apply_single 127 13.6 0.001 0.001 20.195 20.451 multiply_cannon_multrec 9940 15.5 10.424 12.264 17.873 18.017 sum_up_and_integrate 127 10.3 0.319 0.322 16.193 16.281 integrate_v_rspace 127 11.3 0.004 0.004 15.874 15.959 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.559 15.584 calculate_rho_elec 127 8.7 0.479 0.480 15.559 15.583 multiply_cannon_sync_h2d 9940 15.5 15.548 15.557 15.548 15.557 make_images_data 4970 15.5 0.059 0.063 11.088 13.079 init_scf_run 11 5.9 0.000 0.001 12.600 12.601 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.600 12.600 hybrid_alltoall_any 5155 16.4 1.287 3.003 11.016 12.549 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.916 9.964 cp_fm_cholesky_decompose 22 10.9 9.945 9.963 9.945 9.963 wfi_extrapolate 11 7.9 0.001 0.001 9.198 9.198 dbcsr_mm_accdrv_process 20590 16.0 3.774 5.497 7.214 9.148 pw_transfer 1535 11.6 0.089 0.090 8.959 8.975 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.727 8.744 grid_integrate_task_list 127 12.3 8.494 8.675 8.494 8.675 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.010 8.013 mp_alltoall_d11v 2401 14.1 7.553 7.714 7.553 7.714 fft_wrap_pw1pw2_140 519 13.2 0.534 0.538 7.632 7.649 calculate_dm_sparse 127 9.5 0.001 0.001 6.845 6.934 fft3d_ps 1281 14.7 2.741 2.776 6.792 6.805 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.634 6.747 density_rs2pw 127 9.7 0.005 0.005 6.528 6.560 grid_collocate_task_list 127 9.7 6.297 6.334 6.297 6.334 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.231 6.297 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=303.114000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2763.909091, yerr=186.756260 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 1.261752E+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 6799628. MP_ISendRecv 570 19200. MP_Wait 1302 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 0.043 0.065 85.205 85.206 qs_energies 1 2.0 0.000 0.000 84.748 84.755 ls_scf 1 3.0 0.000 0.000 83.834 83.841 dbcsr_multiply_generic 111 6.7 0.018 0.041 72.691 72.870 multiply_cannon 111 7.7 0.041 0.066 55.836 57.512 multiply_cannon_loop 111 8.7 0.211 0.225 52.449 54.270 ls_scf_main 1 4.0 0.000 0.000 52.340 52.341 density_matrix_trs4 2 5.0 0.002 0.003 46.880 46.971 ls_scf_init_scf 1 4.0 0.000 0.000 28.459 28.460 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.345 27.406 mp_waitall_1 11316 10.9 22.367 25.337 22.367 25.337 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.216 25.241 multiply_cannon_multrec 2664 9.7 8.187 8.903 15.483 17.166 multiply_cannon_sync_h2d 2664 9.7 13.718 15.236 13.718 15.236 make_m2s 222 7.7 0.008 0.011 13.201 13.699 make_images 222 8.7 0.099 0.109 13.179 13.679 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.490 12.556 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.456 9.141 make_images_data 222 9.7 0.004 0.005 7.741 8.340 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.628 6.917 7.866 hybrid_alltoall_any 227 10.6 0.216 1.849 6.668 7.736 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.083 6.208 7.083 calculate_norms 4752 9.8 5.546 6.211 5.546 6.211 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.145 mp_sum_l 807 5.4 3.123 4.695 3.123 4.695 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.062 4.406 mp_irecv_dv 6231 10.9 2.045 4.372 2.045 4.372 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.328 3.578 make_images_sizes 222 9.7 0.000 0.000 0.740 3.481 mp_alltoall_i44 222 10.7 0.740 3.481 0.740 3.481 arnoldi_extremal 4 6.8 0.000 0.000 3.414 3.442 arnoldi_normal_ev 4 7.8 0.001 0.004 3.414 3.441 build_subspace 16 8.4 0.009 0.012 3.314 3.317 ls_scf_post 1 4.0 0.000 0.000 3.035 3.041 ls_scf_store_result 1 5.0 0.000 0.000 2.854 2.895 dbcsr_special_finalize 555 9.7 0.005 0.006 2.369 2.769 dbcsr_merge_single_wm 555 10.7 0.454 0.590 2.361 2.761 make_images_pack 222 9.7 2.205 2.635 2.207 2.637 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.375 2.630 dbcsr_sort_data 658 11.4 2.165 2.536 2.165 2.536 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.468 2.067 2.