=== 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: 2616d1f209524a36d9c99e7eecb679e8d7bcccfe ################# 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/01 job id: 41598546 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/02 job id: 41598554 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/03 job id: 41598555 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/04 job id: 41598557 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/05 job id: 41598558 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/06 job id: 41598559 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/07 job id: 41598560 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/08 job id: 41598565 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/09 job id: 41598566 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/10 job id: 41598567 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/11 job id: 41598568 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/12 job id: 41598569 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/13 job id: 41598570 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/14 job id: 41598571 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/15 job id: 41598572 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/16 job id: 41598573 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/17 job id: 41598574 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/18 job id: 41598575 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/19 job id: 41598577 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/20 job id: 41598584 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/21 job id: 41598588 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/22 job id: 41598592 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/23 job id: 41598595 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/24 job id: 41598596 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/25 job id: 41598601 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/26 job id: 41598602 --- 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 0.053 0.062 134.997 134.997 farming_run 1 2.0 134.295 134.311 134.918 134.945 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459757E+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 0.011 0.028 117.073 117.073 qs_energies 1 2.0 0.000 0.000 116.737 116.739 mp2_main 1 3.0 0.000 0.000 114.882 114.883 mp2_gpw_main 1 4.0 0.019 0.024 114.073 114.075 mp2_ri_gpw_compute_in 1 5.0 0.173 0.194 95.441 95.597 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 56.354 56.510 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.174 42.654 47.761 get_2c_integrals 1 6.0 0.000 0.000 38.062 38.915 integrate_v_rspace 273 8.0 0.433 0.448 25.227 30.179 pw_transfer 6555 10.6 0.376 0.399 27.641 28.152 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.270 26.579 grid_integrate_task_list 273 9.0 20.987 26.464 20.987 26.464 fft_wrap_pw1pw2_100 2178 12.4 1.213 1.435 23.836 24.157 compute_2c_integrals 1 7.0 0.002 0.002 20.467 20.468 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.704 20.191 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.445 19.701 20.190 rpa_ri_compute_en 1 5.0 0.000 0.000 18.521 18.687 cp_fm_cholesky_decompose 12 8.2 17.532 18.391 17.532 18.391 cholesky_decomp 1 7.0 0.000 0.000 16.440 17.299 fft3d_s 5443 13.4 16.311 16.765 16.332 16.786 ao_to_mo_and_store_B_mult_1 272 7.0 10.868 15.599 10.868 15.599 calculate_wavefunction 272 8.0 5.457 5.594 12.650 13.329 rpa_num_int 1 6.0 0.000 0.000 10.505 10.505 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.466 10.476 calc_potential_gpw 544 9.5 0.005 0.005 9.692 10.105 calc_mat_Q 8 8.0 0.000 0.000 9.326 9.419 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.971 9.280 contract_S_to_Q 8 9.0 0.000 0.000 8.746 8.839 potential_pw2rs 545 10.0 0.106 0.108 7.760 8.456 parallel_gemm_fm 14 9.1 0.000 0.000 8.317 8.411 parallel_gemm_fm_cosma 14 10.1 8.317 8.411 8.317 8.411 collocate_single_gaussian 272 10.0 0.039 0.042 7.533 7.819 create_integ_mat 1 6.0 0.023 0.027 7.766 7.766 array2fm 1 7.0 0.000 0.000 6.746 7.199 pw_scatter_s 2720 13.7 4.425 4.563 4.425 4.563 pw_gather_s 2722 13.2 3.954 4.411 3.954 4.411 pw_poisson_solve 545 10.5 2.478 2.537 3.553 3.772 array2fm_buffer_send 1 8.0 2.989 3.164 2.989 3.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.074875, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 0.065 0.072 396.127 396.129 farming_run 1 2.0 395.198 395.312 396.050 396.091 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225273E+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 0.012 0.041 210.067 210.068 qs_energies 1 2.0 0.000 0.000 209.831 209.860 scf_env_do_scf 1 3.0 0.000 0.000 106.120 106.120 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.273 105.281 rebuild_ks_matrix 4 6.0 0.000 0.000 105.271 105.279 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.271 105.279 hfx_ks_matrix 4 8.0 0.001 0.001 104.896 104.900 integrate_four_center 4 9.0 0.144 0.461 104.895 104.900 mp2_main 1 3.0 0.004 0.033 103.425 103.453 mp2_gpw_main 1 4.0 0.037 0.078 102.585 102.602 integrate_four_center_main 4 10.0 0.135 0.533 96.789 99.439 integrate_four_center_bin 265 11.0 96.654 99.403 96.654 99.403 init_scf_loop 1 4.0 0.000 0.000 91.994 91.994 mp2_ri_gpw_compute_in 1 5.0 0.066 0.083 75.720 76.719 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.020 54.920 55.921 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.163 42.508 47.688 integrate_v_rspace 95 8.0 0.396 0.565 28.435 33.445 pw_transfer 2240 10.6 0.146 0.167 30.126 30.489 ao_to_mo_and_store_B_mult_1 91 7.0 10.723 29.541 10.723 29.541 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 29.113 29.527 grid_integrate_task_list 95 9.0 23.714 28.925 23.714 28.925 mp2_ri_gpw_compute_en 1 5.0 0.065 0.088 26.698 28.418 fft_wrap_pw1pw2_100 730 12.4 1.278 1.462 26.852 27.310 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.910 24.994 25.005 get_2c_integrals 1 6.0 0.003 0.027 20.722 20.737 compute_2c_integrals 1 7.0 0.003 0.010 19.699 19.706 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 19.283 19.573 mp2_eri_2c_integrate_gpw 1 9.0 1.750 1.875 19.282 19.571 fft3d_s 1823 13.4 18.535 18.961 18.548 18.974 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.125 14.125 calculate_wavefunction 91 8.0 2.024 2.063 9.804 10.093 potential_pw2rs 186 10.0 0.033 0.034 8.688 9.417 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.593 8.789 9.166 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.608 9.053 calc_potential_gpw 182 9.5 0.002 0.002 8.620 8.922 local_gemm 172 8.0 8.233 8.607 8.233 8.607 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.514 7.972 8.433 collocate_single_gaussian 91 10.0 0.017 0.021 7.920 8.228 mp_sendrecv_dm3 2068 8.0 6.014 6.472 6.014 6.472 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.447 6.346 6.447 mp_sync 38 10.4 2.751 5.660 2.751 5.660 pw_gather_s 912 13.2 4.976 5.482 4.976 5.482 pw_scatter_s 910 13.7 3.965 4.299 3.965 4.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.585922, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 449.028096E+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 592243. 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 0.013 0.032 51.626 51.627 qs_mol_dyn_low 1 2.0 0.003 0.004 51.401 51.408 qs_forces 11 3.9 0.002 0.002 51.339 51.340 qs_energies 11 4.9 0.001 0.002 49.859 49.868 scf_env_do_scf 11 5.9 0.000 0.001 43.820 43.820 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.811 41.811 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.372 33.838 qs_scf_new_mos 108 7.5 0.000 0.001 31.705 32.001 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.705 32.001 ot_scf_mini 108 9.5 0.002 0.002 30.094 30.255 multiply_cannon 2286 13.5 0.185 0.194 26.100 27.618 multiply_cannon_loop 2286 14.5 1.524 1.594 25.432 26.998 velocity_verlet 10 3.0 0.001 0.002 25.437 25.437 ot_mini 108 10.5 0.001 0.001 19.326 19.554 qs_ot_get_derivative 108 11.5 0.001 0.001 16.412 16.600 mp_waitall_1 267858 16.1 8.697 14.876 8.697 14.876 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.079 13.206 multiply_cannon_multrec 54864 15.5 4.198 6.505 7.682 11.284 rebuild_ks_matrix 119 8.3 0.000 0.000 7.932 8.057 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.010 7.932 8.057 multiply_cannon_sync_h2d 54864 15.5 5.876 7.803 5.876 7.803 mp_sum_l 7207 12.9 5.407 7.199 5.407 7.199 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.987 7.100 qs_ot_get_p 119 10.4 0.001 0.001 6.178 6.502 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.539 6.011 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.394 5.490 init_scf_run 11 5.9 0.000 0.001 4.822 4.822 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.822 4.822 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.838 3.406 4.772 sum_up_and_integrate 119 10.3 0.008 0.009 4.588 4.596 integrate_v_rspace 119 11.3 0.002 0.002 4.580 4.589 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.111 4.235 calculate_rho_elec 119 8.7 0.011 0.016 4.111 4.234 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.814 3.565 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.408 3.456 rs_pw_transfer 974 11.9 0.012 0.014 2.986 3.071 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.854 3.039 apply_single 119 13.6 0.000 0.000 2.854 3.039 calculate_dm_sparse 119 9.5 0.000 0.000 2.863 3.036 jit_kernel_multiply 13 15.8 2.157 2.872 2.157 2.872 calculate_first_density_matrix 1 7.0 0.000 0.001 2.713 2.724 ot_diis_step 108 11.5 0.006 0.006 2.612 2.613 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.590 2.591 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.417 2.478 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.392 2.392 cp_fm_redistribute_end 50 14.0 2.174 2.371 2.178 2.372 cp_fm_diag_elpa_base 50 14.0 0.193 2.333 0.194 2.346 density_rs2pw 119 9.7 0.003 0.004 2.219 2.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.296 2.298 acc_transpose_blocks 54864 15.5 0.228 0.247 1.768 2.200 grid_integrate_task_list 119 12.3 2.020 2.116 2.020 2.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.031 2.079 wfi_extrapolate 11 7.9 0.001 0.001 2.047 2.047 init_scf_loop 11 6.9 0.000 0.000 1.992 1.993 potential_pw2rs 119 12.3 0.004 0.004 1.969 1.984 mp_sum_d 4129 12.0 1.264 1.848 1.264 1.848 pw_transfer 1439 11.6 0.052 0.055 1.593 1.664 make_m2s 4572 13.5 0.054 0.055 1.583 1.628 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.519 1.591 make_images 4572 14.5 0.133 0.138 1.500 1.545 mp_alltoall_d11v 2130 13.8 1.276 1.471 1.276 1.471 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.414 1.432 fft3d_ps 1201 14.6 0.357 0.461 1.293 1.357 grid_collocate_task_list 119 9.7 1.288 1.353 1.288 1.353 mp_waitany 12084 13.8 1.176 1.288 1.176 1.288 fft_wrap_pw1pw2_140 487 13.2 0.083 0.097 1.176 1.251 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.708 1.101 acc_transpose_blocks_kernels 54864 16.5 0.255 0.390 0.773 1.049 parallel_gemm_fm 81 9.0 0.000 0.000 1.039 1.044 parallel_gemm_fm_cosma 81 10.0 1.039 1.044 1.039 1.044 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.627000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 485.797888E+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 1710877. 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 0.027 0.043 37.807 37.807 qs_mol_dyn_low 1 2.0 0.005 0.017 37.519 37.527 qs_forces 11 3.9 0.014 0.017 37.456 37.460 qs_energies 11 4.9 0.025 0.029 35.781 35.787 scf_env_do_scf 11 5.9 0.000 0.001 30.738 30.738 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.335 28.336 dbcsr_multiply_generic 2286 12.5 0.099 0.104 21.104 21.466 qs_scf_new_mos 108 7.5 0.001 0.001 19.383 19.623 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.382 19.623 ot_scf_mini 108 9.5 0.002 0.003 18.491 18.659 velocity_verlet 10 3.0 0.001 0.002 17.975 17.976 multiply_cannon 2286 13.5 0.207 0.220 16.244 17.802 multiply_cannon_loop 2286 14.5 0.911 0.990 15.139 16.505 ot_mini 108 10.5 0.001 0.001 11.599 11.833 mp_waitall_1 217478 16.2 5.792 10.924 5.792 10.924 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.124 9.396 qs_ot_get_derivative 108 11.5 0.001 0.002 9.161 9.333 multiply_cannon_multrec 27432 15.5 1.973 4.541 5.861 8.796 rebuild_ks_matrix 119 8.3 0.000 0.000 7.005 7.145 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.004 7.144 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.183 6.310 dbcsr_mm_accdrv_process 47894 16.0 3.056 4.996 3.817 5.648 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.515 4.359 qs_ot_get_p 119 10.4 0.001 0.001 4.031 4.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.036 4.109 apply_single 119 13.6 0.000 0.000 3.036 4.109 sum_up_and_integrate 119 10.3 0.014 0.016 4.061 4.068 integrate_v_rspace 119 11.3 0.002 0.002 4.047 4.053 mp_sum_l 7207 12.9 1.996 3.936 1.996 3.936 init_scf_run 11 5.9 0.000 0.001 3.802 3.