=== 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: 1da4a734777d3157e9099d8f027b8c79c676a226 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/01 job id: 44987464 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/02 job id: 44987465 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/03 job id: 44987466 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/04 job id: 44987467 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/05 job id: 44987468 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/06 job id: 44987470 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/07 job id: 44987471 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/08 job id: 44987472 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/09 job id: 44987473 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/10 job id: 44987474 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/11 job id: 44987476 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/12 job id: 44987478 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/13 job id: 44987480 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/14 job id: 44987481 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/15 job id: 44987482 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/16 job id: 44987483 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/17 job id: 44987484 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/18 job id: 44987486 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/19 job id: 44987487 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/20 job id: 44987488 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/21 job id: 44987489 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/22 job id: 44987490 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/23 job id: 44987491 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/24 job id: 44987492 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/25 job id: 44987493 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/26 job id: 44987494 --- 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/1da4a734777d3157e9099d8f027b8c79c676a226_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 0.024 0.034 135.980 135.981 farming_run 1 2.0 135.427 135.428 135.947 135.951 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458459E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.097 116.101 116.166 qs_energies 1 2.0 0.000 0.000 115.876 115.879 mp2_main 1 3.0 0.000 0.000 113.801 113.804 mp2_gpw_main 1 4.0 0.020 0.026 112.804 112.807 mp2_ri_gpw_compute_in 1 5.0 0.172 0.176 94.112 94.500 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.525 55.914 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.173 41.826 47.162 get_2c_integrals 1 6.0 0.000 0.001 37.622 38.411 integrate_v_rspace 273 8.0 0.436 0.449 25.179 30.261 pw_transfer 6555 10.6 0.370 0.391 27.535 28.237 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.180 26.741 grid_integrate_task_list 273 9.0 20.979 26.558 20.979 26.558 fft_wrap_pw1pw2_100 2178 12.4 1.215 1.438 23.692 24.252 compute_2c_integrals 1 7.0 0.003 0.003 19.663 19.664 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.978 19.376 mp2_eri_2c_integrate_gpw 1 9.0 2.362 2.437 18.975 19.375 rpa_ri_compute_en 1 5.0 0.001 0.001 18.581 18.765 cp_fm_cholesky_decompose 12 8.2 17.926 18.762 17.926 18.762 cholesky_decomp 1 7.0 0.000 0.000 16.800 17.625 fft3d_s 5443 13.4 16.226 16.614 16.248 16.636 ao_to_mo_and_store_B_mult_1 272 7.0 10.866 15.602 10.866 15.602 calculate_wavefunction 272 8.0 5.414 5.566 12.565 13.136 rpa_num_int 1 6.0 0.000 0.001 10.542 10.542 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.509 10.539 calc_mat_Q 8 8.0 0.000 0.000 9.335 9.443 contract_S_to_Q 8 9.0 0.000 0.000 8.760 8.862 calc_potential_gpw 544 9.5 0.005 0.006 8.251 8.638 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.275 8.557 potential_pw2rs 545 10.0 0.108 0.110 7.769 8.456 parallel_gemm_fm 14 9.1 0.000 0.000 8.326 8.431 parallel_gemm_fm_cosma 14 10.1 8.326 8.431 8.326 8.431 collocate_single_gaussian 272 10.0 0.039 0.042 7.521 7.728 create_integ_mat 1 6.0 0.022 0.028 7.554 7.554 array2fm 1 7.0 0.000 0.000 6.646 6.969 pw_scatter_s 2720 13.7 4.413 4.634 4.413 4.634 pw_gather_s 2722 13.2 3.905 4.399 3.905 4.399 array2fm_buffer_send 1 8.0 2.996 3.194 2.996 3.194 pw_poisson_solve 545 10.5 1.131 1.183 2.199 2.351 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.807137, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 0.078 0.085 395.415 395.416 farming_run 1 2.0 394.519 394.522 395.322 395.324 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.228255E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 19023437. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.042 208.764 208.764 qs_energies 1 2.0 0.000 0.000 208.522 208.529 scf_env_do_scf 1 3.0 0.000 0.000 106.389 106.389 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.447 105.457 rebuild_ks_matrix 4 6.0 0.000 0.000 105.445 105.455 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.067 105.445 105.455 hfx_ks_matrix 4 8.0 0.001 0.001 105.056 105.059 integrate_four_center 4 9.0 0.144 0.459 105.055 105.059 mp2_main 1 3.0 0.002 0.019 101.834 101.841 mp2_gpw_main 1 4.0 0.037 0.075 100.904 100.916 integrate_four_center_main 4 10.0 0.089 0.519 96.620 98.894 integrate_four_center_bin 263 11.0 96.532 98.888 96.532 98.888 init_scf_loop 1 4.0 0.000 0.000 92.213 92.213 mp2_ri_gpw_compute_in 1 5.0 0.070 0.115 74.783 75.750 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.461 55.426 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.061 47.202 integrate_v_rspace 95 8.0 0.399 0.579 28.437 33.401 pw_transfer 2240 10.6 0.143 0.164 29.886 30.318 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.910 29.359 grid_integrate_task_list 95 9.0 23.729 28.900 23.729 28.900 mp2_ri_gpw_compute_en 1 5.0 0.059 0.109 25.919 27.609 ao_to_mo_and_store_B_mult_1 91 7.0 10.713 27.516 10.713 27.516 fft_wrap_pw1pw2_100 730 12.4 1.253 1.421 26.622 27.044 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.888 24.241 24.252 get_2c_integrals 1 6.0 0.001 0.013 20.214 20.259 compute_2c_integrals 1 7.0 0.003 0.004 19.186 19.202 compute_2c_integrals_loop_lm 1 8.0 0.004 0.027 18.795 19.069 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.862 18.791 19.068 fft3d_s 1823 13.4 18.438 18.819 18.452 18.832 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.175 14.175 calculate_wavefunction 91 8.0 2.024 2.059 9.750 9.998 potential_pw2rs 186 10.0 0.033 0.034 8.626 9.199 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.578 8.672 9.017 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.010 8.194 8.592 local_gemm 172 8.0 8.117 8.448 8.117 8.448 collocate_single_gaussian 91 10.0 0.019 0.065 7.856 8.117 calc_potential_gpw 182 9.5 0.002 0.002 7.868 8.063 mp2_ri_gpw_compute_en_comm 22 7.0 0.523 0.547 7.331 7.695 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.423 6.351 6.423 mp_sync 38 10.4 3.110 5.985 3.110 5.985 mp_sendrecv_dm3 2068 8.0 5.305 5.671 5.305 5.671 pw_gather_s 912 13.2 4.920 5.373 4.920 5.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.904682, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.661248E+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 578641. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 53.848 53.849 qs_mol_dyn_low 1 2.0 0.003 0.006 53.568 53.580 qs_forces 11 3.9 0.002 0.005 53.498 53.499 qs_energies 11 4.9 0.002 0.002 51.999 52.024 scf_env_do_scf 11 5.9 0.000 0.001 45.996 45.997 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.794 43.794 dbcsr_multiply_generic 2286 12.5 0.094 0.096 33.919 34.501 qs_scf_new_mos 108 7.5 0.000 0.001 33.056 33.376 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.055 33.376 ot_scf_mini 108 9.5 0.002 0.002 31.324 31.523 multiply_cannon 2286 13.5 0.192 0.204 26.026 27.819 multiply_cannon_loop 2286 14.5 1.495 1.557 25.230 27.006 velocity_verlet 10 3.0 0.001 0.002 26.837 26.838 ot_mini 108 10.5 0.001 0.001 19.727 19.965 qs_ot_get_derivative 108 11.5 0.001 0.002 16.772 16.946 mp_waitall_1 245248 16.5 8.768 14.907 8.768 14.907 multiply_cannon_metrocomm3 54864 15.5 0.068 0.075 6.030 13.431 multiply_cannon_multrec 54864 15.5 4.182 6.529 7.470 10.825 rebuild_ks_matrix 119 8.3 0.000 0.000 8.241 8.372 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 8.241 8.372 multiply_cannon_sync_h2d 54864 15.5 5.834 7.778 5.834 7.778 mp_sum_l 7207 12.9 5.684 7.432 5.684 7.432 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.296 7.412 qs_ot_get_p 119 10.4 0.001 0.001 6.903 7.203 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.697 6.176 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.742 5.846 sum_up_and_integrate 119 10.3 0.012 0.014 4.773 4.780 integrate_v_rspace 119 11.3 0.002 0.002 4.761 4.769 init_scf_run 11 5.9 0.000 0.001 4.705 4.705 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.705 4.705 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.543 4.690 calculate_rho_elec 119 8.7 0.011 0.017 4.542 4.689 dbcsr_mm_accdrv_process 76910 16.1 1.187 1.832 3.210 4.476 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.805 3.862 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.926 3.479 rs_pw_transfer 974 11.9 0.011 0.012 2.991 3.119 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.914 3.119 apply_single 119 13.6 0.000 0.000 2.914 3.118 calculate_dm_sparse 119 9.5 0.000 0.000 2.946 3.100 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.952 2.952 ot_diis_step 108 11.5 0.006 0.006 2.674 2.675 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.664 2.664 cp_fm_redistribute_end 50 14.0 2.418 2.634 2.423 2.636 density_rs2pw 119 9.7 0.004 0.005 2.495 2.602 jit_kernel_multiply 13 15.8 1.960 2.557 1.960 2.557 cp_fm_diag_elpa_base 50 14.0 0.210 2.532 0.211 2.545 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.480 2.540 calculate_first_density_matrix 1 7.0 0.001 0.002 2.449 2.458 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.245 2.247 wfi_extrapolate 11 7.9 0.001 0.001 2.182 2.182 acc_transpose_blocks 54864 15.5 0.239 0.257 1.766 2.182 init_scf_loop 11 6.9 0.000 0.000 2.179 2.179 grid_integrate_task_list 119 12.3 2.039 2.163 2.039 2.163 pw_transfer 1439 11.6 0.053 0.060 2.070 2.150 potential_pw2rs 119 12.3 0.004 0.004 2.088 2.099 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.994 2.077 mp_sum_d 4127 12.0 1.370 2.000 1.370 2.000 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.932 1.990 make_m2s 4572 13.5 0.054 0.056 1.901 1.943 make_images 4572 14.5 0.134 0.139 1.818 1.859 fft3d_ps 1201 14.6 0.369 0.475 1.765 1.844 mp_alltoall_d11v 2130 13.8 1.457 1.638 1.457 1.638 fft_wrap_pw1pw2_140 487 13.2 0.082 0.096 1.531 1.616 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.382 1.413 mp_waitany 12084 13.8 1.226 1.361 1.226 1.361 grid_collocate_task_list 119 9.7 1.292 1.356 1.292 1.356 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.720 1.127 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.849000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.645184E+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 1030242. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.027 41.439 41.439 qs_mol_dyn_low 1 2.0 0.003 0.003 41.163 41.171 qs_forces 11 3.9 0.002 0.002 41.098 41.099 qs_energies 11 4.9 0.002 0.003 39.343 39.346 scf_env_do_scf 11 5.9 0.000 0.001 33.327 33.328 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.503 30.503 dbcsr_multiply_generic 2286 12.5 0.099 0.103 22.050 22.406 qs_scf_new_mos 108 7.5 0.001 0.001 20.755 20.997 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.755 20.997 ot_scf_mini 108 9.5 0.002 0.003 19.850 20.016 velocity_verlet 10 3.0 0.001 0.001 19.182 19.184 multiply_cannon 2286 13.5 0.242 0.285 16.687 18.378 multiply_cannon_loop 2286 14.5 0.907 0.974 15.464 17.058 ot_mini 108 10.5 0.001 0.001 12.193 12.421 mp_waitall_1 200699 16.5 6.094 11.354 6.094 11.354 qs_ot_get_derivative 108 11.5 0.001 0.001 9.687 9.856 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.295 9.556 multiply_cannon_multrec 27432 15.5 1.969 4.502 5.912 8.843 rebuild_ks_matrix 119 8.3 0.000 0.000 7.722 7.866 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.721 7.866 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.826 6.957 dbcsr_mm_accdrv_process 47894 16.0 3.028 5.023 3.873 5.718 qs_ot_get_p 119 10.4 0.001 0.001 4.706 4.934 init_scf_run 11 5.9 0.000 0.001 4.695 4.695 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.695 4.695 sum_up_and_integrate 119 10.3 0.024 0.026 4.583 4.613 integrate_v_rspace 119 11.3 0.002 0.002 4.559 4.590 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.715 4.563 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.132 4.173 calculate_rho_elec 119 8.7 0.021 0.024 4.132 4.172 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.078 4.156 apply_single 119 13.6 0.000 0.000 3.078 4.156 mp_sum_l 7207 12.9 2.206 4.145 2.206 4.145 rs_pw_transfer 974 11.9 0.010 0.011 3.061 3.574 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.157 3.178 calculate_first_density_matrix 1 7.0 0.001 0.001 3.033 3.035 make_m2s 4572 13.5 0.052 0.054 2.708 2.963 density_rs2pw 119 9.7 0.004 0.004 2.459 2.946 make_images 4572 14.5 0.201 0.237 2.619 2.871 init_scf_loop 11 6.9 0.000 0.000 2.799 2.799 multiply_cannon_sync_h2d 27432 15.5 2.142 2.773 2.142 2.773 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.702 2.703 ot_diis_step 108 11.5 0.011 0.011 2.448 2.449 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.290 2.383 potential_pw2rs 119 12.3 0.006 0.006 2.294 2.335 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.310 2.310 cp_fm_redistribute_end 50 14.0 1.909 2.281 1.914 2.281 pw_transfer 1439 11.6 0.065 0.071 2.221 2.277 cp_fm_diag_elpa_base 50 14.0 0.353 2.159 0.366 2.214 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.129 2.189 calculate_dm_sparse 119 9.5 0.000 0.000 2.100 2.177 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.171 2.