=== 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: e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594 ################# 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/01 job id: 41654664 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/02 job id: 41654665 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/03 job id: 41654666 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/04 job id: 41654667 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/05 job id: 41654668 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/06 job id: 41654669 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/07 job id: 41654670 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/08 job id: 41654671 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/09 job id: 41654672 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/10 job id: 41654673 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/11 job id: 41654674 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/12 job id: 41654675 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/13 job id: 41654676 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/14 job id: 41654678 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/15 job id: 41654679 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/16 job id: 41654680 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/17 job id: 41654681 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/18 job id: 41654682 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/19 job id: 41654683 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/20 job id: 41654684 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/21 job id: 41654685 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/22 job id: 41654686 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/23 job id: 41654687 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/24 job id: 41654688 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/25 job id: 41654689 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/26 job id: 41654690 --- 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 0.040 0.055 135.836 135.837 farming_run 1 2.0 135.245 135.247 135.781 135.788 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452552E+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 0.010 0.027 117.105 117.105 qs_energies 1 2.0 0.000 0.000 116.855 116.857 mp2_main 1 3.0 0.000 0.000 114.918 114.920 mp2_gpw_main 1 4.0 0.033 0.049 114.039 114.041 mp2_ri_gpw_compute_in 1 5.0 0.172 0.189 95.261 95.515 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 56.335 56.575 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.171 42.659 47.872 get_2c_integrals 1 6.0 0.000 0.000 38.281 38.768 integrate_v_rspace 273 8.0 0.433 0.449 25.249 30.189 pw_transfer 6555 10.6 0.371 0.402 27.657 28.153 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.278 26.606 grid_integrate_task_list 273 9.0 21.007 26.473 21.007 26.473 fft_wrap_pw1pw2_100 2178 12.4 1.231 1.427 23.839 24.189 compute_2c_integrals 1 7.0 0.002 0.002 20.429 20.443 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.748 20.106 mp2_eri_2c_integrate_gpw 1 9.0 2.364 2.442 19.745 20.102 rpa_ri_compute_en 1 5.0 0.000 0.000 18.654 18.830 cp_fm_cholesky_decompose 12 8.2 17.843 18.328 17.843 18.328 cholesky_decomp 1 7.0 0.000 0.000 16.699 17.187 fft3d_s 5443 13.4 16.327 16.806 16.348 16.826 ao_to_mo_and_store_B_mult_1 272 7.0 10.848 15.584 10.848 15.584 calculate_wavefunction 272 8.0 5.449 5.599 12.636 13.242 rpa_num_int 1 6.0 0.000 0.000 10.616 10.617 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.555 10.567 calc_potential_gpw 544 9.5 0.005 0.005 9.658 10.039 calc_mat_Q 8 8.0 0.000 0.000 9.368 9.454 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 9.001 9.222 contract_S_to_Q 8 9.0 0.000 0.000 8.789 8.872 potential_pw2rs 545 10.0 0.106 0.109 7.821 8.581 parallel_gemm_fm 14 9.1 0.000 0.000 8.369 8.445 parallel_gemm_fm_cosma 14 10.1 8.369 8.445 8.369 8.445 collocate_single_gaussian 272 10.0 0.039 0.042 7.562 7.773 create_integ_mat 1 6.0 0.022 0.027 7.703 7.703 array2fm 1 7.0 0.000 0.000 6.659 7.159 pw_scatter_s 2720 13.7 4.386 4.555 4.386 4.555 pw_gather_s 2722 13.2 3.967 4.422 3.967 4.422 pw_poisson_solve 545 10.5 2.484 2.531 3.573 3.742 array2fm_buffer_send 1 8.0 2.945 3.160 2.945 3.160 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.040627, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 0.072 0.092 398.499 398.500 farming_run 1 2.0 397.730 397.739 398.404 398.427 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.216369E+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 24433745. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.048 211.667 211.667 qs_energies 1 2.0 0.000 0.000 211.331 211.421 scf_env_do_scf 1 3.0 0.000 0.000 106.948 106.948 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.999 106.007 rebuild_ks_matrix 4 6.0 0.000 0.000 105.998 106.006 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 105.998 106.006 hfx_ks_matrix 4 8.0 0.001 0.001 105.600 105.603 integrate_four_center 4 9.0 0.143 0.465 105.599 105.602 mp2_main 1 3.0 0.000 0.000 104.088 104.177 mp2_gpw_main 1 4.0 0.035 0.069 103.141 103.232 integrate_four_center_main 4 10.0 0.134 0.483 96.800 99.554 integrate_four_center_bin 269 11.0 96.666 99.535 96.666 99.535 init_scf_loop 1 4.0 0.000 0.000 92.283 92.283 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 75.805 76.845 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.947 55.984 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.161 42.557 47.817 integrate_v_rspace 95 8.0 0.396 0.555 28.499 33.514 pw_transfer 2240 10.6 0.147 0.164 30.158 30.453 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 29.143 29.486 grid_integrate_task_list 95 9.0 23.764 28.987 23.764 28.987 mp2_ri_gpw_compute_en 1 5.0 0.086 0.170 27.186 28.901 ao_to_mo_and_store_B_mult_1 91 7.0 10.697 28.705 10.697 28.705 fft_wrap_pw1pw2_100 730 12.4 1.285 1.462 26.871 27.267 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.905 25.471 25.482 get_2c_integrals 1 6.0 0.000 0.000 20.777 20.797 compute_2c_integrals 1 7.0 0.002 0.003 19.754 19.760 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 19.316 19.625 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.892 19.315 19.624 fft3d_s 1823 13.4 18.528 18.879 18.541 18.893 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.664 14.664 calculate_wavefunction 91 8.0 2.026 2.063 9.806 10.143 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.587 8.831 9.431 potential_pw2rs 186 10.0 0.033 0.034 8.710 9.425 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.629 9.094 calc_potential_gpw 182 9.5 0.002 0.002 8.635 8.953 local_gemm 172 8.0 8.274 8.865 8.274 8.865 mp2_ri_gpw_compute_en_comm 22 7.0 0.489 0.510 8.407 8.798 collocate_single_gaussian 91 10.0 0.017 0.020 7.941 8.163 mp_sendrecv_dm3 2068 8.0 6.466 6.906 6.466 6.906 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.432 6.345 6.432 mp_sync 38 10.4 3.443 5.866 3.443 5.866 pw_gather_s 912 13.2 4.990 5.491 4.990 5.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.113619, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 450.437120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.061 56.163 56.164 qs_mol_dyn_low 1 2.0 0.003 0.003 54.536 54.584 qs_forces 11 3.9 0.002 0.002 54.467 54.467 qs_energies 11 4.9 0.001 0.002 52.961 52.976 scf_env_do_scf 11 5.9 0.000 0.001 46.723 46.723 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.412 44.413 dbcsr_multiply_generic 2286 12.5 0.092 0.096 34.710 35.194 qs_scf_new_mos 108 7.5 0.000 0.001 33.583 33.859 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.582 33.859 ot_scf_mini 108 9.5 0.002 0.002 31.870 32.050 multiply_cannon 2286 13.5 0.185 0.194 26.571 28.195 velocity_verlet 10 3.0 0.001 0.002 27.567 27.567 multiply_cannon_loop 2286 14.5 1.521 1.593 25.762 27.378 ot_mini 108 10.5 0.001 0.001 20.523 20.760 qs_ot_get_derivative 108 11.5 0.001 0.001 17.410 17.596 mp_waitall_1 267858 16.1 9.669 15.728 9.669 15.728 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 6.338 13.373 multiply_cannon_multrec 54864 15.5 4.150 6.446 7.483 10.856 rebuild_ks_matrix 119 8.3 0.000 0.000 8.317 8.464 