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.802 3.802 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.487 3.525 calculate_rho_elec 119 8.7 0.021 0.024 3.486 3.524 multiply_cannon_sync_h2d 27432 15.5 2.180 2.802 2.180 2.802 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.642 2.660 make_m2s 4572 13.5 0.052 0.054 2.413 2.649 rs_pw_transfer 974 11.9 0.010 0.011 2.453 2.561 make_images 4572 14.5 0.201 0.239 2.324 2.560 ot_diis_step 108 11.5 0.011 0.013 2.389 2.390 init_scf_loop 11 6.9 0.000 0.000 2.384 2.384 calculate_first_density_matrix 1 7.0 0.000 0.001 2.343 2.345 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.213 2.214 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.083 2.173 calculate_dm_sparse 119 9.5 0.001 0.001 2.027 2.100 density_rs2pw 119 9.7 0.004 0.004 1.850 1.937 grid_integrate_task_list 119 12.3 1.828 1.933 1.828 1.933 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.907 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.904 1.907 cp_fm_redistribute_end 50 14.0 1.576 1.886 1.579 1.887 cp_fm_diag_elpa_base 50 14.0 0.300 1.834 0.307 1.863 potential_pw2rs 119 12.3 0.006 0.006 1.831 1.836 jit_kernel_multiply 9 16.1 0.708 1.792 0.708 1.792 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.718 1.758 pw_transfer 1439 11.6 0.063 0.066 1.692 1.723 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.604 1.636 prepare_preconditioner 11 7.9 0.000 0.000 1.500 1.527 make_preconditioner 11 8.9 0.000 0.000 1.500 1.527 make_images_data 4572 15.5 0.044 0.051 1.127 1.527 acc_transpose_blocks 27432 15.5 0.109 0.114 1.190 1.491 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.970 1.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.396 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.411 1.411 fft3d_ps 1201 14.6 0.499 0.554 1.309 1.336 grid_collocate_task_list 119 9.7 1.241 1.328 1.241 1.328 fft_wrap_pw1pw2_140 487 13.2 0.078 0.084 1.279 1.311 mp_alltoall_d11v 2130 13.8 1.148 1.298 1.148 1.298 mp_allgather_i34 2286 14.5 0.549 1.256 0.549 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.247 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.154 1.204 mp_sum_d 4129 12.0 0.548 0.984 0.548 0.984 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.952 0.955 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.889 acc_transpose_blocks_kernels 27432 16.5 0.182 0.269 0.657 0.867 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.692 0.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=37.807000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.090909, yerr=1.621141 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 517.824512E+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 0.034 0.063 31.492 31.492 qs_mol_dyn_low 1 2.0 0.007 0.023 31.152 31.160 qs_forces 11 3.9 0.003 0.006 31.055 31.057 qs_energies 11 4.9 0.002 0.004 29.303 29.306 scf_env_do_scf 11 5.9 0.001 0.001 24.670 24.670 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 22.179 22.180 dbcsr_multiply_generic 2286 12.5 0.093 0.098 16.133 16.221 velocity_verlet 10 3.0 0.002 0.002 14.827 14.831 qs_scf_new_mos 108 7.5 0.001 0.001 14.369 14.391 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.369 14.391 ot_scf_mini 108 9.5 0.002 0.003 13.657 13.671 multiply_cannon 2286 13.5 0.194 0.200 12.933 13.663 multiply_cannon_loop 2286 14.5 0.641 0.671 12.171 12.977 ot_mini 108 10.5 0.001 0.001 8.547 8.564 qs_ot_get_derivative 108 11.5 0.001 0.002 7.066 7.078 multiply_cannon_multrec 18288 15.5 1.946 2.891 6.750 7.012 rebuild_ks_matrix 119 8.3 0.000 0.000 6.400 6.416 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.399 6.416 dbcsr_mm_accdrv_process 38222 16.0 3.900 5.266 4.719 5.643 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.496 5.510 sum_up_and_integrate 119 10.3 0.018 0.018 4.006 4.033 integrate_v_rspace 119 11.3 0.002 0.003 3.989 4.017 mp_waitall_1 169478 16.3 2.715 3.812 2.715 3.812 init_scf_run 11 5.9 0.000 0.001 3.430 3.430 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.429 3.430 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.652 3.286 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.168 3.175 calculate_rho_elec 119 8.7 0.030 0.031 3.167 3.175 qs_ot_get_p 119 10.4 0.004 0.027 3.009 3.028 init_scf_loop 11 6.9 0.001 0.002 2.474 2.477 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.362 2.357 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.035 2.352 apply_single 119 13.6 0.000 0.000 2.035 2.352 rs_pw_transfer 974 11.9 0.009 0.010 2.225 2.336 calculate_first_density_matrix 1 7.0 0.001 0.002 2.252 2.253 qs_ot_p2m_diag 50 11.0 0.012 0.013 1.961 1.967 grid_integrate_task_list 119 12.3 1.799 1.909 1.799 1.909 make_m2s 4572 13.5 0.044 0.046 1.756 1.903 density_rs2pw 119 9.7 0.003 0.004 1.762 1.886 jit_kernel_multiply 10 15.8 0.767 1.874 0.767 1.874 potential_pw2rs 119 12.3 0.007 0.008 1.809 1.835 pw_transfer 1439 11.6 0.063 0.066 1.804 1.832 make_images 4572 14.5 0.191 0.206 1.670 1.815 calculate_dm_sparse 119 9.5 0.001 0.001 1.796 1.807 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.714 1.743 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.704 1.704 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.693 1.695 prepare_preconditioner 11 7.9 0.000 0.000 1.662 1.664 make_preconditioner 11 8.9 0.000 0.001 1.662 1.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.519 1.600 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.560 1.566 multiply_cannon_sync_h2d 18288 15.5 1.395 1.565 1.395 1.565 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.514 1.525 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.493 1.494 cp_fm_redistribute_end 50 14.0 1.110 1.475 1.111 1.475 ot_diis_step 108 11.5 0.012 0.013 1.462 1.462 cp_fm_diag_elpa_base 50 14.0 0.349 1.419 0.363 1.456 fft_wrap_pw1pw2_140 487 13.2 0.088 0.092 1.410 1.440 fft3d_ps 1201 14.6 0.506 0.522 1.399 1.429 mp_sum_l 7207 12.9 1.059 1.396 1.059 1.396 grid_collocate_task_list 119 9.7 1.204 1.310 1.204 1.310 acc_transpose_blocks 18288 15.5 0.075 0.078 1.238 1.260 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.174 1.178 wfi_extrapolate 11 7.9 0.001 0.001 1.132 1.132 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.978 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.972 0.974 make_images_data 4572 15.5 0.044 0.048 0.797 0.970 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.687 0.893 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.852 0.875 mp_alltoall_z22v 1201 16.6 0.746 0.823 0.746 0.823 acc_transpose_blocks_kernels 18288 16.5 0.209 0.216 0.799 0.812 mp_alltoall_d11v 2130 13.8 0.660 0.763 0.660 0.763 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.657 0.742 cp_fm_cholesky_invert 11 10.9 0.692 0.696 0.692 0.696 rs_pw_transfer_RS2PW_140 130 11.5 0.175 0.181 0.576 0.674 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.492000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=493.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 555.044864E+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 0.012 0.031 34.699 34.700 qs_mol_dyn_low 1 2.0 0.003 0.003 34.470 34.478 qs_forces 11 3.9 0.002 0.002 34.365 34.366 qs_energies 11 4.9 0.001 0.001 32.669 32.676 scf_env_do_scf 11 5.9 0.000 0.001 27.045 27.046 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.695 23.695 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.507 18.635 velocity_verlet 10 3.0 0.002 0.002 17.369 17.373 qs_scf_new_mos 108 7.5 0.001 0.001 15.687 15.748 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.686 15.747 multiply_cannon 2286 13.5 0.232 0.265 15.059 15.474 ot_scf_mini 108 9.5 0.002 0.003 14.763 14.817 multiply_cannon_loop 2286 14.5 0.943 0.975 14.137 14.437 multiply_cannon_multrec 27432 15.5 2.371 3.087 9.231 9.632 ot_mini 108 10.5 0.001 0.001 9.064 9.134 dbcsr_mm_accdrv_process 47916 15.9 5.692 7.719 6.766 8.007 qs_ot_get_derivative 108 11.5 0.001 0.001 7.271 7.327 rebuild_ks_matrix 119 8.3 0.000 0.000 6.386 6.442 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.386 6.441 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.669 5.717 init_scf_run 11 5.9 0.000 0.001 4.311 4.311 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.311 4.311 sum_up_and_integrate 119 10.3 0.020 0.022 3.672 3.679 integrate_v_rspace 119 11.3 0.002 0.002 3.651 3.659 init_scf_loop 11 6.9 0.000 0.000 3.332 3.332 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.222 3.251 calculate_rho_elec 119 8.7 0.040 0.046 3.222 3.250 qs_ot_get_p 119 10.4 0.001 0.001 3.110 3.195 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.715 3.110 calculate_first_density_matrix 1 7.0 0.000 0.001 2.978 2.980 prepare_preconditioner 11 7.9 0.000 0.000 2.509 2.518 make_preconditioner 11 8.9 0.000 0.000 2.509 2.518 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.480 2.481 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.113 2.445 mp_waitall_1 145218 16.4 1.947 2.432 1.947 2.432 jit_kernel_multiply 10 15.8 1.012 2.413 1.012 2.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.318 2.331 make_m2s 4572 13.5 0.054 0.056 2.127 2.259 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.015 2.157 apply_single 119 13.6 0.000 0.000 2.014 2.157 make_images 4572 14.5 0.274 0.336 2.019 2.149 calculate_dm_sparse 119 9.5 0.000 0.000 2.012 2.068 rs_pw_transfer 974 11.9 0.009 0.010 1.923 2.006 grid_integrate_task_list 119 12.3 1.817 1.903 1.817 1.903 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.894 1.903 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.863 1.892 density_rs2pw 119 9.7 0.003 0.004 1.712 1.794 ot_diis_step 108 11.5 0.012 0.012 1.753 1.754 pw_transfer 1439 11.6 0.063 0.066 1.709 1.739 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.621 1.653 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.586 1.586 acc_transpose_blocks 27432 15.5 0.113 0.116 1.469 1.493 potential_pw2rs 119 12.3 0.009 0.009 1.483 1.485 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.341 1.376 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.351 1.351 cp_fm_redistribute_end 50 14.0 0.891 1.330 0.892 1.330 fft3d_ps 1201 14.6 0.534 0.586 1.299 1.324 cp_fm_diag_elpa_base 50 14.0 0.416 1.271 0.436 1.309 grid_collocate_task_list 119 9.7 1.222 1.299 1.222 1.299 wfi_extrapolate 11 7.9 0.001 0.001 1.287 1.287 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.772 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.205 1.215 mp_sum_l 7207 12.9 0.895 1.170 0.895 1.170 cp_fm_upper_to_full 72 13.5 0.811 1.125 0.811 1.125 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.086 1.108 multiply_cannon_sync_h2d 27432 15.5 1.003 1.075 1.003 1.075 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.064 1.065 dbcsr_complete_redistribute 329 12.2 0.117 0.140 0.743 1.018 make_images_data 4572 15.5 0.044 0.048 0.811 0.940 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.693 0.887 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.874 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.842 0.867 mp_alltoall_d11v 2130 13.8 0.704 0.843 0.704 0.843 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.553 0.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.810 mp_alltoall_i22 627 13.8 0.428 0.724 0.428 0.724 cp_fm_cholesky_invert 11 10.9 0.717 0.720 0.717 0.720 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.700000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.727273, yerr=4.412604 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 605.413376E+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 0.035 0.046 27.545 27.546 qs_mol_dyn_low 1 2.0 0.003 0.003 27.297 27.304 qs_forces 11 3.9 0.002 0.002 27.239 27.240 qs_energies 11 4.9 0.001 0.001 25.556 25.558 scf_env_do_scf 11 5.9 0.000 0.001 20.352 20.352 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 17.905 17.906 velocity_verlet 10 3.0 0.002 0.002 13.931 13.934 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.407 12.480 qs_scf_new_mos 108 7.5 0.001 0.001 10.461 10.484 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.461 10.483 multiply_cannon 2286 13.5 0.231 0.239 9.978 10.429 ot_scf_mini 108 9.5 0.002 0.002 9.799 9.825 multiply_cannon_loop 2286 14.5 0.332 0.344 9.088 9.268 multiply_cannon_multrec 9144 15.5 1.568 1.818 6.210 6.505 rebuild_ks_matrix 119 8.3 0.000 0.000 5.736 5.757 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.736 5.756 ot_mini 108 10.5 0.001 0.001 5.464 5.496 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.108 5.126 dbcsr_mm_accdrv_process 12550 15.8 3.142 4.300 4.541 4.719 qs_ot_get_derivative 108 11.5 0.001 0.001 4.194 4.219 init_scf_run 11 5.9 0.000 0.001 3.789 3.789 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.789 3.789 sum_up_and_integrate 119 10.3 0.022 0.023 3.492 3.496 integrate_v_rspace 119 11.3 0.003 0.003 3.469 3.474 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.147 3.152 calculate_rho_elec 119 8.7 0.059 0.061 3.147 3.152 calculate_first_density_matrix 1 7.0 0.001 0.001 2.694 2.695 qs_ot_get_p 119 10.4 0.001 0.001 2.581 2.629 init_scf_loop 11 6.9 0.000 0.000 2.427 2.427 jit_kernel_multiply 10 15.8 1.361 2.365 1.361 2.