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.933 1.973 grid_integrate_task_list 119 12.3 1.832 1.953 1.832 1.953 fft3d_ps 1201 14.6 0.515 0.568 1.829 1.885 jit_kernel_multiply 9 16.2 0.793 1.788 0.793 1.788 prepare_preconditioner 11 7.9 0.000 0.000 1.711 1.737 make_preconditioner 11 8.9 0.000 0.000 1.711 1.737 acc_transpose_blocks 27432 15.5 0.111 0.116 1.321 1.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.605 1.662 fft_wrap_pw1pw2_140 487 13.2 0.081 0.087 1.589 1.650 make_images_data 4572 15.5 0.044 0.050 1.208 1.650 wfi_extrapolate 11 7.9 0.001 0.001 1.587 1.587 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.056 1.557 mp_alltoall_d11v 2130 13.8 1.294 1.487 1.294 1.487 mp_allgather_i34 2286 14.5 0.631 1.440 0.631 1.440 grid_collocate_task_list 119 9.7 1.229 1.367 1.229 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.313 1.321 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.219 1.266 mp_sum_d 4127 12.0 0.650 1.107 0.650 1.107 acc_transpose_blocks_kernels 27432 16.5 0.183 0.273 0.785 1.082 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.159 0.559 1.045 mp_waitany 5720 13.7 0.540 1.028 0.540 1.028 mp_alltoall_z22v 1201 16.6 0.867 0.999 0.867 0.999 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.977 0.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.956 0.969 parallel_gemm_fm 81 9.0 0.000 0.000 0.931 0.935 parallel_gemm_fm_cosma 81 10.0 0.931 0.935 0.931 0.935 make_images_sizes 4572 15.5 0.005 0.005 0.671 0.934 mp_alltoall_i44 4572 16.5 0.666 0.929 0.666 0.929 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.606 0.827 0.874 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.439000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.545455, yerr=1.372697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 522.690560E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.055 33.664 33.665 qs_mol_dyn_low 1 2.0 0.003 0.003 33.360 33.369 qs_forces 11 3.9 0.001 0.002 33.296 33.297 qs_energies 11 4.9 0.001 0.001 31.688 31.690 scf_env_do_scf 11 5.9 0.000 0.001 26.784 26.784 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.129 24.130 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.757 16.852 velocity_verlet 10 3.0 0.001 0.001 15.976 15.977 qs_scf_new_mos 108 7.5 0.001 0.001 15.470 15.484 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.469 15.484 ot_scf_mini 108 9.5 0.002 0.002 14.722 14.732 multiply_cannon 2286 13.5 0.203 0.233 13.167 14.035 multiply_cannon_loop 2286 14.5 0.635 0.668 12.321 13.230 ot_mini 108 10.5 0.001 0.001 9.013 9.033 qs_ot_get_derivative 108 11.5 0.001 0.001 7.479 7.492 multiply_cannon_multrec 18288 15.5 1.919 2.912 6.747 7.128 rebuild_ks_matrix 119 8.3 0.000 0.001 6.658 6.677 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.657 6.677 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.896 5.913 dbcsr_mm_accdrv_process 38222 16.0 4.118 5.415 4.743 5.558 sum_up_and_integrate 119 10.3 0.029 0.030 4.215 4.221 integrate_v_rspace 119 11.3 0.002 0.003 4.186 4.195 mp_waitall_1 158411 16.6 2.922 3.816 2.922 3.816 init_scf_run 11 5.9 0.000 0.001 3.716 3.716 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.716 3.716 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.667 3.675 calculate_rho_elec 119 8.7 0.030 0.031 3.667 3.674 qs_ot_get_p 119 10.4 0.001 0.001 3.468 3.497 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.778 3.418 rs_pw_transfer 974 11.9 0.009 0.010 2.587 2.817 init_scf_loop 11 6.9 0.000 0.000 2.635 2.635 calculate_first_density_matrix 1 7.0 0.001 0.001 2.438 2.439 density_rs2pw 119 9.7 0.004 0.004 2.176 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.057 2.375 apply_single 119 13.6 0.000 0.000 2.056 2.375 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.318 2.325 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.466 2.145 pw_transfer 1439 11.6 0.065 0.071 2.066 2.085 make_m2s 4572 13.5 0.044 0.046 1.930 2.049 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.036 2.036 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.972 1.993 potential_pw2rs 119 12.3 0.007 0.008 1.969 1.990 make_images 4572 14.5 0.191 0.203 1.844 1.962 grid_integrate_task_list 119 12.3 1.801 1.897 1.801 1.897 jit_kernel_multiply 10 16.3 0.573 1.867 0.573 1.867 calculate_dm_sparse 119 9.5 0.000 0.000 1.841 1.849 prepare_preconditioner 11 7.9 0.000 0.000 1.834 1.836 make_preconditioner 11 8.9 0.000 0.000 1.834 1.836 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.778 1.780 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.689 1.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.762 1.763 cp_fm_redistribute_end 50 14.0 1.328 1.757 1.330 1.758 cp_fm_diag_elpa_base 50 14.0 0.411 1.669 0.425 1.709 fft3d_ps 1201 14.6 0.523 0.541 1.652 1.676 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.650 1.658 mp_sum_l 7207 12.9 1.304 1.632 1.304 1.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.570 1.578 multiply_cannon_sync_h2d 18288 15.5 1.377 1.543 1.377 1.543 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.524 1.540 ot_diis_step 108 11.5 0.011 0.011 1.501 1.502 grid_collocate_task_list 119 9.7 1.213 1.373 1.213 1.373 acc_transpose_blocks 18288 15.5 0.077 0.078 1.242 1.267 wfi_extrapolate 11 7.9 0.001 0.001 1.230 1.230 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.193 make_images_data 4572 15.5 0.044 0.048 0.855 1.002 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.951 0.952 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.929 0.950 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.739 0.916 mp_alltoall_d11v 2130 13.8 0.715 0.850 0.715 0.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.836 mp_alltoall_z22v 1201 16.6 0.716 0.832 0.716 0.832 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.800 0.825 mp_waitany 9880 13.7 0.535 0.802 0.535 0.802 cp_fm_cholesky_invert 11 10.9 0.781 0.786 0.781 0.786 rs_pw_transfer_RS2PW_140 130 11.5 0.124 0.138 0.520 0.766 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.665000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 546.091008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 37.038 37.039 qs_mol_dyn_low 1 2.0 0.003 0.004 36.826 36.837 qs_forces 11 3.9 0.002 0.002 36.763 36.764 qs_energies 11 4.9 0.001 0.002 34.992 34.998 scf_env_do_scf 11 5.9 0.000 0.001 29.711 29.712 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.087 26.087 velocity_verlet 10 3.0 0.001 0.001 18.732 18.734 dbcsr_multiply_generic 2286 12.5 0.099 0.103 18.559 18.683 qs_scf_new_mos 108 7.5 0.001 0.001 17.079 17.130 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.078 17.129 ot_scf_mini 108 9.5 0.002 0.003 16.127 16.176 multiply_cannon 2286 13.5 0.239 0.276 14.543 15.056 multiply_cannon_loop 2286 14.5 0.938 0.969 13.503 13.845 ot_mini 108 10.5 0.001 0.001 9.682 9.741 multiply_cannon_multrec 27432 15.5 2.346 3.040 8.597 8.939 qs_ot_get_derivative 108 11.5 0.001 0.001 7.808 7.859 dbcsr_mm_accdrv_process 47916 15.9 5.279 7.208 6.156 7.348 rebuild_ks_matrix 119 8.3 0.000 0.000 6.943 7.007 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.942 7.007 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.174 6.228 sum_up_and_integrate 119 10.3 0.035 0.038 4.068 4.075 integrate_v_rspace 119 11.3 0.002 0.003 4.033 4.040 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.826 3.862 calculate_rho_elec 119 8.7 0.040 0.046 3.825 3.861 qs_ot_get_p 119 10.4 0.001 0.001 3.777 3.848 init_scf_run 11 5.9 0.000 0.001 3.846 3.846 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.845 3.846 init_scf_loop 11 6.9 0.000 0.000 3.602 3.602 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.813 3.279 prepare_preconditioner 11 7.9 0.000 0.000 2.690 2.697 make_preconditioner 11 8.9 0.000 0.000 2.689 2.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.287 2.616 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.142 2.607 apply_single 119 13.6 0.000 0.000 2.142 2.607 mp_waitall_1 137007 16.6 2.079 2.593 2.079 2.593 rs_pw_transfer 974 11.9 0.009 0.009 2.377 2.575 make_m2s 4572 13.5 0.054 0.056 2.430 2.514 density_rs2pw 119 9.7 0.004 0.004 2.231 2.429 make_images 4572 14.5 0.271 0.330 2.323 2.407 calculate_first_density_matrix 1 7.0 0.001 0.001 2.381 2.384 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.287 2.297 pw_transfer 1439 11.6 0.066 0.072 2.245 2.286 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.152 2.197 calculate_dm_sparse 119 9.5 0.000 0.000 2.047 2.111 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.033 2.059 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.962 1.962 grid_integrate_task_list 119 12.3 1.820 1.916 1.820 1.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.905 1.906 fft3d_ps 1201 14.6 0.560 0.611 1.826 1.863 ot_diis_step 108 11.5 0.012 0.012 1.832 1.832 potential_pw2rs 119 12.3 0.008 0.009 1.801 1.810 mp_sum_l 7207 12.9 1.169 1.809 1.169 1.809 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.727 1.778 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.680 1.681 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.641 1.657 cp_fm_redistribute_end 50 14.0 1.111 1.649 1.112 1.649 cp_fm_diag_elpa_base 50 14.0 0.514 1.571 0.534 1.610 acc_transpose_blocks 27432 15.5 0.114 0.116 1.463 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.411 1.411 grid_collocate_task_list 119 9.7 1.220 1.348 1.220 1.348 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.223 1.234 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.813 1.210 dbcsr_complete_redistribute 329 12.2 0.154 0.219 0.929 1.202 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.155 1.171 jit_kernel_multiply 6 16.0 0.815 1.151 0.815 1.151 qs_energies_init_hamiltonians 11 5.9 0.004 0.015 1.147 1.148 cp_fm_upper_to_full 72 13.5 0.804 1.117 0.804 1.117 make_images_data 4572 15.5 0.045 0.048 0.943 1.040 multiply_cannon_sync_h2d 27432 15.5 0.986 1.026 0.986 1.026 mp_alltoall_d11v 2130 13.8 0.840 0.994 0.840 0.994 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.676 0.944 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.812 0.939 mp_alltoall_z22v 1201 16.6 0.851 0.893 0.851 0.893 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.790 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.862 0.869 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.834 0.850 cp_fm_cholesky_invert 11 10.9 0.828 0.832 0.828 0.832 mp_alltoall_i22 627 13.8 0.441 0.747 0.441 0.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.039000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.000000, yerr=3.516196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 604.401664E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 29.985 29.986 qs_mol_dyn_low 1 2.0 0.003 0.003 29.702 29.710 qs_forces 11 3.9 0.001 0.002 29.640 29.641 qs_energies 11 4.9 0.001 0.002 27.911 27.914 scf_env_do_scf 11 5.9 0.000 0.001 22.279 22.279 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.644 19.644 velocity_verlet 10 3.0 0.001 0.001 14.838 14.841 dbcsr_multiply_generic 2286 12.5 0.090 0.091 12.741 12.809 qs_scf_new_mos 108 7.5 0.001 0.001 11.313 11.339 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.312 11.339 ot_scf_mini 108 9.5 0.002 0.002 10.620 10.647 multiply_cannon 2286 13.5 0.244 0.258 10.130 10.568 multiply_cannon_loop 2286 14.5 0.329 0.339 9.190 9.364 multiply_cannon_multrec 9144 15.5 1.683 1.928 6.091 6.327 rebuild_ks_matrix 119 8.3 0.000 0.000 6.199 6.215 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.198 6.214 ot_mini 108 10.5 0.001 0.001 5.827 5.856 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.535 5.549 qs_ot_get_derivative 108 11.5 0.001 0.001 4.508 4.534 dbcsr_mm_accdrv_process 12550 15.8 3.316 4.330 4.305 4.371 init_scf_run 11 5.9 0.000 0.001 4.161 4.161 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.161 4.161 sum_up_and_integrate 119 10.3 0.038 0.041 3.842 3.846 integrate_v_rspace 119 11.3 0.003 0.003 3.804 3.808 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.704 3.712 calculate_rho_elec 119 8.7 0.060 0.061 3.703 3.711 qs_ot_get_p 119 10.4 0.001 0.001 2.980 3.024 calculate_first_density_matrix 1 7.0 0.001 0.001 2.971 2.971 init_scf_loop 11 6.9 0.000 0.000 2.613 2.614 pw_transfer 1439 11.6 0.066 0.069 2.244 2.256 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.150 2.165 density_rs2pw 119 9.7 0.004 0.004 2.016 2.160 mp_waitall_1 115863 16.7 1.544 2.014 1.544 2.014 jit_kernel_multiply 10 15.4 0.950 2.000 0.950 2.000 rs_pw_transfer 974 11.9 0.008 0.008 1.858 1.989 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.982 1.984 make_m2s 4572 13.5 0.034 0.036 1.812 1.951 grid_integrate_task_list 119 12.3 1.856 1.925 1.856 1.