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.010 8.316 8.463 multiply_cannon_sync_h2d 54864 15.5 5.792 7.858 5.792 7.858 mp_sum_l 7207 12.9 5.868 7.557 5.868 7.557 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.360 7.490 qs_ot_get_p 119 10.4 0.001 0.001 6.639 6.932 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.904 6.338 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.975 6.091 init_scf_run 11 5.9 0.000 0.001 4.952 4.952 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.951 4.952 sum_up_and_integrate 119 10.3 0.008 0.009 4.874 4.881 integrate_v_rspace 119 11.3 0.002 0.002 4.866 4.874 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.570 4.707 calculate_rho_elec 119 8.7 0.011 0.016 4.570 4.706 dbcsr_mm_accdrv_process 76910 16.1 1.184 1.825 3.255 4.536 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.154 3.852 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.583 3.648 rs_pw_transfer 974 11.9 0.012 0.013 3.301 3.401 calculate_dm_sparse 119 9.5 0.000 0.000 3.038 3.173 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.912 3.107 apply_single 119 13.6 0.000 0.000 2.912 3.107 ot_diis_step 108 11.5 0.006 0.006 2.768 2.768 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.712 2.713 density_rs2pw 119 9.7 0.003 0.004 2.553 2.635 jit_kernel_multiply 13 15.8 2.009 2.626 2.009 2.626 calculate_first_density_matrix 1 7.0 0.001 0.001 2.589 2.592 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.506 2.585 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.415 2.416 cp_fm_redistribute_end 50 14.0 2.192 2.390 2.197 2.392 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.382 2.386 cp_fm_diag_elpa_base 50 14.0 0.193 2.330 0.194 2.339 wfi_extrapolate 11 7.9 0.001 0.001 2.299 2.299 init_scf_loop 11 6.9 0.000 0.000 2.291 2.291 acc_transpose_blocks 54864 15.5 0.229 0.247 1.785 2.251 potential_pw2rs 119 12.3 0.004 0.004 2.208 2.231 mp_sum_d 4129 12.0 1.483 2.120 1.483 2.120 grid_integrate_task_list 119 12.3 2.016 2.116 2.016 2.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.033 2.095 make_m2s 4572 13.5 0.054 0.056 1.958 2.007 pw_transfer 1439 11.6 0.052 0.057 1.913 1.991 make_images 4572 14.5 0.133 0.139 1.875 1.924 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.839 1.919 mp_alltoall_d11v 2130 13.8 1.627 1.815 1.627 1.815 fft3d_ps 1201 14.6 0.360 0.463 1.614 1.686 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.490 1.523 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.408 1.488 grid_collocate_task_list 119 9.7 1.289 1.360 1.289 1.360 mp_waitany 12084 13.8 1.192 1.357 1.192 1.357 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.857 1.268 create_qs_kind_set 1 2.0 0.000 0.000 1.229 1.234 read_qs_kind 2 3.0 0.022 1.173 1.228 1.234 parser_read_line 2807 4.0 0.001 0.001 1.206 1.225 parser_read_line_low 5 5.0 0.002 0.040 1.205 1.224 broadcast_input_information 5 6.0 0.000 0.000 1.203 1.222 mp_bcast_i 105 3.0 1.198 1.215 1.198 1.215 prepare_preconditioner 11 7.9 0.000 0.000 1.123 1.157 make_preconditioner 11 8.9 0.000 0.000 1.123 1.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.164000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 486.760448E+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 1700725. 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 0.013 0.029 40.021 40.021 qs_mol_dyn_low 1 2.0 0.003 0.003 39.680 39.830 qs_forces 11 3.9 0.002 0.002 39.604 39.606 qs_energies 11 4.9 0.001 0.001 37.847 37.851 scf_env_do_scf 11 5.9 0.000 0.001 32.642 32.642 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.056 30.057 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.940 22.380 qs_scf_new_mos 108 7.5 0.001 0.001 20.576 20.833 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.576 20.833 ot_scf_mini 108 9.5 0.002 0.003 19.643 19.823 velocity_verlet 10 3.0 0.002 0.004 19.199 19.200 multiply_cannon 2286 13.5 0.208 0.220 16.522 18.062 multiply_cannon_loop 2286 14.5 0.911 0.985 15.330 16.978 ot_mini 108 10.5 0.001 0.001 12.092 12.332 mp_waitall_1 217478 16.2 6.570 11.776 6.570 11.776 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.293 9.810 qs_ot_get_derivative 