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.123 2.124 mp_waitall_1 121218 16.5 1.547 2.107 1.547 2.107 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.977 1.985 grid_integrate_task_list 119 12.3 1.848 1.917 1.848 1.917 make_m2s 4572 13.5 0.034 0.036 1.676 1.843 make_images 4572 14.5 0.268 0.303 1.586 1.751 prepare_preconditioner 11 7.9 0.000 0.000 1.710 1.714 make_preconditioner 11 8.9 0.000 0.000 1.710 1.714 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.703 1.706 calculate_dm_sparse 119 9.5 0.000 0.000 1.685 1.702 rs_pw_transfer 974 11.9 0.008 0.009 1.616 1.670 pw_transfer 1439 11.6 0.064 0.066 1.660 1.667 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.600 1.626 density_rs2pw 119 9.7 0.003 0.003 1.569 1.621 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.570 1.577 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.514 1.514 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.475 1.489 grid_collocate_task_list 119 9.7 1.271 1.343 1.271 1.343 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.291 1.300 potential_pw2rs 119 12.3 0.010 0.011 1.296 1.298 ot_diis_step 108 11.5 0.013 0.013 1.259 1.259 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.245 1.246 fft3d_ps 1201 14.6 0.537 0.549 1.234 1.241 cp_fm_redistribute_end 50 14.0 0.619 1.228 0.620 1.228 cp_fm_diag_elpa_base 50 14.0 0.564 1.161 0.607 1.216 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.209 1.210 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.144 1.167 apply_single 119 13.6 0.000 0.000 1.144 1.167 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.116 1.121 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.093 1.106 wfi_extrapolate 11 7.9 0.001 0.001 1.051 1.051 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.760 1.013 make_images_data 4572 15.5 0.038 0.042 0.776 0.974 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 acc_transpose_blocks 9144 15.5 0.039 0.040 0.900 0.909 mp_alltoall_d11v 2130 13.8 0.776 0.856 0.776 0.856 cp_fm_cholesky_invert 11 10.9 0.841 0.844 0.841 0.844 multiply_cannon_sync_h2d 9144 15.5 0.704 0.794 0.704 0.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.748 0.750 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.707 0.715 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.663 0.714 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.340 0.687 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.658 0.663 mp_allgather_i34 2286 14.5 0.224 0.620 0.224 0.620 mp_alltoall_z22v 1201 16.6 0.571 0.601 0.571 0.601 qs_create_task_list 11 7.9 0.001 0.001 0.542 0.566 generate_qs_task_list 11 8.9 0.190 0.212 0.541 0.566 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.311 0.559 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.546000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.363636, yerr=8.844852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 746.242048E+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 0.081 0.110 41.730 41.731 qs_mol_dyn_low 1 2.0 0.003 0.003 41.455 41.463 qs_forces 11 3.9 0.002 0.002 39.530 39.531 qs_energies 11 4.9 0.001 0.002 37.613 37.617 scf_env_do_scf 11 5.9 0.001 0.001 31.481 31.481 velocity_verlet 10 3.0 0.002 0.002 23.830 24.069 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.841 23.841 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.447 17.460 qs_scf_new_mos 108 7.5 0.001 0.001 15.239 15.331 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.238 15.330 multiply_cannon 2286 13.5 0.300 0.307 13.832 14.574 ot_scf_mini 108 9.5 0.002 0.002 14.179 14.273 multiply_cannon_loop 2286 14.5 0.344 0.349 12.620 13.397 multiply_cannon_multrec 9144 15.5 3.716 5.342 8.959 9.106 ot_mini 108 10.5 0.001 0.001 8.523 8.635 init_scf_loop 11 6.9 0.000 0.000 7.615 7.616 rebuild_ks_matrix 119 8.3 0.000 0.000 6.676 6.811 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.676 6.810 prepare_preconditioner 11 7.9 0.000 0.000 6.719 6.732 make_preconditioner 11 8.9 0.000 0.000 6.719 6.731 qs_ot_get_derivative 108 11.5 0.001 0.001 6.544 6.640 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.275 6.615 dbcsr_mm_accdrv_process 12550 15.8 4.061 5.591 5.119 6.560 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.036 6.158 cp_fm_upper_to_full 72 14.2 3.162 4.555 3.162 4.555 init_scf_run 11 5.9 0.000 0.001 4.099 4.099 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.099 4.099 sum_up_and_integrate 119 10.3 0.038 0.038 3.547 3.554 integrate_v_rspace 119 11.3 0.003 0.003 3.509 3.516 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.506 3.513 calculate_rho_elec 119 8.7 0.118 0.121 3.505 3.512 mp_waitall_1 97218 16.6 2.232 3.225 2.232 3.225 qs_ot_get_p 119 10.4 0.001 0.001 2.978 3.110 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.374 2.801 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.959 2.791 calculate_first_density_matrix 1 7.0 0.001 0.001 2.770 2.771 mp_alltoall_i22 627 13.8 1.634 2.636 1.634 2.636 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.657 2.476 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.429 2.429 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.149 2.406 apply_single 119 13.6 0.000 0.000 2.149 2.405 make_m2s 4572 13.5 0.038 0.038 2.116 2.259 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.439 2.251 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.993 2.174 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.160 make_images 4572 14.5 0.359 0.387 1.995 2.138 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.209 2.126 grid_integrate_task_list 119 12.3 1.990 2.003 1.990 2.003 ot_diis_step 108 11.5 0.014 0.014 1.954 1.955 pw_transfer 1439 11.6 0.067 0.068 1.949 1.953 mp_sum_dm 438 4.9 1.689 1.941 1.689 1.941 md_write_output 11 3.9 0.005 0.036 0.241 1.914 md_output 10 3.0 0.000 0.000 0.241 1.911 update_particle_set 20 4.0 0.000 0.000 1.670 1.909 write_trajectory 44 4.9 0.001 0.006 0.235 1.877 write_particle_coordinates 11 5.9 0.234 1.871 0.234 1.871 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.854 1.858 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.787 1.788 qs_energies_init_hamiltonians 11 5.9 0.004 0.006 1.777 1.778 density_rs2pw 119 9.7 0.003 0.003 1.707 1.723 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.630 1.680 mp_sum_l 7207 12.9 0.960 1.653 0.960 1.653 jit_kernel_multiply 8 15.5 1.031 1.649 1.031 1.649 fft_wrap_pw1pw2_140 487 13.2 0.094 0.095 1.554 1.558 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.514 1.514 fft3d_ps 1201 14.6 0.567 0.578 1.480 1.484 grid_collocate_task_list 119 9.7 1.435 1.449 1.435 1.449 cp_fm_cholesky_invert 11 10.9 1.378 1.382 1.378 1.382 rs_pw_transfer 974 11.9 0.009 0.009 1.333 1.371 wfi_extrapolate 11 7.9 0.001 0.001 1.276 1.276 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.261 1.262 cp_fm_diag_elpa_base 50 14.0 1.118 1.170 1.259 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.225 1.234 hybrid_alltoall_any 4725 16.4 0.087 0.148 0.967 1.158 potential_pw2rs 119 12.3 0.014 0.014 1.157 1.158 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.113 1.133 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.115 1.132 make_images_data 4572 15.5 0.042 0.045 0.954 1.130 multiply_cannon_sync_h2d 9144 15.5 1.039 1.043 1.039 1.043 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.030 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.827 1.024 qs_create_task_list 11 7.9 0.017 0.028 0.965 0.979 generate_qs_task_list 11 8.9 0.372 0.393 0.948 0.978 acc_transpose_blocks 9144 15.5 0.039 0.039 0.913 0.935 mp_alltoall_d11v 2130 13.8 0.916 0.935 0.916 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.731000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=702.000000, yerr=12.982506 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 500.121600E+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 1666469. 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 0.038 0.041 84.066 84.067 qs_mol_dyn_low 1 2.0 0.003 0.003 83.708 83.718 qs_forces 11 3.9 0.003 0.003 83.640 83.641 qs_energies 11 4.9 0.002 0.005 80.735 80.749 scf_env_do_scf 11 5.9 0.000 0.001 71.578 71.581 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.933 65.934 dbcsr_multiply_generic 2055 12.4 0.106 0.109 52.759 53.108 qs_scf_new_mos 99 7.5 0.000 0.001 48.434 48.535 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.433 48.535 ot_scf_mini 99 9.5 0.002 0.002 46.009 46.121 multiply_cannon 2055 13.4 0.178 0.181 42.969 43.862 multiply_cannon_loop 2055 14.4 1.570 1.614 41.936 42.923 velocity_verlet 10 3.0 0.001 0.002 42.708 42.709 ot_mini 99 10.5 0.001 0.001 27.964 28.063 qs_ot_get_derivative 99 11.5 0.001 0.001 21.111 21.227 multiply_cannon_multrec 49320 15.4 12.146 12.947 17.332 18.288 rebuild_ks_matrix 110 8.3 0.000 0.000 14.814 14.919 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.814 14.919 mp_waitall_1 241148 16.1 12.153 13.132 12.153 13.132 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.981 13.081 multiply_cannon_sync_h2d 49320 15.4 9.978 10.546 9.978 10.546 qs_ot_get_p 110 10.4 0.001 0.001 9.441 9.602 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.032 8.048 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.169 7.732 apply_single 110 13.6 0.000 0.001 7.169 7.732 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.193 7.708 sum_up_and_integrate 110 10.3 0.023 0.028 7.332 7.347 integrate_v_rspace 110 11.3 0.002 0.003 7.309 7.328 init_scf_run 11 5.9 0.000 0.001 7.118 7.118 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.117 7.118 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.876 6.928 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.522 6.683 calculate_rho_elec 110 8.6 0.020 0.024 6.521 6.683 ot_diis_step 99 11.5 0.006 0.006 6.553 6.553 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.106 6.131 mp_sum_l 6514 12.8 5.351 6.127 5.351 6.127 init_scf_loop 11 6.9 0.000 0.000 5.613 5.614 dbcsr_mm_accdrv_process 87628 16.1 2.059 2.174 5.065 5.475 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.061 5.061 rs_pw_transfer 902 11.9 0.012 0.014 4.352 4.578 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.556 4.558 cp_fm_redistribute_end 48 14.0 3.965 4.527 3.970 4.528 cp_fm_diag_elpa_base 48 14.0 0.552 4.427 0.556 4.456 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.763 4.186 make_m2s 4110 13.4 0.060 0.065 3.979 4.125 wfi_extrapolate 11 7.9 0.001 0.001 4.061 4.061 make_images 4110 14.4 0.178 0.193 3.883 4.033 calculate_dm_sparse 110 9.5 0.000 0.001 3.786 3.913 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.908 3.912 density_rs2pw 110 9.6 0.004 0.004 3.562 3.781 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.564 3.604 prepare_preconditioner 11 7.9 0.000 0.000 3.430 3.454 make_preconditioner 11 8.9 0.000 0.000 3.430 3.454 grid_integrate_task_list 110 12.3 3.249 3.426 3.249 3.426 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.323 3.380 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.214 3.261 pw_transfer 1331 11.6 0.054 0.065 2.919 2.994 calculate_first_density_matrix 1 7.0 0.004 0.005 2.967 2.972 potential_pw2rs 110 12.3 0.006 0.006 2.928 2.959 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.827 2.900 jit_kernel_multiply 13 15.9 2.723 2.849 2.723 2.849 mp_alltoall_d11v 2046 13.8 2.176 2.565 2.176 2.565 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.360 2.437 fft3d_ps 1111 14.6 0.751 0.840 2.366 2.429 acc_transpose_blocks 49320 15.4 0.214 0.227 2.186 2.275 grid_collocate_task_list 110 9.6 2.088 2.214 2.088 2.214 mp_waitany 14300 13.8 1.734 2.005 1.734 2.005 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.971 1.989 make_images_data 4110 15.4 0.043 0.047 1.818 1.962 mp_sum_d 3883 11.9 1.473 1.914 1.473 1.914 cp_fm_cholesky_invert 11 10.9 1.884 1.888 1.884 1.888 hybrid_alltoall_any 4261 16.3 0.082 0.478 1.576 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.067000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.181818, yerr=2.328409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.829824E+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 1985669. 