=== 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: 948733f1c7cd1fbbd9135e2cb108ad12e592a17c ################# 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 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/01 job id: 41469274 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/02 job id: 41469279 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/03 job id: 41469282 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/04 job id: 41469285 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/05 job id: 41469288 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/06 job id: 41469292 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/07 job id: 41469295 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/08 job id: 41469296 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/09 job id: 41469299 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/10 job id: 41469302 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/11 job id: 41469305 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/12 job id: 41469308 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/13 job id: 41469309 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/14 job id: 41469312 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/15 job id: 41469315 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/16 job id: 41469318 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/17 job id: 41469322 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/18 job id: 41469325 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/19 job id: 41469327 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/20 job id: 41469329 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/21 job id: 41469332 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/22 job id: 41469334 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/23 job id: 41469335 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/24 job id: 41469336 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/25 job id: 41469337 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/26 job id: 41469338 --- 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/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 0.022 0.037 134.892 134.893 farming_run 1 2.0 134.320 134.322 134.862 134.865 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455378E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 116.888 116.888 qs_energies 1 2.0 0.000 0.000 116.662 116.664 mp2_main 1 3.0 0.000 0.000 114.728 114.730 mp2_gpw_main 1 4.0 0.019 0.025 113.796 113.797 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 94.372 94.653 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 56.273 56.555 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.169 42.568 47.771 get_2c_integrals 1 6.0 0.000 0.000 37.273 37.925 integrate_v_rspace 273 8.0 0.439 0.451 25.228 30.191 pw_transfer 6555 10.6 0.376 0.389 27.555 27.863 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.221 26.574 grid_integrate_task_list 273 9.0 20.985 26.487 20.985 26.487 fft_wrap_pw1pw2_100 2178 12.4 1.168 1.241 23.782 24.140 compute_2c_integrals 1 7.0 0.002 0.002 20.001 20.003 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 19.593 19.731 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.434 19.590 19.727 rpa_ri_compute_en 1 5.0 0.000 0.000 19.312 19.478 cp_fm_cholesky_decompose 12 8.2 17.278 17.922 17.278 17.922 fft3d_s 5443 13.4 16.264 16.809 16.286 16.830 cholesky_decomp 1 7.0 0.000 0.000 16.124 16.739 ao_to_mo_and_store_B_mult_1 272 7.0 10.861 15.588 10.861 15.588 calculate_wavefunction 272 8.0 5.388 5.543 12.566 13.231 rpa_num_int 1 6.0 0.000 0.000 10.620 10.620 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.566 10.607 calc_potential_gpw 544 9.5 0.005 0.005 9.670 10.031 calc_mat_Q 8 8.0 0.000 0.000 9.372 9.469 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.914 9.153 contract_S_to_Q 8 9.0 0.000 0.000 8.793 8.891 parallel_gemm_fm 14 9.1 0.000 0.000 8.381 8.451 parallel_gemm_fm_cosma 14 10.1 8.381 8.451 8.381 8.451 potential_pw2rs 545 10.0 0.106 0.108 7.720 8.446 create_integ_mat 1 6.0 0.023 0.028 8.317 8.317 collocate_single_gaussian 272 10.0 0.040 0.042 7.470 7.679 array2fm 1 7.0 0.000 0.000 7.183 7.650 pw_scatter_s 2720 13.7 4.459 4.584 4.459 4.584 pw_gather_s 2722 13.2 3.962 4.436 3.962 4.436 pw_poisson_solve 545 10.5 2.479 2.511 3.516 3.639 array2fm_buffer_send 1 8.0 3.267 3.447 3.267 3.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.794779, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.038 398.376 398.377 farming_run 1 2.0 397.632 397.641 398.341 398.344 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222210E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.050 211.441 211.442 qs_energies 1 2.0 0.000 0.000 211.222 211.228 scf_env_do_scf 1 3.0 0.000 0.000 107.446 107.446 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.584 106.594 rebuild_ks_matrix 4 6.0 0.000 0.000 106.583 106.592 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.064 106.583 106.592 hfx_ks_matrix 4 8.0 0.001 0.001 106.200 106.205 integrate_four_center 4 9.0 0.144 0.468 106.199 106.204 mp2_main 1 3.0 0.000 0.000 103.485 103.491 mp2_gpw_main 1 4.0 0.035 0.067 102.647 102.654 integrate_four_center_main 4 10.0 0.105 0.576 97.417 100.202 integrate_four_center_bin 266 11.0 97.312 100.099 97.312 100.099 init_scf_loop 1 4.0 0.000 0.000 93.217 93.217 mp2_ri_gpw_compute_in 1 5.0 0.064 0.072 75.774 76.813 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.919 55.954 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.513 47.697 integrate_v_rspace 95 8.0 0.399 0.573 28.457 33.486 pw_transfer 2240 10.6 0.145 0.163 30.163 30.594 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 29.137 29.608 ao_to_mo_and_store_B_mult_1 91 7.0 10.718 29.214 10.718 29.214 grid_integrate_task_list 95 9.0 23.728 28.944 23.728 28.944 mp2_ri_gpw_compute_en 1 5.0 0.055 0.062 26.722 28.383 fft_wrap_pw1pw2_100 730 12.4 1.308 1.446 26.876 27.334 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.882 25.032 25.042 get_2c_integrals 1 6.0 0.000 0.000 20.771 20.794 compute_2c_integrals 1 7.0 0.003 0.003 19.744 19.754 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 19.394 19.602 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.878 19.392 19.601 fft3d_s 1823 13.4 18.503 18.815 18.516 18.827 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.228 14.228 calculate_wavefunction 91 8.0 2.020 2.068 9.796 10.115 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.577 8.812 9.517 potential_pw2rs 186 10.0 0.033 0.034 8.720 9.441 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.677 9.055 calc_potential_gpw 182 9.5 0.002 0.002 8.663 8.958 local_gemm 172 8.0 8.255 8.948 8.255 8.948 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.526 7.993 8.430 collocate_single_gaussian 91 10.0 0.017 0.019 7.978 8.349 mp_sync 38 10.4 3.443 7.086 3.443 7.086 mp_sendrecv_dm3 2068 8.0 6.014 6.456 6.014 6.456 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.431 6.350 6.431 pw_gather_s 912 13.2 4.994 5.493 4.994 5.493 pw_scatter_s 910 13.7 3.968 4.233 3.968 4.233 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.648106, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 449.183744E+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 204504. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 52.924 52.925 qs_mol_dyn_low 1 2.0 0.003 0.004 52.686 52.693 qs_forces 11 3.9 0.002 0.002 52.238 52.238 qs_energies 11 4.9 0.002 0.003 50.762 50.776 scf_env_do_scf 11 5.9 0.001 0.001 44.212 44.213 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.076 42.076 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.974 34.459 qs_scf_new_mos 108 7.5 0.000 0.001 31.835 32.148 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.834 32.148 ot_scf_mini 108 9.5 0.002 0.002 30.173 30.373 multiply_cannon 2286 13.5 0.183 0.193 26.427 28.108 multiply_cannon_loop 2286 14.5 1.479 1.555 25.718 27.433 velocity_verlet 10 3.0 0.001 0.001 25.872 25.877 ot_mini 108 10.5 0.001 0.001 19.271 19.522 qs_ot_get_derivative 108 11.5 0.001 0.001 16.307 16.501 mp_waitall_1 267858 16.1 8.683 15.489 8.683 15.489 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 6.065 14.087 multiply_cannon_multrec 54864 15.5 4.240 6.553 7.777 11.269 rebuild_ks_matrix 119 8.3 0.000 0.000 8.008 8.157 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.010 8.008 8.157 multiply_cannon_sync_h2d 54864 15.5 5.929 7.475 5.929 7.475 mp_sum_l 7207 12.9 5.515 7.282 5.515 7.282 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.066 7.199 qs_ot_get_p 119 10.4 0.001 0.001 6.289 6.573 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.397 5.858 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.493 5.609 init_scf_run 11 5.9 0.000 0.001 5.331 5.331 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.331 5.331 dbcsr_mm_accdrv_process 76910 16.1 1.153 1.810 3.460 4.811 sum_up_and_integrate 119 10.3 0.008 0.009 4.653 4.660 integrate_v_rspace 119 11.3 0.002 0.002 4.645 4.652 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.239 4.341 calculate_rho_elec 119 8.7 0.011 0.016 4.238 4.340 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.491 3.567 calculate_dm_sparse 119 9.5 0.000 0.000 3.077 3.204 calculate_first_density_matrix 1 7.0 0.000 0.000 3.147 3.160 rs_pw_transfer 974 11.9 0.012 0.013 3.075 3.160 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.891 3.083 apply_single 119 13.6 0.000 0.000 2.891 3.083 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.694 2.957 jit_kernel_multiply 13 15.8 2.245 2.919 2.245 2.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.757 2.759 acc_transpose_blocks 54864 15.5 0.220 0.241 2.091 2.688 ot_diis_step 108 11.5 0.006 0.006 2.671 2.671 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.643 2.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.496 2.537 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.411 2.469 density_rs2pw 119 9.7 0.004 0.004 2.340 2.453 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.401 2.402 cp_fm_redistribute_end 50 14.0 2.179 2.377 2.184 2.378 cp_fm_diag_elpa_base 50 14.0 0.193 2.329 0.194 2.341 grid_integrate_task_list 119 12.3 2.032 2.127 2.032 2.127 init_scf_loop 11 6.9 0.000 0.000 2.118 2.119 wfi_extrapolate 11 7.9 0.001 0.001 2.110 2.110 potential_pw2rs 119 12.3 0.004 0.004 2.010 2.024 mp_sum_d 4129 12.0 1.301 1.934 1.301 1.934 make_m2s 4572 13.5 0.053 0.055 1.720 1.768 pw_transfer 1439 11.6 0.052 0.055 1.672 1.743 make_images 4572 14.5 0.133 0.138 1.638 1.685 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.598 1.670 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.582 1.603 acc_transpose_blocks_kernels 54864 16.5 0.249 0.381 1.128 1.550 mp_alltoall_d11v 2130 13.8 1.338 1.503 1.338 1.503 fft3d_ps 1201 14.6 0.357 0.460 1.373 1.437 grid_collocate_task_list 119 9.7 1.290 1.351 1.290 1.351 mp_waitany 12084 13.8 1.187 1.325 1.187 1.325 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.233 1.302 jit_kernel_transpose 5 15.5 0.879 1.177 0.879 1.177 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.733 1.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.925000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 486.178816E+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 1244183. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.141 38.498 38.500 qs_mol_dyn_low 1 2.0 0.006 0.026 38.150 38.157 qs_forces 11 3.9 0.003 0.004 38.009 38.010 qs_energies 11 4.9 0.004 0.025 36.220 36.242 scf_env_do_scf 11 5.9 0.001 0.013 30.806 30.806 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.344 28.346 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.111 21.467 qs_scf_new_mos 108 7.5 0.001 0.001 19.499 19.741 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.499 19.741 ot_scf_mini 108 9.5 0.002 0.003 18.618 18.790 velocity_verlet 10 3.0 0.004 0.028 18.114 18.119 multiply_cannon 2286 13.5 0.207 0.215 16.214 17.889 multiply_cannon_loop 2286 14.5 0.903 0.979 15.125 16.510 ot_mini 108 10.5 0.001 0.001 11.648 11.885 mp_waitall_1 217478 16.2 5.936 11.055 5.936 11.055 multiply_cannon_metrocomm3 27432 15.5 0.069 0.071 4.145 9.429 qs_ot_get_derivative 108 11.5 0.001 0.001 9.191 9.365 multiply_cannon_multrec 27432 15.5 1.956 4.336 5.836 8.713 rebuild_ks_matrix 119 8.3 0.000 0.000 7.112 7.254 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.019 7.112 7.253 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.230 6.357 dbcsr_mm_accdrv_process 47894 16.0 2.966 5.324 3.812 5.651 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.570 4.385 qs_ot_get_p 119 10.4 0.001 0.001 4.091 4.323 sum_up_and_integrate 119 10.3 0.016 0.018 4.151 4.162 integrate_v_rspace 119 11.3 0.002 0.003 4.135 4.148 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.047 4.113 apply_single 119 13.6 0.000 0.000 3.047 4.113 init_scf_run 11 5.9 0.000 0.001 4.073 4.074 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.073 4.073 mp_sum_l 7207 12.9 2.099 4.062 2.099 4.062 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.560 3.594 calculate_rho_elec 119 8.7 0.021 0.024 3.559 3.593 multiply_cannon_sync_h2d 27432 15.5 2.192 2.933 2.192 2.933 qs_ot_p2m_diag 50 11.0 0.009 0.012 2.680 2.699 rs_pw_transfer 974 11.9 0.010 0.011 2.578 2.678 make_m2s 4572 13.5 0.052 0.054 2.432 2.655 calculate_first_density_matrix 1 7.0 0.000 0.000 2.571 2.573 make_images 4572 14.5 0.200 0.238 2.344 2.564 init_scf_loop 11 6.9 0.000 0.000 2.427 2.430 ot_diis_step 108 11.5 0.011 0.013 2.406 2.407 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.254 2.255 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.082 2.173 calculate_dm_sparse 119 9.5 0.000 0.001 2.035 2.111 density_rs2pw 119 9.7 0.004 0.004 1.932 2.021 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.945 1.946 grid_integrate_task_list 119 12.3 1.829 1.941 1.829 1.941 potential_pw2rs 119 12.3 0.006 0.006 1.919 1.935 cp_fm_redistribute_end 50 14.0 1.606 1.923 1.609 1.924 cp_fm_diag_elpa_base 50 14.0 0.306 1.867 0.313 1.897 jit_kernel_multiply 10 16.1 0.794 1.890 0.794 1.