108 11.5 0.001 0.001 9.542 9.721 multiply_cannon_multrec 27432 15.5 1.955 4.512 5.851 8.590 rebuild_ks_matrix 119 8.3 0.000 0.000 7.403 7.546 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.403 7.545 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.518 6.647 dbcsr_mm_accdrv_process 47894 16.0 3.047 5.241 3.826 5.668 qs_ot_get_p 119 10.4 0.001 0.001 4.525 4.769 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.679 4.532 sum_up_and_integrate 119 10.3 0.015 0.016 4.368 4.373 integrate_v_rspace 119 11.3 0.002 0.002 4.354 4.360 mp_sum_l 7207 12.9 2.264 4.271 2.264 4.271 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.090 4.202 apply_single 119 13.6 0.000 0.000 3.090 4.202 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.917 3.966 calculate_rho_elec 119 8.7 0.021 0.024 3.916 3.966 init_scf_run 11 5.9 0.000 0.001 3.958 3.959 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.958 3.958 rs_pw_transfer 974 11.9 0.010 0.011 2.943 3.064 make_m2s 4572 13.5 0.052 0.054 2.721 2.974 make_images 4572 14.5 0.201 0.238 2.632 2.884 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.795 2.817 multiply_cannon_sync_h2d 27432 15.5 2.165 2.775 2.165 2.775 init_scf_loop 11 6.9 0.000 0.000 2.556 2.557 ot_diis_step 108 11.5 0.011 0.011 2.490 2.491 calculate_first_density_matrix 1 7.0 0.000 0.001 2.347 2.349 density_rs2pw 119 9.7 0.004 0.004 2.241 2.348 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.346 2.346 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.221 2.317 calculate_dm_sparse 119 9.5 0.000 0.000 2.081 2.157 potential_pw2rs 119 12.3 0.006 0.006 2.098 2.113 jit_kernel_multiply 11 16.2 0.726 2.039 0.726 2.039 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.994 1.997 pw_transfer 1439 11.6 0.064 0.066 1.910 1.942 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.928 1.929 grid_integrate_task_list 119 12.3 1.822 1.914 1.822 1.914 cp_fm_redistribute_end 50 14.0 1.583 1.893 1.587 1.895 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.821 1.856 cp_fm_diag_elpa_base 50 14.0 0.298 1.823 0.306 1.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.732 1.774 prepare_preconditioner 11 7.9 0.000 0.000 1.655 1.683 make_preconditioner 11 8.9 0.000 0.000 1.655 1.683 make_images_data 4572 15.5 0.044 0.051 1.213 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.549 1.609 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.053 1.567 wfi_extrapolate 11 7.9 0.001 0.001 1.558 1.558 fft3d_ps 1201 14.6 0.497 0.553 1.528 1.555 mp_alltoall_d11v 2130 13.8 1.356 1.505 1.356 1.505 acc_transpose_blocks 27432 15.5 0.109 0.115 1.195 1.501 mp_allgather_i34 2286 14.5 0.635 1.430 0.635 1.430 fft_wrap_pw1pw2_140 487 13.2 0.077 0.084 1.392 1.427 grid_collocate_task_list 119 9.7 1.239 1.303 1.239 1.303 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.237 1.284 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.260 1.268 mp_sum_d 4129 12.0 0.663 1.080 0.663 1.080 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.968 0.969 make_images_sizes 4572 15.5 0.005 0.005 0.692 0.965 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.949 0.964 mp_alltoall_i44 4572 16.5 0.688 0.960 0.688 0.960 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.663 0.879 mp_alltoall_z22v 1201 16.6 0.787 0.858 0.787 0.858 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.605 0.803 0.847 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.021000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.363636, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.444800E+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 0.159 0.230 33.816 33.818 qs_mol_dyn_low 1 2.0 0.006 0.012 32.313 32.332 qs_forces 11 3.9 0.008 0.029 32.222 32.226 qs_energies 11 4.9 0.006 0.015 30.606 30.616 scf_env_do_scf 11 5.9 0.001 0.001 25.612 25.612 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 22.943 22.944 dbcsr_multiply_generic 2286 12.5 0.093 0.096 16.428 16.519 velocity_verlet 10 3.0 0.024 0.036 15.358 15.360 qs_scf_new_mos 108 7.5 0.001 0.001 14.749 14.769 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.749 14.769 ot_scf_mini 