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 0.067 0.105 68.766 68.766 qs_mol_dyn_low 1 2.0 0.003 0.003 68.222 68.231 qs_forces 11 3.9 0.002 0.002 68.154 68.155 qs_energies 11 4.9 0.002 0.006 64.810 64.813 scf_env_do_scf 11 5.9 0.001 0.001 56.181 56.185 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.529 48.530 dbcsr_multiply_generic 2055 12.4 0.112 0.115 37.547 37.762 velocity_verlet 10 3.0 0.001 0.002 36.170 36.172 qs_scf_new_mos 99 7.5 0.001 0.001 32.407 32.539 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.406 32.538 multiply_cannon 2055 13.4 0.221 0.242 30.994 32.250 ot_scf_mini 99 9.5 0.003 0.003 30.755 30.881 multiply_cannon_loop 2055 14.4 0.937 0.960 29.676 30.575 ot_mini 99 10.5 0.001 0.001 18.131 18.264 multiply_cannon_multrec 24660 15.4 7.665 9.077 13.879 15.417 rebuild_ks_matrix 110 8.3 0.000 0.001 13.791 13.901 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.791 13.901 qs_ot_get_derivative 99 11.5 0.001 0.001 12.345 12.476 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.147 12.243 mp_waitall_1 186928 16.3 7.798 10.001 7.798 10.001 multiply_cannon_sync_h2d 24660 15.4 7.039 8.035 7.039 8.035 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.082 7.732 init_scf_loop 11 6.9 0.000 0.002 7.613 7.615 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.487 7.227 apply_single 110 13.6 0.000 0.001 6.487 7.227 sum_up_and_integrate 110 10.3 0.028 0.033 6.688 6.698 integrate_v_rspace 110 11.3 0.002 0.003 6.660 6.672 dbcsr_mm_accdrv_process 52282 16.1 4.768 5.832 6.053 6.332 init_scf_run 11 5.9 0.000 0.001 6.237 6.238 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.237 6.238 qs_ot_get_p 110 10.4 0.001 0.001 5.990 6.181 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.930 5.940 calculate_rho_elec 110 8.6 0.039 0.047 5.929 5.940 ot_diis_step 99 11.5 0.010 0.011 5.738 5.738 prepare_preconditioner 11 7.9 0.000 0.000 5.577 5.595 make_preconditioner 11 8.9 0.000 0.001 5.577 5.595 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.650 5.340 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.161 5.313 make_m2s 4110 13.4 0.057 0.060 4.160 4.629 make_images 4110 14.4 0.399 0.444 4.051 4.517 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.101 4.123 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.674 3.674 wfi_extrapolate 11 7.9 0.001 0.001 3.527 3.527 pw_transfer 1331 11.6 0.067 0.074 3.346 3.486 rs_pw_transfer 902 11.9 0.012 0.014 3.278 3.467 density_rs2pw 110 9.6 0.004 0.004 3.220 3.411 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.235 3.376 grid_integrate_task_list 110 12.3 3.151 3.369 3.151 3.369 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.308 3.310 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.176 3.247 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.130 3.132 cp_fm_redistribute_end 48 14.0 2.333 3.102 2.336 3.104 cp_fm_diag_elpa_base 48 14.0 0.734 2.975 0.765 3.062 calculate_dm_sparse 110 9.5 0.001 0.001 2.969 3.002 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.953 2.987 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 2.685 2.826 make_images_data 4110 15.4 0.047 0.051 2.319 2.791 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.078 2.742 fft3d_ps 1111 14.6 1.069 1.265 2.591 2.715 calculate_first_density_matrix 1 7.0 0.001 0.009 2.624 2.626 cp_fm_cholesky_invert 11 10.9 2.591 2.598 2.591 2.598 potential_pw2rs 110 12.3 0.009 0.009 2.567 2.579 mp_sum_l 6514 12.8 1.778 2.399 1.778 2.399 grid_collocate_task_list 110 9.6 2.122 2.252 2.122 2.252 mp_alltoall_d11v 2046 13.8 1.798 2.010 1.798 2.010 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.945 1.969 jit_kernel_multiply 10 16.3 0.932 1.923 0.932 1.923 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.880 1.881 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.808 1.821 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.778 1.666 mp_allgather_i34 2055 14.4 0.631 1.630 0.631 1.630 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.560 acc_transpose_blocks 24660 15.4 0.108 0.111 1.531 1.555 mp_irecv_dv 57340 16.2 0.654 1.550 0.654 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.477 dbcsr_complete_redistribute 325 12.2 0.244 0.299 1.160 1.424 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.766000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.181818, yerr=7.720318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.394560E+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 0.066 0.221 60.801 60.801 qs_mol_dyn_low 1 2.0 0.004 0.008 60.067 60.077 qs_forces 11 3.9 0.004 0.020 59.989 59.999 qs_energies 11 4.9 0.003 0.008 56.789 56.805 scf_env_do_scf 11 5.9 0.001 0.001 48.313 48.313 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.685 39.686 velocity_verlet 10 3.0 0.001 0.002 32.525 32.527 dbcsr_multiply_generic 2055 12.4 0.116 0.124 28.562 28.841 qs_scf_new_mos 99 7.5 0.001 0.001 25.013 25.099 qs_scf_loop_do_ot 99 8.5 0.001 0.002 25.012 25.099 ot_scf_mini 99 9.5 0.003 0.004 23.795 23.900 multiply_cannon 2055 13.4 0.212 0.227 22.248 23.427 multiply_cannon_loop 2055 14.4 0.622 0.636 21.070 22.092 ot_mini 99 10.5 0.001 0.002 13.456 13.563 rebuild_ks_matrix 110 8.3 0.000 0.000 12.274 12.407 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.274 12.407 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.838 10.956 multiply_cannon_multrec 16440 15.4 3.953 4.840 9.944 10.747 mp_waitall_1 146766 16.3 7.137 10.256 7.137 10.256 qs_ot_get_derivative 99 11.5 0.001 0.001 9.055 9.161 init_scf_loop 11 6.9 0.002 0.009 8.590 8.593 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.228 7.232 prepare_preconditioner 11 7.9 0.000 0.000 6.796 6.813 make_preconditioner 11 8.9 0.000 0.001 6.796 6.813 sum_up_and_integrate 110 10.3 0.033 0.034 6.481 6.497 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.135 6.483 integrate_v_rspace 110 11.3 0.002 0.003 6.447 6.464 dbcsr_mm_accdrv_process 34862 16.1 5.120 5.508 5.844 5.951 init_scf_run 11 5.9 0.000 0.001 5.787 5.788 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.787 5.787 qs_ot_get_p 110 10.4 0.001 0.001 5.521 5.652 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.634 5.642 calculate_rho_elec 110 8.6 0.058 0.058 5.633 5.641 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.915 5.339 apply_single 110 13.6 0.000 0.000 4.914 5.339 make_m2s 4110 13.4 0.050 0.052 4.241 4.597 make_images 4110 14.4 0.392 0.507 4.125 4.479 ot_diis_step 99 11.5 0.011 0.014 4.374 4.375 multiply_cannon_sync_h2d 16440 15.4 3.695 4.203 3.695 4.203 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.876 3.880 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.117 3.757 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.539 3.539 grid_integrate_task_list 110 12.3 3.169 3.406 3.169 3.406 rs_pw_transfer 902 11.9 0.010 0.011 3.132 3.360 pw_transfer 1331 11.6 0.066 0.073 3.188 3.197 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.076 3.089 density_rs2pw 110 9.6 0.004 0.004 2.914 3.076 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.041 3.043 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.003 3.005 cp_fm_redistribute_end 48 14.0 1.866 2.974 1.868 2.974 wfi_extrapolate 11 7.9 0.001 0.001 2.965 2.965 make_images_data 4110 15.4 0.043 0.047 2.465 2.946 cp_fm_diag_elpa_base 48 14.0 1.044 2.839 1.102 2.944 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.201 2.844 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.678 2.738 calculate_first_density_matrix 1 7.0 0.001 0.001 2.737 2.737 cp_fm_cholesky_invert 11 10.9 2.687 2.693 2.687 2.693 fft_wrap_pw1pw2_140 451 13.1 0.214 0.219 2.619 2.635 calculate_dm_sparse 110 9.5 0.001 0.002 2.545 2.575 fft3d_ps 1111 14.6 1.061 1.068 2.387 2.398 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.310 2.380 grid_collocate_task_list 110 9.6 2.177 2.361 2.177 2.361 potential_pw2rs 110 12.3 0.011 0.011 2.331 2.343 mp_alltoall_d11v 2046 13.8 1.804 2.338 1.804 2.338 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.865 2.301 qs_energies_init_hamiltonians 11 5.9 0.007 0.013 2.244 2.257 mp_irecv_dv 48980 15.7 0.795 2.178 0.795 2.178 mp_sum_l 6514 12.8 1.453 2.056 1.453 2.056 dbcsr_complete_redistribute 325 12.2 0.334 0.383 1.504 1.956 cp_fm_upper_to_full 70 13.6 1.381 1.853 1.381 1.853 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.659 cp_fm_cholesky_decompose 22 10.9 1.610 1.628 1.610 1.628 mp_allgather_i34 2055 14.4 0.502 1.622 0.502 1.622 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.510 1.518 mp_waitany 17072 13.8 1.335 1.514 1.335 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.482 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.013 1.449 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.329 1.440 rs_gather_matrices 110 12.3 0.137 0.151 0.897 1.412 rs_pw_transfer_RS2PW_140 121 11.5 0.260 0.267 1.139 1.354 jit_kernel_multiply 6 16.8 0.335 1.337 0.335 1.337 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.271 1.279 acc_transpose_blocks 16440 15.4 0.073 0.076 1.229 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.801000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=8.214078 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 730.226688E+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 0.033 0.099 64.996 64.997 qs_mol_dyn_low 1 2.0 0.004 0.005 64.659 64.667 qs_forces 11 3.9 0.003 0.007 64.587 64.589 qs_energies 11 4.9 0.002 0.007 61.199 61.204 scf_env_do_scf 11 5.9 0.001 0.001 52.450 52.453 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.676 40.677 velocity_verlet 10 3.0 0.002 0.002 36.675 36.678 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.814 30.029 qs_scf_new_mos 99 7.5 0.001 0.001 26.177 26.282 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.176 26.281 ot_scf_mini 99 9.5 0.003 0.004 24.558 24.658 multiply_cannon 2055 13.4 0.243 0.263 22.857 24.133 multiply_cannon_loop 2055 14.4 0.885 0.906 21.495 22.079 ot_mini 99 10.5 0.001 0.001 14.016 14.141 multiply_cannon_multrec 24660 15.4 4.239 6.877 12.696 13.899 rebuild_ks_matrix 110 8.3 0.000 0.000 11.985 12.090 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 11.985 12.090 init_scf_loop 11 6.9 0.001 0.004 11.731 11.733 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.599 10.691 prepare_preconditioner 11 7.9 0.000 0.000 9.981 9.999 make_preconditioner 11 8.9 0.000 0.001 9.981 9.999 qs_ot_get_derivative 99 11.5 0.001 0.002 9.882 9.988 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.168 9.646 dbcsr_mm_accdrv_process 52304 16.0 7.100 8.478 8.309 9.246 mp_waitall_1 126806 16.4 4.450 6.424 4.450 6.424 sum_up_and_integrate 110 10.3 0.038 0.041 6.355 6.367 integrate_v_rspace 110 11.3 0.003 0.006 6.317 6.329 init_scf_run 11 5.9 0.000 0.001 5.968 5.968 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.967 5.968 qs_ot_get_p 110 10.4 0.001 0.005 5.574 5.732 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.695 5.703 calculate_rho_elec 110 8.6 0.077 0.081 5.694 5.703 make_m2s 4110 13.4 0.059 0.061 5.383 5.673 make_images 4110 14.4 0.578 0.696 5.243 5.527 cp_fm_upper_to_full 70 13.8 3.347 4.758 3.347 4.758 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 3.999 4.097 apply_single 110 13.6 0.000 0.000 3.999 4.097 ot_diis_step 99 11.5 0.012 0.013 4.097 4.097 dbcsr_complete_redistribute 325 12.2 0.418 0.461 2.701 3.849 qs_ot_p2m_diag 48 11.0 0.055 0.065 3.791 3.805 grid_integrate_task_list 110 12.3 3.268 3.448 3.268 3.448 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.379 3.392 multiply_cannon_sync_h2d 24660 15.4 3.198 3.352 3.198 3.352 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.315 3.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.300 3.302 pw_transfer 1331 11.6 0.066 0.073 3.271 3.300 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.160 3.298 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.243 3.295 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.160 3.193 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.010 3.068 density_rs2pw 110 9.6 0.004 0.004 2.864 3.037 make_images_data 4110 15.4 0.046 0.049 2.715 3.035 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.940 2.979 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.249 2.970 wfi_extrapolate 11 7.9 0.001 0.002 2.951 2.951 calculate_dm_sparse 110 9.5 0.001 0.001 2.914 2.945 calculate_first_density_matrix 1 7.0 0.001 0.004 2.926 2.928 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.804 2.923 mp_alltoall_i22 605 13.7 1.661 2.864 1.661 2.864 rs_pw_transfer 902 11.9 0.011 0.011 2.608 2.783 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.774 2.777 cp_fm_redistribute_end 48 14.0 1.382 2.747 1.384 2.748 cp_fm_diag_elpa_base 48 14.0 1.281 2.616 1.361 2.726 cp_fm_cholesky_invert 11 10.9 2.709 2.717 2.709 2.717 fft_wrap_pw1pw2_140 451 13.1 0.206 0.217 2.674 2.712 fft3d_ps 1111 14.6 1.061 1.092 2.458 2.476 grid_collocate_task_list 110 9.6 2.226 2.446 2.226 2.446 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.280 2.283 jit_kernel_multiply 12 16.0 0.880 2.189 0.880 2.189 potential_pw2rs 110 12.3 0.013 0.013 2.163 2.172 acc_transpose_blocks 24660 15.4 0.105 0.107 1.981 2.019 mp_alltoall_d11v 2046 13.8 1.788 2.003 1.788 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.681 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.699 cp_fm_cholesky_decompose 22 10.9 1.655 1.697 1.655 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.643 mp_allgather_i34 2055 14.4 0.464 1.557 0.464 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.546 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.827 1.502 mp_sum_l 6514 12.8 0.891 1.447 0.891 1.447 mp_irecv_dv 62702 16.1 0.728 1.425 0.728 1.425 acc_transpose_blocks_kernels 24660 16.4 0.302 0.308 1.340 1.372 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.230 1.