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.856 1.859 pw_transfer 1439 11.6 0.064 0.066 1.729 1.765 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.698 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.640 1.678 make_images_data 4572 15.5 0.045 0.051 1.140 1.565 prepare_preconditioner 11 7.9 0.000 0.000 1.528 1.555 make_preconditioner 11 8.9 0.000 0.001 1.528 1.555 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.422 1.480 acc_transpose_blocks 27432 15.5 0.106 0.110 1.181 1.463 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.992 1.456 wfi_extrapolate 11 7.9 0.001 0.002 1.452 1.452 mp_alltoall_d11v 2130 13.8 1.164 1.411 1.164 1.411 fft3d_ps 1201 14.6 0.496 0.550 1.346 1.381 fft_wrap_pw1pw2_140 487 13.2 0.077 0.086 1.291 1.328 grid_collocate_task_list 119 9.7 1.242 1.322 1.242 1.322 mp_allgather_i34 2286 14.5 0.535 1.278 0.535 1.278 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.163 1.203 mp_sum_d 4129 12.0 0.646 1.104 0.646 1.104 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.970 0.971 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.938 0.952 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.652 0.856 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.704 0.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.500000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.272727, yerr=1.285649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 519.573504E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 31.438 31.438 qs_mol_dyn_low 1 2.0 0.003 0.004 31.180 31.188 qs_forces 11 3.9 0.002 0.002 31.125 31.125 qs_energies 11 4.9 0.001 0.001 29.559 29.560 scf_env_do_scf 11 5.9 0.000 0.001 24.765 24.765 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.240 22.241 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.401 16.498 velocity_verlet 10 3.0 0.001 0.001 14.861 14.862 qs_scf_new_mos 108 7.5 0.001 0.001 14.469 14.481 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.468 14.480 multiply_cannon 2286 13.5 0.194 0.200 13.274 14.032 ot_scf_mini 108 9.5 0.002 0.003 13.758 13.767 multiply_cannon_loop 2286 14.5 0.639 0.663 12.488 13.260 ot_mini 108 10.5 0.001 0.001 8.594 8.607 multiply_cannon_multrec 18288 15.5 1.963 2.861 7.014 7.233 qs_ot_get_derivative 108 11.5 0.001 0.001 7.108 7.119 rebuild_ks_matrix 119 8.3 0.000 0.000 6.236 6.251 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.235 6.250 dbcsr_mm_accdrv_process 38222 16.0 4.206 5.591 4.966 5.937 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.508 5.522 sum_up_and_integrate 119 10.3 0.019 0.019 3.880 3.887 integrate_v_rspace 119 11.3 0.002 0.003 3.862 3.870 mp_waitall_1 169478 16.3 2.689 3.840 2.689 3.840 init_scf_run 11 5.9 0.000 0.001 3.631 3.631 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.631 3.631 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.694 3.295 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.116 3.125 calculate_rho_elec 119 8.7 0.030 0.031 3.115 3.124 qs_ot_get_p 119 10.4 0.001 0.001 2.998 3.018 init_scf_loop 11 6.9 0.000 0.000 2.508 2.509 calculate_first_density_matrix 1 7.0 0.000 0.000 2.461 2.461 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.404 2.451 rs_pw_transfer 974 11.9 0.009 0.010 2.261 2.397 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.013 2.333 apply_single 119 13.6 0.000 0.000 2.012 2.333 jit_kernel_multiply 10 16.2 0.708 2.076 0.708 2.076 qs_ot_p2m_diag 50 11.0 0.012 0.012 1.980 1.984 grid_integrate_task_list 119 12.3 1.818 1.926 1.818 1.926 calculate_dm_sparse 119 9.5 0.000 0.000 1.900 1.913 make_m2s 4572 13.5 0.044 0.045 1.711 1.869 density_rs2pw 119 9.7 0.004 0.004 1.729 1.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.809 1.811 make_images 4572 14.5 0.190 0.203 1.626 1.782 prepare_preconditioner 11 7.9 0.000 0.000 1.731 1.734 make_preconditioner 11 8.9 0.000 0.000 1.731 1.734 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.721 1.722 potential_pw2rs 119 12.3 0.007 0.008 1.689 1.693 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.583 1.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.661 pw_transfer 1439 11.6 0.063 0.066 1.572 1.579 multiply_cannon_sync_h2d 18288 15.5 1.431 1.575 1.431 1.575 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.528 1.535 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.504 1.504 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.482 1.491 cp_fm_redistribute_end 50 14.0 1.117 1.485 1.118 1.485 cp_fm_diag_elpa_base 50 14.0 0.352 1.428 0.365 1.467 ot_diis_step 108 11.5 0.011 0.011 1.465 1.465 mp_sum_l 7207 12.9 1.032 1.419 1.032 1.419 grid_collocate_task_list 119 9.7 1.204 1.325 1.204 1.325 acc_transpose_blocks 18288 15.5 0.073 0.075 1.283 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.281 1.286 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.173 1.183 fft3d_ps 1201 14.6 0.499 0.514 1.167 1.176 wfi_extrapolate 11 7.9 0.001 0.001 1.127 1.127 qs_energies_init_hamiltonians 11 5.9 0.002 0.010 0.942 0.943 make_images_data 4572 15.5 0.044 0.049 0.762 0.938 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.870 0.892 acc_transpose_blocks_kernels 18288 16.5 0.208 0.217 0.845 0.865 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.655 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.796 0.798 mp_alltoall_d11v 2130 13.8 0.619 0.728 0.619 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.723 cp_fm_cholesky_invert 11 10.9 0.703 0.707 0.703 0.707 rs_pw_transfer_RS2PW_140 130 11.5 0.176 0.180 0.561 0.695 jit_kernel_transpose 5 15.6 0.637 0.648 0.637 0.648 mp_waitany 9880 13.7 0.519 0.638 0.519 0.638 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.438000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.090909, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 554.844160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.073 35.246 35.247 qs_mol_dyn_low 1 2.0 0.003 0.006 35.005 35.012 qs_forces 11 3.9 0.002 0.002 34.909 34.910 qs_energies 11 4.9 0.002 0.007 33.150 33.156 scf_env_do_scf 11 5.9 0.001 0.001 28.098 28.099 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.544 24.545 dbcsr_multiply_generic 2286 12.5 0.100 0.107 18.219 18.396 velocity_verlet 10 3.0 0.001 0.002 18.026 18.030 qs_scf_new_mos 108 7.5 0.001 0.001 16.259 16.310 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.259 16.309 ot_scf_mini 108 9.5 0.002 0.003 15.325 15.372 multiply_cannon 2286 13.5 0.236 0.274 14.538 14.982 multiply_cannon_loop 2286 14.5 0.937 0.968 13.571 13.962 ot_mini 108 10.5 0.001 0.001 9.492 9.552 multiply_cannon_multrec 27432 15.5 2.332 3.055 8.586 8.934 qs_ot_get_derivative 108 11.5 0.001 0.001 7.672 7.720 dbcsr_mm_accdrv_process 47916 15.9 5.252 6.901 6.158 7.361 rebuild_ks_matrix 119 8.3 0.000 0.000 6.587 6.635 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.586 6.635 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.828 5.873 sum_up_and_integrate 119 10.3 0.021 0.024 3.840 3.849 integrate_v_rspace 119 11.3 0.002 0.002 3.818 3.828 init_scf_run 11 5.9 0.000 0.001 3.639 3.640 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.639 3.640 init_scf_loop 11 6.9 0.000 0.000 3.528 3.528 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.363 3.393 calculate_rho_elec 119 8.7 0.040 0.046 3.363 3.393 qs_ot_get_p 119 10.4 0.001 0.001 3.205 3.278 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.805 3.197 mp_waitall_1 145218 16.4 2.344 2.888 2.344 2.888 prepare_preconditioner 11 7.9 0.000 0.000 2.687 2.695 make_preconditioner 11 8.9 0.000 0.001 2.687 2.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.283 2.609 make_m2s 4572 13.5 0.054 0.056 2.277 2.384 rs_pw_transfer 974 11.9 0.009 0.010 2.155 2.279 make_images 4572 14.5 0.274 0.333 2.170 2.275 calculate_first_density_matrix 1 7.0 0.000 0.000 2.242 2.244 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.047 2.171 apply_single 119 13.6 0.000 0.000 2.047 2.170 calculate_dm_sparse 119 9.5 0.000 0.000 2.026 2.076 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.938 1.963 density_rs2pw 119 9.7 0.003 0.004 1.847 1.953 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.931 1.940 grid_integrate_task_list 119 12.3 1.831 1.896 1.831 1.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.840 1.841 pw_transfer 1439 11.6 0.063 0.066 1.806 1.837 jit_kernel_multiply 10 16.0 0.845 1.817 0.845 1.817 ot_diis_step 108 11.5 0.012 0.016 1.780 1.781 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.718 1.751 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.635 1.649 potential_pw2rs 119 12.3 0.009 0.009 1.628 1.634 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.614 1.615 acc_transpose_blocks 27432 15.5 0.110 0.113 1.439 1.465 fft_wrap_pw1pw2_140 487 13.2 0.086 0.095 1.394 1.428 fft3d_ps 1201 14.6 0.534 0.586 1.396 1.424 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.876 1.412 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.368 1.369 wfi_extrapolate 11 7.9 0.001 0.003 1.347 1.348 cp_fm_redistribute_end 50 14.0 0.903 1.345 0.904 1.346 cp_fm_diag_elpa_base 50 14.0 0.420 1.281 0.439 1.321 grid_collocate_task_list 119 9.7 1.220 1.314 1.220 1.314 mp_sum_l 7207 12.9 0.984 1.282 0.984 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.220 qs_energies_init_hamiltonians 11 5.9 0.004 0.024 1.138 1.139 cp_fm_upper_to_full 72 13.5 0.817 1.131 0.817 1.131 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.110 1.130 dbcsr_complete_redistribute 329 12.2 0.125 0.153 0.828 1.103 multiply_cannon_sync_h2d 27432 15.5 1.000 1.046 1.000 1.046 make_images_data 4572 15.5 0.045 0.049 0.907 1.022 hybrid_alltoall_any 4725 16.4 0.063 0.153 0.791 0.970 mp_alltoall_d11v 2130 13.8 0.791 0.923 0.791 0.923 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.804 0.883 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.597 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.848 0.853 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.226 0.837 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.817 0.830 cp_fm_cholesky_invert 11 10.9 0.821 0.824 0.821 0.824 mp_alltoall_z22v 1201 16.6 0.744 0.773 0.744 0.773 mp_alltoall_i22 627 13.8 0.441 0.743 0.441 0.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.247000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.181818, yerr=3.785575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 597.532672E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.077 27.943 27.943 qs_mol_dyn_low 1 2.0 0.003 0.004 27.508 27.515 qs_forces 11 3.9 0.002 0.002 27.008 27.008 qs_energies 11 4.9 0.001 0.001 25.324 25.326 scf_env_do_scf 11 5.9 0.001 0.002 20.510 20.510 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.081 18.082 velocity_verlet 10 3.0 0.001 0.001 14.286 14.316 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.114 12.188 qs_scf_new_mos 108 7.5 0.001 0.001 10.732 10.757 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.731 10.756 ot_scf_mini 108 9.5 0.002 0.002 10.062 10.091 multiply_cannon 2286 13.5 0.231 0.239 9.678 10.060 multiply_cannon_loop 2286 14.5 0.331 0.341 8.778 8.980 multiply_cannon_multrec 9144 15.5 1.670 1.921 5.953 6.162 ot_mini 108 10.5 0.001 0.001 5.678 5.710 rebuild_ks_matrix 119 8.3 0.000 0.000 5.660 5.684 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.660 5.683 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.041 5.062 qs_ot_get_derivative 108 11.5 0.001 0.001 4.399 4.426 dbcsr_mm_accdrv_process 12550 15.8 3.196 3.841 4.183 4.262 sum_up_and_integrate 119 10.3 0.023 0.024 3.443 3.446 integrate_v_rspace 119 11.3 0.002 0.003 3.419 3.423 init_scf_run 11 5.9 0.000 0.001 3.290 3.290 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.290 3.290 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.108 3.113 calculate_rho_elec 119 8.7 0.059 0.061 3.107 3.113 qs_ot_get_p 119 10.4 0.001 0.001 2.651 2.689 init_scf_loop 11 6.9 0.000 0.000 2.409 2.410 calculate_first_density_matrix 1 7.0 0.000 0.000 2.205 2.206 grid_integrate_task_list 119 12.3 1.851 1.921 1.851 1.921 mp_waitall_1 121218 16.5 1.422 1.897 1.422 1.897 jit_kernel_multiply 10 15.9 0.949 1.892 0.949 1.892 make_m2s 4572 13.5 0.034 0.035 1.672 1.832 calculate_dm_sparse 119 9.5 0.000 0.000 1.765 1.781 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.769 1.773 make_images 4572 14.5 0.267 0.303 1.583 1.742 prepare_preconditioner 11 7.9 0.000 0.000 1.700 1.706 make_preconditioner 11 8.9 0.000 0.000 1.700 1.706 rs_pw_transfer 974 11.9 0.008 0.009 1.607 1.697 density_rs2pw 119 9.7 0.003 0.004 1.530 1.