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.905 1.906 make_images 4572 14.5 0.269 0.304 1.722 1.859 prepare_preconditioner 11 7.9 0.000 0.000 1.843 1.846 make_preconditioner 11 8.9 0.000 0.000 1.843 1.846 calculate_dm_sparse 119 9.5 0.000 0.000 1.816 1.836 fft3d_ps 1201 14.6 0.562 0.576 1.807 1.817 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.781 1.781 fft_wrap_pw1pw2_140 487 13.2 0.088 0.092 1.748 1.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.733 1.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.734 1.742 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.596 1.610 potential_pw2rs 119 12.3 0.010 0.011 1.560 1.567 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.463 1.465 cp_fm_redistribute_end 50 14.0 0.730 1.436 0.731 1.437 cp_fm_diag_elpa_base 50 14.0 0.661 1.360 0.704 1.415 grid_collocate_task_list 119 9.7 1.271 1.391 1.271 1.391 ot_diis_step 108 11.5 0.013 0.013 1.304 1.304 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.250 1.251 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.228 acc_transpose_blocks 9144 15.5 0.038 0.040 1.178 1.198 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.170 1.190 apply_single 119 13.6 0.000 0.000 1.170 1.189 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.154 1.165 wfi_extrapolate 11 7.9 0.001 0.001 1.141 1.141 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.819 1.042 make_images_data 4572 15.5 0.038 0.042 0.847 1.011 cp_fm_cholesky_invert 11 10.9 0.967 0.970 0.967 0.970 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.936 0.957 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.914 mp_alltoall_d11v 2130 13.8 0.815 0.906 0.815 0.906 jit_kernel_transpose 5 15.6 0.818 0.841 0.818 0.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.788 0.789 multiply_cannon_sync_h2d 9144 15.5 0.718 0.783 0.718 0.783 parallel_gemm_fm 81 9.0 0.000 0.000 0.765 0.766 parallel_gemm_fm_cosma 81 10.0 0.765 0.766 0.765 0.766 mp_alltoall_z22v 1201 16.6 0.693 0.755 0.693 0.755 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.705 0.754 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.738 0.746 make_basis_sm 11 9.8 0.000 0.000 0.738 0.739 mp_allgather_i34 2286 14.5 0.254 0.690 0.254 0.690 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.317 0.661 yz_to_x 606 15.1 0.265 0.276 0.624 0.643 x_to_yz 595 16.2 0.275 0.286 0.610 0.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.986000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.181818, yerr=2.551940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 762.404864E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.036 42.171 42.172 qs_mol_dyn_low 1 2.0 0.003 0.003 41.884 41.894 qs_forces 11 3.9 0.002 0.002 41.818 41.819 qs_energies 11 4.9 0.002 0.002 39.817 39.822 scf_env_do_scf 11 5.9 0.001 0.001 34.127 34.127 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.104 26.105 velocity_verlet 10 3.0 0.001 0.001 23.796 23.802 dbcsr_multiply_generic 2286 12.5 0.098 0.100 17.436 17.685 qs_scf_new_mos 108 7.5 0.001 0.001 16.085 16.181 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.084 16.180 ot_scf_mini 108 9.5 0.002 0.002 14.993 15.095 multiply_cannon 2286 13.5 0.303 0.313 13.489 14.503 multiply_cannon_loop 2286 14.5 0.345 0.350 12.248 13.242 ot_mini 108 10.5 0.001 0.001 8.858 8.978 multiply_cannon_multrec 9144 15.5 3.407 4.760 8.548 8.673 init_scf_loop 11 6.9 0.000 0.000 7.997 7.999 rebuild_ks_matrix 119 8.3 0.000 0.000 7.458 7.603 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.457 7.602 prepare_preconditioner 11 7.9 0.000 0.000 7.013 7.026 make_preconditioner 11 8.9 0.000 0.000 7.013 7.026 qs_ot_get_derivative 108 11.5 0.001 0.001 6.836 6.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.538 6.900 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.741 6.872 dbcsr_mm_accdrv_process 12550 15.8 3.986 5.338 5.015 6.391 cp_fm_upper_to_full 72 14.2 3.238 4.655 3.238 4.655 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.370 4.379 calculate_rho_elec 119 8.7 0.118 0.121 4.370 4.378 sum_up_and_integrate 119 10.3 0.064 0.066 4.118 4.124 integrate_v_rspace 119 11.3 0.003 0.003 4.053 4.060 init_scf_run 11 5.9 0.000 0.001 3.644 3.645 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.644 3.645 qs_ot_get_p 119 10.4 0.001 0.001 3.385 3.526 mp_waitall_1 94719 16.7 2.343 3.277 2.343 3.277 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.487 2.933 dbcsr_complete_redistribute 329 12.2 0.312 0.316 2.053 2.908 pw_transfer 1439 11.6 0.068 0.068 2.890 2.905 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.793 2.812 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.732 2.566 make_m2s 4572 13.5 0.038 0.038 2.345 2.489 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.194 2.463 apply_single 119 13.6 0.000 0.000 2.194 2.462 fft3d_ps 1201 14.6 0.600 0.624 2.418 2.438 fft_wrap_pw1pw2_140 487 13.2 0.094 0.097 2.357 2.378 make_images 4572 14.5 0.354 0.385 2.221 2.366 mp_alltoall_i22 627 13.8 1.450 2.297 1.450 2.297 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.470 2.296 density_rs2pw 119 9.7 0.004 0.004 2.265 2.281 calculate_dm_sparse 119 9.5 0.000 0.000 2.168 2.187 calculate_first_density_matrix 1 7.0 0.001 0.001 2.174 2.176 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.250 2.140 grid_integrate_task_list 119 12.3 2.060 2.091 2.060 2.091 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.082 2.084 ot_diis_step 108 11.5 0.014 0.015 1.991 1.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.847 1.847 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.804 1.805 mp_sum_l 7207 12.9 1.064 1.789 1.064 1.789 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.773 1.775 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.709 1.761 jit_kernel_multiply 10 15.8 1.001 1.742 1.001 1.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.554 1.589 rs_pw_transfer 974 11.9 0.009 0.010 1.494 1.514 grid_collocate_task_list 119 9.7 1.492 1.506 1.492 1.506 cp_fm_cholesky_invert 11 10.9 1.496 1.499 1.496 1.499 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.496 1.497 cp_fm_diag_elpa_base 50 14.0 1.342 1.399 1.494 1.495 potential_pw2rs 119 12.3 0.014 0.014 1.485 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.403 1.404 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.169 1.394 make_images_data 4572 15.5 0.042 0.046 1.110 1.316 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.228 1.238 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.153 1.175 mp_alltoall_d11v 2130 13.8 1.142 1.169 1.142 1.169 qs_env_update_s_mstruct 11 6.9 0.002 0.005 1.104 1.116 multiply_cannon_sync_h2d 9144 15.5 1.042 1.047 1.042 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.025 yz_to_x 606 15.1 0.461 0.477 0.975 0.985 qs_create_task_list 11 7.9 0.001 0.001 0.940 0.951 generate_qs_task_list 11 8.9 0.371 0.390 0.940 0.951 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.944 acc_transpose_blocks 9144 15.5 0.039 0.039 0.908 0.915 mp_alltoall_z22v 1201 16.6 0.867 0.914 0.867 0.914 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.172000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.636364, yerr=14.593104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.964800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 85.226 85.227 qs_mol_dyn_low 1 2.0 0.003 0.004 84.974 84.985 qs_forces 11 3.9 0.002 0.003 84.903 84.907 qs_energies 11 4.9 0.001 0.002 81.981 82.001 scf_env_do_scf 11 5.9 0.000 0.001 72.540 72.542 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.968 66.968 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.953 53.196 qs_scf_new_mos 99 7.5 0.000 0.001 49.226 49.333 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.225 49.332 ot_scf_mini 99 9.5 0.002 0.002 46.742 46.826 multiply_cannon 2055 13.4 0.178 0.185 42.828 43.619 velocity_verlet 10 3.0 0.001 0.001 43.076 43.076 multiply_cannon_loop 2055 14.4 1.545 1.588 41.863 42.692 ot_mini 99 10.5 0.001 0.001 28.391 28.480 qs_ot_get_derivative 99 11.5 0.001 0.001 21.557 21.642 multiply_cannon_multrec 49320 15.4 12.069 12.676 17.120 17.809 rebuild_ks_matrix 110 8.3 0.000 0.001 14.812 14.908 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.811 14.908 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.998 13.081 mp_waitall_1 220248 16.4 11.984 12.947 11.984 12.947 multiply_cannon_sync_h2d 49320 15.4 9.891 10.465 9.891 10.465 qs_ot_get_p 110 10.4 0.001 0.001 9.691 9.792 multiply_cannon_metrocomm3 49320 15.4 0.078 0.083 7.212 8.278 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.442 7.952 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.172 7.687 apply_single 110 13.6 0.000 0.001 7.171 7.687 init_scf_run 11 5.9 0.000 0.001 7.392 7.393 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.392 7.392 sum_up_and_integrate 110 10.3 0.036 0.042 7.255 7.269 integrate_v_rspace 110 11.3 0.003 0.003 7.218 7.241 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.912 6.983 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.767 6.901 calculate_rho_elec 110 8.6 0.021 0.026 6.766 6.900 ot_diis_step 99 11.5 0.005 0.006 6.534 6.535 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.403 6.455 mp_sum_l 6514 12.8 5.585 6.412 5.585 6.412 init_scf_loop 11 6.9 0.000 0.000 5.543 5.544 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.365 5.365 dbcsr_mm_accdrv_process 87628 16.1 2.096 2.222 4.929 5.256 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.879 4.880 cp_fm_redistribute_end 48 14.0 4.259 4.854 4.262 4.855 cp_fm_diag_elpa_base 48 14.0 0.587 4.720 0.590 4.742 rs_pw_transfer 902 11.9 0.012 0.013 3.782 4.460 density_rs2pw 110 9.6 0.005 0.005 3.550 4.224 make_m2s 4110 13.4 0.061 0.065 4.077 4.187 make_images 4110 14.4 0.178 0.192 3.980 4.094 wfi_extrapolate 11 7.9 0.001 0.001 4.070 4.070 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.809 3.961 calculate_dm_sparse 110 9.5 0.000 0.001 3.855 3.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.601 3.607 pw_transfer 1331 11.6 0.054 0.068 3.406 3.476 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.380 3.438 grid_integrate_task_list 110 12.3 3.243 3.396 3.243 3.396 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.318 3.392 prepare_preconditioner 11 7.9 0.000 0.000 3.358 3.383 make_preconditioner 11 8.9 0.000 0.000 3.358 3.383 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.254 3.300 calculate_first_density_matrix 1 7.0 0.001 0.001 3.228 3.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.146 3.208 fft3d_ps 1111 14.6 0.777 0.878 2.845 2.915 fft_wrap_pw1pw2_140 451 13.1 0.172 0.192 2.799 2.875 potential_pw2rs 110 12.3 0.006 0.007 2.757 2.778 jit_kernel_multiply 13 15.9 2.546 2.572 2.546 2.572 mp_waitany 14300 13.8 1.830 2.533 1.830 2.533 mp_alltoall_d11v 2046 13.8 2.141 2.471 2.141 2.471 grid_collocate_task_list 110 9.6 2.082 2.300 2.082 2.300 acc_transpose_blocks 49320 15.4 0.231 0.240 2.213 2.291 mp_sum_d 3881 11.9 1.528 2.063 1.528 2.063 make_images_data 4110 15.4 0.042 0.045 1.852 2.007 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.956 1.977 parallel_gemm_fm 81 9.0 0.000 0.000 1.854 1.858 parallel_gemm_fm_cosma 81 10.0 1.854 1.858 1.854 1.858 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.606 1.847 cp_fm_cholesky_invert 11 10.9 1.837 1.840 1.837 1.840 mp_alltoall_z22v 1111 16.6 1.425 1.720 1.425 1.720 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.227000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.545455, yerr=2.230517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.603392E+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 3251944. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.070 71.548 71.559 qs_mol_dyn_low 1 2.0 0.003 0.004 71.243 71.265 qs_forces 11 3.9 0.002 0.003 71.161 71.162 qs_energies 11 4.9 0.002 0.006 67.749 67.753 scf_env_do_scf 11 5.9 0.001 0.001 58.473 58.476 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 50.304 50.305 dbcsr_multiply_generic 2055 12.4 0.114 0.119 38.114 38.365 velocity_verlet 10 3.0 0.001 0.002 37.341 37.343 qs_scf_new_mos 99 7.5 0.001 0.001 33.461 33.598 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.461 33.597 multiply_cannon 2055 13.4 0.221 0.242 31.140 32.558 ot_scf_mini 99 9.5 0.003 0.004 31.786 31.901 multiply_cannon_loop 2055 14.4 0.929 0.965 29.772 30.548 ot_mini 99 10.5 0.001 0.001 18.571 18.696 multiply_cannon_multrec 24660 15.4 7.589 9.073 13.790 15.160 rebuild_ks_matrix 110 8.3 0.000 0.001 14.170 14.309 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.169 14.308 qs_ot_get_derivative 99 11.5 0.001 0.002 12.696 12.817 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.454 12.580 mp_waitall_1 176588 16.5 7.959 10.432 7.959 10.432 init_scf_loop 11 6.9 0.000 0.002 8.125 8.129 multiply_cannon_metrocomm3 24660 15.4 0.071 0.075 5.308 8.084 multiply_cannon_sync_h2d 24660 15.4 6.981 7.774 6.981 7.774 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.538 7.226 apply_single 110 13.6 0.000 0.001 6.538 7.225 sum_up_and_integrate 110 10.3 0.051 0.057 6.896 6.907 init_scf_run 11 5.9 0.000 0.001 6.856 6.857 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.856 6.857 integrate_v_rspace 110 11.3 0.002 0.003 6.844 6.855 qs_ot_get_p 110 10.4 0.001 0.002 6.597 6.747 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.452 6.465 calculate_rho_elec 110 8.6 0.039 0.047 6.452 6.464 dbcsr_mm_accdrv_process 52282 16.1 4.693 5.533 6.043 6.359 prepare_preconditioner 11 7.9 0.000 0.000 5.922 5.946 make_preconditioner 11 8.9 0.000 0.002 5.922 5.946 ot_diis_step 99 11.5 0.010 0.010 5.816 5.816 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.477 5.632 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.745 5.480 make_m2s 4110 13.4 0.057 0.060 4.444 4.893 make_images 4110 14.4 0.400 0.448 4.334 4.779 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.578 4.600 density_rs2pw 110 9.6 0.004 0.005 3.549 4.153 pw_transfer 1331 11.6 0.066 0.074 3.986 4.145 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.132 4.132 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.879 4.041 rs_pw_transfer 902 11.9 0.012 0.013 3.179 3.844 wfi_extrapolate 11 7.9 0.001 0.001 3.602 3.602 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.546 3.548 cp_fm_redistribute_end 48 14.0 2.648 3.514 2.650 3.514 cp_fm_diag_elpa_base 48 14.0 0.828 3.361 0.860 3.449 fft_wrap_pw1pw2_140 451 13.1 0.203 0.219 3.216 3.382 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.305 3.376 fft3d_ps 1111 14.6 1.107 1.331 3.228 3.372 grid_integrate_task_list 110 12.3 3.144 3.351 3.144 3.351 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.335 3.337 calculate_first_density_matrix 1 7.0 0.001 0.005 3.159 3.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.998 3.041 calculate_dm_sparse 110 9.5 0.001 0.001 2.994 3.025 make_images_data 4110 15.4 0.046 0.050 2.449 2.910 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.153 2.825 cp_fm_cholesky_invert 11 10.9 2.788 2.796 2.788 2.796 potential_pw2rs 110 12.3 0.008 0.008 2.651 2.667 mp_sum_l 6514 12.8 1.926 2.539 1.926 2.539 grid_collocate_task_list 110 9.6 2.045 2.489 2.045 2.489 jit_kernel_multiply 13 16.3 0.996 2.461 0.996 2.461 mp_alltoall_d11v 2046 13.8 1.928 2.288 1.928 2.288 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.800 2.117 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.986 2.006 mp_waitany 10164 13.8 1.249 1.944 1.249 1.944 mp_irecv_dv 57340 16.2 0.675 1.897 0.675 1.897 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.885 1.895 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.891 1.892 mp_allgather_i34 2055 14.4 0.698 1.798 0.698 1.798 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.216 0.968 1.607 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.568 acc_transpose_blocks 24660 15.4 0.113 0.116 1.534 1.557 mp_alltoall_z22v 1111 16.6 1.432 1.546 1.432 1.546 dbcsr_complete_redistribute 325 12.2 0.240 0.294 1.222 1.498 make_basis_sm 11 9.8 0.000 0.000 1.485 1.488 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.474 parallel_gemm_fm 81 9.0 0.000 0.000 1.456 1.463 parallel_gemm_fm_cosma 81 10.0 1.456 1.462 1.456 1.462 cp_fm_cholesky_decompose 22 10.9 1.427 1.433 1.427 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.559000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.727273, yerr=7.337214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.378176E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.060 62.379 62.383 qs_mol_dyn_low 1 2.0 0.003 0.004 62.006 62.017 qs_forces 11 3.9 0.003 0.008 61.929 61.930 qs_energies 11 4.9 0.003 0.006 58.684 58.687 scf_env_do_scf 11 5.9 0.001 0.002 50.587 50.587 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 41.694 41.694 velocity_verlet 10 3.0 0.001 0.001 33.701 33.703 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.877 29.159 qs_scf_new_mos 99 7.5 0.001 0.001 26.105 26.196 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.105 26.196 ot_scf_mini 99 9.5 0.003 0.003 24.866 24.974 multiply_cannon 2055 13.4 0.211 0.222 22.320 23.600 multiply_cannon_loop 2055 14.4 0.614 0.630 21.039 22.274 ot_mini 99 10.5 0.001 0.001 13.951 14.064 rebuild_ks_matrix 110 8.3 0.000 0.001 12.750 12.876 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 12.749 12.875 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.242 11.354 multiply_cannon_multrec 16440 15.4 3.843 5.130 9.630 10.899 mp_waitall_1 139946 16.5 7.338 10.189 7.338 10.189 qs_ot_get_derivative 99 11.5 0.001 0.002 9.400 9.506 init_scf_loop 11 6.9 0.001 0.004 8.849 8.849 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.527 7.354 prepare_preconditioner 11 7.9 0.000 0.001 7.028 7.043 make_preconditioner 11 8.9 0.001 0.006 7.028 7.043 sum_up_and_integrate 110 10.3 0.059 0.061 6.856 6.873 integrate_v_rspace 110 11.3 0.003 0.004 6.797 6.814 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.315 6.678 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.248 6.260 calculate_rho_elec 110 8.6 0.058 0.059 6.247 6.260 qs_ot_get_p 110 10.4 0.001 0.001 6.035 6.172 dbcsr_mm_accdrv_process 34862 16.1 4.686 5.172 5.642 5.867 init_scf_run 11 5.9 0.000 0.001 5.628 5.628 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.628 5.628 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.051 5.489 apply_single 110 13.6 0.000 0.000 5.051 5.489 make_m2s 4110 13.4 0.050 0.052 4.280 4.665 density_rs2pw 110 9.6 0.004 0.005 3.360 4.604 make_images 4110 14.4 0.395 0.509 4.164 4.550 ot_diis_step 99 11.5 0.010 0.011 4.515 4.515 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.289 4.295 multiply_cannon_sync_h2d 16440 15.4 3.632 4.269 3.632 4.269 rs_pw_transfer 902 11.9 0.011 0.015 3.037 4.241 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.927 3.928 pw_transfer 1331 11.6 0.066 0.073 3.846 3.859 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.217 3.847 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.739 3.754 grid_integrate_task_list 110 12.3 3.148 3.385 3.148 3.385 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.336 3.338 cp_fm_redistribute_end 48 14.0 2.085 3.306 2.088 3.307 cp_fm_diag_elpa_base 48 14.0 1.155 3.134 1.213 3.243 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.185 3.205 wfi_extrapolate 11 7.9 0.001 0.002 3.065 3.065 fft3d_ps 1111 14.6 1.091 1.100 3.039 3.051 make_images_data 4110 15.4 0.043 0.048 2.425 2.903 cp_fm_cholesky_invert 11 10.9 2.823 2.829 2.823 2.829 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.827 2.828 hybrid_alltoall_any 4261 16.3 0.105 0.370 2.150 2.785 potential_pw2rs 110 12.3 0.011 0.011 2.549 2.564 calculate_dm_sparse 110 9.5 0.001 0.001 2.531 2.562 mp_waitany 17072 13.8 1.266 2.536 1.266 2.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.483 2.529 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.454 2.512 calculate_first_density_matrix 1 7.0 0.002 0.006 2.476 2.477 grid_collocate_task_list 110 9.6 2.077 2.475 2.077 2.475 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.893 2.467 mp_alltoall_d11v 2046 13.8 1.904 2.372 1.904 2.372 mp_sum_l 6514 12.8 1.663 2.344 1.663 2.344 mp_irecv_dv 48980 15.7 0.822 2.343 0.822 2.343 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.189 0.921 2.132 dbcsr_complete_redistribute 325 12.2 0.322 0.358 1.564 2.013 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.987 1.988 cp_fm_upper_to_full 70 13.6 1.392 1.862 1.392 1.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.709 1.723 cp_fm_cholesky_decompose 22 10.9 1.668 1.685 1.668 1.685 mp_allgather_i34 2055 14.4 0.608 1.653 0.608 1.653 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.051 1.509 jit_kernel_multiply 8 16.6 0.565 1.507 0.565 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.483 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.466 rs_gather_matrices 110 12.3 0.232 0.258 1.021 1.416 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.322 1.328 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.383000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.727273, yerr=8.410933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 726.487040E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.049 65.846 65.847 qs_mol_dyn_low 1 2.0 0.003 0.004 65.517 65.530 qs_forces 11 3.9 0.003 0.007 65.437 65.441 qs_energies 11 4.9 0.003 0.011 61.980 61.987 scf_env_do_scf 11 5.9 0.001 0.002 53.645 53.648 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.826 41.827 velocity_verlet 10 3.0 0.001 0.001 37.449 37.452 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.458 29.638 qs_scf_new_mos 99 7.5 0.001 0.001 26.392 26.507 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.391 26.506 ot_scf_mini 99 9.5 0.003 0.005 24.754 24.870 multiply_cannon 2055 13.4 0.241 0.257 22.360 23.235 multiply_cannon_loop 2055 14.4 0.884 0.903 21.008 21.568 ot_mini 99 10.5 0.001 0.001 14.125 14.255 multiply_cannon_multrec 24660 15.4 4.191 6.757 12.616 13.777 rebuild_ks_matrix 110 8.3 0.000 0.000 12.453 12.561 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.452 12.560 init_scf_loop 11 6.9 0.001 0.003 11.778 11.778 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.998 11.089 qs_ot_get_derivative 99 11.5 0.001 0.001 9.961 10.079 prepare_preconditioner 11 7.9 0.000 0.000 9.999 10.014 make_preconditioner 11 8.9 0.000 0.001 9.999 10.014 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.226 9.673 dbcsr_mm_accdrv_process 52304 16.0 6.615 7.956 8.278 9.187 sum_up_and_integrate 110 10.3 0.067 0.069 6.713 6.725 integrate_v_rspace 110 11.3 0.003 0.003 6.646 6.658 mp_waitall_1 121746 16.5 4.422 6.363 4.422 6.363 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.325 6.334 calculate_rho_elec 110 8.6 0.077 0.081 6.325 6.333 qs_ot_get_p 110 10.4 0.001 0.001 5.632 5.776 make_m2s 4110 13.4 0.060 0.061 5.468 5.691 init_scf_run 11 5.9 0.000 0.001 5.558 5.558 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.557 5.558 make_images 4110 14.4 0.576 0.697 5.328 5.546 cp_fm_upper_to_full 70 13.8 3.270 4.647 3.270 4.647 ot_diis_step 99 11.5 0.011 0.011 4.125 4.125 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.043 4.114 apply_single 110 13.6 0.000 0.000 4.043 4.114 pw_transfer 1331 11.6 0.065 0.074 4.026 4.058 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.921 3.955 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.842 3.856 dbcsr_complete_redistribute 325 12.2 0.412 0.453 2.710 3.843 density_rs2pw 110 9.6 0.004 0.004 3.237 3.755 grid_integrate_task_list 110 12.3 3.254 3.407 3.254 3.407 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 3.349 3.388 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.360 3.360 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.294 3.338 multiply_cannon_sync_h2d 24660 15.4 3.178 3.322 3.178 3.322 multiply_cannon_metrocomm3 24660 15.4 0.035 0.037 1.483 3.300 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.125 3.246 fft3d_ps 1111 14.6 1.095 1.134 3.213 3.235 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.067 3.133 rs_pw_transfer 902 11.9 0.010 0.011 2.556 3.098 make_images_data 4110 15.4 0.046 0.050 2.756 3.042 wfi_extrapolate 11 7.9 0.001 0.006 3.017 3.017 calculate_dm_sparse 110 9.5 0.001 0.001 2.931 2.966 hybrid_alltoall_any 4261 16.3 0.119 0.453 2.378 2.927 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.764 2.869 mp_alltoall_i22 605 13.7 1.651 2.822 1.651 2.