108 9.5 0.002 0.003 14.026 14.045 multiply_cannon 2286 13.5 0.194 0.202 13.100 13.915 multiply_cannon_loop 2286 14.5 0.643 0.674 12.304 13.143 ot_mini 108 10.5 0.001 0.001 8.715 8.728 qs_ot_get_derivative 108 11.5 0.001 0.001 7.206 7.221 multiply_cannon_multrec 18288 15.5 1.914 2.888 6.762 7.099 rebuild_ks_matrix 119 8.3 0.000 0.000 6.463 6.482 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.463 6.481 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.689 5.705 dbcsr_mm_accdrv_process 38222 16.0 4.105 5.337 4.763 5.601 mp_waitall_1 169478 16.3 3.073 4.300 3.073 4.300 sum_up_and_integrate 119 10.3 0.017 0.018 4.067 4.076 integrate_v_rspace 119 11.3 0.002 0.003 4.049 4.060 init_scf_run 11 5.9 0.000 0.001 3.629 3.629 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.629 3.629 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.389 3.398 calculate_rho_elec 119 8.7 0.030 0.031 3.389 3.397 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.668 3.299 qs_ot_get_p 119 10.4 0.001 0.001 3.151 3.175 init_scf_loop 11 6.9 0.001 0.002 2.652 2.655 rs_pw_transfer 974 11.9 0.009 0.010 2.535 2.653 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.387 2.477 calculate_first_density_matrix 1 7.0 0.001 0.001 2.382 2.383 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.045 2.371 apply_single 119 13.6 0.000 0.000 2.045 2.371 density_rs2pw 119 9.7 0.003 0.004 1.980 2.071 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.032 2.037 make_m2s 4572 13.5 0.044 0.045 1.846 1.986 jit_kernel_multiply 10 16.2 0.606 1.919 0.606 1.919 calculate_dm_sparse 119 9.5 0.001 0.001 1.907 1.916 make_images 4572 14.5 0.191 0.204 1.760 1.898 grid_integrate_task_list 119 12.3 1.809 1.882 1.809 1.882 potential_pw2rs 119 12.3 0.007 0.008 1.843 1.855 prepare_preconditioner 11 7.9 0.000 0.000 1.844 1.847 make_preconditioner 11 8.9 0.000 0.001 1.844 1.847 pw_transfer 1439 11.6 0.063 0.066 1.790 1.806 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.695 1.778 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.766 1.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.749 1.751 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.700 1.717 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.592 1.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.551 1.556 multiply_cannon_sync_h2d 18288 15.5 1.411 1.554 1.411 1.554 mp_sum_l 7207 12.9 1.165 1.519 1.165 1.519 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.508 1.509 cp_fm_redistribute_end 50 14.0 1.119 1.486 1.121 1.486 ot_diis_step 108 11.5 0.011 0.013 1.478 1.478 cp_fm_diag_elpa_base 50 14.0 0.350 1.417 0.364 1.457 fft3d_ps 1201 14.6 0.505 0.521 1.386 1.402 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.327 1.344 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.275 1.278 acc_transpose_blocks 18288 15.5 0.076 0.077 1.246 1.270 grid_collocate_task_list 119 9.7 1.203 1.269 1.203 1.269 wfi_extrapolate 11 7.9 0.001 0.001 1.199 1.199 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.023 1.032 create_qs_kind_set 1 2.0 0.001 0.002 0.987 1.006 read_qs_kind 2 3.0 0.052 0.925 0.986 1.006 parser_read_line 2807 4.0 0.001 0.001 0.934 0.996 parser_read_line_low 5 5.0 0.003 0.054 0.933 0.996 broadcast_input_information 5 6.0 0.000 0.000 0.930 0.995 mp_bcast_i 105 3.0 0.922 0.987 0.922 0.987 make_images_data 4572 15.5 0.044 0.047 0.812 0.984 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.702 0.911 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.885 0.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.846 mp_alltoall_d11v 2130 13.8 0.727 0.834 0.727 0.834 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.804 0.817 cp_fm_cholesky_invert 11 10.9 0.801 0.805 0.801 0.805 mp_alltoall_z22v 1201 16.6 0.718 0.785 0.718 0.785 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.659 0.747 rs_pw_transfer_RS2PW_140 130 11.5 0.175 0.180 0.631 0.744 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.461 0.713 mp_waitany 9880 13.7 0.582 0.692 0.582 0.692 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.818000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=3.098920 