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.594 1.619 pw_transfer 1439 11.6 0.063 0.065 1.589 1.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.584 1.586 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.579 1.580 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.537 1.553 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.500 1.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.430 1.436 grid_collocate_task_list 119 9.7 1.270 1.363 1.270 1.363 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.314 1.314 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.303 1.303 cp_fm_redistribute_end 50 14.0 0.648 1.285 0.649 1.285 cp_fm_diag_elpa_base 50 14.0 0.593 1.215 0.635 1.271 ot_diis_step 108 11.5 0.013 0.013 1.268 1.268 potential_pw2rs 119 12.3 0.010 0.011 1.264 1.266 fft_wrap_pw1pw2_140 487 13.2 0.084 0.088 1.225 1.233 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.188 1.193 fft3d_ps 1201 14.6 0.534 0.546 1.163 1.172 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.145 1.167 apply_single 119 13.6 0.000 0.000 1.145 1.167 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.103 1.116 wfi_extrapolate 11 7.9 0.001 0.001 1.042 1.042 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.755 0.989 acc_transpose_blocks 9144 15.5 0.038 0.039 0.952 0.958 make_images_data 4572 15.5 0.038 0.042 0.765 0.946 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.929 cp_fm_cholesky_invert 11 10.9 0.841 0.844 0.841 0.844 mp_alltoall_d11v 2130 13.8 0.751 0.840 0.751 0.840 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.757 0.800 multiply_cannon_sync_h2d 9144 15.5 0.706 0.788 0.706 0.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.739 0.741 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.712 0.714 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.701 0.710 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.328 0.692 qs_create_task_list 11 7.9 0.032 0.087 0.630 0.653 generate_qs_task_list 11 8.9 0.191 0.213 0.597 0.637 mp_allgather_i34 2286 14.5 0.235 0.617 0.235 0.617 jit_kernel_transpose 5 15.6 0.595 0.598 0.595 0.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.943000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.000000, yerr=4.369314 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 746.000384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 39.659 39.659 qs_mol_dyn_low 1 2.0 0.003 0.004 39.463 39.470 qs_forces 11 3.9 0.002 0.002 39.407 39.407 qs_energies 11 4.9 0.001 0.001 37.485 37.487 scf_env_do_scf 11 5.9 0.001 0.001 32.016 32.016 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.239 24.239 velocity_verlet 10 3.0 0.001 0.001 22.439 22.444 dbcsr_multiply_generic 2286 12.5 0.102 0.104 17.293 17.543 qs_scf_new_mos 108 7.5 0.001 0.001 15.595 15.704 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.594 15.703 ot_scf_mini 108 9.5 0.002 0.002 14.532 14.641 multiply_cannon 2286 13.5 0.298 0.303 13.535 14.463 multiply_cannon_loop 2286 14.5 0.344 0.351 12.284 13.225 ot_mini 108 10.5 0.001 0.001 8.746 8.870 multiply_cannon_multrec 9144 15.5 3.370 4.690 8.529 8.648 init_scf_loop 11 6.9 0.000 0.000 7.750 7.750 rebuild_ks_matrix 119 8.3 0.000 0.000 6.710 6.869 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.710 6.868 prepare_preconditioner 11 7.9 0.000 0.000 6.844 6.857 make_preconditioner 11 8.9 0.000 0.000 6.844 6.857 qs_ot_get_derivative 108 11.5 0.001 0.001 6.704 6.812 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.379 6.736 dbcsr_mm_accdrv_process 12550 15.8 4.219 5.623 5.035 6.340 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.069 6.212 cp_fm_upper_to_full 72 14.2 3.182 4.594 3.182 4.594 sum_up_and_integrate 119 10.3 0.038 0.038 3.557 3.564 integrate_v_rspace 119 11.3 0.003 0.003 3.519 3.527 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.519 3.525 calculate_rho_elec 119 8.7 0.118 0.121 3.519 3.525 init_scf_run 11 5.9 0.000 0.001 3.452 3.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.452 3.452 mp_waitall_1 97218 16.6 2.457 3.450 2.457 3.450 qs_ot_get_p 119 10.4 0.001 0.001 3.073 3.222 dbcsr_complete_redistribute 329 12.2 0.296 0.301 1.990 2.825 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.429 2.806 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.501 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.154 2.443 apply_single 119 13.6 0.000 0.000 2.154 2.443 make_m2s 4572 13.5 0.038 0.038 2.204 2.386 mp_alltoall_i22 627 13.8 1.434 2.288 1.434 2.288 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.460 2.272 make_images 4572 14.5 0.352 0.382 2.085 2.267 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.299 2.198 calculate_dm_sparse 119 9.5 0.000 0.000 2.166 2.182 calculate_first_density_matrix 1 7.0 0.000 0.000 2.110 2.111 ot_diis_step 108 11.5 0.014 0.014 2.017 2.017 grid_integrate_task_list 119 12.3 1.998 2.015 1.998 2.015 pw_transfer 1439 11.6 0.067 0.068 1.903 1.906 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.808 1.811 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.788 1.789 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.758 1.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.747 1.748 mp_sum_l 7207 12.9 1.014 1.728 1.014 1.728 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.667 1.721 density_rs2pw 119 9.7 0.003 0.003 1.705 1.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.524 1.570 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.523 1.523 fft_wrap_pw1pw2_140 487 13.2 0.087 0.088 1.518 1.523 grid_collocate_task_list 119 9.7 1.439 1.449 1.439 1.449 fft3d_ps 1201 14.6 0.564 0.575 1.441 1.443 cp_fm_cholesky_invert 11 10.9 1.433 1.436 1.433 1.436 rs_pw_transfer 974 11.9 0.009 0.009 1.371 1.406 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.044 1.310 wfi_extrapolate 11 7.9 0.001 0.001 1.288 1.288 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.273 1.273 make_images_data 4572 15.5 0.042 0.045 1.051 1.273 cp_fm_diag_elpa_base 50 14.0 1.124 1.179 1.271 1.271 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.255 potential_pw2rs 119 12.3 0.014 0.014 1.164 1.165 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.116 1.134 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.084 1.099 multiply_cannon_sync_h2d 9144 15.5 1.038 1.041 1.038 1.041 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.990 1.029 jit_kernel_multiply 6 15.5 0.789 1.029 0.789 1.029 mp_alltoall_d11v 2130 13.8 0.937 0.957 0.937 0.957 qs_create_task_list 11 7.9 0.001 0.001 0.941 0.953 generate_qs_task_list 11 8.9 0.372 0.392 0.940 0.953 acc_transpose_blocks 9144 15.5 0.038 0.038 0.918 0.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.851 0.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=39.659000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.363636, yerr=13.158286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 498.999296E+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 661251. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.041 83.044 83.045 qs_mol_dyn_low 1 2.0 0.003 0.006 82.751 82.761 qs_forces 11 3.9 0.002 0.003 82.684 82.685 qs_energies 11 4.9 0.002 0.007 79.828 79.848 scf_env_do_scf 11 5.9 0.000 0.001 70.807 70.809 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.144 65.145 dbcsr_multiply_generic 2055 12.4 0.105 0.108 51.851 52.133 qs_scf_new_mos 99 7.5 0.000 0.001 47.721 47.867 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.721 47.867 ot_scf_mini 99 9.5 0.002 0.002 45.255 45.364 multiply_cannon 2055 13.4 0.178 0.184 42.216 43.256 multiply_cannon_loop 2055 14.4 1.550 1.594 41.225 42.236 velocity_verlet 10 3.0 0.001 0.001 42.190 42.191 ot_mini 99 10.5 0.001 0.002 27.344 27.440 qs_ot_get_derivative 99 11.5 0.001 0.002 20.593 20.688 multiply_cannon_multrec 49320 15.4 12.109 12.887 17.104 17.795 rebuild_ks_matrix 110 8.3 0.000 0.001 14.780 14.927 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.780 14.926 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.003 13.134 mp_waitall_1 241148 16.1 11.535 12.594 11.535 12.594 multiply_cannon_sync_h2d 49320 15.4 9.978 10.535 9.978 10.535 qs_ot_get_p 110 10.4 0.001 0.002 9.411 9.526 multiply_cannon_metrocomm3 49320 15.4 0.076 0.079 6.781 8.140 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.167 7.660 apply_single 110 13.6 0.000 0.000 7.166 7.660 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.969 7.510 sum_up_and_integrate 110 10.3 0.023 0.030 7.315 7.329 integrate_v_rspace 110 11.3 0.003 0.004 7.291 7.312 init_scf_run 11 5.9 0.000 0.001 6.831 6.831 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.830 6.831 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.670 6.734 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.455 6.596 calculate_rho_elec 110 8.6 0.020 0.024 6.455 6.596 ot_diis_step 99 11.5 0.007 0.015 6.512 6.513 mp_sum_l 6514 12.8 5.218 6.138 5.218 6.138 qs_ot_p2m_diag 48 11.0 0.012 0.021 6.040 6.078 init_scf_loop 11 6.9 0.000 0.001 5.634 5.635 dbcsr_mm_accdrv_process 87628 16.1 2.058 2.189 4.875 5.212 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.056 5.056 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.573 4.574 cp_fm_redistribute_end 48 14.0 3.981 4.545 3.986 4.547 cp_fm_diag_elpa_base 48 14.0 0.555 4.460 0.558 4.482 rs_pw_transfer 902 11.9 0.012 0.014 4.277 4.437 make_m2s 4110 13.4 0.060 0.065 3.966 4.094 wfi_extrapolate 11 7.9 0.001 0.001 4.048 4.048 make_images 4110 14.4 0.178 0.193 3.871 4.002 calculate_dm_sparse 110 9.5 0.000 0.001 3.833 3.955 density_rs2pw 110 9.6 0.004 0.009 3.505 3.659 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.542 3.610 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.591 3.596 prepare_preconditioner 11 7.9 0.000 0.000 3.435 3.468 make_preconditioner 11 8.9 0.000 0.001 3.435 3.468 grid_integrate_task_list 110 12.3 3.269 3.424 3.269 3.424 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.261 3.310 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.256 3.296 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.176 3.221 pw_transfer 1331 11.6 0.054 0.063 2.977 3.039 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.885 2.951 potential_pw2rs 110 12.3 0.006 0.006 2.889 2.907 calculate_first_density_matrix 1 7.0 0.000 0.001 2.666 2.676 jit_kernel_multiply 13 15.9 2.534 2.544 2.534 2.544 mp_alltoall_d11v 2046 13.8 2.208 2.530 2.208 2.530 fft_wrap_pw1pw2_140 451 13.1 0.170 0.192 2.421 2.486 fft3d_ps 1111 14.6 0.753 0.839 2.422 2.478 acc_transpose_blocks 49320 15.4 0.209 0.225 2.202 2.301 grid_collocate_task_list 110 9.6 2.088 2.235 2.088 2.235 mp_waitany 14300 13.8 1.839 2.093 1.839 2.093 mp_sum_d 3883 11.9 1.494 2.022 1.494 2.022 make_images_data 4110 15.4 0.043 0.046 1.836 1.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.939 cp_fm_cholesky_invert 11 10.9 1.858 1.862 1.858 1.862 hybrid_alltoall_any 4261 16.3 0.083 0.483 1.590 1.845 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.642 1.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.045000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.818182, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.740288E+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 2151582. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.031 68.833 68.834 qs_mol_dyn_low 1 2.0 0.003 0.006 68.493 68.503 qs_forces 11 3.9 0.002 0.005 68.427 68.428 qs_energies 11 4.9 0.002 0.004 65.118 65.122 scf_env_do_scf 11 5.9 0.001 0.001 55.830 55.833 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.266 48.267 dbcsr_multiply_generic 2055 12.4 0.118 0.126 38.355 38.500 velocity_verlet 10 3.0 0.001 0.002 35.685 35.687 multiply_cannon 2055 13.4 0.221 0.243 31.811 32.983 qs_scf_new_mos 99 7.5 0.001 0.001 32.506 32.666 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.505 32.665 multiply_cannon_loop 2055 14.4 0.933 0.956 30.505 31.319 ot_scf_mini 99 9.5 0.003 0.006 30.855 31.007 ot_mini 99 10.5 0.001 0.001 18.187 18.342 multiply_cannon_multrec 24660 15.4 7.653 9.537 14.111 15.777 rebuild_ks_matrix 110 8.3 0.000 0.000 13.607 13.743 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.607 13.742 qs_ot_get_derivative 99 11.5 0.001 0.002 12.400 12.551 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.006 12.137 mp_waitall_1 186928 16.3 7.714 10.427 7.714 10.427 multiply_cannon_sync_h2d 24660 15.4 7.110 8.526 7.110 8.526 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.141 7.820 init_scf_loop 11 6.9 0.000 0.000 7.528 7.531 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.508 7.160 apply_single 110 13.6 0.000 0.001 6.508 7.160 init_scf_run 11 5.9 0.000 0.001 6.821 6.822 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.821 6.822 dbcsr_mm_accdrv_process 52282 16.1 4.831 5.546 6.298 6.592 sum_up_and_integrate 110 10.3 0.029 0.034 6.486 6.498 integrate_v_rspace 110 11.3 0.002 0.004 6.456 6.467 qs_ot_get_p 110 10.4 0.001 0.001 6.032 6.221 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.732 5.745 calculate_rho_elec 110 8.6 0.039 0.047 5.732 5.744 ot_diis_step 99 11.5 0.010 0.012 5.740 5.740 prepare_preconditioner 11 7.9 0.000 0.000 5.470 5.488 make_preconditioner 11 8.9 0.000 0.001 5.470 5.488 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.658 5.376 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.049 5.205 make_m2s 4110 13.4 0.056 0.060 4.115 4.621 make_images 4110 14.4 0.399 0.445 4.007 4.509 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.120 4.143 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.805 3.806 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.686 3.687 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.481 3.522 wfi_extrapolate 11 7.9 0.001 0.003 3.486 3.486 grid_integrate_task_list 110 12.3 3.159 3.316 3.159 3.316 rs_pw_transfer 902 11.9 0.012 0.014 3.141 3.287 pw_transfer 1331 11.6 0.067 0.071 3.137 3.280 calculate_first_density_matrix 1 7.0 0.000 0.001 3.241 3.242 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.163 3.240 density_rs2pw 110 9.6 0.004 0.009 3.049 3.211 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.026 3.169 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.127 3.128 cp_fm_redistribute_end 48 14.0 2.331 3.099 2.333 3.100 calculate_dm_sparse 110 9.5 0.001 0.001 3.071 3.099 cp_fm_diag_elpa_base 48 14.0 0.733 2.978 0.764 3.061 make_images_data 4110 15.4 0.047 0.051 2.288 2.807 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.015 2.765 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 2.534 2.677 cp_fm_cholesky_invert 11 10.9 2.526 2.532 2.526 2.532 mp_sum_l 6514 12.8 1.805 2.520 1.805 2.520 fft3d_ps 1111 14.6 1.073 1.271 2.381 2.520 potential_pw2rs 110 12.3 0.008 0.009 2.415 2.426 grid_collocate_task_list 110 9.6 2.122 2.245 2.122 2.245 acc_transpose_blocks 24660 15.4 0.105 0.108 2.030 2.052 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.945 1.964 jit_kernel_multiply 9 16.5 1.117 1.948 1.117 1.948 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.901 1.901 mp_alltoall_d11v 2046 13.8 1.699 1.877 1.699 1.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.764 1.785 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.754 1.712 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.651 1.663 mp_allgather_i34 2055 14.4 0.638 1.625 0.638 1.625 mp_irecv_dv 57340 16.2 0.631 1.596 0.631 1.596 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.387 1.499 acc_transpose_blocks_kernels 24660 16.4 0.307 0.316 1.419 1.435 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.834000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.363636, yerr=7.022985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.856832E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.050 60.237 60.238 qs_mol_dyn_low 1 2.0 0.004 0.016 59.847 59.857 qs_forces 11 3.9 0.002 0.002 59.768 59.769 qs_energies 11 4.9 0.002 0.006 56.568 56.572 scf_env_do_scf 11 5.9 0.001 0.003 48.090 48.090 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.625 39.626 velocity_verlet 10 3.0 0.001 0.002 32.171 32.173 dbcsr_multiply_generic 2055 12.4 0.108 0.114 29.056 29.308 qs_scf_new_mos 99 7.5 0.001 0.001 25.106 25.202 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.105 25.201 multiply_cannon 2055 13.4 0.212 0.225 22.712 24.141 ot_scf_mini 99 9.5 0.002 0.003 23.887 23.998 multiply_cannon_loop 2055 14.4 0.620 0.631 21.502 22.711 ot_mini 99 10.5 0.001 0.001 13.679 13.790 rebuild_ks_matrix 110 8.3 0.000 0.000 12.228 12.375 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.022 12.228 12.375 multiply_cannon_multrec 16440 15.4 3.906 5.076 9.955 10.945 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.787 10.927 mp_waitall_1 146766 16.3 7.321 10.422 7.321 10.422 qs_ot_get_derivative 99 11.5 0.001 0.002 9.196 9.307 init_scf_loop 11 6.9 0.000 0.000 8.430 8.431 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.432 7.364 prepare_preconditioner 11 7.9 0.000 0.000 6.680 6.699 make_preconditioner 11 8.9 0.000 0.001 6.680 6.699 sum_up_and_integrate 110 10.3 0.034 0.035 6.444 6.459 integrate_v_rspace 110 11.3 0.002 0.003 6.410 6.425 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.003 6.342 dbcsr_mm_accdrv_process 34862 16.1 5.122 5.685 5.904 6.123 init_scf_run 11 5.9 0.000 0.001 6.036 6.037 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.036 6.036 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.505 5.515 calculate_rho_elec 110 8.6 0.058 0.058 5.505 5.514 qs_ot_get_p 110 10.4 0.001 0.001 5.380 5.514 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.039 5.463 apply_single 110 13.6 0.000 0.000 5.038 5.463 make_m2s 4110 13.4 0.050 0.052 4.193 4.594 make_images 4110 14.4 0.392 0.509 4.079 4.480 ot_diis_step 99 11.5 0.011 0.013 4.455 4.455 multiply_cannon_sync_h2d 16440 15.4 3.692 4.244 3.692 4.244 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.211 3.804 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.791 3.795 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.442 3.443 grid_integrate_task_list 110 12.3 3.199 3.370 3.199 3.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.167 3.169 pw_transfer 1331 11.6 0.066 0.073 3.083 3.092 rs_pw_transfer 902 11.9 0.010 0.011 2.877 3.063 calculate_first_density_matrix 1 7.0 0.000 0.000 3.009 3.010 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.971 2.983 make_images_data 4110 15.4 0.044 0.048 2.498 2.980 density_rs2pw 110 9.6 0.004 0.004 2.821 2.974 wfi_extrapolate 11 7.9 0.001 0.003 2.944 2.944 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.185 2.928 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.911 2.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.850 2.912 cp_fm_redistribute_end 48 14.0 1.809 2.884 1.811 2.884 cp_fm_diag_elpa_base 48 14.0 1.011 2.754 1.069 2.855 cp_fm_cholesky_invert 11 10.9 2.633 2.639 2.633 2.639 calculate_dm_sparse 110 9.5 0.001 0.001 2.581 2.608 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.547 2.558 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.886 2.438 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.309 2.375 grid_collocate_task_list 110 9.6 2.174 2.348 2.174 2.348 mp_irecv_dv 48980 15.7 0.816 2.318 0.816 2.318 potential_pw2rs 110 12.3 0.011 0.011 2.292 2.301 fft3d_ps 1111 14.6 1.062 1.071 2.282 2.293 mp_alltoall_d11v 2046 13.8 1.746 2.243 1.746 2.243 mp_sum_l 6514 12.8 1.543 2.135 1.543 2.135 qs_energies_init_hamiltonians 11 5.9 0.001 0.007 2.005 2.006 dbcsr_complete_redistribute 325 12.2 0.330 0.357 1.397 1.858 cp_fm_upper_to_full 70 13.6 1.382 1.845 1.382 1.845 mp_allgather_i34 2055 14.4 0.533 1.696 0.533 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.653 jit_kernel_multiply 8 16.6 0.396 1.609 0.396 1.609 cp_fm_cholesky_decompose 22 10.9 1.555 1.574 1.555 1.574 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.559 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.375 1.514 acc_transpose_blocks 16440 15.4 0.071 0.073 1.404 1.420 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.969 1.407 mp_waitany 17072 13.8 1.159 1.393 1.159 1.393 rs_gather_matrices 110 12.3 0.138 0.151 0.872 1.357 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.260 1.268 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.238000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.727273, yerr=10.028060 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 736.718848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.093 63.921 63.922 qs_mol_dyn_low 1 2.0 0.003 0.004 63.564 63.577 qs_forces 11 3.9 0.002 0.002 63.432 63.432 qs_energies 11 4.9 0.002 0.005 60.067 60.070 scf_env_do_scf 11 5.9 0.001 0.001 51.579 51.582 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.967 39.968 velocity_verlet 10 3.0 0.001 0.001 36.340 36.345 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.306 29.536 qs_scf_new_mos 99 7.5 0.001 0.001 25.753 25.858 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.752 25.858 ot_scf_mini 99 9.5 0.003 0.003 24.142 24.248 multiply_cannon 2055 13.4 0.244 0.265 22.457 23.767 multiply_cannon_loop 2055 14.4 0.883 0.904 21.141 21.726 multiply_cannon_multrec 24660 15.4 4.199 6.601 12.855 14.146 ot_mini 99 10.5 0.001 0.001 13.813 13.921 rebuild_ks_matrix 110 8.3 0.000 0.000 11.842 11.966 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.842 11.966 init_scf_loop 11 6.9 0.000 0.000 11.573 11.573 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.482 10.590 prepare_preconditioner 11 7.9 0.000 0.000 9.854 9.870 make_preconditioner 11 8.9 0.000 0.001 9.854 9.870 qs_ot_get_derivative 99 11.5 0.001 0.001 9.680 9.780 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.076 9.534 dbcsr_mm_accdrv_process 52304 16.0 7.138 8.440 8.508 9.429 sum_up_and_integrate 110 10.3 0.039 0.040 6.248 6.265 integrate_v_rspace 110 11.3 0.003 0.003 6.209 6.225 mp_waitall_1 126806 16.4 4.216 6.130 4.216 6.130 init_scf_run 11 5.9 0.000 0.001 5.631 5.631 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.631 5.631 make_m2s 4110 13.4 0.059 0.060 5.302 5.593 qs_ot_get_p 110 10.4 0.001 0.001 5.353 5.516 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.504 5.514 calculate_rho_elec 110 8.6 0.077 0.081 5.503 5.513 make_images 4110 14.4 0.578 0.699 5.162 5.448 cp_fm_upper_to_full 70 13.8 3.319 4.714 3.319 4.714 ot_diis_step 99 11.5 0.011 0.011 4.097 4.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.000 4.094 apply_single 110 13.6 0.000 0.000 3.999 4.094 dbcsr_complete_redistribute 325 12.2 0.430 0.475 2.639 3.799 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.639 3.653 grid_integrate_task_list 110 12.3 3.288 3.440 3.288 3.440 multiply_cannon_sync_h2d 24660 15.4 3.200 3.375 3.200 3.375 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.227 3.276 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.117 3.273 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.353 3.192 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.172 3.172 pw_transfer 1331 11.6 0.066 0.073 3.097 3.121 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.986 3.012 calculate_dm_sparse 110 9.5 0.001 0.001 2.977 3.003 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.947 3.000 make_images_data 4110 15.4 0.046 0.050 2.675 2.978 hybrid_alltoall_any 4261 16.3 0.120 0.462 2.274 2.939 wfi_extrapolate 11 7.9 0.001 0.001 2.912 2.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.910 2.912 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.769 2.901 mp_alltoall_i22 605 13.7 1.637 2.852 1.637 2.852 density_rs2pw 110 9.6 0.004 0.004 2.703 2.839 rs_pw_transfer 902 11.9 0.011 0.011 2.492 2.663 cp_fm_cholesky_invert 11 10.9 2.654 2.662 2.654 2.662 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.660 2.662 cp_fm_redistribute_end 48 14.0 1.325 2.634 1.326 2.635 calculate_first_density_matrix 1 7.0 0.000 0.000 2.631 2.633 cp_fm_diag_elpa_base 48 14.0 1.228 2.509 1.306 2.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.562 2.603 fft_wrap_pw1pw2_140 451 13.1 0.206 0.215 2.553 2.581 grid_collocate_task_list 110 9.6 2.222 2.421 2.222 2.421 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.341 2.342 fft3d_ps 1111 14.6 1.057 1.091 2.284 2.302 potential_pw2rs 110 12.3 0.013 0.013 2.074 2.078 jit_kernel_multiply 10 16.0 1.041 2.054 1.041 2.054 mp_alltoall_d11v 2046 13.8 1.681 1.813 1.681 1.813 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.687 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.617 1.719 cp_fm_cholesky_decompose 22 10.9 1.636 1.690 1.636 1.690 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.603 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.600 1.617 mp_allgather_i34 2055 14.4 0.417 1.555 0.417 1.555 acc_transpose_blocks 24660 15.4 0.103 0.106 1.518 1.543 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.852 1.491 mp_sum_l 6514 12.8 0.895 1.416 0.895 1.416 mp_irecv_dv 62702 16.1 0.754 1.410 0.754 1.410 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.230 1.