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.997000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.272727, yerr=9.955272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 806.649856E+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 0.088 0.095 55.191 55.192 qs_mol_dyn_low 1 2.0 0.003 0.003 54.602 54.611 qs_forces 11 3.9 0.014 0.015 54.536 54.537 qs_energies 11 4.9 0.001 0.001 50.905 50.911 scf_env_do_scf 11 5.9 0.000 0.001 42.353 42.353 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.622 34.623 velocity_verlet 10 3.0 0.001 0.002 30.617 30.620 dbcsr_multiply_generic 2055 12.4 0.104 0.105 23.159 23.262 qs_scf_new_mos 99 7.5 0.001 0.001 20.471 20.523 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.471 20.522 ot_scf_mini 99 9.5 0.002 0.002 19.239 19.265 multiply_cannon 2055 13.4 0.246 0.263 17.713 19.088 multiply_cannon_loop 2055 14.4 0.324 0.336 16.350 16.613 rebuild_ks_matrix 110 8.3 0.000 0.000 11.537 11.569 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.536 11.569 ot_mini 99 10.5 0.001 0.001 10.361 10.380 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.276 10.306 multiply_cannon_multrec 8220 15.4 3.392 4.727 7.980 8.961 init_scf_loop 11 6.9 0.000 0.000 7.681 7.681 mp_waitall_1 106626 16.5 5.981 7.654 5.981 7.654 qs_ot_get_derivative 99 11.5 0.001 0.001 6.579 6.603 sum_up_and_integrate 110 10.3 0.048 0.049 6.186 6.199 integrate_v_rspace 110 11.3 0.003 0.003 6.138 6.150 prepare_preconditioner 11 7.9 0.000 0.000 6.017 6.022 make_preconditioner 11 8.9 0.000 0.000 6.017 6.022 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.613 5.683 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.639 5.654 calculate_rho_elec 110 8.6 0.115 0.115 5.639 5.654 dbcsr_mm_accdrv_process 17442 15.9 3.069 4.259 4.457 5.515 init_scf_run 11 5.9 0.000 0.001 5.432 5.432 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.432 5.432 qs_ot_get_p 110 10.4 0.001 0.001 4.959 4.986 make_m2s 4110 13.4 0.039 0.040 4.216 4.511 make_images 4110 14.4 0.634 0.688 4.086 4.380 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.881 4.247 ot_diis_step 99 11.5 0.012 0.012 3.763 3.763 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.681 3.704 apply_single 110 13.6 0.000 0.000 3.681 3.704 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.592 3.596 grid_integrate_task_list 110 12.3 3.369 3.509 3.369 3.509 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.291 3.291 pw_transfer 1331 11.6 0.066 0.071 3.241 3.250 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.130 3.143 multiply_cannon_sync_h2d 8220 15.4 2.923 3.015 2.923 3.015 cp_fm_cholesky_invert 11 10.9 2.925 2.929 2.925 2.929 make_images_data 4110 15.4 0.038 0.043 2.419 2.867 density_rs2pw 110 9.6 0.004 0.004 2.704 2.816 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.763 2.764 hybrid_alltoall_any 4261 16.3 0.199 0.857 2.323 2.740 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.736 2.737 cp_fm_redistribute_end 48 14.0 0.700 2.733 0.704 2.734 wfi_extrapolate 11 7.9 0.001 0.001 2.714 2.714 cp_fm_diag_elpa_base 48 14.0 1.849 2.537 2.022 2.701 fft_wrap_pw1pw2_140 451 13.1 0.217 0.221 2.680 2.697 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.667 2.669 calculate_dm_sparse 110 9.5 0.001 0.001 2.624 2.666 calculate_first_density_matrix 1 7.0 0.000 0.001 2.629 2.630 rs_pw_transfer 902 11.9 0.010 0.010 2.291 2.446 grid_collocate_task_list 110 9.6 2.327 2.424 2.327 2.424 fft3d_ps 1111 14.6 1.114 1.137 2.383 2.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.388 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.771 1.985 potential_pw2rs 110 12.3 0.015 0.015 1.977 1.983 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.897 1.911 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.767 1.780 mp_alltoall_d11v 2046 13.8 1.588 1.741 1.588 1.741 cp_fm_cholesky_decompose 22 10.9 1.684 1.704 1.684 1.704 mp_allgather_i34 2055 14.4 0.550 1.670 0.550 1.670 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.509 1.631 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.594 1.604 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.593 1.597 dbcsr_complete_redistribute 325 12.2 0.556 0.577 1.438 1.525 qs_create_task_list 11 7.9 0.001 0.001 1.223 1.324 generate_qs_task_list 11 8.9 0.379 0.446 1.222 1.323 jit_kernel_multiply 6 15.7 1.080 1.213 1.080 1.213 mp_waitany 9240 13.8 1.023 1.188 1.023 1.188 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.730 1.162 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.124 1.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.192000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=764.636364, yerr=8.282212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 1.226260E+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 0.055 0.065 85.319 85.319 qs_mol_dyn_low 1 2.0 0.003 0.003 84.805 84.813 qs_forces 11 3.9 0.002 0.002 84.739 84.740 qs_energies 11 4.9 0.001 0.001 80.676 80.677 scf_env_do_scf 11 5.9 0.001 0.001 70.469 70.469 velocity_verlet 10 3.0 0.002 0.002 54.358 54.364 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.180 42.180 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.608 28.677 init_scf_loop 11 6.9 0.000 0.000 28.215 28.216 prepare_preconditioner 11 7.9 0.000 0.000 26.275 26.282 make_preconditioner 11 8.9 0.000 0.000 26.275 26.282 qs_scf_new_mos 99 7.5 0.001 0.001 25.976 26.003 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.976 26.002 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.446 25.740 ot_scf_mini 99 9.5 0.002 0.002 24.260 24.291 multiply_cannon 2055 13.4 0.340 0.374 21.629 22.638 multiply_cannon_loop 2055 14.4 0.343 0.346 19.796 20.138 cp_fm_upper_to_full 70 14.2 12.718 18.253 12.718 18.253 ot_mini 99 10.5 0.001 0.001 13.446 13.479 rebuild_ks_matrix 110 8.3 0.001 0.001 13.118 13.156 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.117 13.156 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.928 11.964 dbcsr_complete_redistribute 325 12.2 1.021 1.046 7.419 10.761 multiply_cannon_multrec 8220 15.4 4.347 4.538 9.675 9.815 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.374 9.713 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.813 9.122 qs_ot_get_derivative 99 11.5 0.001 0.001 8.862 8.893 mp_alltoall_i22 605 13.7 5.425 8.777 5.425 8.777 mp_waitall_1 87304 16.6 7.822 8.731 7.822 8.731 sum_up_and_integrate 110 10.3 0.090 0.091 6.504 6.520 integrate_v_rspace 110 11.3 0.003 0.003 6.414 6.431 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.297 6.332 calculate_rho_elec 110 8.6 0.226 0.227 6.296 6.332 init_scf_run 11 5.9 0.000 0.001 5.944 5.944 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.944 5.944 make_m2s 4110 13.4 0.043 0.043 5.263 5.838 make_images 4110 14.4 0.878 0.933 5.076 5.650 qs_ot_get_p 110 10.4 0.001 0.001 5.607 5.639 dbcsr_mm_accdrv_process 11614 15.7 3.219 3.697 5.184 5.486 cp_fm_cholesky_invert 11 10.9 5.225 5.230 5.225 5.230 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.697 5.144 apply_single 110 13.6 0.000 0.000 4.697 5.144 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.761 5.098 ot_diis_step 99 11.5 0.015 0.016 4.564 4.564 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.116 4.122 multiply_cannon_sync_h2d 8220 15.4 3.946 3.952 3.946 3.952 grid_integrate_task_list 110 12.3 3.655 3.710 3.655 3.710 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.803 3.685 pw_transfer 1331 11.6 0.073 0.074 3.677 3.682 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.674 3.675 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.672 3.672 make_images_data 4110 15.4 0.041 0.044 2.917 3.631 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.553 3.558 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.813 3.253 calculate_dm_sparse 110 9.5 0.001 0.001 3.184 3.214 wfi_extrapolate 11 7.9 0.001 0.001 3.180 3.180 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.122 3.123 cp_fm_diag_elpa_base 48 14.0 2.595 2.789 3.120 3.121 fft_wrap_pw1pw2_140 451 13.1 0.218 0.219 3.079 3.085 density_rs2pw 110 9.6 0.004 0.004 2.942 2.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.947 2.950 fft3d_ps 1111 14.6 1.266 1.275 2.760 2.766 calculate_first_density_matrix 1 7.0 0.001 0.001 2.660 2.661 grid_collocate_task_list 110 9.6 2.626 2.649 2.626 2.649 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.492 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.384 2.400 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.179 2.228 rs_pw_transfer 902 11.9 0.011 0.011 2.154 2.190 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.177 potential_pw2rs 110 12.3 0.021 0.021 2.039 2.045 cp_fm_cholesky_decompose 22 10.9 1.959 1.980 1.959 1.980 jit_kernel_multiply 10 15.4 1.764 1.946 1.764 1.946 qs_create_task_list 11 7.9 0.000 0.001 1.901 1.942 generate_qs_task_list 11 8.9 0.740 0.796 1.901 1.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.795 mp_alltoall_d11v 2046 13.8 1.742 1.794 1.742 1.794 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.761 1.778 mp_allgather_i34 2055 14.4 0.624 1.764 0.624 1.764 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.676 1.720 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.319000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1128.090909, yerr=35.209644 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 625.831936E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 56839. MP_Allreduce 11238 786. MP_Sync 168 MP_Alltoall 2210 2534587. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.114 202.668 202.669 qs_mol_dyn_low 1 2.0 0.003 0.005 202.064 202.077 qs_forces 11 3.9 0.007 0.030 201.964 201.966 qs_energies 11 4.9 0.002 0.005 196.289 196.307 scf_env_do_scf 11 5.9 0.001 0.001 179.599 179.603 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 158.888 158.890 dbcsr_multiply_generic 2485 12.5 0.173 0.176 123.063 124.040 velocity_verlet 10 3.0 0.001 0.003 121.527 121.529 qs_scf_new_mos 116 7.6 0.001 0.001 119.979 120.269 qs_scf_loop_do_ot 116 8.6 0.001 0.001 119.978 120.268 ot_scf_mini 116 9.6 0.003 0.004 113.471 113.709 multiply_cannon 2485 13.5 0.235 0.244 99.957 102.272 multiply_cannon_loop 2485 14.5 2.109 2.179 97.801 100.126 ot_mini 116 10.6 0.001 0.002 64.964 65.225 multiply_cannon_multrec 59640 15.5 32.868 35.427 41.234 43.520 qs_ot_get_derivative 116 11.6 0.001 0.002 40.244 40.493 rebuild_ks_matrix 127 8.3 0.001 0.001 33.487 33.878 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.019 33.486 33.877 mp_waitall_1 288940 16.2 28.034 31.201 28.034 31.201 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.080 30.434 multiply_cannon_sync_h2d 59640 15.5 27.227 29.727 27.227 29.727 qs_ot_get_p 127 10.4 0.001 0.001 26.807 27.105 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.150 24.958 apply_single 127 13.6 0.001 0.001 24.149 24.958 ot_diis_step 116 11.6 0.008 0.010 24.389 24.390 init_scf_loop 11 6.9 0.000 0.001 20.635 20.636 qs_ot_p2m_diag 82 11.4 0.076 0.090 20.260 20.395 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.793 19.015 multiply_cannon_metrocomm3 59640 15.5 0.110 0.117 15.319 18.107 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.708 17.709 prepare_preconditioner 11 7.9 0.000 0.000 16.028 16.094 make_preconditioner 11 8.9 0.000 0.000 16.028 16.094 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.254 15.422 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.957 14.963 cp_fm_redistribute_end 82 14.4 11.810 14.887 11.824 14.893 cp_fm_diag_elpa_base 82 14.4 3.023 14.608 3.055 14.717 sum_up_and_integrate 127 10.3 0.052 0.064 14.180 14.195 integrate_v_rspace 127 11.3 0.003 0.004 14.127 14.143 make_m2s 4970 13.5 0.102 0.111 13.672 14.091 make_images 4970 14.5 0.398 0.416 13.493 13.919 qs_rho_update_rho_low 127 7.7 0.001 0.002 12.725 12.830 calculate_rho_elec 127 8.7 0.044 0.063 12.724 12.829 init_scf_run 11 5.9 0.000 0.001 12.439 12.439 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.438 12.439 mp_sum_l 7804 13.0 8.220 9.350 8.220 9.350 cp_fm_cholesky_invert 11 10.9 9.168 9.176 9.168 9.176 wfi_extrapolate 11 7.9 0.001 0.003 9.071 9.072 calculate_dm_sparse 127 9.5 0.001 0.001 8.385 8.521 dbcsr_mm_accdrv_process 123452 16.2 3.283 3.439 7.934 8.477 multiply_cannon_metrocomm1 59640 15.5 0.085 0.089 5.958 8.449 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.829 7.943 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.729 7.825 make_images_data 4970 15.5 0.068 0.077 6.756 7.729 density_rs2pw 127 9.7 0.006 0.006 6.568 7.538 grid_integrate_task_list 127 12.3 7.047 7.455 7.047 7.455 rs_pw_transfer 1038 11.9 0.017 0.019 6.343 7.361 hybrid_alltoall_any 5155 16.4 0.288 2.244 5.882 7.113 pw_transfer 1535 11.6 0.074 0.101 6.684 6.