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.920256E+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 0.229 0.349 38.222 38.224 qs_mol_dyn_low 1 2.0 0.005 0.019 36.576 36.585 qs_forces 11 3.9 0.002 0.003 36.442 36.443 qs_energies 11 4.9 0.002 0.004 34.658 34.673 scf_env_do_scf 11 5.9 0.001 0.001 29.260 29.261 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.754 25.755 dbcsr_multiply_generic 2286 12.5 0.102 0.115 19.010 19.190 velocity_verlet 10 3.0 0.005 0.027 18.790 18.795 qs_scf_new_mos 108 7.5 0.001 0.001 17.208 17.270 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.207 17.269 ot_scf_mini 108 9.5 0.002 0.003 16.243 16.294 multiply_cannon 2286 13.5 0.226 0.285 14.792 15.249 multiply_cannon_loop 2286 14.5 0.945 0.976 13.741 14.220 ot_mini 108 10.5 0.001 0.002 10.036 10.101 multiply_cannon_multrec 27432 15.5 2.363 3.232 8.659 9.188 qs_ot_get_derivative 108 11.5 0.001 0.002 8.181 8.232 dbcsr_mm_accdrv_process 47916 15.9 5.284 6.980 6.202 7.399 rebuild_ks_matrix 119 8.3 0.000 0.000 6.731 6.774 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.730 6.773 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.975 6.011 sum_up_and_integrate 119 10.3 0.020 0.022 3.932 3.939 integrate_v_rspace 119 11.3 0.002 0.003 3.912 3.919 init_scf_run 11 5.9 0.000 0.001 3.859 3.860 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.859 3.860 qs_ot_get_p 119 10.4 0.001 0.001 3.516 3.595 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.480 3.519 calculate_rho_elec 119 8.7 0.040 0.046 3.480 3.518 init_scf_loop 11 6.9 0.000 0.002 3.481 3.483 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.894 3.347 mp_waitall_1 145218 16.4 2.636 3.293 2.636 3.293 make_m2s 4572 13.5 0.055 0.060 2.594 2.713 prepare_preconditioner 11 7.9 0.000 0.000 2.624 2.633 make_preconditioner 11 8.9 0.000 0.001 2.624 2.633 make_images 4572 14.5 0.275 0.337 2.486 2.603 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.125 2.575 apply_single 119 13.6 0.000 0.000 2.125 2.575 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.216 2.556 calculate_first_density_matrix 1 7.0 0.001 0.001 2.472 2.474 rs_pw_transfer 974 11.9 0.009 0.010 2.331 2.462 calculate_dm_sparse 119 9.5 0.001 0.001 2.124 2.177 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.050 2.076 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.060 2.069 density_rs2pw 119 9.7 0.003 0.004 1.958 2.060 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.917 1.919 grid_integrate_task_list 119 12.3 1.833 1.913 1.833 1.913 pw_transfer 1439 11.6 0.063 0.067 1.869 1.905 jit_kernel_multiply 10 15.9 0.858 1.844 0.858 1.844 mp_sum_l 7207 12.9 1.262 1.832 1.262 1.832 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.780 1.819 ot_diis_step 108 11.5 0.012 0.014 1.811 1.812 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.735 1.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.720 potential_pw2rs 119 12.3 0.009 0.010 1.683 1.689 acc_transpose_blocks 27432 15.5 0.113 0.115 1.481 1.503 fft3d_ps 1201 14.6 0.535 0.585 1.454 1.478 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.475 1.476 fft_wrap_pw1pw2_140 487 13.2 0.091 0.104 1.416 1.456 cp_fm_redistribute_end 50 14.0 0.969 1.444 0.971 1.445 cp_fm_diag_elpa_base 50 14.0 0.450 1.377 0.471 1.416 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.814 1.405 wfi_extrapolate 11 7.9 0.001 0.001 1.334 1.334 grid_collocate_task_list 119 9.7 1.222 1.321 1.222 1.321 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.281 1.290 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.190 1.192 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.169 1.189 make_images_data 4572 15.5 0.044 0.048 1.011 1.168 dbcsr_complete_redistribute 329 12.2 0.151 0.175 0.873 1.148 create_qs_kind_set 1 2.0 0.001 0.001 1.115 1.144 read_qs_kind 2 3.0 0.059 1.054 1.114 1.144 cp_fm_upper_to_full 72 13.5 0.818 1.143 0.818 1.143 parser_read_line 2807 4.0 0.001 0.001 1.055 1.136 parser_read_line_low 5 5.0 0.003 0.042 1.054 1.136 broadcast_input_information 5 6.0 0.000 0.000 1.051 1.134 mp_bcast_i 105 3.0 1.045 1.131 1.045 1.131 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.851 1.066 multiply_cannon_sync_h2d 27432 15.5 0.986 1.051 0.986 1.051 mp_alltoall_d11v 2130 13.8 0.849 0.955 0.849 0.955 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.637 0.906 acc_transpose_blocks_kernels 27432 16.5 0.268 0.274 0.853 0.874 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.848 0.854 multiply_cannon_metrocomm1 27432 15.5 0.032 0.033 0.349 0.844 mp_alltoall_z22v 1201 16.6 0.801 0.830 0.801 0.830 cp_fm_cholesky_invert 11 10.9 0.783 0.786 0.783 0.