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=63.922000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.909091, yerr=10.273130 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 833.839104E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 54.791 54.792 qs_mol_dyn_low 1 2.0 0.003 0.003 54.534 54.544 qs_forces 11 3.9 0.002 0.002 54.470 54.471 qs_energies 11 4.9 0.001 0.002 50.818 50.822 scf_env_do_scf 11 5.9 0.000 0.001 42.560 42.561 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.746 34.746 velocity_verlet 10 3.0 0.001 0.001 31.014 31.016 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.005 23.121 qs_scf_new_mos 99 7.5 0.001 0.001 20.500 20.559 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.500 20.558 ot_scf_mini 99 9.5 0.002 0.002 19.253 19.303 multiply_cannon 2055 13.4 0.242 0.250 17.491 18.778 multiply_cannon_loop 2055 14.4 0.322 0.335 16.143 16.357 rebuild_ks_matrix 110 8.3 0.000 0.000 11.594 11.636 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.594 11.636 ot_mini 99 10.5 0.001 0.001 10.536 10.582 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.321 10.357 multiply_cannon_multrec 8220 15.4 3.200 4.548 7.416 8.625 mp_waitall_1 106626 16.5 6.445 8.146 6.445 8.146 init_scf_loop 11 6.9 0.000 0.000 7.766 7.768 qs_ot_get_derivative 99 11.5 0.001 0.001 6.684 6.733 sum_up_and_integrate 110 10.3 0.048 0.048 6.248 6.262 integrate_v_rspace 110 11.3 0.003 0.003 6.200 6.213 prepare_preconditioner 11 7.9 0.000 0.000 6.115 6.124 make_preconditioner 11 8.9 0.000 0.000 6.115 6.124 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.686 5.764 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.678 5.689 calculate_rho_elec 110 8.6 0.115 0.116 5.677 5.688 init_scf_run 11 5.9 0.000 0.001 5.109 5.109 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.108 5.109 dbcsr_mm_accdrv_process 17442 15.9 2.820 4.187 4.086 4.997 qs_ot_get_p 110 10.4 0.001 0.001 4.773 4.834 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.209 4.570 make_m2s 4110 13.4 0.038 0.040 4.233 4.500 make_images 4110 14.4 0.634 0.691 4.102 4.372 ot_diis_step 99 11.5 0.012 0.012 3.829 3.829 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.750 3.793 apply_single 110 13.6 0.000 0.000 3.750 3.793 grid_integrate_task_list 110 12.3 3.371 3.506 3.371 3.506 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.447 3.450 pw_transfer 1331 11.6 0.066 0.071 3.301 3.310 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.191 3.203 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.141 3.142 multiply_cannon_sync_h2d 8220 15.4 2.900 3.079 2.900 3.079 cp_fm_cholesky_invert 11 10.9 2.971 2.974 2.971 2.974 make_images_data 4110 15.4 0.038 0.043 2.438 2.864 density_rs2pw 110 9.6 0.004 0.004 2.736 2.842 wfi_extrapolate 11 7.9 0.001 0.001 2.789 2.789 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.721 2.731 hybrid_alltoall_any 4261 16.3 0.199 0.863 2.361 2.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.694 2.694 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.625 2.625 cp_fm_redistribute_end 48 14.0 0.665 2.600 0.670 2.601 cp_fm_diag_elpa_base 48 14.0 1.752 2.415 1.925 2.570 grid_collocate_task_list 110 9.6 2.326 2.549 2.326 2.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.539 2.541 calculate_dm_sparse 110 9.5 0.001 0.001 2.487 2.524 fft3d_ps 1111 14.6 1.109 1.118 2.447 2.456 rs_pw_transfer 902 11.9 0.010 0.010 2.310 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.227 2.228 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.158 2.176 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.788 2.018 potential_pw2rs 110 12.3 0.015 0.015 2.008 2.011 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.906 1.923 mp_alltoall_d11v 2046 13.8 1.671 1.876 1.671 1.876 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.829 1.857 jit_kernel_multiply 10 15.8 0.959 1.747 0.959 1.747 cp_fm_cholesky_decompose 22 10.9 1.675 1.701 1.675 1.701 mp_allgather_i34 2055 14.4 0.537 1.662 0.537 1.662 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.514 1.628 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.604 1.610 dbcsr_complete_redistribute 325 12.2 0.572 0.592 1.500 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.444 1.453 qs_create_task_list 11 7.9 0.001 0.001 1.230 1.327 generate_qs_task_list 11 8.9 0.379 0.448 1.229 1.327 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.477 1.305 mp_irecv_dv 24056 15.7 0.453 1.262 0.453 1.262 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.168 1.188 mp_waitany 9240 13.8 1.042 1.167 1.042 1.167 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.847 1.126 mp_alltoall_z22v 1111 16.6 1.106 1.122 1.106 1.122 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.792000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.454545, yerr=10.671659 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 1.189933E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.055 85.565 85.566 qs_mol_dyn_low 1 2.0 0.003 0.004 85.181 85.190 qs_forces 11 3.9 0.002 0.003 85.115 85.116 qs_energies 11 4.9 0.001 0.002 81.024 81.027 scf_env_do_scf 11 5.9 0.001 0.001 70.669 70.670 velocity_verlet 10 3.0 0.001 0.001 54.451 54.457 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.609 42.609 dbcsr_multiply_generic 2055 12.4 0.119 0.122 29.115 29.223 init_scf_loop 11 6.9 0.000 0.000 27.986 27.987 qs_scf_new_mos 99 7.5 0.001 0.001 26.280 26.369 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.279 26.369 prepare_preconditioner 11 7.9 0.000 0.000 26.031 26.041 make_preconditioner 11 8.9 0.000 0.000 26.031 26.041 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.425 25.484 ot_scf_mini 99 9.5 0.002 0.002 24.560 24.635 multiply_cannon 2055 13.4 0.338 0.366 22.027 22.786 multiply_cannon_loop 2055 14.4 0.340 0.344 20.229 20.511 cp_fm_upper_to_full 70 14.2 12.582 17.886 12.582 17.886 ot_mini 99 10.5 0.001 0.001 13.784 13.854 rebuild_ks_matrix 110 8.3 0.000 0.001 13.241 13.319 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.240 13.318 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.032 12.103 dbcsr_complete_redistribute 325 12.2 1.047 1.069 7.212 10.269 multiply_cannon_multrec 8220 15.4 4.448 4.630 9.830 9.965 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.166 9.220 qs_ot_get_derivative 99 11.5 0.001 0.001 9.084 9.161 mp_waitall_1 87304 16.6 7.903 8.890 7.903 8.890 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.590 8.622 mp_alltoall_i22 605 13.7 5.208 8.236 5.208 8.236 sum_up_and_integrate 110 10.3 0.090 0.091 6.621 6.639 integrate_v_rspace 110 11.3 0.003 0.004 6.531 6.549 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.335 6.371 calculate_rho_elec 110 8.6 0.226 0.227 6.334 6.370 init_scf_run 11 5.9 0.000 0.001 6.070 6.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.069 6.071 make_m2s 4110 13.4 0.043 0.043 5.283 5.821 qs_ot_get_p 110 10.4 0.001 0.001 5.609 5.697 make_images 4110 14.4 0.875 0.921 5.097 5.634 dbcsr_mm_accdrv_process 11614 15.7 3.256 3.639 5.239 5.467 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.790 5.307 apply_single 110 13.6 0.000 0.000 4.790 5.306 cp_fm_cholesky_invert 11 10.9 5.259 5.263 5.259 5.263 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.847 5.161 ot_diis_step 99 11.5 0.015 0.016 4.678 4.678 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.053 4.060 multiply_cannon_sync_h2d 8220 15.4 3.947 3.954 3.947 3.954 pw_transfer 1331 11.6 0.073 0.073 3.756 3.765 grid_integrate_task_list 110 12.3 3.663 3.719 3.663 3.719 qs_energies_init_hamiltonians 11 5.9 0.030 0.031 3.685 3.687 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.632 3.640 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.609 3.609 hybrid_alltoall_any 4261 16.3 0.256 0.553 2.833 3.599 make_images_data 4110 15.4 0.041 0.044 2.916 3.585 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.906 3.375 calculate_dm_sparse 110 9.5 0.001 0.001 3.195 3.224 wfi_extrapolate 11 7.9 0.001 0.001 3.214 3.214 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.165 3.168 fft_wrap_pw1pw2_140 451 13.1 0.219 0.221 3.147 3.155 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.053 3.054 cp_fm_diag_elpa_base 48 14.0 2.515 2.708 3.052 3.052 density_rs2pw 110 9.6 0.004 0.004 2.953 2.971 fft3d_ps 1111 14.6 1.261 1.267 2.837 2.844 calculate_first_density_matrix 1 7.0 0.000 0.000 2.745 2.746 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.670 2.696 grid_collocate_task_list 110 9.6 2.627 2.653 2.627 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.389 2.431 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.166 2.231 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.116 2.220 rs_pw_transfer 902 11.9 0.011 0.011 2.165 2.212 potential_pw2rs 110 12.3 0.021 0.021 2.109 2.114 cp_fm_cholesky_decompose 22 10.9 2.034 2.047 2.034 2.047 jit_kernel_multiply 10 15.3 1.782 1.984 1.782 1.984 qs_create_task_list 11 7.9 0.001 0.001 1.900 1.945 generate_qs_task_list 11 8.9 0.737 0.793 1.899 1.944 mp_alltoall_d11v 2046 13.8 1.835 1.872 1.835 1.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.811 1.818 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.774 1.789 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.717 1.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.566000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1117.000000, yerr=22.478272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/15/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 1.094965E+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 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 630.005760E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3266098. MP_ISendRecv 48640 18752. MP_Wait 66796 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 0.020 0.039 203.639 203.640 qs_mol_dyn_low 1 2.0 0.003 0.004 203.233 203.247 qs_forces 11 3.9 0.003 0.004 203.128 203.130 qs_energies 11 4.9 0.002 0.004 197.586 197.597 scf_env_do_scf 11 5.9 0.001 0.002 180.657 180.661 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.939 159.941 dbcsr_multiply_generic 2507 12.6 0.174 0.177 124.261 125.087 velocity_verlet 10 3.0 0.001 0.001 122.379 122.380 qs_scf_new_mos 117 7.6 0.001 0.001 120.923 121.222 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.923 121.221 ot_scf_mini 117 9.6 0.003 0.003 114.319 114.620 multiply_cannon 2507 13.6 0.238 0.251 101.344 103.983 multiply_cannon_loop 2507 14.6 2.100 2.166 99.167 101.713 ot_mini 117 10.6 0.001 0.001 65.378 65.663 multiply_cannon_multrec 60168 15.6 33.240 35.830 41.564 44.018 qs_ot_get_derivative 117 11.6 0.001 0.001 40.401 40.693 rebuild_ks_matrix 128 8.3 0.001 0.001 33.572 33.996 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.021 33.572 33.996 mp_waitall_1 291448 16.2 28.288 30.884 28.288 30.884 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.205 30.572 multiply_cannon_sync_h2d 60168 15.6 27.439 29.902 27.439 29.902 qs_ot_get_p 128 10.4 0.001 0.001 27.095 27.562 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.380 25.504 apply_single 128 13.6 0.001 0.001 24.379 25.503 ot_diis_step 117 11.6 0.007 0.008 24.649 24.650 init_scf_loop 11 6.9 0.000 0.000 20.645 20.647 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.429 20.523 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.718 18.973 multiply_cannon_metrocomm3 60168 15.6 0.110 0.116 15.538 18.732 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.840 17.841 prepare_preconditioner 11 7.9 0.000 0.000 16.038 16.079 make_preconditioner 11 8.9 0.000 0.001 16.038 16.079 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.290 15.483 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.046 15.051 cp_fm_redistribute_end 83 14.4 11.882 14.976 11.895 14.981 cp_fm_diag_elpa_base 83 14.4 3.040 14.685 3.072 14.797 sum_up_and_integrate 128 10.3 0.052 0.064 14.204 14.218 integrate_v_rspace 128 11.3 0.003 0.004 14.151 14.171 make_m2s 5014 13.6 0.104 0.113 13.671 14.013 make_images 5014 14.6 0.402 0.420 13.490 13.840 init_scf_run 11 5.9 0.000 0.001 12.808 12.809 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.808 12.809 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.656 12.748 calculate_rho_elec 128 8.7 0.045 0.063 12.655 12.747 mp_sum_l 7870 13.0 7.986 9.579 7.986 9.579 cp_fm_cholesky_invert 11 10.9 9.275 9.283 9.275 9.283 wfi_extrapolate 11 7.9 0.001 0.001 8.971 8.971 calculate_dm_sparse 128 9.5 0.001 0.001 8.396 8.482 multiply_cannon_metrocomm1 60168 15.6 0.086 0.090 6.016 8.472 dbcsr_mm_accdrv_process 124484 16.2 3.290 3.432 7.883 8.436 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.806 7.982 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.793 7.936 make_images_data 5014 15.6 0.068 0.075 6.708 7.558 grid_integrate_task_list 128 12.3 7.059 7.542 7.059 7.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.287 7.294 density_rs2pw 128 9.7 0.006 0.007 6.542 7.180 hybrid_alltoall_any 5200 16.5 0.291 2.260 5.886 7.106 rs_pw_transfer 1046 11.9 0.017 0.020 6.334 6.938 pw_transfer 1547 11.6 0.075 0.103 6.595 6.865 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.382 6.