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.748 6.756 fft_wrap_pw1pw2 1281 12.7 0.010 0.014 6.471 6.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.884 5.984 fft_wrap_pw1pw2_140 519 13.2 0.436 0.481 5.612 5.797 fft3d_ps 1281 14.7 2.080 2.577 5.280 5.497 mp_alltoall_d11v 2401 14.1 4.204 5.394 4.204 5.394 grid_collocate_task_list 127 9.7 4.686 5.046 4.686 5.046 potential_pw2rs 127 12.3 0.009 0.010 4.763 4.782 cp_fm_cholesky_decompose 22 10.9 4.663 4.677 4.663 4.677 mp_sum_d 4443 12.1 3.703 4.574 3.703 4.574 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.669000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.818182, yerr=6.364935 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 824.868864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57173. MP_Allreduce 11257 946. MP_Sync 168 MP_Alltoall 1955 4872950. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.057 186.565 186.565 qs_mol_dyn_low 1 2.0 0.003 0.006 186.133 186.147 qs_forces 11 3.9 0.004 0.009 186.036 186.037 qs_energies 11 4.9 0.002 0.005 179.306 179.317 scf_env_do_scf 11 5.9 0.001 0.001 162.507 162.517 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 130.243 130.246 velocity_verlet 10 3.0 0.002 0.003 116.650 116.651 dbcsr_multiply_generic 2485 12.5 0.183 0.191 96.935 98.053 qs_scf_new_mos 116 7.6 0.001 0.001 92.066 92.466 qs_scf_loop_do_ot 116 8.6 0.001 0.002 92.065 92.466 ot_scf_mini 116 9.6 0.004 0.005 87.312 87.809 multiply_cannon 2485 13.5 0.471 0.524 77.370 81.868 multiply_cannon_loop 2485 14.5 1.255 1.302 74.119 76.502 ot_mini 116 10.6 0.001 0.002 49.079 49.505 mp_waitall_1 224796 16.4 24.281 38.223 24.281 38.223 multiply_cannon_multrec 29820 15.5 21.929 27.089 31.941 37.293 rebuild_ks_matrix 127 8.3 0.001 0.001 32.550 33.022 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.023 32.550 33.021 init_scf_loop 11 6.9 0.001 0.002 32.176 32.181 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.281 29.711 multiply_cannon_metrocomm3 29820 15.5 0.095 0.100 15.562 28.094 qs_ot_get_derivative 116 11.6 0.001 0.002 27.393 27.882 prepare_preconditioner 11 7.9 0.000 0.000 27.785 27.852 make_preconditioner 11 8.9 0.000 0.001 27.785 27.852 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.483 27.033 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 21.785 22.956 apply_single 127 13.6 0.001 0.001 21.785 22.955 multiply_cannon_sync_h2d 29820 15.5 19.189 22.876 19.189 22.876 ot_diis_step 116 11.6 0.015 0.017 21.516 21.518 qs_ot_get_p 127 10.4 0.001 0.002 20.211 20.783 cp_fm_cholesky_invert 11 10.9 16.235 16.247 16.235 16.247 make_m2s 4970 13.5 0.089 0.096 13.905 15.551 qs_ot_p2m_diag 82 11.4 0.185 0.213 15.495 15.528 make_images 4970 14.5 1.146 1.329 13.697 15.343 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.378 14.379 sum_up_and_integrate 127 10.3 0.061 0.070 14.147 14.178 integrate_v_rspace 127 11.3 0.003 0.004 14.085 14.119 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.751 12.785 calculate_rho_elec 127 8.7 0.087 0.105 12.751 12.784 init_scf_run 11 5.9 0.000 0.001 11.997 11.999 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.997 11.998 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.515 11.521 cp_fm_redistribute_end 82 14.4 6.714 11.462 6.726 11.463 cp_fm_diag_elpa_base 82 14.4 4.510 10.989 4.722 11.372 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.714 11.055 multiply_cannon_metrocomm4 27335 15.5 0.096 0.113 3.693 10.472 make_images_data 4970 15.5 0.065 0.073 8.422 10.378 dbcsr_mm_accdrv_process 61726 16.2 4.443 5.207 9.470 10.131 mp_irecv_dv 68888 16.3 3.500 10.082 3.500 10.082 hybrid_alltoall_any 5155 16.4 0.338 1.489 7.183 10.033 wfi_extrapolate 11 7.9 0.001 0.003 8.285 8.285 pw_transfer 1535 11.6 0.083 0.098 7.590 7.666 grid_integrate_task_list 127 12.3 7.100 7.481 7.100 7.481 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.362 7.439 density_rs2pw 127 9.7 0.006 0.006 6.614 7.068 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.161 6.970 cp_fm_cholesky_decompose 22 10.9 6.711 6.800 6.711 6.800 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.764 6.774 fft_wrap_pw1pw2_140 519 13.2 0.458 0.496 6.477 6.563 calculate_dm_sparse 127 9.5 0.001 0.002 6.401 6.537 rs_pw_transfer 1038 11.9 0.014 0.015 5.611 6.082 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.892 6.060 fft3d_ps 1281 14.7 2.753 2.925 5.752 5.805 mp_sum_l 7804 13.0 3.910 5.705 3.910 5.705 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.250 5.314 grid_collocate_task_list 127 9.7 4.869 5.246 4.869 5.246 mp_alltoall_d11v 2401 14.1 3.927 4.882 3.927 4.882 mp_allgather_i34 2485 14.5 1.843 4.821 1.843 4.821 potential_pw2rs 127 12.3 0.016 0.018 4.735 4.753 mp_sum_d 4447 12.1 2.582 3.895 2.582 3.895 dbcsr_complete_redistribute 393 12.7 0.776 0.898 2.952 3.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=186.565000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.909091, yerr=0.287480 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 2.906046E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 940.056576E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172869264 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.148 173.327 173.328 qs_mol_dyn_low 1 2.0 0.004 0.005 172.703 172.714 qs_forces 11 3.9 0.003 0.008 172.606 172.609 qs_energies 11 4.9 0.005 0.011 166.047 166.059 scf_env_do_scf 11 5.9 0.001 0.001 150.386 150.388 scf_env_do_scf_inner_loop 116 6.6 0.028 0.068 115.801 115.803 velocity_verlet 10 3.0 0.002 0.002 110.602 110.607 dbcsr_multiply_generic 2485 12.5 0.181 0.188 81.032 82.172 qs_scf_new_mos 116 7.6 0.001 0.001 79.476 79.846 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.475 79.845 ot_scf_mini 116 9.6 0.004 0.005 75.395 75.808 multiply_cannon 2485 13.5 0.496 0.531 61.395 66.594 multiply_cannon_loop 2485 14.5 0.857 0.886 58.453 61.114 ot_mini 116 10.6 0.001 0.002 42.049 42.483 mp_waitall_1 176908 16.5 25.103 35.240 25.103 35.240 init_scf_loop 11 6.9 0.001 0.002 34.486 34.489 prepare_preconditioner 11 7.9 0.000 0.000 30.475 30.527 make_preconditioner 11 8.9 0.000 0.001 30.475 30.527 rebuild_ks_matrix 127 8.3 0.001 0.001 30.014 30.432 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.020 30.013 30.432 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.129 29.505 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.014 27.405 multiply_cannon_multrec 19880 15.5 13.416 16.783 22.196 25.608 multiply_cannon_metrocomm3 19880 15.5 0.057 0.060 15.097 24.972 qs_ot_get_derivative 116 11.6 0.007 0.085 22.591 22.996 apply_preconditioner_dbcsr 127 12.6 0.002 0.027 19.591 20.572 apply_single 127 13.6 0.001 0.001 19.589 20.572 qs_ot_get_p 127 10.4 0.001 0.001 18.881 19.402 ot_diis_step 116 11.6 0.018 0.021 19.364 19.364 multiply_cannon_sync_h2d 19880 15.5 14.252 15.879 14.252 15.879 make_m2s 4970 13.5 0.081 0.087 14.675 15.597 make_images 4970 14.5 1.167 1.255 14.443 15.357 qs_ot_p2m_diag 82 11.4 0.263 0.279 14.675 14.694 cp_fm_cholesky_invert 11 10.9 14.195 14.203 14.195 14.203 sum_up_and_integrate 127 10.3 0.070 0.077 14.027 14.050 integrate_v_rspace 127 11.3 0.003 0.004 13.956 13.981 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.661 13.663 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.907 12.959 calculate_rho_elec 127 8.7 0.131 0.145 12.907 12.958 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.746 10.747 cp_fm_redistribute_end 82 14.4 4.057 10.696 4.071 10.698 cp_fm_diag_elpa_base 82 14.4 6.212 10.130 6.610 10.600 init_scf_run 11 5.9 0.000 0.001 10.520 10.521 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.520 10.520 make_images_data 4970 15.5 0.060 0.068 9.136 10.434 hybrid_alltoall_any 5155 16.4 0.429 1.947 7.945 9.888 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.832 9.128 multiply_cannon_metrocomm4 17395 15.5 0.062 0.072 3.388 9.004 mp_irecv_dv 49801 16.2 3.265 8.758 3.265 8.758 dbcsr_mm_accdrv_process 41158 16.2 4.569 5.359 8.228 8.366 pw_transfer 1535 11.6 0.083 0.100 7.591 7.698 grid_integrate_task_list 127 12.3 7.283 7.647 7.283 7.647 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.367 7.480 cp_fm_upper_to_full 104 14.5 5.696 7.371 5.696 7.371 wfi_extrapolate 11 7.9 0.004 0.013 7.351 7.351 cp_fm_cholesky_decompose 22 10.9 7.172 7.199 7.172 7.199 density_rs2pw 127 9.7 0.005 0.006 6.376 6.680 fft_wrap_pw1pw2_140 519 13.2 0.468 0.510 6.519 6.641 dbcsr_complete_redistribute 393 12.7 1.167 1.201 4.425 6.160 calculate_dm_sparse 127 9.5 0.001 0.002 5.859 5.978 fft3d_ps 1281 14.7 2.676 2.896 5.655 5.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.472 5.478 rs_pw_transfer 1038 11.9 0.013 0.014 5.118 5.468 grid_collocate_task_list 127 9.7 5.028 5.363 5.028 5.363 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.523 5.211 mp_alltoall_d11v 2401 14.1 4.182 5.100 4.182 5.100 mp_sum_l 7804 13.0 3.184 5.061 3.184 5.061 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.270 5.004 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.572 4.686 mp_allgather_i34 2485 14.5 1.481 4.618 1.481 4.618 potential_pw2rs 127 12.3 0.021 0.023 4.490 4.498 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.326 4.031 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.954 3.986 mp_alltoall_i22 712 14.1 1.935 3.810 1.935 3.810 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.784 3.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.508 3.545 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.328000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.363636, yerr=7.619028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.135632E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57164. MP_Allreduce 11256 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.111 185.317 185.318 qs_mol_dyn_low 1 2.0 0.003 0.005 184.778 184.791 qs_forces 11 3.9 0.003 0.004 184.645 184.650 qs_energies 11 4.9 0.002 0.005 177.653 177.662 scf_env_do_scf 11 5.9 0.001 0.001 160.207 160.217 velocity_verlet 10 3.0 0.002 0.002 122.271 122.274 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.054 114.056 dbcsr_multiply_generic 2485 12.5 0.190 0.205 79.874 80.592 qs_scf_new_mos 116 7.6 0.001 0.001 78.837 79.166 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.836 79.165 ot_scf_mini 116 9.6 0.003 0.004 74.188 74.497 multiply_cannon 2485 13.5 0.553 0.588 55.404 58.133 multiply_cannon_loop 2485 14.5 1.175 1.209 51.515 53.343 init_scf_loop 11 6.9 0.000 0.001 46.032 46.033 ot_mini 116 10.6 0.001 0.002 41.960 42.290 prepare_preconditioner 11 7.9 0.000 0.000 41.958 41.987 make_preconditioner 11 8.9 0.000 0.001 41.958 41.987 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.580 40.601 multiply_cannon_multrec 29820 15.5 13.976 19.104 26.671 31.323 rebuild_ks_matrix 127 8.3 0.001 0.001 28.798 29.066 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.798 29.066 mp_waitall_1 152434 16.5 17.406 26.642 17.406 26.642 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.010 26.248 qs_ot_get_derivative 116 11.6 0.002 0.002 22.509 22.825 make_m2s 4970 13.5 0.095 0.099 20.123 21.337 make_images 4970 14.5 1.952 2.281 19.819 21.037 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.897 19.430 apply_single 127 13.6 0.001 0.001 18.897 19.429 ot_diis_step 116 11.6 0.018 0.020 19.318 19.320 qs_ot_get_p 127 10.4 0.001 0.001 18.616 18.971 cp_fm_upper_to_full 104 14.7 11.111 16.458 11.111 16.458 cp_fm_cholesky_invert 11 10.9 15.952 15.961 15.952 15.961 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.430 15.622 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.538 14.590 sum_up_and_integrate 127 10.3 0.079 0.087 13.852 13.872 integrate_v_rspace 127 11.3 0.003 0.004 13.773 13.796 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.287 13.288 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.028 13.057 calculate_rho_elec 127 8.7 0.173 0.189 13.027 13.056 make_images_data 4970 15.5 0.062 0.067 10.842 12.767 dbcsr_mm_accdrv_process 61748 16.2 7.326 8.345 12.274 12.726 dbcsr_complete_redistribute 393 12.7 1.507 1.636 8.916 12.704 multiply_cannon_sync_h2d 29820 15.5 11.604 12.530 11.604 12.530 init_scf_run 11 5.9 0.000 0.001 11.521 11.522 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.521 11.522 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.579 11.350 hybrid_alltoall_any 5155 16.4 0.521 2.192 9.737 11.109 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.329 10.332 cp_fm_redistribute_end 82 14.4 1.774 10.273 1.788 10.278 cp_fm_diag_elpa_base 82 14.4 7.885 9.671 8.463 10.170 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.355 10.029 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.298 9.519 mp_alltoall_i22 712 14.1 5.599 9.328 5.599 9.328 grid_integrate_task_list 127 12.3 7.491 7.803 7.491 7.803 pw_transfer 1535 11.6 0.082 0.094 7.666 7.746 wfi_extrapolate 11 7.9 0.001 0.001 7.527 7.527 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.440 7.525 cp_fm_cholesky_decompose 22 10.9 7.419 7.524 7.419 7.524 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.753 6.759 fft_wrap_pw1pw2_140 519 13.2 0.470 0.481 6.605 6.710 mp_irecv_dv 75445 16.2 2.610 6.502 2.610 6.502 calculate_dm_sparse 127 9.5 0.001 0.002 6.354 6.452 density_rs2pw 127 9.7 0.005 0.006 6.166 6.388 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.322 6.353 fft3d_ps 1281 14.7 2.773 2.843 5.697 5.759 grid_collocate_task_list 127 9.7 5.164 5.615 5.164 5.615 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.251 5.345 mp_alltoall_d11v 2401 14.1 4.802 5.089 4.802 5.089 rs_pw_transfer 1038 11.9 0.013 0.014 4.627 4.868 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.374 4.468 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.440 4.441 potential_pw2rs 127 12.3 0.023 0.024 4.201 4.216 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.117 4.194 calculate_first_density_matrix 1 7.0 0.001 0.002 3.809 3.