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.227 2.308 buffer_matrices_ensure_size 222 8.7 1.742 2.076 1.742 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.790 1.792 rebuild_ks_matrix 3 7.3 0.000 0.000 1.781 1.782 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.781 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.206000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 2.105151E+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_ISendRecv 282 57600. MP_Wait 828 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 0.056 0.161 92.595 92.596 qs_energies 1 2.0 0.000 0.001 91.832 91.846 ls_scf 1 3.0 0.004 0.016 90.506 90.511 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.278 76.737 multiply_cannon 111 7.7 0.030 0.044 53.543 57.109 ls_scf_main 1 4.0 0.003 0.027 55.155 55.162 multiply_cannon_loop 111 8.7 0.116 0.123 50.185 53.036 density_matrix_trs4 2 5.0 0.002 0.003 49.388 49.627 ls_scf_init_scf 1 4.0 0.001 0.003 31.772 31.773 mp_waitall_1 9246 10.9 21.575 31.048 21.575 31.048 ls_scf_init_matrix_S 1 5.0 0.202 0.809 30.575 30.674 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.330 27.347 multiply_cannon_multrec 1332 9.7 13.089 16.657 22.273 26.995 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.991 21.542 make_m2s 222 7.7 0.006 0.007 15.463 16.076 make_images 222 8.7 1.574 1.961 15.433 16.047 dbcsr_mm_accdrv_process 4041 10.4 0.269 0.431 8.787 10.322 dbcsr_mm_accdrv_process_sort 4041 11.4 8.382 9.890 8.382 9.890 make_images_data 222 9.7 0.004 0.004 8.911 9.825 hybrid_alltoall_any 227 10.6 0.521 2.474 8.314 9.351 mp_sum_l 807 5.4 6.002 9.121 6.002 9.121 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.233 7.722 mp_irecv_dv 3311 11.0 3.214 7.674 3.214 7.674 calculate_norms 2376 9.8 5.997 6.757 5.997 6.757 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.051 6.314 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.758 6.179 multiply_cannon_sync_h2d 1332 9.7 4.756 6.008 4.756 6.008 arnoldi_extremal 4 6.8 0.000 0.000 4.682 4.712 arnoldi_normal_ev 4 7.8 0.003 0.009 4.682 4.711 build_subspace 16 8.4 0.014 0.021 4.420 4.424 ls_scf_post 1 4.0 0.002 0.008 3.575 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.284 3.402 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.160 3.384 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.218 2.741 3.220 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.291 3.135 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.534 2.615 make_images_pack 222 9.7 2.025 2.425 2.027 2.427 mp_allgather_i34 111 8.7 1.057 2.385 1.057 2.385 dbcsr_sort_data 436 11.2 1.827 2.099 1.827 2.099 make_images_sizes 222 9.7 0.000 0.000 0.602 1.917 mp_alltoall_i44 222 10.7 0.602 1.917 0.602 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.596000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 2.717327E+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_ISendRecv 186 57600. MP_Wait 732 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 0.037 0.051 93.923 93.925 qs_energies 1 2.0 0.000 0.000 93.276 93.279 ls_scf 1 3.0 0.000 0.001 91.865 91.868 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.719 76.986 ls_scf_main 1 4.0 0.000 0.001 57.553 57.558 multiply_cannon 111 7.7 0.050 0.080 52.809 56.527 multiply_cannon_loop 111 8.7 0.099 0.106 49.252 53.115 density_matrix_trs4 2 5.0 0.002 0.003 51.636 51.871 mp_waitall_1 7374 11.0 23.803 32.960 23.803 32.960 ls_scf_init_scf 1 4.0 0.000 0.002 30.753 30.755 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.614 29.679 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.216 27.237 multiply_cannon_multrec 888 9.7 12.508 15.587 21.023 24.654 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.076 22.449 make_m2s 222 7.7 0.006 0.007 17.109 18.335 make_images 222 8.7 1.970 2.293 17.071 18.295 hybrid_alltoall_any 227 10.6 0.621 2.875 9.260 10.847 make_images_data 222 9.7 0.003 0.004 9.662 10.716 dbcsr_mm_accdrv_process 3754 10.4 0.251 0.418 8.030 9.229 mp_sum_l 807 5.4 5.259 9.195 5.259 9.195 dbcsr_mm_accdrv_process_sort 3754 11.4 7.661 8.811 7.661 8.811 multiply_cannon_sync_h2d 888 9.7 6.015 7.339 6.015 7.339 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.685 7.186 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.961 7.058 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.464 6.881 mp_irecv_dv 2335 11.1 2.450 6.836 2.450 6.836 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.986 5.221 arnoldi_extremal 4 6.8 0.000 0.000 5.137 5.166 arnoldi_normal_ev 4 7.8 0.001 0.006 5.137 5.166 build_subspace 16 8.4 0.014 0.020 4.826 4.835 calculate_norms 1584 9.8 4.294 4.737 4.294 4.737 mp_allgather_i34 111 8.7 1.385 3.810 1.385 3.810 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.455 3.781 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.591 3.020 3.593 ls_scf_post 1 4.0 0.000 0.000 3.558 3.563 ls_scf_store_result 1 5.0 0.000 0.000 3.309 3.386 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.720 2.813 make_images_sizes 222 9.7 0.000 0.000 1.151 2.397 mp_alltoall_i44 222 10.7 1.151 2.396 1.151 2.396 dbcsr_sort_data 325 11.1 1.854 2.134 1.854 2.134 make_images_pack 222 9.7 1.816 2.123 1.819 2.125 dbcsr_data_release 9322 10.9 1.312 1.896 1.312 1.