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.430 6.513 fft_wrap_pw1pw2_140 523 13.2 0.438 0.487 5.520 5.725 fft3d_ps 1291 14.7 2.092 2.583 5.182 5.386 mp_alltoall_d11v 2415 14.1 4.008 5.130 4.008 5.130 grid_collocate_task_list 128 9.7 4.723 5.092 4.723 5.092 potential_pw2rs 128 12.3 0.009 0.010 4.810 4.834 cp_fm_cholesky_decompose 22 10.9 4.609 4.622 4.609 4.622 mp_sum_d 4459 12.1 3.628 4.366 3.628 4.366 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.640000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.181818, yerr=7.259078 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/16/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 2.199914E+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 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 825.049088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5325881. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 188.649 188.650 qs_mol_dyn_low 1 2.0 0.003 0.003 188.292 188.305 qs_forces 11 3.9 0.003 0.003 188.212 188.213 qs_energies 11 4.9 0.001 0.002 181.501 181.512 scf_env_do_scf 11 5.9 0.001 0.001 165.087 165.098 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.408 132.411 velocity_verlet 10 3.0 0.001 0.001 119.074 119.075 dbcsr_multiply_generic 2527 12.6 0.186 0.193 97.635 98.987 qs_scf_new_mos 118 7.6 0.001 0.001 93.873 94.298 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.872 94.297 ot_scf_mini 118 9.6 0.004 0.004 89.047 89.568 multiply_cannon 2527 13.6 0.480 0.531 77.945 81.839 multiply_cannon_loop 2527 14.6 1.271 1.313 74.865 77.661 ot_mini 118 10.6 0.001 0.001 49.631 50.114 mp_waitall_1 228564 16.4 24.180 37.791 24.180 37.791 multiply_cannon_multrec 30324 15.6 22.387 26.959 32.043 37.090 rebuild_ks_matrix 129 8.3 0.001 0.001 32.741 33.321 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.740 33.321 init_scf_loop 11 6.9 0.000 0.000 32.591 32.592 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.539 30.067 prepare_preconditioner 11 7.9 0.000 0.000 28.268 28.347 make_preconditioner 11 8.9 0.000 0.000 28.267 28.347 qs_ot_get_derivative 118 11.6 0.001 0.002 27.648 28.179 multiply_cannon_metrocomm3 30324 15.6 0.097 0.102 15.435 27.614 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.017 27.541 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.059 23.226 apply_single 129 13.6 0.001 0.001 22.058 23.226 multiply_cannon_sync_h2d 30324 15.6 19.629 22.290 19.629 22.290 ot_diis_step 118 11.6 0.014 0.015 21.804 21.806 qs_ot_get_p 129 10.4 0.001 0.001 21.075 21.611 cp_fm_cholesky_invert 11 10.9 16.527 16.539 16.527 16.539 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.349 16.383 make_m2s 5054 13.6 0.090 0.096 14.038 15.443 make_images 5054 14.6 1.164 1.351 13.829 15.233 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.217 15.218 sum_up_and_integrate 129 10.3 0.062 0.072 14.227 14.257 integrate_v_rspace 129 11.3 0.003 0.003 14.165 14.196 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.790 12.819 calculate_rho_elec 129 8.7 0.088 0.106 12.789 12.818 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.184 12.190 cp_fm_redistribute_end 83 14.4 7.129 12.126 7.141 12.127 cp_fm_diag_elpa_base 83 14.4 4.754 11.583 4.972 12.033 init_scf_run 11 5.9 0.000 0.001 11.636 11.638 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.636 11.637 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.808 11.164 multiply_cannon_metrocomm4 27797 15.6 0.098 0.113 3.730 10.444 make_images_data 5054 15.6 0.067 0.073 8.444 10.226 mp_irecv_dv 70031 16.3 3.536 10.062 3.536 10.062 dbcsr_mm_accdrv_process 62734 16.2 4.527 5.253 9.114 9.682 hybrid_alltoall_any 5240 16.5 0.344 1.524 7.257 9.431 wfi_extrapolate 11 7.9 0.001 0.001 8.250 8.250 grid_integrate_task_list 129 12.3 7.232 7.569 7.232 7.569 pw_transfer 1559 11.6 0.084 0.098 7.454 7.519 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.223 7.285 density_rs2pw 129 9.7 0.006 0.006 6.637 7.101 cp_fm_cholesky_decompose 22 10.9 6.963 7.052 6.963 7.052 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.255 7.020 calculate_dm_sparse 129 9.5 0.001 0.001 6.474 6.624 fft_wrap_pw1pw2_140 527 13.2 0.463 0.502 6.347 6.422 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.390 6.399 rs_pw_transfer 1054 12.0 0.014 0.015 5.695 6.222 mp_sum_l 7930 13.1 3.853 5.921 3.853 5.921 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.564 5.734 fft3d_ps 1301 14.7 2.790 2.958 5.587 5.627 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.383 5.464 grid_collocate_task_list 129 9.7 4.950 5.309 4.950 5.309 potential_pw2rs 129 12.3 0.016 0.018 4.750 4.769 mp_allgather_i34 2527 14.6 1.654 4.554 1.654 4.554 mp_alltoall_d11v 2423 14.1 3.816 4.145 3.816 4.145 mp_sum_d 4493 12.2 2.580 4.021 2.580 4.021 dbcsr_complete_redistribute 395 12.7 0.785 0.867 2.977 3.827 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.650000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.000000, yerr=1.809068 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 940.486656E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.120 0.538 174.089 174.090 qs_mol_dyn_low 1 2.0 0.003 0.005 173.166 173.200 qs_forces 11 3.9 0.003 0.003 172.984 172.986 qs_energies 11 4.9 0.004 0.012 166.453 166.461 scf_env_do_scf 11 5.9 0.001 0.002 151.062 151.070 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.375 116.376 velocity_verlet 10 3.0 0.001 0.002 111.429 111.435 dbcsr_multiply_generic 2507 12.6 0.182 0.191 80.987 82.005 qs_scf_new_mos 117 7.6 0.001 0.001 80.021 80.336 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.020 80.335 ot_scf_mini 117 9.6 0.004 0.004 75.918 76.311 multiply_cannon 2507 13.6 0.503 0.529 61.464 66.115 multiply_cannon_loop 2507 14.6 0.862 0.889 58.406 60.955 ot_mini 117 10.6 0.001 0.002 42.119 42.519 init_scf_loop 11 6.9 0.000 0.000 34.588 34.592 mp_waitall_1 178456 16.5 24.769 34.057 24.769 34.057 prepare_preconditioner 11 7.9 0.000 0.000 30.582 30.628 make_preconditioner 11 8.9 0.000 0.000 30.582 30.628 rebuild_ks_matrix 128 8.3 0.001 0.001 30.003 30.512 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 30.002 30.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.191 29.577 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.082 27.539 multiply_cannon_multrec 20056 15.6 13.626 16.271 22.249 24.983 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.006 24.707 qs_ot_get_derivative 117 11.6 0.001 0.002 22.398 22.794 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.817 20.754 apply_single 128 13.6 0.001 0.001 19.817 20.754 qs_ot_get_p 128 10.4 0.001 0.003 19.230 19.649 ot_diis_step 117 11.6 0.019 0.027 19.616 19.618 multiply_cannon_sync_h2d 20056 15.6 14.251 15.692 14.251 15.692 make_m2s 5014 13.6 0.080 0.087 14.524 15.487 make_images 5014 14.6 1.186 1.288 14.291 15.252 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.063 15.071 cp_fm_cholesky_invert 11 10.9 14.160 14.169 14.160 14.169 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.075 14.077 sum_up_and_integrate 128 10.3 0.072 0.077 14.006 14.033 integrate_v_rspace 128 11.3 0.004 0.006 13.934 13.962 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.897 12.937 calculate_rho_elec 128 8.7 0.132 0.146 12.896 12.936 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.113 11.114 cp_fm_redistribute_end 83 14.4 4.190 11.058 4.204 11.061 cp_fm_diag_elpa_base 83 14.4 6.430 10.487 6.837 10.960 init_scf_run 11 5.9 0.000 0.001 10.330 10.331 scf_env_initial_rho_setup 11 6.9 0.002 0.009 10.330 10.330 make_images_data 5014 15.6 0.060 0.069 8.962 10.260 hybrid_alltoall_any 5200 16.5 0.428 1.942 7.710 9.642 multiply_cannon_metrocomm4 17549 15.6 0.063 0.073 3.416 9.161 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.757 9.041 mp_irecv_dv 50230 16.2 3.293 8.913 3.293 8.913 dbcsr_mm_accdrv_process 41502 16.2 4.302 5.171 8.074 8.178 pw_transfer 1547 11.6 0.083 0.100 7.522 7.636 grid_integrate_task_list 128 12.3 7.322 7.636 7.322 7.636 cp_fm_upper_to_full 105 14.5 5.738 7.482 5.738 7.482 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.296 7.415 wfi_extrapolate 11 7.9 0.001 0.003 7.313 7.314 cp_fm_cholesky_decompose 22 10.9 7.218 7.245 7.218 7.245 density_rs2pw 128 9.7 0.006 0.006 6.386 6.680 fft_wrap_pw1pw2_140 523 13.2 0.471 0.513 6.460 6.590 dbcsr_complete_redistribute 395 12.7 1.200 1.247 4.481 6.258 calculate_dm_sparse 128 9.5 0.001 0.001 5.690 5.774 fft3d_ps 1291 14.7 2.702 2.915 5.565 5.662 grid_collocate_task_list 128 9.7 5.066 5.609 5.066 5.609 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.489 5.494 rs_pw_transfer 1046 11.9 0.013 0.014 5.137 5.453 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.471 5.201 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.325 5.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.582 4.740 mp_allgather_i34 2507 14.6 1.588 4.529 1.588 4.529 mp_sum_l 7870 13.0 3.324 4.511 3.324 4.511 potential_pw2rs 128 12.3 0.021 0.023 4.462 4.477 mp_alltoall_d11v 2415 14.1 4.073 4.420 4.073 4.420 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.372 4.120 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.017 4.046 mp_alltoall_i22 716 14.1 1.950 3.870 1.950 3.870 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.793 3.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.430 3.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.090000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.454545, yerr=12.658907 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.134223E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509575. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.058 191.462 191.464 qs_mol_dyn_low 1 2.0 0.003 0.004 190.932 190.945 qs_forces 11 3.9 0.004 0.020 190.793 190.807 qs_energies 11 4.9 0.006 0.025 183.631 183.647 scf_env_do_scf 11 5.9 0.008 0.021 166.107 166.123 velocity_verlet 10 3.0 0.001 0.002 127.572 127.577 scf_env_do_scf_inner_loop 118 6.6 0.029 0.073 118.365 118.373 qs_scf_new_mos 118 7.6 0.001 0.001 81.680 81.874 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.680 81.873 dbcsr_multiply_generic 2529 12.6 0.192 0.197 80.806 81.451 ot_scf_mini 118 9.6 0.004 0.004 77.103 77.377 multiply_cannon 2529 13.6 0.560 0.591 55.494 58.011 multiply_cannon_loop 2529 14.6 1.193 1.223 51.716 53.162 init_scf_loop 11 6.9 0.000 0.000 47.607 47.608 prepare_preconditioner 11 7.9 0.000 0.000 43.404 43.426 make_preconditioner 11 8.9 0.000 0.001 43.404 43.426 ot_mini 118 10.6 0.009 0.024 42.940 43.209 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.990 42.080 multiply_cannon_multrec 30348 15.6 14.447 19.622 26.721 31.457 rebuild_ks_matrix 129 8.3 0.001 0.001 29.772 29.990 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.022 29.771 29.990 mp_waitall_1 155106 16.5 18.146 28.232 18.146 28.232 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.937 27.134 qs_ot_get_derivative 118 11.6 0.002 0.002 23.029 23.307 make_m2s 5058 13.6 0.095 0.100 20.821 21.807 make_images 5058 14.6 1.971 2.304 20.514 21.497 qs_ot_get_p 129 10.4 0.001 0.001 20.115 20.426 ot_diis_step 118 11.6 0.059 0.138 19.764 19.769 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.121 19.698 apply_single 129 13.6 0.001 0.001 19.120 19.698 cp_fm_cholesky_invert 11 10.9 16.898 16.908 16.898 16.908 cp_fm_upper_to_full 106 14.7 11.323 16.762 11.323 16.762 qs_ot_p2m_diag 84 11.4 0.347 0.395 15.782 15.835 multiply_cannon_metrocomm3 30348 15.6 0.047 0.050 6.274 15.356 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.496 14.497 sum_up_and_integrate 129 10.3 0.080 0.087 14.437 14.465 integrate_v_rspace 129 11.3 0.003 0.004 14.357 14.387 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.698 13.747 calculate_rho_elec 129 8.7 0.176 0.192 13.697 13.746 make_images_data 5058 15.6 0.064 0.068 11.330 13.288 multiply_cannon_sync_h2d 30348 15.6 11.889 13.176 11.889 13.176 dbcsr_complete_redistribute 397 12.7 1.534 1.657 9.161 13.018 dbcsr_mm_accdrv_process 62780 16.2 7.416 8.352 11.848 12.389 hybrid_alltoall_any 5245 16.5 0.530 2.222 9.926 12.291 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.645 11.496 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.372 11.375 cp_fm_redistribute_end 84 14.4 1.954 11.302 1.970 11.306 init_scf_run 11 5.9 0.000 0.001 11.291 11.292 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.291 11.292 cp_fm_diag_elpa_base 84 14.4 8.711 10.680 9.304 11.184 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.391 10.118 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.416 9.626 mp_alltoall_i22 720 14.1 5.596 9.430 5.596 9.430 pw_transfer 1559 11.6 0.085 0.098 8.084 8.165 grid_integrate_task_list 129 12.3 7.587 7.981 7.587 7.981 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.853 7.937 cp_fm_cholesky_decompose 22 10.9 7.689 7.791 7.689 7.791 wfi_extrapolate 11 7.9 0.001 0.002 7.742 7.742 density_rs2pw 129 9.7 0.006 0.006 6.690 7.058 multiply_cannon_metrocomm4 25290 15.6 0.075 0.085 2.704 7.051 fft_wrap_pw1pw2_140 527 13.2 0.476 0.483 6.885 6.982 mp_irecv_dv 76751 16.2 2.562 6.785 2.562 6.785 calculate_dm_sparse 129 9.5 0.001 0.001 6.368 6.474 fft3d_ps 1301 14.7 2.819 2.876 6.080 6.146 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.727 5.765 grid_collocate_task_list 129 9.7 5.246 5.743 5.246 5.743 mp_alltoall_d11v 2429 14.1 5.083 5.630 5.083 5.630 rs_pw_transfer 1054 12.0 0.013 0.015 5.228 5.612 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.570 4.650 potential_pw2rs 129 12.3 0.024 0.024 4.581 4.598 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.577 4.579 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.492 4.561 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.250 4.