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.318000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.090909, yerr=17.333386 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.495736E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.084 171.002 171.002 qs_mol_dyn_low 1 2.0 0.003 0.003 170.376 170.388 qs_forces 11 3.9 0.015 0.018 170.271 170.277 qs_energies 11 4.9 0.002 0.002 162.957 162.968 scf_env_do_scf 11 5.9 0.001 0.001 145.836 145.849 velocity_verlet 10 3.0 0.002 0.002 112.257 112.261 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 110.256 110.257 dbcsr_multiply_generic 2529 12.6 0.182 0.187 74.404 74.898 qs_scf_new_mos 118 7.6 0.001 0.001 74.331 74.467 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.331 74.467 ot_scf_mini 118 9.6 0.004 0.004 69.852 69.923 multiply_cannon 2529 13.6 0.591 0.630 54.535 58.653 multiply_cannon_loop 2529 14.6 0.454 0.464 49.911 50.846 ot_mini 118 10.6 0.001 0.001 39.596 39.671 init_scf_loop 11 6.9 0.000 0.000 35.428 35.429 mp_waitall_1 130746 16.6 26.108 33.383 26.108 33.383 prepare_preconditioner 11 7.9 0.000 0.000 31.576 31.601 make_preconditioner 11 8.9 0.000 0.000 31.576 31.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.476 29.729 rebuild_ks_matrix 129 8.3 0.001 0.001 28.855 28.957 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.854 28.957 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.375 26.469 multiply_cannon_multrec 10116 15.6 10.448 14.244 17.785 20.813 ot_diis_step 118 11.6 0.020 0.023 20.019 20.019 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.516 19.837 apply_single 129 13.6 0.001 0.001 19.515 19.836 multiply_cannon_metrocomm3 10116 15.6 0.022 0.024 12.340 19.818 qs_ot_get_derivative 118 11.6 0.002 0.002 19.510 19.586 make_m2s 5058 13.6 0.067 0.070 16.316 18.542 make_images 5058 14.6 2.348 2.870 16.006 18.227 cp_fm_cholesky_invert 11 10.9 17.967 17.974 17.967 17.974 qs_ot_get_p 129 10.4 0.001 0.001 17.246 17.329 sum_up_and_integrate 129 10.3 0.103 0.111 14.007 14.058 integrate_v_rspace 129 11.3 0.003 0.004 13.903 13.960 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.582 13.599 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.176 13.214 calculate_rho_elec 129 8.7 0.259 0.270 13.175 13.213 make_images_data 5058 15.6 0.052 0.060 9.941 12.557 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.447 12.448 multiply_cannon_sync_h2d 10116 15.6 11.737 12.098 11.737 12.098 hybrid_alltoall_any 5245 16.5 0.812 3.833 9.729 11.977 init_scf_run 11 5.9 0.000 0.001 10.308 10.309 scf_env_initial_rho_setup 11 6.9 0.011 0.034 10.308 10.308 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.430 9.434 cp_fm_diag_elpa_base 84 14.4 9.183 9.257 9.420 9.423 grid_integrate_task_list 129 12.3 7.848 8.216 7.848 8.216 cp_fm_cholesky_decompose 22 10.9 7.953 8.072 7.953 8.072 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.908 7.966 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.408 7.818 pw_transfer 1559 11.6 0.083 0.089 7.739 7.758 dbcsr_mm_accdrv_process 20934 16.1 2.651 3.460 6.962 7.599 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.512 7.535 wfi_extrapolate 11 7.9 0.001 0.001 7.240 7.240 mp_allgather_i34 2529 14.6 2.820 7.089 2.820 7.089 fft_wrap_pw1pw2_140 527 13.2 0.497 0.517 6.635 6.663 density_rs2pw 129 9.7 0.005 0.006 5.988 6.345 calculate_dm_sparse 129 9.5 0.001 0.001 6.166 6.228 grid_collocate_task_list 129 9.7 5.517 5.829 5.517 5.829 fft3d_ps 1301 14.7 2.744 2.829 5.667 5.683 dbcsr_complete_redistribute 397 12.7 2.125 2.185 5.127 5.519 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.283 5.285 mp_alltoall_d11v 2429 14.1 4.598 5.249 4.598 5.249 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 5.188 5.197 rs_pw_transfer 1054 12.0 0.013 0.013 4.185 4.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.117 4.148 potential_pw2rs 129 12.3 0.027 0.028 4.085 4.094 multiply_cannon_metrocomm4 7587 15.6 0.024 0.026 1.865 4.066 mp_irecv_dv 29102 15.9 1.828 3.994 1.828 3.994 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.868 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.696 3.726 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.354 3.658 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.531 3.624 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.002000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1394.727273, yerr=41.767954 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.318660E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59299. MP_Allreduce 10975 1515. MP_Sync 86 MP_Alltoall 1700 36954339. 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 0.031 0.054 286.452 286.453 qs_mol_dyn_low 1 2.0 0.003 0.003 285.921 285.932 qs_forces 11 3.9 0.003 0.003 285.825 285.828 qs_energies 11 4.9 0.005 0.006 277.178 277.192 scf_env_do_scf 11 5.9 0.001 0.001 255.116 255.131 velocity_verlet 10 3.0 0.002 0.002 206.277 206.285 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.166 131.168 init_scf_loop 11 6.9 0.000 0.000 123.700 123.702 prepare_preconditioner 11 7.9 0.000 0.000 119.059 119.092 make_preconditioner 11 8.9 0.000 0.000 119.059 119.092 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.296 116.233 qs_scf_new_mos 116 7.6 0.001 0.001 89.674 89.786 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.673 89.785 ot_scf_mini 116 9.6 0.004 0.004 84.965 85.012 dbcsr_multiply_generic 2485 12.5 0.213 0.220 83.044 83.612 cp_fm_upper_to_full 104 14.8 52.972 75.777 52.972 75.777 multiply_cannon 2485 13.5 0.683 0.719 59.112 59.511 multiply_cannon_loop 2485 14.5 0.471 0.479 55.429 56.752 ot_mini 116 10.6 0.001 0.001 44.070 44.119 dbcsr_complete_redistribute 393 12.7 3.989 4.035 29.335 42.006 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.998 38.653 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.718 36.275 mp_alltoall_i22 712 14.1 21.556 34.167 21.556 34.167 rebuild_ks_matrix 127 8.3 0.001 0.001 33.291 33.329 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.290 33.328 cp_fm_cholesky_invert 11 10.9 33.235 33.241 33.235 33.241 mp_waitall_1 104546 16.7 28.163 31.811 28.163 31.811 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.031 31.068 qs_ot_get_p 127 10.4 0.001 0.001 25.540 25.572 qs_ot_get_derivative 116 11.6 0.002 0.002 23.832 23.878 qs_ot_p2m_diag 82 11.4 0.869 0.877 21.610 21.638 make_m2s 4970 13.5 0.076 0.078 19.363 20.406 ot_diis_step 116 11.6 0.022 0.022 20.207 20.208 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 19.019 20.186 make_images 4970 14.5 3.758 3.872 18.891 19.937 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.900 19.901 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.475 19.678 apply_single 127 13.6 0.001 0.001 19.475 19.677 multiply_cannon_multrec 9940 15.5 10.386 12.082 18.343 18.404 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.794 16.797 cp_fm_diag_elpa_base 82 14.4 12.405 13.963 16.790 16.793 multiply_cannon_sync_h2d 9940 15.5 15.533 15.539 15.533 15.539 sum_up_and_integrate 127 10.3 0.193 0.196 15.357 15.451 integrate_v_rspace 127 11.3 0.004 0.004 15.165 15.259 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.686 14.711 calculate_rho_elec 127 8.7 0.479 0.479 14.685 14.710 init_scf_run 11 5.9 0.000 0.001 12.225 12.225 scf_env_initial_rho_setup 11 6.9 0.003 0.003 12.224 12.225 hybrid_alltoall_any 5155 16.4 1.293 3.022 10.373 12.028 make_images_data 4970 15.5 0.059 0.064 10.276 11.967 dbcsr_mm_accdrv_process 20590 16.0 3.716 5.620 7.722 9.449 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.275 9.316 cp_fm_cholesky_decompose 22 10.9 8.849 8.897 8.849 8.897 grid_integrate_task_list 127 12.3 8.482 8.685 8.482 8.685 wfi_extrapolate 11 7.9 0.001 0.001 8.529 8.529 pw_transfer 1535 11.6 0.092 0.094 8.103 8.111 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 7.883 7.884 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 7.866 7.873 fft_wrap_pw1pw2_140 519 13.2 0.520 0.523 6.973 6.983 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.930 6.976 mp_alltoall_d11v 2401 14.1 6.663 6.817 6.663 6.817 calculate_dm_sparse 127 9.5 0.001 0.001 6.535 6.631 grid_collocate_task_list 127 9.7 6.291 6.324 6.291 6.324 copy_dbcsr_to_fm 185 11.7 0.004 0.004 5.937 6.046 fft3d_ps 1281 14.7 2.736 2.746 5.941 5.958 density_rs2pw 127 9.7 0.005 0.005 5.892 5.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.453000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2161.818182, yerr=61.916099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 1.260315E+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 0.066 0.083 85.408 85.409 qs_energies 1 2.0 0.000 0.000 84.826 84.840 ls_scf 1 3.0 0.000 0.000 83.918 83.932 dbcsr_multiply_generic 111 6.7 0.028 0.043 73.103 73.261 multiply_cannon 111 7.7 0.017 0.020 56.325 57.484 multiply_cannon_loop 111 8.7 0.209 0.224 52.919 54.190 ls_scf_main 1 4.0 0.000 0.000 52.199 52.200 density_matrix_trs4 2 5.0 0.002 0.004 46.736 46.820 ls_scf_init_scf 1 4.0 0.000 0.000 28.681 28.682 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.608 27.659 mp_waitall_1 11316 10.9 22.898 26.438 22.898 26.438 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.515 25.537 multiply_cannon_multrec 2664 9.7 8.177 8.902 15.559 17.553 multiply_cannon_sync_h2d 2664 9.7 13.624 15.059 13.624 15.059 make_m2s 222 7.7 0.008 0.011 13.096 13.549 make_images 222 8.7 0.098 0.109 13.074 13.528 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.919 13.068 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.632 6.998 8.354 make_images_data 222 9.7 0.004 0.005 7.656 8.279 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.649 8.096 hybrid_alltoall_any 227 10.6 0.215 1.833 6.555 7.769 dbcsr_mm_accdrv_process_sort 4760 11.4 6.211 7.095 6.211 7.095 calculate_norms 4752 9.8 5.510 6.223 5.510 6.223 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.014 5.175 mp_sum_l 807 5.4 3.140 4.377 3.140 4.377 make_images_sizes 222 9.7 0.000 0.000 0.768 3.523 mp_alltoall_i44 222 10.7 0.768 3.523 0.768 3.523 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.367 3.377 arnoldi_extremal 4 6.8 0.000 0.000 3.201 3.224 arnoldi_normal_ev 4 7.8 0.001 0.004 3.201 3.224 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.043 3.153 mp_irecv_dv 6231 10.9 2.026 3.129 2.026 3.129 build_subspace 16 8.4 0.009 0.012 3.107 3.109 ls_scf_post 1 4.0 0.000 0.000 3.038 3.051 ls_scf_store_result 1 5.0 0.000 0.000 2.850 2.886 dbcsr_special_finalize 555 9.7 0.005 0.006 2.320 2.771 dbcsr_merge_single_wm 555 10.7 0.464 0.609 2.312 2.763 make_images_pack 222 9.7 2.207 2.623 2.209 2.625 dbcsr_matrix_vector_mult 304 9.0 0.003 0.009 2.314 2.557 dbcsr_sort_data 658 11.4 2.108 2.512 2.108 2.512 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.463 2.071 2.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.226 2.303 buffer_matrices_ensure_size 222 8.7 1.779 2.182 1.779 2.182 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.756 1.757 rebuild_ks_matrix 3 7.3 0.000 0.000 1.746 1.748 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.746 1.