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.819 2.822 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.803 2.805 cp_fm_cholesky_invert 11 10.9 2.792 2.800 2.792 2.800 cp_fm_redistribute_end 48 14.0 1.403 2.773 1.405 2.774 cp_fm_diag_elpa_base 48 14.0 1.288 2.633 1.366 2.743 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.483 calculate_first_density_matrix 1 7.0 0.001 0.003 2.439 2.441 grid_collocate_task_list 110 9.6 2.181 2.425 2.181 2.425 potential_pw2rs 110 12.3 0.012 0.013 2.351 2.363 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.257 2.261 jit_kernel_multiply 12 15.6 1.329 2.197 1.329 2.197 mp_alltoall_d11v 2046 13.8 1.802 2.123 1.802 2.123 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.721 1.755 cp_fm_cholesky_decompose 22 10.9 1.691 1.723 1.691 1.723 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.715 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.595 1.707 mp_sum_l 6514 12.8 0.949 1.607 0.949 1.607 mp_waitany 13376 13.8 1.084 1.596 1.084 1.596 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.549 acc_transpose_blocks 24660 15.4 0.108 0.109 1.522 1.544 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.857 1.388 mp_alltoall_z22v 1111 16.6 1.260 1.369 1.260 1.369 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.179 0.813 1.348 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.216 1.321 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.847000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.000000, yerr=7.920055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 832.405504E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 57.353 57.354 qs_mol_dyn_low 1 2.0 0.003 0.003 57.118 57.130 qs_forces 11 3.9 0.002 0.002 56.810 56.812 qs_energies 11 4.9 0.001 0.002 53.079 53.086 scf_env_do_scf 11 5.9 0.000 0.001 44.679 44.680 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 36.503 36.505 velocity_verlet 10 3.0 0.001 0.001 32.186 32.204 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.220 23.362 qs_scf_new_mos 99 7.5 0.001 0.001 21.084 21.150 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.083 21.150 ot_scf_mini 99 9.5 0.002 0.002 19.842 19.880 multiply_cannon 2055 13.4 0.249 0.268 17.591 18.959 multiply_cannon_loop 2055 14.4 0.319 0.331 16.195 16.621 rebuild_ks_matrix 110 8.3 0.000 0.000 12.183 12.240 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.182 12.239 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.826 10.876 ot_mini 99 10.5 0.001 0.001 10.682 10.717 multiply_cannon_multrec 8220 15.4 3.262 4.413 7.592 8.391 init_scf_loop 11 6.9 0.000 0.000 8.125 8.127 mp_waitall_1 103326 16.6 6.184 7.792 6.184 7.792 qs_ot_get_derivative 99 11.5 0.001 0.001 6.857 6.894 sum_up_and_integrate 110 10.3 0.079 0.080 6.663 6.677 integrate_v_rspace 110 11.3 0.003 0.003 6.584 6.596 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.460 6.479 calculate_rho_elec 110 8.6 0.115 0.116 6.460 6.478 prepare_preconditioner 11 7.9 0.000 0.000 6.435 6.441 make_preconditioner 11 8.9 0.000 0.000 6.435 6.441 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.094 qs_ot_get_p 110 10.4 0.001 0.001 5.205 5.248 init_scf_run 11 5.9 0.000 0.001 5.241 5.242 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.241 5.241 dbcsr_mm_accdrv_process 17442 15.9 2.861 3.921 4.200 5.144 make_m2s 4110 13.4 0.039 0.040 4.264 4.571 make_images 4110 14.4 0.639 0.695 4.134 4.443 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.067 4.425 pw_transfer 1331 11.6 0.066 0.071 4.209 4.217 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.102 4.114 ot_diis_step 99 11.5 0.012 0.012 3.799 3.800 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.767 3.771 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.706 3.760 apply_single 110 13.6 0.000 0.000 3.706 3.759 density_rs2pw 110 9.6 0.004 0.004 3.202 3.575 grid_integrate_task_list 110 12.3 3.354 3.573 3.354 3.573 fft_wrap_pw1pw2_140 451 13.1 0.217 0.219 3.529 3.549 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.455 3.456 fft3d_ps 1111 14.6 1.148 1.176 3.344 3.351 cp_fm_cholesky_invert 11 10.9 3.246 3.250 3.246 3.250 multiply_cannon_sync_h2d 8220 15.4 2.916 2.957 2.916 2.957 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.896 2.897 cp_fm_redistribute_end 48 14.0 0.737 2.863 0.742 2.864 make_images_data 4110 15.4 0.038 0.042 2.417 2.860 wfi_extrapolate 11 7.9 0.001 0.001 2.832 2.832 cp_fm_diag_elpa_base 48 14.0 1.943 2.666 2.114 2.826 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.347 2.769 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.686 2.687 rs_pw_transfer 902 11.9 0.010 0.010 2.282 2.679 calculate_dm_sparse 110 9.5 0.001 0.001 2.541 2.591 grid_collocate_task_list 110 9.6 2.280 2.540 2.280 2.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.523 2.524 calculate_first_density_matrix 1 7.0 0.001 0.003 2.310 2.313 potential_pw2rs 110 12.3 0.015 0.015 2.204 2.208 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.143 2.159 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.970 1.996 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.983 mp_alltoall_d11v 2046 13.8 1.660 1.955 1.660 1.955 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.861 1.881 cp_fm_cholesky_decompose 22 10.9 1.771 1.790 1.771 1.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.691 1.698 mp_allgather_i34 2055 14.4 0.581 1.672 0.581 1.672 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.522 1.654 dbcsr_complete_redistribute 325 12.2 0.557 0.580 1.457 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.504 1.519 mp_waitany 9240 13.8 1.090 1.499 1.090 1.499 jit_kernel_multiply 8 15.8 1.027 1.492 1.027 1.492 qs_create_task_list 11 7.9 0.000 0.001 1.229 1.334 generate_qs_task_list 11 8.9 0.381 0.449 1.228 1.334 rs_gather_matrices 110 12.3 0.324 0.364 0.939 1.228 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.798 1.199 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.160 1.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.354000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=786.545455, yerr=10.421687 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 1.348215E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.066 88.331 88.335 qs_mol_dyn_low 1 2.0 0.003 0.003 88.005 88.014 qs_forces 11 3.9 0.002 0.002 87.932 87.933 qs_energies 11 4.9 0.001 0.001 83.710 83.712 scf_env_do_scf 11 5.9 0.001 0.001 73.705 73.706 velocity_verlet 10 3.0 0.001 0.001 56.732 56.739 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.583 44.584 init_scf_loop 11 6.9 0.000 0.000 29.049 29.051 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.452 28.537 prepare_preconditioner 11 7.9 0.000 0.000 27.009 27.016 make_preconditioner 11 8.9 0.000 0.000 27.009 27.016 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.022 26.493 qs_scf_new_mos 99 7.5 0.001 0.001 26.085 26.132 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.084 26.132 ot_scf_mini 99 9.5 0.002 0.002 24.330 24.366 multiply_cannon 2055 13.4 0.342 0.359 21.513 22.302 multiply_cannon_loop 2055 14.4 0.342 0.344 19.729 20.005 cp_fm_upper_to_full 70 14.2 13.141 18.857 13.141 18.857 rebuild_ks_matrix 110 8.3 0.000 0.001 14.279 14.324 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.279 14.323 ot_mini 99 10.5 0.001 0.001 13.538 13.575 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.936 12.976 dbcsr_complete_redistribute 325 12.2 1.030 1.057 7.553 10.895 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.528 9.873 multiply_cannon_multrec 8220 15.4 4.340 4.528 9.535 9.673 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.971 9.281 qs_ot_get_derivative 99 11.5 0.001 0.001 8.973 9.013 mp_alltoall_i22 605 13.7 5.597 8.960 5.597 8.960 mp_waitall_1 84994 16.7 7.558 8.441 7.558 8.441 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.836 7.866 calculate_rho_elec 110 8.6 0.227 0.227 7.836 7.865 sum_up_and_integrate 110 10.3 0.150 0.151 7.411 7.423 integrate_v_rspace 110 11.3 0.004 0.004 7.261 7.273 init_scf_run 11 5.9 0.000 0.001 5.714 5.714 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.714 5.714 make_m2s 4110 13.4 0.043 0.044 5.205 5.707 qs_ot_get_p 110 10.4 0.001 0.001 5.606 5.646 make_images 4110 14.4 0.884 0.938 5.017 5.519 pw_transfer 1331 11.6 0.075 0.075 5.349 5.359 cp_fm_cholesky_invert 11 10.9 5.327 5.331 5.327 5.331 dbcsr_mm_accdrv_process 11614 15.7 3.233 3.647 5.053 5.296 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.233 5.242 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.844 5.152 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.672 5.141 apply_single 110 13.6 0.000 0.000 4.672 5.140 fft_wrap_pw1pw2_140 451 13.1 0.222 0.222 4.550 4.566 ot_diis_step 99 11.5 0.015 0.016 4.543 4.543 fft3d_ps 1111 14.6 1.293 1.303 4.428 4.438 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.060 4.068 multiply_cannon_sync_h2d 8220 15.4 3.945 3.951 3.945 3.951 density_rs2pw 110 9.6 0.004 0.004 3.856 3.884 grid_integrate_task_list 110 12.3 3.662 3.730 3.662 3.730 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.692 3.693 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.616 3.616 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.886 3.544 make_images_data 4110 15.4 0.042 0.044 2.819 3.459 wfi_extrapolate 11 7.9 0.001 0.001 3.326 3.326 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.846 3.286 calculate_dm_sparse 110 9.5 0.001 0.001 3.136 3.159 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.015 3.016 cp_fm_diag_elpa_base 48 14.0 2.467 2.675 3.014 3.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.837 2.840 grid_collocate_task_list 110 9.6 2.633 2.657 2.633 2.657 potential_pw2rs 110 12.3 0.021 0.021 2.520 2.522 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.406 2.428 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.351 2.364 rs_pw_transfer 902 11.9 0.011 0.011 2.293 2.315 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.214 2.271 calculate_first_density_matrix 1 7.0 0.001 0.003 2.263 2.266 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.191 cp_fm_cholesky_decompose 22 10.9 2.049 2.069 2.049 2.069 mp_alltoall_d11v 2046 13.8 1.939 2.030 1.939 2.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.948 1.952 qs_create_task_list 11 7.9 0.001 0.001 1.896 1.944 generate_qs_task_list 11 8.9 0.736 0.787 1.895 1.943 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.751 1.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.335000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1214.727273, yerr=55.200611 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.352768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.080 209.557 209.560 qs_mol_dyn_low 1 2.0 0.007 0.034 209.071 209.086 qs_forces 11 3.9 0.003 0.003 208.892 208.894 qs_energies 11 4.9 0.020 0.067 203.065 203.099 scf_env_do_scf 11 5.9 0.001 0.003 186.099 186.103 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.523 164.526 velocity_verlet 10 3.0 0.005 0.024 126.624 126.625 dbcsr_multiply_generic 2507 12.6 0.178 0.183 125.198 125.964 qs_scf_new_mos 117 7.6 0.001 0.001 123.347 123.674 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.346 123.673 ot_scf_mini 117 9.6 0.003 0.005 116.808 117.086 multiply_cannon 2507 13.6 0.238 0.245 101.389 103.475 multiply_cannon_loop 2507 14.6 2.083 2.183 99.099 101.225 ot_mini 117 10.6 0.001 0.001 66.770 67.045 multiply_cannon_multrec 60168 15.6 33.307 35.831 41.569 43.955 qs_ot_get_derivative 117 11.6 0.001 0.003 41.685 41.954 rebuild_ks_matrix 128 8.3 0.001 0.001 34.584 34.997 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 34.583 34.997 mp_waitall_1 267128 16.5 28.938 32.089 28.938 32.089 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.003 31.381 multiply_cannon_sync_h2d 60168 15.6 27.391 29.521 27.391 29.521 qs_ot_get_p 128 10.4 0.001 0.006 28.124 28.489 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.457 25.347 apply_single 128 13.6 0.001 0.001 24.457 25.347 ot_diis_step 117 11.6 0.007 0.008 24.705 24.707 init_scf_loop 11 6.9 0.001 0.005 21.495 21.500 qs_ot_p2m_diag 83 11.4 0.077 0.092 21.241 21.324 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.392 19.582 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 15.910 19.090 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.537 18.539 prepare_preconditioner 11 7.9 0.000 0.001 16.694 16.749 make_preconditioner 11 8.9 0.001 0.008 16.694 16.749 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.902 16.072 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.469 15.475 cp_fm_redistribute_end 83 14.4 12.217 15.376 12.231 15.380 cp_fm_diag_elpa_base 83 14.4 3.102 14.988 3.136 15.092 sum_up_and_integrate 128 10.3 0.089 0.107 14.919 14.944 integrate_v_rspace 128 11.3 0.004 0.004 14.829 14.857 make_m2s 5014 13.6 0.105 0.114 14.059 14.476 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.292 14.373 calculate_rho_elec 128 8.7 0.045 0.064 14.292 14.372 make_images 5014 14.6 0.404 0.423 13.875 14.302 init_scf_run 11 5.9 0.000 0.001 12.608 12.609 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.608 12.609 density_rs2pw 128 9.7 0.007 0.008 7.640 10.941 mp_sum_l 7870 13.0 8.507 9.929 8.507 9.929 rs_pw_transfer 1046 11.9 0.017 0.019 6.282 9.537 cp_fm_cholesky_invert 11 10.9 9.484 9.492 9.484 9.492 wfi_extrapolate 11 7.9 0.001 0.001 9.361 9.361 pw_transfer 1547 11.6 0.076 0.097 8.519 8.791 multiply_cannon_metrocomm1 60168 15.6 0.089 0.097 6.265 8.634 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.314 8.588 dbcsr_mm_accdrv_process 124484 16.2 3.199 3.467 7.814 8.510 calculate_dm_sparse 128 9.5 0.001 0.001 8.363 8.471 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.119 8.282 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.826 7.914 make_images_data 5014 15.6 0.065 0.071 6.787 7.687 grid_integrate_task_list 128 12.3 7.035 7.623 7.035 7.623 fft_wrap_pw1pw2_140 523 13.2 0.444 0.512 7.040 7.367 fft3d_ps 1291 14.7 2.103 2.891 7.101 7.276 hybrid_alltoall_any 5200 16.5 0.292 2.272 5.941 7.063 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.703 6.712 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.811 5.962 mp_waitany 16020 13.9 2.713 5.949 2.713 5.949 grid_collocate_task_list 128 9.7 4.567 5.775 4.567 5.775 mp_alltoall_d11v 2415 14.1 4.401 5.682 4.401 5.682 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.302 2.140 5.460 potential_pw2rs 128 12.3 0.009 0.011 5.254 5.292 cp_fm_cholesky_decompose 22 10.9 5.004 5.020 5.004 5.020 mp_sum_d 4462 12.2 3.790 4.561 3.790 4.561 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.560000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=7.755563 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 827.187200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57145. MP_Allreduce 11262 946. MP_Sync 168 MP_Alltoall 1955 5616511. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 193.128 193.129 qs_mol_dyn_low 1 2.0 0.003 0.004 192.790 192.806 qs_forces 11 3.9 0.003 0.003 192.706 192.707 qs_energies 11 4.9 0.002 0.003 185.816 185.830 scf_env_do_scf 11 5.9 0.001 0.001 169.230 169.240 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 135.160 135.163 velocity_verlet 10 3.0 0.001 0.001 121.535 121.537 dbcsr_multiply_generic 2485 12.5 0.184 0.190 97.540 98.736 qs_scf_new_mos 116 7.6 0.001 0.001 94.654 95.142 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.654 95.142 ot_scf_mini 116 9.6 0.004 0.004 89.865 90.455 multiply_cannon 2485 13.5 0.477 0.533 76.989 81.406 multiply_cannon_loop 2485 14.5 1.243 1.283 73.542 76.227 ot_mini 116 10.6 0.001 0.001 49.911 50.436 mp_waitall_1 212858 16.6 25.150 41.021 25.150 41.021 multiply_cannon_multrec 29820 15.5 21.563 26.069 31.122 36.040 rebuild_ks_matrix 127 8.3 0.001 0.001 33.640 34.294 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 33.639 34.293 init_scf_loop 11 6.9 0.000 0.000 33.962 33.963 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.227 30.821 multiply_cannon_metrocomm3 29820 15.5 0.096 0.102 16.017 30.142 prepare_preconditioner 11 7.9 0.000 0.000 29.476 29.539 make_preconditioner 11 8.9 0.000 0.000 29.476 29.539 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.187 28.738 qs_ot_get_derivative 116 11.6 0.001 0.002 28.038 28.625 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.942 23.057 apply_single 127 13.6 0.001 0.001 21.942 23.057 qs_ot_get_p 127 10.4 0.001 0.001 21.702 22.364 ot_diis_step 116 11.6 0.014 0.015 21.699 21.701 multiply_cannon_sync_h2d 29820 15.5 18.924 21.205 18.924 21.205 cp_fm_cholesky_invert 11 10.9 17.291 17.304 17.291 17.304 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.783 16.825 make_m2s 4970 13.5 0.089 0.094 14.612 16.297 make_images 4970 14.5 1.151 1.350 14.402 16.082 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.579 15.581 sum_up_and_integrate 127 10.3 0.113 0.130 14.974 15.001 integrate_v_rspace 127 11.3 0.003 0.004 14.860 14.892 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.335 14.375 calculate_rho_elec 127 8.7 0.087 0.105 14.334 14.374 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.389 12.404 cp_fm_redistribute_end 82 14.4 7.233 12.321 7.243 12.321 cp_fm_diag_elpa_base 82 14.4 4.843 11.807 5.054 12.162 init_scf_run 11 5.9 0.000 0.001 11.732 11.733 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.732 11.733 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.016 11.439 make_images_data 4970 15.5 0.062 0.071 8.922 10.895 multiply_cannon_metrocomm4 27335 15.5 0.097 0.112 3.801 10.613 hybrid_alltoall_any 5155 16.4 0.338 1.481 7.642 10.242 mp_irecv_dv 68888 16.3 3.606 10.232 3.606 10.232 density_rs2pw 127 9.7 0.006 0.007 7.745 10.139 dbcsr_mm_accdrv_process 61726 16.2 4.609 5.426 9.028 9.632 pw_transfer 1535 11.6 0.086 0.112 9.167 9.246 fft_wrap_pw1pw2 1281 12.7 0.011 0.013 8.944 9.028 wfi_extrapolate 11 7.9 0.001 0.001 8.598 8.598 rs_pw_transfer 1038 11.9 0.014 0.016 5.988 8.405 fft_wrap_pw1pw2_140 519 13.2 0.471 0.532 7.827 7.931 grid_integrate_task_list 127 12.3 7.087 7.503 7.087 7.503 fft3d_ps 1281 14.7 2.730 2.900 7.320 7.360 cp_fm_cholesky_decompose 22 10.9 7.226 7.317 7.226 7.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.283 7.056 calculate_dm_sparse 127 9.5 0.001 0.001 6.451 6.574 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.175 6.184 mp_sum_l 7804 13.0 4.168 6.005 4.168 6.005 grid_collocate_task_list 127 9.7 4.670 5.811 4.670 5.811 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.382 5.457 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.281 5.456 potential_pw2rs 127 12.3 0.015 0.018 5.290 5.304 mp_allgather_i34 2485 14.5 2.028 5.274 2.028 5.274 mp_alltoall_d11v 2401 14.1 4.253 5.079 4.253 5.079 mp_waitany 11660 13.9 2.465 4.834 2.465 4.834 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.371 2.046 4.435 dbcsr_complete_redistribute 393 12.7 0.774 0.870 3.291 4.142 mp_sum_d 4446 12.1 2.661 4.103 2.661 4.103 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.822 3.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.129000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.090909, yerr=4.813892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 939.638784E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 180.077 180.079 qs_mol_dyn_low 1 2.0 0.003 0.003 179.690 179.707 qs_forces 11 3.9 0.003 0.003 179.556 179.561 qs_energies 11 4.9 0.001 0.002 172.891 172.900 scf_env_do_scf 11 5.9 0.001 0.001 157.311 157.312 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 121.409 121.410 velocity_verlet 10 3.0 0.001 0.001 115.432 115.435 dbcsr_multiply_generic 2507 12.6 0.181 0.185 81.934 82.948 qs_scf_new_mos 117 7.6 0.001 0.001 82.562 82.939 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.561 82.939 ot_scf_mini 117 9.6 0.003 0.004 78.426 78.826 multiply_cannon 2507 13.6 0.501 0.520 61.923 66.407 multiply_cannon_loop 2507 14.6 0.852 0.875 58.780 61.466 ot_mini 117 10.6 0.001 0.001 42.796 43.219 init_scf_loop 11 6.9 0.000 0.000 35.805 35.806 mp_waitall_1 170520 16.6 25.161 34.387 25.161 34.387 prepare_preconditioner 11 7.9 0.000 0.000 31.692 31.736 make_preconditioner 11 8.9 0.000 0.000 31.692 31.736 rebuild_ks_matrix 128 8.3 0.001 0.001 31.158 31.721 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.157 31.720 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.352 30.700 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.085 28.595 multiply_cannon_multrec 20056 15.6 13.450 16.217 22.199 25.010 multiply_cannon_metrocomm3 20056 15.6 0.060 0.065 15.322 24.292 qs_ot_get_derivative 117 11.6 0.001 0.002 23.074 23.481 qs_ot_get_p 128 10.4 0.001 0.001 20.961 21.473 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.845 20.786 apply_single 128 13.6 0.001 0.001 19.845 20.786 ot_diis_step 117 11.6 0.018 0.018 19.606 19.606 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.578 16.591 multiply_cannon_sync_h2d 20056 15.6 14.279 15.997 14.279 15.997 make_m2s 5014 13.6 0.081 0.086 14.784 15.710 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.546 15.548 make_images 5014 14.6 1.187 1.275 14.549 15.467 sum_up_and_integrate 128 10.3 0.132 0.142 14.939 14.968 cp_fm_cholesky_invert 11 10.9 14.896 14.906 14.896 14.906 integrate_v_rspace 128 11.3 0.004 0.004 14.807 14.832 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.627 14.674 calculate_rho_elec 128 8.7 0.132 0.147 14.627 14.674 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.312 12.314 cp_fm_redistribute_end 83 14.4 4.654 12.237 4.671 12.240 cp_fm_diag_elpa_base 83 14.4 7.138 11.624 7.544 12.089 init_scf_run 11 5.9 0.000 0.001 10.576 10.576 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.575 10.576 make_images_data 5014 15.6 0.060 0.069 9.048 10.418 hybrid_alltoall_any 5200 16.5 0.434 1.989 7.811 9.732 density_rs2pw 128 9.7 0.006 0.006 7.518 9.483 pw_transfer 1547 11.6 0.086 0.108 9.328 9.453 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.117 9.413 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.103 9.233 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.422 9.070 mp_irecv_dv 50230 16.2 3.300 8.825 3.300 8.825 dbcsr_mm_accdrv_process 41502 16.2 4.367 4.991 8.212 8.383 fft_wrap_pw1pw2_140 523 13.2 0.479 0.521 7.991 8.131 grid_integrate_task_list 128 12.3 7.272 7.835 7.272 7.835 cp_fm_cholesky_decompose 22 10.9 7.706 7.743 7.706 7.743 cp_fm_upper_to_full 105 14.5 5.728 7.493 5.728 7.493 wfi_extrapolate 11 7.9 0.001 0.001 7.472 7.472 fft3d_ps 1291 14.7 2.672 2.909 7.365 7.463 rs_pw_transfer 1046 11.9 0.013 0.015 5.388 7.354 dbcsr_complete_redistribute 395 12.7 1.166 1.196 4.568 6.330 calculate_dm_sparse 128 9.5 0.001 0.001 5.733 5.851 grid_collocate_task_list 128 9.7 4.893 5.838 4.893 5.838 mp_alltoall_d11v 2415 14.1 4.382 5.554 4.382 5.554 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.522 5.529 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.609 5.293 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.388 5.141 mp_sum_l 7870 13.0 3.405 5.069 3.405 5.069 potential_pw2rs 128 12.3 0.020 0.022 5.007 5.031 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.645 4.769 mp_allgather_i34 2507 14.6 1.672 4.615 1.672 4.615 mp_waitany 11748 13.9 2.340 4.434 2.340 4.434 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.082 4.110 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.321 4.018 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.348 1.930 3.896 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.763 3.765 mp_alltoall_i22 716 14.1 1.898 3.760 1.898 3.760 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.585 3.639 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.079000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.272727, yerr=11.786671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 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 1963542011904 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.155596E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4073 57357. MP_Allreduce 11221 1069. MP_Sync 168 MP_Alltoall 1700 12496361. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 189.208 189.209 qs_mol_dyn_low 1 2.0 0.003 0.003 188.824 188.839 qs_forces 11 3.9 0.003 0.003 188.669 188.673 qs_energies 11 4.9 0.001 0.002 181.423 181.434 scf_env_do_scf 11 5.9 0.001 0.001 164.579 164.586 velocity_verlet 10 3.0 0.001 0.001 125.308 125.312 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.431 117.432 qs_scf_new_mos 116 7.6 0.001 0.001 79.740 80.073 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.740 80.073 dbcsr_multiply_generic 2485 12.5 0.188 0.193 78.817 79.547 ot_scf_mini 116 9.6 0.003 0.004 75.288 75.582 multiply_cannon 2485 13.5 0.550 0.583 54.344 57.189 multiply_cannon_loop 2485 14.5 1.168 1.198 50.641 52.404 init_scf_loop 11 6.9 0.000 0.000 47.023 47.024 prepare_preconditioner 11 7.9 0.000 0.000 42.853 42.876 make_preconditioner 11 8.9 0.000 0.000 42.853 42.876 ot_mini 116 10.6 0.001 0.001 42.139 42.433 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.417 41.431 multiply_cannon_multrec 29820 15.5 13.964 19.196 25.894 30.538 rebuild_ks_matrix 127 8.3 0.001 0.001 30.038 30.321 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 30.038 30.321 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.005 27.252 mp_waitall_1 146592 16.7 17.234 26.949 17.234 26.949 qs_ot_get_derivative 116 11.6 0.001 0.002 22.745 23.048 make_m2s 4970 13.5 0.095 0.098 20.010 21.020 make_images 4970 14.5 1.921 2.253 19.707 20.715 qs_ot_get_p 127 10.4 0.001 0.001 19.545 19.843 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.788 19.352 apply_single 127 13.6 0.001 0.001 18.787 19.352 ot_diis_step 116 11.6 0.017 0.018 19.275 19.276 cp_fm_cholesky_invert 11 10.9 16.573 16.582 16.573 16.582 cp_fm_upper_to_full 104 14.7 11.047 16.316 11.047 16.316 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.373 15.425 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.434 15.249 sum_up_and_integrate 127 10.3 0.139 0.152 14.942 14.969 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.814 14.858 calculate_rho_elec 127 8.7 0.174 0.189 14.813 14.857 integrate_v_rspace 127 11.3 0.004 0.004 14.803 14.834 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.081 14.083 dbcsr_complete_redistribute 393 12.7 1.496 1.621 9.033 12.777 multiply_cannon_sync_h2d 29820 15.5 11.561 12.425 11.561 12.425 make_images_data 4970 15.5 0.063 0.067 10.741 12.352 dbcsr_mm_accdrv_process 61748 16.2 7.246 8.227 11.507 11.976 hybrid_alltoall_any 5155 16.4 0.520 2.193 9.478 11.520 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.636 11.362 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.981 10.984 init_scf_run 11 5.9 0.000 0.001 10.947 10.948 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.946 10.948 cp_fm_redistribute_end 82 14.4 1.899 10.891 1.918 10.898 cp_fm_diag_elpa_base 82 14.4 8.370 10.270 8.942 10.750 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.413 10.076 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.449 9.660 pw_transfer 1535 11.6 0.088 0.107 9.588 9.648 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.362 9.429 mp_alltoall_i22 712 14.1 5.591 9.423 5.591 9.423 density_rs2pw 127 9.7 0.006 0.006 7.253 8.586 fft_wrap_pw1pw2_140 519 13.2 0.480 0.490 8.254 8.343 cp_fm_cholesky_decompose 22 10.9 7.750 7.841 7.750 7.841 grid_integrate_task_list 127 12.3 7.455 7.826 7.455 7.826 wfi_extrapolate 11 7.9 0.001 0.001 7.763 7.763 fft3d_ps 1281 14.7 2.755 2.837 7.599 7.648 multiply_cannon_metrocomm4 24850 15.5 0.076 0.088 2.780 7.068 mp_irecv_dv 75445 16.2 2.636 6.804 2.636 6.804 calculate_dm_sparse 127 9.5 0.001 0.001 6.133 6.230 rs_pw_transfer 1038 11.9 0.013 0.014 4.805 6.186 grid_collocate_task_list 127 9.7 5.039 5.666 5.039 5.666 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.517 5.551 mp_alltoall_d11v 2401 14.1 4.980 5.546 4.980 5.546 potential_pw2rs 127 12.3 0.022 0.023 4.799 4.822 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.466 4.530 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.379 4.474 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.466 4.467 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.126 4.