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.464000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.090909, yerr=17.085105 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/19/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 5.865088E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.486406E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838186. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.094 178.221 178.221 qs_mol_dyn_low 1 2.0 0.003 0.004 177.590 177.604 qs_forces 11 3.9 0.003 0.003 177.479 177.482 qs_energies 11 4.9 0.002 0.002 169.675 169.682 scf_env_do_scf 11 5.9 0.001 0.007 151.573 151.590 velocity_verlet 10 3.0 0.001 0.001 118.471 118.475 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.196 114.197 qs_scf_new_mos 117 7.6 0.001 0.001 77.539 77.685 qs_scf_loop_do_ot 117 8.6 0.001 0.001 77.538 77.684 dbcsr_multiply_generic 2507 12.6 0.181 0.186 76.936 77.442 ot_scf_mini 117 9.6 0.004 0.004 72.974 73.108 multiply_cannon 2507 13.6 0.585 0.618 56.027 60.152 multiply_cannon_loop 2507 14.6 0.450 0.460 50.794 51.848 ot_mini 117 10.6 0.001 0.001 40.979 41.135 init_scf_loop 11 6.9 0.000 0.000 37.217 37.219 mp_waitall_1 129618 16.6 28.101 37.019 28.101 37.019 prepare_preconditioner 11 7.9 0.000 0.000 33.214 33.242 make_preconditioner 11 8.9 0.000 0.000 33.214 33.242 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.878 31.232 rebuild_ks_matrix 128 8.3 0.001 0.001 29.565 29.693 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 29.564 29.692 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.688 26.813 multiply_cannon_multrec 10028 15.6 10.404 14.742 17.718 22.488 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 13.167 20.941 qs_ot_get_derivative 117 11.6 0.002 0.002 20.642 20.782 ot_diis_step 117 11.6 0.020 0.020 20.259 20.261 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.744 19.983 apply_single 128 13.6 0.001 0.001 19.744 19.983 make_m2s 5014 13.6 0.066 0.071 16.949 19.856 make_images 5014 14.6 2.311 2.874 16.646 19.556 cp_fm_cholesky_invert 11 10.9 19.081 19.087 19.081 19.087 qs_ot_get_p 128 10.4 0.001 0.001 18.486 18.678 sum_up_and_integrate 128 10.3 0.103 0.110 14.514 14.560 integrate_v_rspace 128 11.3 0.004 0.005 14.410 14.455 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.383 14.398 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.718 13.755 calculate_rho_elec 128 8.7 0.257 0.268 13.718 13.755 make_images_data 5014 15.6 0.052 0.060 10.417 13.379 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.236 13.238 hybrid_alltoall_any 5200 16.5 0.814 3.564 10.151 12.820 multiply_cannon_sync_h2d 10028 15.6 11.481 12.412 11.481 12.412 init_scf_run 11 5.9 0.000 0.001 11.058 11.058 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.058 11.058 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.099 10.108 cp_fm_diag_elpa_base 83 14.4 9.850 9.925 10.092 10.101 multiply_cannon_metrocomm1 10028 15.6 0.030 0.030 4.804 8.295 cp_fm_cholesky_decompose 22 10.9 8.149 8.254 8.149 8.254 pw_transfer 1547 11.6 0.082 0.089 8.185 8.210 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.075 8.190 grid_integrate_task_list 128 12.3 7.793 8.041 7.793 8.041 mp_allgather_i34 2507 14.6 3.447 8.019 3.447 8.019 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.960 7.990 wfi_extrapolate 11 7.9 0.001 0.001 7.951 7.951 dbcsr_mm_accdrv_process 20762 16.1 2.584 3.429 6.948 7.557 fft_wrap_pw1pw2_140 523 13.2 0.496 0.513 6.974 7.005 density_rs2pw 128 9.7 0.005 0.006 6.350 6.718 calculate_dm_sparse 128 9.5 0.001 0.001 6.329 6.438 fft3d_ps 1291 14.7 2.727 2.797 6.124 6.159 dbcsr_complete_redistribute 395 12.7 2.165 2.237 5.691 6.052 mp_alltoall_d11v 2415 14.1 5.457 5.958 5.457 5.958 grid_collocate_task_list 128 9.7 5.494 5.726 5.494 5.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.580 5.633 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.867 5.535 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.470 5.471 mp_irecv_dv 28860 15.9 1.833 5.448 1.833 5.448 rs_pw_transfer 1046 11.9 0.013 0.013 4.722 5.113 potential_pw2rs 128 12.3 0.026 0.028 4.485 4.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.338 4.377 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.742 4.105 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.926 3.956 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.600 3.925 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.788 3.909 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.838 3.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.776 3.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.221000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1390.000000, yerr=38.333465 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019193401344 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019193401344 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820223E+12 0.0% 0.0% 100.0% flops max/rank 11.609361E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705860672 0.0% 0.0% 100.0% number of processed stacks 1951856 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.6 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 2.274476E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99840 MPI messages size (bytes): total size 1.127791E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.295984E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45208 35114713088 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108831376 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 59878. MP_Allreduce 10875 1518. MP_Sync 86 MP_Alltoall 1700 36954490. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.068 304.875 304.876 qs_mol_dyn_low 1 2.0 0.003 0.004 304.274 304.294 qs_forces 11 3.9 0.003 0.003 304.142 304.144 qs_energies 11 4.9 0.026 0.061 295.252 295.262 scf_env_do_scf 11 5.9 0.001 0.002 271.377 271.388 velocity_verlet 10 3.0 0.001 0.002 218.824 218.837 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 143.219 143.222 init_scf_loop 11 6.9 0.000 0.000 127.895 127.899 prepare_preconditioner 11 7.9 0.000 0.000 122.825 122.862 make_preconditioner 11 8.9 0.000 0.000 122.825 122.862 make_full_inverse_cholesky 11 9.9 0.000 0.000 99.228 119.958 qs_scf_new_mos 116 7.6 0.001 0.001 97.987 98.103 qs_scf_loop_do_ot 116 8.6 0.001 0.001 97.986 98.102 ot_scf_mini 116 9.6 0.004 0.004 93.209 93.284 dbcsr_multiply_generic 2496 12.6 0.212 0.220 88.827 89.298 cp_fm_upper_to_full 104 14.8 52.865 75.492 52.865 75.492 multiply_cannon 2496 13.6 0.685 0.764 61.963 63.758 multiply_cannon_loop 2496 14.6 0.469 0.479 57.966 59.196 ot_mini 116 10.6 0.001 0.001 47.531 47.602 dbcsr_complete_redistribute 393 12.7 4.088 4.134 29.700 42.354 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.065 38.721 cp_fm_cholesky_invert 11 10.9 36.604 36.611 36.604 36.611 mp_waitall_1 104986 16.7 32.168 36.328 32.168 36.328 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.552 36.118 rebuild_ks_matrix 127 8.3 0.001 0.001 35.751 35.789 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 35.751 35.788 mp_alltoall_i22 712 14.1 21.421 33.908 21.421 33.908 qs_ks_update_qs_env 127 7.6 0.001 0.001 33.170 33.216 qs_ot_get_p 127 10.4 0.001 0.001 29.368 29.415 qs_ot_get_derivative 116 11.6 0.002 0.002 26.077 26.143 qs_ot_p2m_diag 82 11.4 0.867 0.873 24.734 24.768 cp_dbcsr_syevd 82 12.4 0.005 0.006 22.961 22.964 make_m2s 4992 13.6 0.076 0.078 21.583 22.868 make_images 4992 14.6 3.745 3.870 21.115 22.403 multiply_cannon_metrocomm3 9984 15.6 0.023 0.024 20.847 21.989 ot_diis_step 116 11.6 0.022 0.022 21.379 21.385 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.391 20.603 apply_single 127 13.6 0.001 0.001 20.390 20.603 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.717 19.720 cp_fm_diag_elpa_base 82 14.4 15.249 16.954 19.712 19.715 multiply_cannon_multrec 9984 15.6 10.566 12.333 18.233 18.488 sum_up_and_integrate 127 10.3 0.193 0.194 16.708 16.816 integrate_v_rspace 127 11.3 0.004 0.004 16.515 16.622 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.443 16.462 calculate_rho_elec 127 8.7 0.479 0.479 16.442 16.462 multiply_cannon_sync_h2d 9984 15.6 15.544 15.552 15.544 15.552 make_images_data 4992 15.6 0.059 0.063 11.927 14.228 hybrid_alltoall_any 5177 16.4 1.295 3.009 11.820 14.110 init_scf_run 11 5.9 0.000 0.001 13.606 13.606 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.605 13.606 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.068 10.127 pw_transfer 1535 11.6 0.092 0.094 9.911 9.926 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 9.673 9.687 wfi_extrapolate 11 7.9 0.001 0.001 9.503 9.503 cp_fm_cholesky_decompose 22 10.9 9.439 9.499 9.439 9.499 dbcsr_mm_accdrv_process 20634 16.0 3.848 5.983 7.432 9.413 grid_integrate_task_list 127 12.3 8.489 8.659 8.489 8.659 fft_wrap_pw1pw2_140 519 13.2 0.522 0.528 8.402 8.425 mp_alltoall_d11v 2401 14.1 8.072 8.194 8.072 8.194 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.090 8.092 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.730 7.846 fft3d_ps 1281 14.7 2.753 2.778 7.740 7.762 density_rs2pw 127 9.7 0.005 0.005 7.218 7.248 calculate_dm_sparse 127 9.5 0.001 0.001 6.775 6.881 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.402 6.522 grid_collocate_task_list 127 9.7 6.288 6.331 6.288 6.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=304.876000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2142.000000, yerr=49.559881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 1.260073E+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 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.038 89.476 89.478 qs_energies 1 2.0 0.000 0.001 89.008 89.022 ls_scf 1 3.0 0.003 0.013 88.082 88.096 dbcsr_multiply_generic 111 6.7 0.015 0.019 75.900 76.113 multiply_cannon 111 7.7 0.021 0.051 57.388 58.564 ls_scf_main 1 4.0 0.000 0.002 55.419 55.430 multiply_cannon_loop 111 8.7 0.209 0.222 53.917 55.400 density_matrix_trs4 2 5.0 0.002 0.003 49.614 49.697 ls_scf_init_scf 1 4.0 0.000 0.002 29.576 29.599 mp_waitall_1 11316 10.9 25.424 28.271 25.424 28.271 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.518 27.583 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.401 25.428 multiply_cannon_multrec 2664 9.7 8.156 8.959 15.445 17.384 multiply_cannon_sync_h2d 2664 9.7 13.368 15.614 13.368 15.614 make_m2s 222 7.7 0.008 0.012 14.484 15.008 make_images 222 8.7 0.111 0.210 14.462 14.988 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 11.249 14.073 make_images_data 222 9.7 0.004 0.005 8.792 9.396 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.686 9.307 hybrid_alltoall_any 227 10.6 0.215 1.829 7.511 8.960 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.601 6.908 8.008 dbcsr_mm_accdrv_process_sort 4760 11.4 6.192 7.130 6.192 7.130 calculate_norms 4752 9.8 5.491 6.172 5.491 6.172 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.158 5.323 mp_sum_l 807 5.4 3.512 5.109 3.512 5.109 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.624 3.899 make_images_sizes 222 9.7 0.000 0.000 0.997 3.677 mp_alltoall_i44 222 10.7 0.996 3.676 0.996 3.676 arnoldi_extremal 4 6.8 0.000 0.003 3.321 3.350 arnoldi_normal_ev 4 7.8 0.001 0.007 3.320 3.350 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.051 3.316 mp_irecv_dv 6231 10.9 2.035 3.297 2.035 3.297 build_subspace 16 8.4 0.009 0.013 3.143 3.147 ls_scf_post 1 4.0 0.000 0.002 3.085 3.098 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.929 dbcsr_special_finalize 555 9.7 0.005 0.006 2.337 2.725 dbcsr_merge_single_wm 555 10.7 0.459 0.572 2.329 2.717 make_images_pack 222 9.7 2.200 2.629 2.202 2.630 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.326 2.568 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.408 2.524 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.479 2.073 2.481 dbcsr_sort_data 658 11.4 2.130 2.462 2.130 2.462 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.254 2.280 rebuild_ks_matrix 3 7.3 0.000 0.000 2.244 2.270 qs_ks_build_kohn_sham_matrix 3 8.3 0.012 0.053 2.244 2.270 