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.409000, 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/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 2.101150E+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 0.028 0.049 91.343 91.343 qs_energies 1 2.0 0.000 0.000 90.821 90.825 ls_scf 1 3.0 0.000 0.002 89.496 89.499 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.613 75.955 multiply_cannon 111 7.7 0.030 0.044 53.267 57.266 ls_scf_main 1 4.0 0.000 0.002 55.195 55.196 multiply_cannon_loop 111 8.7 0.116 0.122 49.990 53.303 density_matrix_trs4 2 5.0 0.002 0.003 49.531 49.706 ls_scf_init_scf 1 4.0 0.000 0.001 30.817 30.819 mp_waitall_1 9246 10.9 21.393 30.143 21.393 30.143 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.668 29.770 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.273 27.287 multiply_cannon_multrec 1332 9.7 13.137 16.686 22.344 27.029 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.856 20.769 make_m2s 222 7.7 0.006 0.007 15.559 16.305 make_images 222 8.7 1.574 1.935 15.529 16.274 dbcsr_mm_accdrv_process 4041 10.4 0.292 0.447 8.808 10.382 make_images_data 222 9.7 0.004 0.004 8.958 9.993 dbcsr_mm_accdrv_process_sort 4041 11.4 8.402 9.935 8.402 9.935 hybrid_alltoall_any 227 10.6 0.518 2.420 8.342 9.464 mp_sum_l 807 5.4 5.489 8.764 5.489 8.764 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.231 7.617 mp_irecv_dv 3311 11.0 3.211 7.554 3.211 7.554 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.257 6.999 calculate_norms 2376 9.8 6.003 6.773 6.003 6.773 multiply_cannon_sync_h2d 1332 9.7 4.777 5.991 4.777 5.991 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.011 5.209 arnoldi_extremal 4 6.8 0.000 0.000 4.637 4.659 arnoldi_normal_ev 4 7.8 0.001 0.004 4.637 4.659 build_subspace 16 8.4 0.014 0.021 4.382 4.386 ls_scf_post 1 4.0 0.000 0.001 3.483 3.488 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.137 3.365 ls_scf_store_result 1 5.0 0.000 0.000 3.193 3.301 dbcsr_matrix_vector_mult_local 304 10.0 2.757 3.226 2.759 3.228 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.198 3.033 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.538 2.623 mp_allgather_i34 111 8.7 0.979 2.469 0.979 2.469 make_images_pack 222 9.7 2.018 2.422 2.020 2.424 dbcsr_sort_data 436 11.2 1.815 2.053 1.815 2.053 dbcsr_data_new 4174 10.1 1.610 1.837 1.610 1.837 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.827 1.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.343000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 2.702500E+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 0.033 0.050 94.408 94.409 qs_energies 1 2.0 0.000 0.000 93.787 93.790 ls_scf 1 3.0 0.001 0.011 92.387 92.391 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.249 77.597 ls_scf_main 1 4.0 0.000 0.001 58.205 58.209 multiply_cannon 111 7.7 0.043 0.100 53.097 57.103 multiply_cannon_loop 111 8.7 0.100 0.107 49.556 53.780 density_matrix_trs4 2 5.0 0.002 0.003 52.336 52.485 mp_waitall_1 7374 11.0 24.445 34.836 24.445 34.836 ls_scf_init_scf 1 4.0 0.000 0.001 30.579 30.582 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.451 29.516 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.055 27.064 multiply_cannon_multrec 888 9.7 12.462 15.035 20.974 24.017 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.458 23.718 make_m2s 222 7.7 0.006 0.007 17.270 18.456 make_images 222 8.7 1.969 2.279 17.232 18.416 hybrid_alltoall_any 227 10.6 0.621 2.862 9.625 11.039 make_images_data 222 9.7 0.003 0.004 9.958 11.038 dbcsr_mm_accdrv_process 3754 10.4 0.257 0.413 8.034 9.247 mp_sum_l 807 5.4 5.332 9.021 5.332 9.021 dbcsr_mm_accdrv_process_sort 3754 11.4 7.666 8.833 7.666 8.833 multiply_cannon_sync_h2d 888 9.7 6.119 7.334 6.119 7.334 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.648 7.105 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.127 7.098 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.462 7.069 mp_irecv_dv 2335 11.1 2.446 7.017 2.446 7.017 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.924 5.114 arnoldi_extremal 4 6.8 0.000 0.002 5.092 5.104 arnoldi_normal_ev 4 7.8 0.001 0.008 5.092 5.104 build_subspace 16 8.4 0.014 0.020 4.784 4.794 calculate_norms 1584 9.8 4.261 4.584 4.261 4.584 mp_allgather_i34 111 8.7 1.385 3.909 1.385 3.909 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.454 3.776 ls_scf_post 1 4.0 0.000 0.003 3.602 3.607 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.604 3.039 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.304 3.423 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.751 2.853 dbcsr_sort_data 325 11.1 1.872 2.127 1.872 2.127 make_images_pack 222 9.7 1.814 2.121 1.817 2.124 make_images_sizes 222 9.7 0.000 0.000 0.995 2.080 mp_alltoall_i44 222 10.7 0.994 2.080 0.994 2.080 dbcsr_data_release 9322 10.9 1.305 1.908 1.305 1.908 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.409000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 3.341914E+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 0.059 0.114 97.612 97.613 qs_energies 1 2.0 0.000 0.000 96.594 96.601 ls_scf 1 3.0 0.000 0.002 94.878 94.885 dbcsr_multiply_generic 111 6.7 0.018 0.021 78.722 78.976 ls_scf_main 1 4.0 0.001 0.005 58.994 58.996 multiply_cannon 111 7.7 0.052 0.120 51.743 56.304 density_matrix_trs4 2 5.0 0.002 0.003 52.962 53.076 multiply_cannon_loop 111 8.7 0.115 0.126 46.697 50.181 ls_scf_init_scf 1 4.0 0.000 0.001 32.640 32.642 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.458 31.560 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.995 29.008 mp_waitall_1 6438 11.0 22.893 28.810 22.893 28.810 multiply_cannon_multrec 1332 9.7 14.149 17.121 21.975 24.258 make_m2s 222 7.7 0.007 0.008 21.039 22.461 make_images 222 8.7 3.130 3.589 20.988 22.412 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.334 17.388 make_images_data 222 9.7 0.004 0.005 11.751 13.331 hybrid_alltoall_any 227 10.6 0.794 3.716 11.211 12.910 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.400 7.469 8.981 dbcsr_mm_accdrv_process_sort 3641 11.4 7.098 8.564 7.098 8.564 mp_sum_l 807 5.4 4.372 8.425 4.372 8.425 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.400 6.752 multiply_cannon_sync_h2d 1332 9.7 5.525 6.217 5.525 6.217 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.096 6.127 mp_irecv_dv 3229 10.9 2.072 6.043 2.072 6.043 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.634 5.519 arnoldi_extremal 4 6.8 0.000 0.002 5.246 5.258 arnoldi_normal_ev 4 7.8 0.002 0.005 5.246 5.258 build_subspace 16 8.4 0.015 0.021 4.894 4.906 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.536 4.744 mp_allgather_i34 111 8.7 2.182 4.661 2.182 4.661 calculate_norms 2376 9.8 4.186 4.566 4.186 4.566 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.601 3.897 dbcsr_matrix_vector_mult_local 304 10.0 3.221 3.699 3.223 3.701 dbcsr_sort_data 658 11.4 3.097 3.367 3.097 3.367 ls_scf_post 1 4.0 0.001 0.003 3.244 3.250 dbcsr_special_finalize 555 9.7 0.006 0.007 2.850 3.161 dbcsr_merge_single_wm 555 10.7 0.540 0.668 2.841 3.153 ls_scf_store_result 1 5.0 0.000 0.000 2.972 3.038 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.913 2.963 dbcsr_data_release 10477 10.7 1.578 2.398 1.578 2.398 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.613000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2753.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 4.628046E+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 0.091 0.108 101.746 101.747 qs_energies 1 2.0 0.000 0.000 100.897 100.903 ls_scf 1 3.0 0.000 0.000 98.982 98.988 dbcsr_multiply_generic 111 6.7 0.017 0.018 80.803 81.031 multiply_cannon 111 7.7 0.074 0.157 58.362 63.494 ls_scf_main 1 4.0 0.000 0.000 61.620 61.621 multiply_cannon_loop 111 8.7 0.069 0.077 53.820 55.564 density_matrix_trs4 2 5.0 0.002 0.003 54.643 54.732 ls_scf_init_scf 1 4.0 0.000 0.000 33.756 33.761 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.576 32.615 mp_waitall_1 5481 11.0 26.316 32.126 26.316 32.126 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.394 30.405 multiply_cannon_multrec 444 9.7 13.980 16.130 20.999 23.650 make_m2s 222 7.7 0.004 0.005 17.612 20.162 make_images 222 8.7 3.714 4.416 17.551 20.102 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.142 16.305 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.098 15.349 make_images_data 222 9.7 0.003 0.004 9.874 12.404 hybrid_alltoall_any 227 10.6 0.788 3.764 9.688 12.205 multiply_cannon_sync_h2d 444 9.7 6.551 7.956 6.551 7.956 dbcsr_mm_accdrv_process 3003 10.4 0.163 0.349 6.719 7.846 dbcsr_mm_accdrv_process_sort 3003 11.4 6.404 7.503 6.404 7.503 mp_allgather_i34 111 8.7 2.779 6.944 2.779 6.944 arnoldi_extremal 4 6.8 0.000 0.000 5.792 5.803 arnoldi_normal_ev 4 7.8 0.002 0.005 5.792 5.803 build_subspace 16 8.4 0.015 0.020 5.399 5.409 mp_sum_l 807 5.4 2.892 5.070 2.892 5.070 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.616 4.766 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.171 4.369 dbcsr_matrix_vector_mult_local 304 10.0 3.717 4.181 3.719 4.183 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.627 3.966 mp_irecv_dv 1241 11.2 1.607 3.943 1.607 3.943 compute_matrix_preconditioner 1 6.0 0.001 0.002 3.860 3.868 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.963 3.732 calculate_norms 792 9.8 3.538 3.687 3.538 3.687 ls_scf_post 1 4.0 0.000 0.000 3.606 3.611 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.367 3.479 acc_transpose_blocks 444 9.7 0.002 0.002 3.406 3.455 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.427 make_images_sizes 222 9.7 0.000 0.000 1.102 3.418 mp_alltoall_i44 222 10.7 1.101 3.418 1.101 3.418 acc_transpose_blocks_kernels 444 10.7 0.006 0.007 3.368 3.416 jit_kernel_transpose 1 13.0 3.362 3.410 3.362 3.410 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.253 dbcsr_merge_all 275 8.9 0.474 0.518 2.044 2.090 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.747000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3616.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2616d1f209524a36d9c99e7eecb679e8d7bcccfe_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 8.719323E+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 0.084 0.115 107.534 107.534 qs_energies 1 2.0 0.000 0.000 105.916 105.931 ls_scf 1 3.0 0.000 0.000 103.043 103.057 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.327 77.426 ls_scf_main 1 4.0 0.000 0.000 65.576 65.577 density_matrix_trs4 2 5.0 0.002 0.004 56.759 56.804 multiply_cannon 111 7.7 0.128 0.188 49.737 51.484 multiply_cannon_loop 111 8.7 0.068 0.071 46.418 47.319 ls_scf_init_scf 1 4.0 0.000 0.000 33.798 33.799 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.456 32.469 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.638 29.652 mp_waitall_1 4569 11.1 22.404 25.656 22.404 25.656 make_m2s 222 7.7 0.005 0.005 23.966 24.848 make_images 222 8.7 4.577 4.946 23.860 24.740 multiply_cannon_multrec 444 9.7 17.792 18.695 22.377 23.072 hybrid_alltoall_any 227 10.6 1.656 3.615 13.037 15.812 make_images_data 222 9.7 0.003 0.003 13.230 15.492 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.554 11.089 multiply_cannon_sync_h2d 444 9.7 8.826 8.886 8.826 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.323 7.342 arnoldi_normal_ev 4 7.8 0.003 0.009 7.323 7.342 build_subspace 16 8.4 0.026 0.036 6.772 6.782 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.458 5.596 dbcsr_matrix_vector_mult_local 304 10.0 5.041 5.325 5.043 5.327 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.010 5.226 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.786 4.878 dbcsr_mm_accdrv_process 1814 10.4 0.252 0.315 4.416 4.532 dbcsr_mm_accdrv_process_sort 1814 11.4 4.118 4.240 4.118 4.240 ls_scf_post 1 4.0 0.000 0.000 3.669 3.683 ls_scf_store_result 1 5.0 0.000 0.000 3.425 3.432 make_images_sizes 222 9.7 0.000 0.000 1.429 3.425 mp_alltoall_i44 222 10.7 1.428 3.425 1.428 3.425 calculate_norms 792 9.8 3.262 3.324 3.262 3.324 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.139 mp_allgather_i34 111 8.7 0.983 3.013 0.983 3.013 dbcsr_merge_all 275 8.9 0.892 0.916 2.865 2.916 dbcsr_complete_redistribute 5 7.6 1.428 1.464 2.757 2.877 dbcsr_data_release 12724 10.6 2.320 2.857 2.320 2.857 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.843 2.843 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.529 dbcsr_sort_data 325 11.1 2.438 2.498 2.438 2.498 dbcsr_new_transposed 4 7.5 0.243 0.251 2.279 2.292 dbcsr_frobenius_norm 74 6.6 2.055 2.137 2.200 2.244 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.223 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.203 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.203 rebuild_ks_matrix 3 7.3 0.000 0.000 2.156 2.159 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.156 2.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.534000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6813.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 2616d1f209524a36d9c99e7eecb679e8d7bcccfe Summary: empty Status: OK