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.925000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2167.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 3.344245E+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_ISendRecv 138 86400. MP_Wait 600 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 0.073 0.161 97.321 97.322 qs_energies 1 2.0 0.000 0.000 96.538 96.557 ls_scf 1 3.0 0.000 0.001 94.891 94.910 dbcsr_multiply_generic 111 6.7 0.017 0.023 78.697 78.963 ls_scf_main 1 4.0 0.001 0.008 58.961 58.962 multiply_cannon 111 7.7 0.063 0.136 51.683 56.067 density_matrix_trs4 2 5.0 0.002 0.003 52.884 53.045 multiply_cannon_loop 111 8.7 0.114 0.126 46.583 49.644 ls_scf_init_scf 1 4.0 0.000 0.002 32.657 32.660 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.503 31.570 mp_waitall_1 6438 11.0 22.801 29.621 22.801 29.621 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.009 29.021 multiply_cannon_multrec 1332 9.7 14.206 17.354 22.032 24.595 make_m2s 222 7.7 0.007 0.008 21.152 22.605 make_images 222 8.7 3.130 3.593 21.101 22.557 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.325 17.626 make_images_data 222 9.7 0.004 0.004 11.823 13.491 hybrid_alltoall_any 227 10.6 0.796 3.777 11.122 12.941 dbcsr_mm_accdrv_process 3641 10.4 0.194 0.373 7.469 8.972 dbcsr_mm_accdrv_process_sort 3641 11.4 7.096 8.555 7.096 8.555 mp_sum_l 807 5.4 4.277 7.525 4.277 7.525 multiply_cannon_sync_h2d 1332 9.7 5.507 6.646 5.507 6.646 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.057 6.011 mp_irecv_dv 3229 10.9 2.035 5.936 2.035 5.936 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.249 5.875 arnoldi_extremal 4 6.8 0.000 0.000 5.302 5.317 arnoldi_normal_ev 4 7.8 0.001 0.005 5.302 5.317 build_subspace 16 8.4 0.014 0.021 4.959 4.965 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.596 4.824 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.475 4.814 mp_allgather_i34 111 8.7 2.218 4.544 2.218 4.544 calculate_norms 2376 9.8 4.171 4.455 4.171 4.455 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.626 3.936 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.683 3.189 3.685 dbcsr_sort_data 658 11.4 3.073 3.406 3.073 3.406 ls_scf_post 1 4.0 0.005 0.020 3.272 3.292 dbcsr_special_finalize 555 9.7 0.006 0.007 2.815 3.130 dbcsr_merge_single_wm 555 10.7 0.537 0.668 2.807 3.122 ls_scf_store_result 1 5.0 0.000 0.000 2.998 3.070 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.915 2.960 dbcsr_data_release 10477 10.7 1.585 2.453 1.585 2.453 dbcsr_finalize 304 7.8 0.049 0.061 1.807 2.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.322000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 4.640678E+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_ISendRecv 90 115200. MP_Wait 573 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 0.054 0.074 98.987 98.988 qs_energies 1 2.0 0.000 0.000 98.145 98.150 ls_scf 1 3.0 0.000 0.000 96.231 96.235 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.895 78.045 ls_scf_main 1 4.0 0.000 0.000 61.927 61.928 multiply_cannon 111 7.7 0.084 0.124 55.155 61.030 density_matrix_trs4 2 5.0 0.002 0.003 54.924 55.021 multiply_cannon_loop 111 8.7 0.069 0.079 50.554 52.673 mp_waitall_1 5481 11.0 26.052 32.970 26.052 32.970 ls_scf_init_scf 1 4.0 0.000 0.000 30.707 30.713 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.495 29.564 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.329 27.342 multiply_cannon_multrec 444 9.7 14.093 16.325 21.104 24.241 make_m2s 222 7.7 0.004 0.005 17.463 20.008 make_images 222 8.7 3.714 4.423 17.401 19.947 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.089 16.267 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.966 14.660 make_images_data 222 9.7 0.003 0.004 9.800 12.201 hybrid_alltoall_any 