ls_scf_initial_guess 1 5.0 0.000 0.000 2.058 2.162 ls_scf_qs_atomic_guess 1 6.0 0.001 0.011 2.058 2.162 buffer_matrices_ensure_size 222 8.7 1.749 2.073 1.749 2.073 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=89.478000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 2.078052E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.062 94.857 94.860 qs_energies 1 2.0 0.000 0.001 94.298 94.302 ls_scf 1 3.0 0.000 0.002 92.954 92.957 dbcsr_multiply_generic 111 6.7 0.015 0.015 78.439 78.798 multiply_cannon 111 7.7 0.029 0.043 54.803 58.473 ls_scf_main 1 4.0 0.000 0.002 57.457 57.465 multiply_cannon_loop 111 8.7 0.115 0.122 51.481 55.107 density_matrix_trs4 2 5.0 0.002 0.003 51.387 51.586 mp_waitall_1 9246 10.9 23.811 32.711 23.811 32.711 ls_scf_init_scf 1 4.0 0.000 0.001 31.904 31.906 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.728 30.818 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 28.209 28.220 multiply_cannon_multrec 1332 9.7 12.969 16.401 22.146 26.753 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 13.311 22.279 make_m2s 222 7.7 0.006 0.007 16.314 16.856 make_images 222 8.7 1.569 1.969 16.284 16.826 make_images_data 222 9.7 0.004 0.004 9.703 10.626 dbcsr_mm_accdrv_process 4041 10.4 0.267 0.455 8.780 10.382 hybrid_alltoall_any 227 10.6 0.520 2.455 9.004 10.233 dbcsr_mm_accdrv_process_sort 4041 11.4 8.367 9.926 8.367 9.926 mp_sum_l 807 5.4 6.048 8.973 6.048 8.973 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.247 7.751 mp_irecv_dv 3311 11.0 3.228 7.704 3.228 7.704 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.673 7.178 calculate_norms 2376 9.8 5.982 6.774 5.982 6.774 multiply_cannon_sync_h2d 1332 9.7 4.696 5.952 4.696 5.952 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.526 5.740 arnoldi_extremal 4 6.8 0.000 0.000 5.070 5.099 arnoldi_normal_ev 4 7.8 0.005 0.014 5.070 5.099 build_subspace 16 8.4 0.014 0.021 4.770 4.773 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.395 3.677 ls_scf_post 1 4.0 0.000 0.003 3.592 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.304 3.410 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.408 3.248 dbcsr_matrix_vector_mult_local 304 10.0 2.729 3.203 2.731 3.205 mp_allgather_i34 111 8.7 1.020 2.728 1.020 2.728 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.589 2.716 make_images_pack 222 9.7 2.035 2.416 2.038 2.418 dbcsr_sort_data 436 11.2 1.796 2.035 1.796 2.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.860000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1700.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 2.716824E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.091 0.198 96.264 96.267 qs_energies 1 2.0 0.000 0.001 95.458 95.467 ls_scf 1 3.0 0.003 0.010 94.035 94.045 dbcsr_multiply_generic 111 6.7 0.016 0.021 78.433 78.704 ls_scf_main 1 4.0 0.000 0.002 58.881 58.892 multiply_cannon 111 7.7 0.040 0.108 53.659 58.301 multiply_cannon_loop 111 8.7 0.100 0.107 50.103 53.946 density_matrix_trs4 2 5.0 0.002 0.003 52.948 53.108 mp_waitall_1 7374 11.0 25.527 35.616 25.527 35.616 ls_scf_init_scf 1 4.0 0.000 0.001 31.450 31.462 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.199 30.280 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.685 27.695 multiply_cannon_multrec 888 9.7 12.511 15.010 21.025 24.137 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.811 23.606 make_m2s 222 7.7 0.006 0.007 18.007 19.286 make_images 222 8.7 1.971 2.305 17.969 19.247 hybrid_alltoall_any 227 10.6 0.621 2.847 10.186 11.676 make_images_data 222 9.7 0.004 0.005 10.580 11.658 dbcsr_mm_accdrv_process 3754 10.4 0.249 0.413 8.032 9.217 mp_sum_l 807 5.4 5.283 8.989 5.283 8.989 dbcsr_mm_accdrv_process_sort 3754 11.4 7.658 8.803 7.658 8.803 multiply_cannon_sync_h2d 888 9.7 6.092 7.331 6.092 7.331 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.482 7.166 mp_irecv_dv 2335 11.1 2.466 7.121 2.466 7.121 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.750 6.972 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.988 6.964 arnoldi_extremal 4 6.8 0.000 0.000 5.261 5.274 arnoldi_normal_ev 4 7.8 0.001 0.005 5.261 5.274 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.998 5.215 build_subspace 16 8.4 0.014 0.020 4.956 4.963 calculate_norms 1584 9.8 4.261 4.619 4.261 4.619 mp_allgather_i34 111 8.7 1.400 4.011 1.400 4.011 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.554 3.880 ls_scf_post 1 4.0 0.007 0.055 3.700 3.712 dbcsr_matrix_vector_mult_local 304 10.0 2.992 3.563 2.994 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.501 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.780 2.877 make_images_sizes 222 9.7 0.000 0.000 1.105 2.514 mp_alltoall_i44 222 10.7 1.104 2.513 1.104 2.513 make_images_pack 222 9.7 1.820 2.133 1.823 2.135 dbcsr_sort_data 325 11.1 1.895 2.115 1.895 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.929 1.940 dbcsr_data_release 9322 10.9 1.308 1.930 1.308 1.930 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.267000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2165.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 3.372560E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.054 100.084 100.085 qs_energies 1 2.0 0.000 0.000 99.451 99.459 ls_scf 1 3.0 0.000 0.000 97.771 97.777 dbcsr_multiply_generic 111 6.7 0.016 0.017 80.996 81.256 ls_scf_main 1 4.0 0.000 0.000 60.681 60.688 multiply_cannon 111 7.7 0.050 0.120 52.644 56.987 density_matrix_trs4 2 5.0 0.002 0.003 54.347 54.481 multiply_cannon_loop 111 8.7 0.114 0.126 47.555 50.474 ls_scf_init_scf 1 4.0 0.000 0.000 33.715 33.743 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.540 32.616 mp_waitall_1 6438 11.0 24.676 31.716 24.676 31.716 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.046 30.056 multiply_cannon_multrec 1332 9.7 14.197 17.239 22.024 24.524 make_m2s 222 7.7 0.006 0.007 22.110 23.486 make_images 222 8.7 3.143 3.599 22.060 23.438 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.933 17.728 make_images_data 222 9.7 0.004 0.004 12.686 14.426 hybrid_alltoall_any 227 10.6 0.797 3.792 12.005 14.053 dbcsr_mm_accdrv_process 3641 10.4 0.214 0.412 7.466 8.979 dbcsr_mm_accdrv_process_sort 3641 11.4 7.083 8.559 7.083 8.559 mp_sum_l 807 5.4 4.652 7.581 4.652 7.581 multiply_cannon_sync_h2d 1332 9.7 5.475 6.095 5.475 6.095 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.563 5.956 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.076 5.937 mp_irecv_dv 3229 10.9 2.052 5.865 2.052 5.865 arnoldi_extremal 4 6.8 0.000 0.000 5.565 5.605 arnoldi_normal_ev 4 7.8 0.002 0.005 5.565 5.605 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.880 5.563 build_subspace 16 8.4 0.014 0.021 5.057 5.075 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.717 4.892 mp_allgather_i34 111 8.7 2.211 4.651 2.211 4.651 calculate_norms 2376 9.8 4.170 4.570 4.170 4.570 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.643 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.149 3.644 3.151 3.646 dbcsr_sort_data 658 11.4 3.134 3.460 3.134 3.460 ls_scf_post 1 4.0 0.000 0.000 3.376 3.383 ls_scf_store_result 1 5.0 0.000 0.000 3.098 3.172 dbcsr_special_finalize 555 9.7 0.006 0.006 2.873 3.172 dbcsr_merge_single_wm 555 10.7 0.536 0.653 2.864 3.164 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.050 3.104 dbcsr_data_release 10477 10.7 1.586 2.475 1.586 2.475 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.085000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2667.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 4.638777E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.058 101.260 101.262 qs_energies 1 2.0 0.000 0.000 100.462 100.466 ls_scf 1 3.0 0.000 0.000 98.502 98.507 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.569 79.812 ls_scf_main 1 4.0 0.000 0.000 63.226 63.227 multiply_cannon 111 7.7 0.068 0.138 56.247 61.796 density_matrix_trs4 2 5.0 0.002 0.003 56.015 56.116 multiply_cannon_loop 111 8.7 0.069 0.078 51.854 53.428 mp_waitall_1 5481 11.0 28.128 33.418 28.128 33.418 ls_scf_init_scf 1 4.0 0.000 0.000 31.582 31.587 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.320 30.363 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.072 28.089 multiply_cannon_multrec 444 9.7 13.979 16.612 21.011 23.260 make_m2s 222 7.7 0.004 0.005 18.390 20.764 make_images 222 8.7 3.716 4.412 18.328 20.705 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 12.189 17.718 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.226 14.305 hybrid_alltoall_any 227 10.6 0.787 3.750 10.532 13.008 make_images_data 222 9.7 0.003 0.004 10.507 12.964 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.344 6.737 7.897 multiply_cannon_sync_h2d 444 9.7 6.545 7.713 6.545 7.713 dbcsr_mm_accdrv_process_sort 3003 11.4 6.401 7.516 6.401 7.516 mp_allgather_i34 111 8.7 2.629 7.036 2.629 7.036 arnoldi_extremal 4 6.8 0.000 0.000 6.073 6.092 arnoldi_normal_ev 4 7.8 0.003 0.008 6.073 6.092 build_subspace 16 8.4 0.015 0.020 5.546 5.581 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.802 4.949 mp_sum_l 807 5.4 2.981 4.740 2.981 4.740 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.115 4.316 dbcsr_matrix_vector_mult_local 304 10.0 3.623 4.100 3.625 4.103 make_images_sizes 222 9.7 0.000 0.000 1.242 3.989 mp_alltoall_i44 222 10.7 1.242 3.988 1.242 3.988 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.655 3.959 mp_irecv_dv 1241 11.2 1.636 3.925 1.636 3.925 ls_scf_post 1 4.0 0.000 0.000 3.694 3.699 calculate_norms 792 9.8 3.557 3.683 3.557 3.683 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.430 3.538 ls_scf_store_result 1 5.0 0.000 0.000 3.457 3.519 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.029 3.376 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.317 dbcsr_merge_all 275 8.9 0.473 0.537 2.048 2.149 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.262000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3632.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/948733f1c7cd1fbbd9135e2cb108ad12e592a17c_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 8.730513E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.102 110.341 110.342 qs_energies 1 2.0 0.000 0.000 108.926 108.935 ls_scf 1 3.0 0.000 0.000 105.912 105.920 dbcsr_multiply_generic 111 6.7 0.023 0.027 79.381 79.507 ls_scf_main 1 4.0 0.000 0.000 67.196 67.197 density_matrix_trs4 2 5.0 0.002 0.003 58.116 58.161 multiply_cannon 111 7.7 0.133 0.199 51.346 53.497 multiply_cannon_loop 111 8.7 0.067 0.071 47.843 48.603 ls_scf_init_scf 1 4.0 0.000 0.000 34.966 34.968 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.601 33.616 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.625 30.634 mp_waitall_1 4569 11.1 24.143 27.904 24.143 27.904 make_m2s 222 7.7 0.005 0.005 24.481 25.508 make_images 222 8.7 4.573 4.952 24.375 25.400 multiply_cannon_multrec 444 9.7 17.827 18.652 22.421 23.054 hybrid_alltoall_any 227 10.6 1.659 3.626 13.527 16.356 make_images_data 222 9.7 0.003 0.003 13.646 16.208 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.847 12.354 multiply_cannon_sync_h2d 444 9.7 8.847 8.891 8.847 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.571 7.576 arnoldi_normal_ev 4 7.8 0.007 0.016 7.571 7.576 build_subspace 16 8.4 0.026 0.036 6.996 7.005 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.351 5.500 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.003 5.236 dbcsr_matrix_vector_mult_local 304 10.0 4.872 5.166 4.875 5.169 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.031 5.125 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.311 4.414 4.527 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.241 4.114 4.241 make_images_sizes 222 9.7 0.000 0.000 1.527 3.820 mp_alltoall_i44 222 10.7 1.526 3.820 1.526 3.820 ls_scf_post 1 4.0 0.000 0.000 3.750 3.759 mp_allgather_i34 111 8.7 1.160 3.664 1.160 3.664 ls_scf_store_result 1 5.0 0.000 0.000 3.498 3.507 calculate_norms 792 9.8 3.241 3.274 3.241 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.123 qs_energies_init_hamiltonians 1 3.0 0.055 0.091 2.983 2.983 dbcsr_complete_redistribute 5 7.6 1.468 1.510 2.796 2.936 dbcsr_merge_all 275 8.9 0.892 0.916 2.863 2.906 dbcsr_data_release 12724 10.6 2.324 2.841 2.324 2.841 matrix_ls_to_qs 2 6.0 0.000 0.000 2.434 2.588 dbcsr_sort_data 325 11.1 2.437 2.499 2.437 2.499 dbcsr_new_transposed 4 7.5 0.243 0.252 2.368 2.384 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.322 2.325 mp_alltoall_d11v 48 9.2 2.254 2.315 2.254 2.315 rebuild_ks_matrix 3 7.3 0.000 0.000 2.256 2.258 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.256 2.258 dbcsr_frobenius_norm 74 6.6 2.054 2.130 2.196 2.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.342000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6882.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 948733f1c7cd1fbbd9135e2cb108ad12e592a17c Summary: empty Status: OK