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.209000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.090909, yerr=25.614110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 1.510511E+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_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.037 175.261 175.263 qs_mol_dyn_low 1 2.0 0.003 0.003 174.876 174.892 qs_forces 11 3.9 0.003 0.003 174.636 174.641 qs_energies 11 4.9 0.001 0.002 167.066 167.075 scf_env_do_scf 11 5.9 0.001 0.001 149.902 149.905 velocity_verlet 10 3.0 0.001 0.001 115.312 115.324 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 114.221 114.222 dbcsr_multiply_generic 2527 12.6 0.183 0.189 74.647 75.107 qs_scf_new_mos 118 7.6 0.001 0.001 74.726 74.819 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.726 74.818 ot_scf_mini 118 9.6 0.004 0.004 70.224 70.300 multiply_cannon 2527 13.6 0.592 0.630 54.688 59.455 multiply_cannon_loop 2527 14.6 0.444 0.457 50.076 50.953 ot_mini 118 10.6 0.001 0.001 39.921 39.997 init_scf_loop 11 6.9 0.000 0.000 35.532 35.534 mp_waitall_1 126780 16.7 26.132 33.675 26.132 33.675 prepare_preconditioner 11 7.9 0.000 0.000 31.559 31.590 make_preconditioner 11 8.9 0.000 0.000 31.559 31.590 rebuild_ks_matrix 129 8.3 0.001 0.001 30.437 30.532 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 30.436 30.532 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.489 29.780 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.707 27.794 multiply_cannon_multrec 10108 15.6 10.486 14.620 17.835 20.859 ot_diis_step 118 11.6 0.020 0.025 20.022 20.023 qs_ot_get_derivative 118 11.6 0.002 0.002 19.831 19.911 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.533 19.855 apply_single 129 13.6 0.001 0.001 19.533 19.855 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 12.340 19.347 make_m2s 5054 13.6 0.067 0.072 16.309 18.421 make_images 5054 14.6 2.291 2.616 16.000 18.117 cp_fm_cholesky_invert 11 10.9 18.037 18.043 18.037 18.043 qs_ot_get_p 129 10.4 0.001 0.001 17.309 17.399 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.743 15.799 calculate_rho_elec 129 8.7 0.261 0.271 15.743 15.799 sum_up_and_integrate 129 10.3 0.181 0.192 15.359 15.404 integrate_v_rspace 129 11.3 0.004 0.007 15.177 15.232 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.566 13.585 make_images_data 5054 15.6 0.051 0.060 9.972 12.449 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.426 12.426 hybrid_alltoall_any 5240 16.5 0.829 3.702 9.771 12.197 multiply_cannon_sync_h2d 10108 15.6 11.709 12.124 11.709 12.124 init_scf_run 11 5.9 0.000 0.001 10.392 10.392 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.391 10.392 pw_transfer 1559 11.6 0.086 0.092 10.121 10.162 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.897 9.942 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.409 9.419 cp_fm_diag_elpa_base 83 14.4 9.163 9.254 9.405 9.416 fft_wrap_pw1pw2_140 527 13.2 0.500 0.526 8.721 8.768 grid_integrate_task_list 129 12.3 7.803 8.227 7.803 8.227 density_rs2pw 129 9.7 0.006 0.006 7.229 8.123 fft3d_ps 1301 14.7 2.727 2.802 8.044 8.063 cp_fm_cholesky_decompose 22 10.9 7.930 8.040 7.930 8.040 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.935 7.992 multiply_cannon_metrocomm1 10108 15.6 0.029 0.029 4.557 7.659 mp_allgather_i34 2527 14.6 2.806 7.650 2.806 7.650 dbcsr_mm_accdrv_process 20926 16.1 2.746 3.442 6.972 7.601 wfi_extrapolate 11 7.9 0.001 0.001 7.471 7.472 calculate_dm_sparse 129 9.5 0.001 0.001 6.169 6.257 grid_collocate_task_list 129 9.7 5.434 6.105 5.434 6.105 mp_alltoall_d11v 2423 14.1 5.074 5.950 5.074 5.950 dbcsr_complete_redistribute 395 12.7 2.110 2.163 5.094 5.488 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.297 5.298 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.128 5.151 rs_pw_transfer 1054 12.0 0.012 0.013 4.198 5.130 potential_pw2rs 129 12.3 0.026 0.027 4.683 4.696 multiply_cannon_metrocomm4 7581 15.6 0.025 0.027 1.872 4.302 mp_irecv_dv 29086 15.9 1.836 4.226 1.836 4.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.062 4.087 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.566 3.878 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.723 3.753 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.328 3.663 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.641 3.649 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.491 3.575 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.548 3.571 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.263000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.090909, yerr=51.142323 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 2.943889E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 304.908 304.909 qs_mol_dyn_low 1 2.0 0.003 0.003 304.358 304.369 qs_forces 11 3.9 0.003 0.003 304.243 304.247 qs_energies 11 4.9 0.002 0.002 294.775 294.791 scf_env_do_scf 11 5.9 0.001 0.001 272.266 272.279 velocity_verlet 10 3.0 0.001 0.001 220.072 220.083 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 145.516 145.517 init_scf_loop 11 6.9 0.000 0.000 126.494 126.496 prepare_preconditioner 11 7.9 0.000 0.000 121.340 121.372 make_preconditioner 11 8.9 0.000 0.000 121.340 121.372 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.139 118.557 qs_scf_new_mos 118 7.6 0.001 0.001 93.279 93.404 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.278 93.404 ot_scf_mini 118 9.6 0.004 0.004 88.447 88.481 dbcsr_multiply_generic 2534 12.6 0.216 0.224 84.810 85.274 cp_fm_upper_to_full 106 14.8 53.926 77.257 53.926 77.257 multiply_cannon 2534 13.6 0.713 0.764 59.772 60.395 multiply_cannon_loop 2534 14.6 0.476 0.484 56.088 57.310 ot_mini 118 10.6 0.001 0.001 45.482 45.517 dbcsr_complete_redistribute 397 12.7 3.963 4.017 29.891 43.221 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.482 39.831 rebuild_ks_matrix 129 8.3 0.001 0.001 39.343 39.361 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.343 39.360 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.156 37.377 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.276 36.296 mp_alltoall_i22 720 14.1 22.030 35.296 22.030 35.296 cp_fm_cholesky_invert 11 10.9 34.044 34.051 34.044 34.051 mp_waitall_1 104780 16.8 29.012 32.713 29.012 32.713 qs_ot_get_p 129 10.4 0.001 0.001 27.201 27.257 qs_ot_get_derivative 118 11.6 0.002 0.002 24.958 24.999 qs_ot_p2m_diag 84 11.4 0.889 0.894 23.073 23.104 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.302 21.304 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.135 21.154 calculate_rho_elec 129 8.7 0.487 0.488 21.135 21.154 make_m2s 5068 13.6 0.077 0.080 20.207 21.114 multiply_cannon_metrocomm3 10136 15.6 0.024 0.024 19.495 20.658 make_images 5068 14.6 3.809 3.905 19.723 20.634 ot_diis_step 118 11.6 0.022 0.023 20.474 20.475 sum_up_and_integrate 129 10.3 0.323 0.325 20.227 20.303 integrate_v_rspace 129 11.3 0.004 0.004 19.904 19.982 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.665 19.852 apply_single 129 13.6 0.001 0.001 19.665 19.851 multiply_cannon_multrec 10136 15.6 10.706 12.596 18.235 18.298 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.017 18.017 cp_fm_diag_elpa_base 84 14.4 13.469 15.114 18.013 18.013 multiply_cannon_sync_h2d 10136 15.6 15.783 15.816 15.783 15.816 hybrid_alltoall_any 5255 16.5 1.311 3.060 10.855 12.663 pw_transfer 1559 11.6 0.093 0.094 12.609 12.616 make_images_data 5068 15.6 0.060 0.064 10.727 12.449 init_scf_run 11 5.9 0.000 0.001 12.439 12.439 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.438 12.439 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 12.372 12.379 fft_wrap_pw1pw2_140 527 13.2 0.546 0.549 10.962 10.973 fft3d_ps 1301 14.7 2.764 2.775 10.397 10.411 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.776 9.803 wfi_extrapolate 11 7.9 0.001 0.001 9.333 9.333 mp_alltoall_d11v 2429 14.1 8.544 9.317 8.544 9.317 dbcsr_mm_accdrv_process 20954 16.1 3.910 5.556 7.288 9.152 cp_fm_cholesky_decompose 22 10.9 9.115 9.151 9.115 9.151 grid_integrate_task_list 129 12.3 8.645 8.817 8.645 8.817 density_rs2pw 129 9.7 0.005 0.006 8.576 8.708 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.066 8.069 calculate_dm_sparse 129 9.5 0.001 0.001 6.701 6.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.413 6.469 grid_collocate_task_list 129 9.7 6.410 6.449 6.410 6.449 rs_scatter_matrices 140 9.7 3.634 4.622 6.116 6.351 copy_dbcsr_to_fm 187 11.8 0.004 0.005 6.101 6.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=304.909000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2601.454545, yerr=179.573922 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 1.261920E+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 7111212. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.072 84.864 84.872 qs_energies 1 2.0 0.000 0.000 84.389 84.397 ls_scf 1 3.0 0.000 0.000 83.473 83.483 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.499 72.644 multiply_cannon 111 7.7 0.018 0.021 55.814 56.985 multiply_cannon_loop 111 8.7 0.206 0.221 52.394 53.595 ls_scf_main 1 4.0 0.000 0.000 52.211 52.211 density_matrix_trs4 2 5.0 0.002 0.003 46.697 46.779 ls_scf_init_scf 1 4.0 0.000 0.000 28.234 28.236 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.122 27.177 mp_waitall_1 11031 10.9 22.584 25.351 22.584 25.351 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.036 25.061 multiply_cannon_multrec 2664 9.7 8.130 8.869 15.494 17.210 multiply_cannon_sync_h2d 2664 9.7 13.523 14.721 13.523 14.721 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.689 14.123 make_m2s 222 7.7 0.008 0.011 13.097 13.548 make_images 222 8.7 0.099 0.109 13.075 13.528 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.557 8.886 make_images_data 222 9.7 0.004 0.005 7.663 8.241 hybrid_alltoall_any 227 10.6 0.215 1.834 6.563 8.182 dbcsr_mm_accdrv_process 4760 10.4 0.514 0.662 6.982 7.977 dbcsr_mm_accdrv_process_sort 4760 11.4 6.269 7.165 6.269 7.165 calculate_norms 4752 9.8 5.502 6.087 5.502 6.087 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.998 5.126 mp_sum_l 807 5.4 3.055 4.465 3.055 4.465 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.044 3.617 mp_irecv_dv 6231 10.9 2.026 3.593 2.026 3.593 make_images_sizes 222 9.7 0.000 0.000 0.807 3.547 mp_alltoall_i44 222 10.7 0.807 3.547 0.807 3.547 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.288 3.430 arnoldi_extremal 4 6.8 0.000 0.000 3.239 3.266 arnoldi_normal_ev 4 7.8 0.001 0.003 3.239 3.266 build_subspace 16 8.4 0.009 0.012 3.135 3.140 ls_scf_post 1 4.0 0.000 0.000 3.028 3.038 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.884 dbcsr_special_finalize 555 9.7 0.005 0.005 2.272 2.678 dbcsr_merge_single_wm 555 10.7 0.463 0.596 2.265 2.670 make_images_pack 222 9.7 2.209 2.636 2.211 2.638 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.316 2.563 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.460 2.065 2.462 dbcsr_sort_data 658 11.4 2.063 2.441 2.063 2.441 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.311 2.368 buffer_matrices_ensure_size 222 8.7 1.781 2.098 1.781 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.801 1.802 rebuild_ks_matrix 3 7.3 0.000 0.000 1.792 1.793 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.791 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.872000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 2.082443E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.044 91.175 91.177 qs_energies 1 2.0 0.000 0.000 90.675 90.679 ls_scf 1 3.0 0.000 0.001 89.346 89.351 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.305 75.665 multiply_cannon 