227 10.6 0.793 3.771 9.605 12.027 multiply_cannon_sync_h2d 444 9.7 6.465 8.650 6.465 8.650 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.335 6.715 7.857 dbcsr_mm_accdrv_process_sort 3003 11.4 6.401 7.507 6.401 7.507 mp_allgather_i34 111 8.7 2.829 7.008 2.829 7.008 mp_sum_l 807 5.4 3.327 6.375 3.327 6.375 arnoldi_extremal 4 6.8 0.000 0.000 5.897 5.910 arnoldi_normal_ev 4 7.8 0.002 0.005 5.897 5.910 build_subspace 16 8.4 0.015 0.020 5.506 5.516 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.397 4.941 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.614 4.772 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.674 4.522 mp_irecv_dv 1241 11.2 1.657 4.503 1.657 4.503 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.219 4.428 dbcsr_matrix_vector_mult_local 304 10.0 3.693 4.161 3.695 4.163 calculate_norms 792 9.8 3.540 3.660 3.540 3.660 ls_scf_post 1 4.0 0.000 0.000 3.597 3.602 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.394 3.477 ls_scf_store_result 1 5.0 0.000 0.000 3.373 3.425 make_images_sizes 222 9.7 0.000 0.000 1.017 3.306 mp_alltoall_i44 222 10.7 1.017 3.306 1.017 3.306 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.322 dbcsr_merge_all 275 8.9 0.474 0.535 2.053 2.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.988000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3655.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fdee58b7a5e240d23ead32de5f87cde65b0b6616_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 8.734573E+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_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.098 108.006 108.007 qs_energies 1 2.0 0.000 0.000 106.590 106.610 ls_scf 1 3.0 0.000 0.000 103.704 103.722 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.816 77.932 ls_scf_main 1 4.0 0.000 0.000 65.962 65.964 density_matrix_trs4 2 5.0 0.002 0.003 57.086 57.147 multiply_cannon 111 7.7 0.140 0.231 49.931 51.484 multiply_cannon_loop 111 8.7 0.067 0.069 46.498 47.619 ls_scf_init_scf 1 4.0 0.000 0.000 34.066 34.067 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.733 32.749 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.922 29.935 mp_waitall_1 4569 11.1 22.381 26.320 22.381 26.320 make_m2s 222 7.7 0.005 0.005 24.071 25.030 make_images 222 8.7 4.579 4.946 23.965 24.922 multiply_cannon_multrec 444 9.7 17.870 18.606 22.450 23.135 hybrid_alltoall_any 227 10.6 1.657 3.614 13.035 15.871 make_images_data 222 9.7 0.003 0.003 13.277 15.681 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.505 11.238 multiply_cannon_sync_h2d 444 9.7 8.851 8.914 8.851 8.914 arnoldi_extremal 4 6.8 0.000 0.000 7.443 7.458 arnoldi_normal_ev 4 7.8 0.004 0.011 7.443 7.457 build_subspace 16 8.4 0.026 0.036 6.887 6.901 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.533 5.680 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.353 5.045 5.355 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.010 5.259 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.832 4.927 dbcsr_mm_accdrv_process 1814 10.4 0.184 0.314 4.418 4.542 dbcsr_mm_accdrv_process_sort 1814 11.4 4.121 4.248 4.121 4.248 ls_scf_post 1 4.0 0.000 0.000 3.675 3.693 make_images_sizes 222 9.7 0.000 0.000 1.465 3.680 mp_alltoall_i44 222 10.7 1.464 3.680 1.464 3.680 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.431 mp_allgather_i34 111 8.7 1.069 3.296 1.069 3.296 calculate_norms 792 9.8 3.239 3.277 3.239 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.089 3.149 dbcsr_merge_all 275 8.9 0.893 0.918 2.875 2.924 dbcsr_complete_redistribute 5 7.6 1.451 1.486 2.786 2.920 dbcsr_data_release 12724 10.6 2.330 2.870 2.330 2.870 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.856 2.856 matrix_ls_to_qs 2 6.0 0.000 0.000 2.436 2.590 dbcsr_sort_data 325 11.1 2.441 2.504 2.441 2.504 dbcsr_new_transposed 4 7.5 0.244 0.252 2.265 2.287 dbcsr_frobenius_norm 74 6.6 2.059 2.141 2.204 2.249 mp_sum_l 807 5.4 1.497 2.235 1.497 2.235 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.207 dbcsr_add_anytype 103 7.2 0.859 0.892 2.134 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.193 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.007000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6968.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: fdee58b7a5e240d23ead32de5f87cde65b0b6616 Summary: empty Status: OK