111 7.7 0.029 0.041 53.161 56.814 ls_scf_main 1 4.0 0.000 0.000 54.980 54.987 multiply_cannon_loop 111 8.7 0.116 0.122 49.942 53.229 density_matrix_trs4 2 5.0 0.002 0.003 49.174 49.392 ls_scf_init_scf 1 4.0 0.000 0.002 30.870 30.871 mp_waitall_1 9105 10.9 21.025 30.392 21.025 30.392 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.652 29.746 multiply_cannon_multrec 1332 9.7 13.110 16.931 22.375 27.311 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.263 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.621 21.043 make_m2s 222 7.7 0.006 0.007 15.406 16.024 make_images 222 8.7 1.586 1.980 15.376 15.994 dbcsr_mm_accdrv_process 4041 10.4 0.279 0.451 8.865 10.424 dbcsr_mm_accdrv_process_sort 4041 11.4 8.458 9.980 8.458 9.980 make_images_data 222 9.7 0.004 0.004 8.763 9.692 hybrid_alltoall_any 227 10.6 0.524 2.498 8.120 9.112 mp_sum_l 807 5.4 5.442 8.525 5.442 8.525 multiply_cannon_metrocomm4 1221 9.7 0.006 0.009 3.217 7.829 mp_irecv_dv 3311 11.0 3.198 7.766 3.198 7.766 calculate_norms 2376 9.8 6.017 6.835 6.017 6.835 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.192 6.657 multiply_cannon_sync_h2d 1332 9.7 4.864 6.136 4.864 6.136 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.126 5.359 arnoldi_extremal 4 6.8 0.000 0.000 4.713 4.732 arnoldi_normal_ev 4 7.8 0.001 0.005 4.713 4.732 build_subspace 16 8.4 0.014 0.021 4.446 4.449 ls_scf_post 1 4.0 0.000 0.000 3.496 3.501 dbcsr_matrix_vector_mult 304 9.0 0.009 0.022 3.156 3.390 ls_scf_store_result 1 5.0 0.000 0.000 3.209 3.317 dbcsr_matrix_vector_mult_local 304 10.0 2.731 3.204 2.733 3.206 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.270 2.789 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.620 2.726 make_images_pack 222 9.7 2.031 2.408 2.033 2.410 mp_allgather_i34 111 8.7 0.929 2.322 0.929 2.322 dbcsr_sort_data 436 11.2 1.852 2.099 1.852 2.099 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.886 1.889 rebuild_ks_matrix 3 7.3 0.000 0.000 1.874 1.876 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.874 1.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.177000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1713.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 2.689737E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.110 94.240 94.242 qs_energies 1 2.0 0.000 0.000 93.570 93.580 ls_scf 1 3.0 0.000 0.001 92.149 92.160 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.769 77.046 ls_scf_main 1 4.0 0.000 0.001 57.684 57.689 multiply_cannon 111 7.7 0.041 0.088 52.664 56.976 multiply_cannon_loop 111 8.7 0.099 0.105 49.051 52.896 density_matrix_trs4 2 5.0 0.002 0.003 51.682 51.802 mp_waitall_1 7281 11.0 23.962 34.613 23.962 34.613 ls_scf_init_scf 1 4.0 0.000 0.002 30.787 30.789 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.595 29.651 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.152 27.164 multiply_cannon_multrec 888 9.7 12.584 15.138 21.157 24.283 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.974 23.260 make_m2s 222 7.7 0.006 0.007 17.414 18.615 make_images 222 8.7 1.972 2.316 17.376 18.578 hybrid_alltoall_any 227 10.6 0.619 2.874 9.649 11.052 make_images_data 222 9.7 0.003 0.004 10.040 11.003 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.414 8.106 9.342 dbcsr_mm_accdrv_process_sort 3754 11.4 7.736 8.928 7.736 8.928 mp_sum_l 807 5.4 5.145 8.653 5.145 8.653 multiply_cannon_sync_h2d 888 9.7 6.001 7.399 6.001 7.399 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.476 6.856 mp_irecv_dv 2335 11.1 2.460 6.811 2.460 6.811 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.870 6.788 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.571 6.054 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.983 5.160 arnoldi_extremal 4 6.8 0.000 0.000 5.131 5.149 arnoldi_normal_ev 4 7.8 0.001 0.005 5.131 5.149 build_subspace 16 8.4 0.014 0.020 4.811 4.818 calculate_norms 1584 9.8 4.252 4.595 4.252 4.595 mp_allgather_i34 111 8.7 1.458 3.981 1.458 3.981 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.438 3.785 ls_scf_post 1 4.0 0.000 0.000 3.678 3.688 dbcsr_matrix_vector_mult_local 304 10.0 3.005 3.581 3.007 3.583 ls_scf_store_result 1 5.0 0.000 0.000 3.366 3.471 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.863 2.958 make_images_sizes 222 9.7 0.000 0.000 1.049 2.332 mp_alltoall_i44 222 10.7 1.049 2.332 1.049 2.332 dbcsr_sort_data 325 11.1 1.887 2.142 1.887 2.142 make_images_pack 222 9.7 1.817 2.120 1.820 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.940 1.942 rebuild_ks_matrix 3 7.3 0.000 0.000 1.922 1.924 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.922 1.924 dbcsr_data_release 9322 10.9 1.305 1.903 1.305 1.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.242000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2222.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 3.319783E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.062 97.411 97.412 qs_energies 1 2.0 0.000 0.000 96.777 96.783 ls_scf 1 3.0 0.000 0.000 95.094 95.101 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.785 79.087 ls_scf_main 1 4.0 0.000 0.000 59.192 59.192 multiply_cannon 111 7.7 0.072 0.136 51.884 56.839 density_matrix_trs4 2 5.0 0.002 0.003 53.019 53.149 multiply_cannon_loop 111 8.7 0.114 0.127 46.822 49.791 ls_scf_init_scf 1 4.0 0.000 0.000 32.641 32.642 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.401 31.484 mp_waitall_1 6369 11.0 23.110 28.989 23.110 28.989 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.912 28.926 multiply_cannon_multrec 1332 9.7 14.130 17.144 22.015 25.506 make_m2s 222 7.7 0.007 0.008 21.171 22.636 make_images 222 8.7 3.151 3.615 21.121 22.587 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.443 18.004 make_images_data 222 9.7 0.004 0.004 11.844 13.539 hybrid_alltoall_any 227 10.6 0.799 3.853 11.142 12.847 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.413 7.527 9.044 dbcsr_mm_accdrv_process_sort 3641 11.4 7.153 8.625 7.153 8.625 mp_sum_l 807 5.4 4.150 7.553 4.150 7.553 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.652 6.479 multiply_cannon_sync_h2d 1332 9.7 5.500 6.203 5.500 6.203 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.067 5.966 mp_irecv_dv 3229 10.9 2.045 5.902 2.045 5.902 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.131 5.866 arnoldi_extremal 4 6.8 0.000 0.000 5.249 5.269 arnoldi_normal_ev 4 7.8 0.001 0.005 5.249 5.269 build_subspace 16 8.4 0.014 0.021 4.901 4.907 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.575 4.799 calculate_norms 2376 9.8 4.197 4.550 4.197 4.550 mp_allgather_i34 111 8.7 2.169 4.483 2.169 4.483 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.586 3.910 dbcsr_matrix_vector_mult_local 304 10.0 3.168 3.671 3.171 3.673 dbcsr_sort_data 658 11.4 3.049 3.562 3.049 3.562 ls_scf_post 1 4.0 0.000 0.000 3.261 3.268 dbcsr_special_finalize 555 9.7 0.006 0.007 2.791 3.264 dbcsr_merge_single_wm 555 10.7 0.532 0.659 2.782 3.256 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.056 3.132 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.063 dbcsr_data_release 10477 10.7 1.592 2.494 1.592 2.494 dbcsr_finalize 304 7.8 0.049 0.061 1.804 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.975 1.977 rebuild_ks_matrix 3 7.3 0.000 0.000 1.952 1.954 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.952 1.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.412000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 4.624830E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.074 100.076 100.078 qs_energies 1 2.0 0.000 0.000 99.287 99.292 ls_scf 1 3.0 0.000 0.000 97.343 97.348 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.512 78.723 ls_scf_main 1 4.0 0.000 0.000 62.822 62.822 multiply_cannon 111 7.7 0.100 0.199 55.624 60.444 density_matrix_trs4 2 5.0 0.002 0.003 55.468 55.539 multiply_cannon_loop 111 8.7 0.069 0.076 51.175 53.404 mp_waitall_1 5436 11.0 27.201 32.685 27.201 32.685 ls_scf_init_scf 1 4.0 0.000 0.000 30.913 30.915 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.543 29.578 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.356 27.365 multiply_cannon_multrec 444 9.7 13.961 16.319 21.041 23.540 make_m2s 222 7.7 0.004 0.005 17.896 20.335 make_images 222 8.7 3.737 4.417 17.834 20.276 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.808 16.291 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.961 14.804 make_images_data 222 9.7 0.003 0.004 10.232 12.535 hybrid_alltoall_any 227 10.6 0.790 3.769 10.009 12.385 dbcsr_mm_accdrv_process 3003 10.4 0.189 0.350 6.785 7.929 multiply_cannon_sync_h2d 444 9.7 6.539 7.911 6.539 7.911 dbcsr_mm_accdrv_process_sort 3003 11.4 6.468 7.581 6.468 7.581 mp_allgather_i34 111 8.7 2.687 7.037 2.687 7.037 arnoldi_extremal 4 6.8 0.000 0.000 5.836 5.847 arnoldi_normal_ev 4 7.8 0.002 0.005 5.836 5.847 build_subspace 16 8.4 0.015 0.020 5.434 5.442 mp_sum_l 807 5.4 3.042 5.087 3.042 5.087 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.597 4.709 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.181 4.381 dbcsr_matrix_vector_mult_local 304 10.0 3.667 4.143 3.669 4.146 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.730 3.850 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.121 3.848 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.639 3.830 mp_irecv_dv 1241 11.2 1.619 3.802 1.619 3.802 calculate_norms 792 9.8 3.544 3.718 3.544 3.718 ls_scf_post 1 4.0 0.000 0.000 3.609 3.614 make_images_sizes 222 9.7 0.000 0.000 1.001 3.603 mp_alltoall_i44 222 10.7 1.001 3.603 1.001 3.603 ls_scf_store_result 1 5.0 0.000 0.000 3.384 3.427 dbcsr_finalize 304 7.8 0.062 0.078 2.203 2.299 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.227 2.229 rebuild_ks_matrix 3 7.3 0.000 0.000 2.195 2.197 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.195 2.197 dbcsr_merge_all 275 8.9 0.474 0.525 2.050 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.078000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3627.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1da4a734777d3157e9099d8f027b8c79c676a226_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 8.764826E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.098 0.129 108.804 108.804 qs_energies 1 2.0 0.000 0.000 107.271 107.291 ls_scf 1 3.0 0.000 0.000 104.303 104.323 dbcsr_multiply_generic 111 6.7 0.023 0.025 77.608 77.720 ls_scf_main 1 4.0 0.000 0.000 66.333 66.334 density_matrix_trs4 2 5.0 0.002 0.003 56.963 57.009 multiply_cannon 111 7.7 0.154 0.237 50.184 52.173 multiply_cannon_loop 111 8.7 0.067 0.070 46.696 47.479 ls_scf_init_scf 1 4.0 0.000 0.000 34.315 34.316 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.638 32.649 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.758 29.765 mp_waitall_1 4527 11.1 22.434 26.092 22.434 26.092 make_m2s 222 7.7 0.005 0.005 23.877 24.876 make_images 222 8.7 4.603 4.981 23.771 24.768 multiply_cannon_multrec 444 9.7 17.855 18.493 22.488 23.131 hybrid_alltoall_any 227 10.6 1.659 3.620 12.853 15.611 make_images_data 222 9.7 0.003 0.003 13.095 15.506 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.750 11.328 multiply_cannon_sync_h2d 444 9.7 8.816 8.874 8.816 8.874 arnoldi_extremal 4 6.8 0.000 0.000 7.372 7.378 arnoldi_normal_ev 4 7.8 0.003 0.009 7.371 7.378 build_subspace 16 8.4 0.026 0.037 6.786 6.799 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.442 5.595 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.329 5.413 dbcsr_matrix_vector_mult_local 304 10.0 4.975 5.289 4.977 5.292 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.028 5.259 dbcsr_mm_accdrv_process 1814 10.4 0.261 0.348 4.470 4.627 dbcsr_mm_accdrv_process_sort 1814 11.4 4.162 4.288 4.162 4.288 ls_scf_post 1 4.0 0.000 0.000 3.654 3.675 make_images_sizes 222 9.7 0.000 0.000 1.439 3.553 mp_alltoall_i44 222 10.7 1.438 3.553 1.438 3.553 mp_allgather_i34 111 8.7 1.130 3.519 1.130 3.519 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.413 calculate_norms 792 9.8 3.269 3.322 3.269 3.322 dbcsr_finalize 304 7.8 0.082 0.090 3.090 3.163 dbcsr_merge_all 275 8.9 0.892 0.916 2.873 2.944 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.938 2.938 dbcsr_complete_redistribute 5 7.6 1.439 1.470 2.812 2.925 dbcsr_data_release 12724 10.6 2.329 2.905 2.329 2.905 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.629 2.631 rebuild_ks_matrix 3 7.3 0.000 0.000 2.561 2.563 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.561 2.563 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.544 dbcsr_sort_data 325 11.1 2.441 2.498 2.441 2.498 dbcsr_new_transposed 4 7.5 0.244 0.254 2.306 2.325 dbcsr_frobenius_norm 74 6.6 2.058 2.129 2.186 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.204 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.804000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6829.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1da4a734777d3157e9099d8f027b8c79c676a226 Summary: empty Status: OK