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.224000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.363636, yerr=4.183794 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 599.072768E+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 0.028 0.071 30.711 30.711 qs_mol_dyn_low 1 2.0 0.003 0.003 30.462 30.470 qs_forces 11 3.9 0.001 0.002 30.393 30.394 qs_energies 11 4.9 0.001 0.001 28.616 28.618 scf_env_do_scf 11 5.9 0.000 0.001 22.676 22.676 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.041 20.042 velocity_verlet 10 3.0 0.011 0.014 14.969 14.971 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.558 13.628 qs_scf_new_mos 108 7.5 0.001 0.001 11.893 11.924 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.893 11.923 multiply_cannon 2286 13.5 0.229 0.234 10.680 11.241 ot_scf_mini 108 9.5 0.002 0.002 11.192 11.222 multiply_cannon_loop 2286 14.5 0.333 0.344 9.739 9.894 rebuild_ks_matrix 119 8.3 0.000 0.000 6.163 6.191 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.163 6.190 ot_mini 108 10.5 0.001 0.001 6.025 6.062 multiply_cannon_multrec 9144 15.5 1.577 1.823 5.790 6.052 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.454 5.480 qs_ot_get_derivative 108 11.5 0.001 0.001 4.659 4.691 init_scf_run 11 5.9 0.000 0.001 4.465 4.465 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.464 4.465 dbcsr_mm_accdrv_process 12550 15.8 3.120 4.120 4.111 4.209 sum_up_and_integrate 119 10.3 0.022 0.023 3.802 3.808 integrate_v_rspace 119 11.3 0.003 0.003 3.780 3.786 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.566 3.573 calculate_rho_elec 119 8.7 0.059 0.061 3.566 3.573 qs_ot_get_p 119 10.4 0.001 0.001 3.352 3.393 calculate_first_density_matrix 1 7.0 0.000 0.001 3.195 3.196 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.622 2.623 init_scf_loop 11 6.9 0.000 0.000 2.613 2.614 mp_waitall_1 121218 16.5 2.086 2.569 2.086 2.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.444 2.453 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.261 2.263 rs_pw_transfer 974 11.9 0.008 0.009 2.040 2.138 make_m2s 4572 13.5 0.035 0.036 1.936 2.092 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.048 2.048 make_images 4572 14.5 0.268 0.304 1.846 2.001 density_rs2pw 119 9.7 0.003 0.004 1.935 2.000 acc_transpose_blocks 9144 15.5 0.039 0.040 1.947 1.985 grid_integrate_task_list 119 12.3 1.856 1.928 1.856 1.928 pw_transfer 1439 11.6 0.063 0.065 1.902 1.917 prepare_preconditioner 11 7.9 0.000 0.000 1.849 1.853 make_preconditioner 11 8.9 0.000 0.000 1.849 1.853 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.812 1.828 jit_kernel_multiply 10 15.6 0.954 1.788 0.954 1.788 calculate_dm_sparse 119 9.5 0.000 0.000 1.757 1.778 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.728 1.758 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 1.705 1.746 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.707 1.708 cp_fm_redistribute_end 50 14.0 0.856 1.667 0.858 1.669 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.645 1.664 jit_kernel_transpose 5 15.6 1.589 1.630 1.589 1.630 cp_fm_diag_elpa_base 50 14.0 0.765 1.562 0.809 1.619 potential_pw2rs 119 12.3 0.010 0.011 1.544 1.550 fft3d_ps 1201 14.6 0.541 0.553 1.477 1.490 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.422 1.439 grid_collocate_task_list 119 9.7 1.277 1.363 1.277 1.363 ot_diis_step 108 11.5 0.013 0.013 1.346 1.346 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.270 1.285 qs_energies_init_hamiltonians 11 5.9 0.025 0.030 1.252 1.252 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.190 1.225 apply_single 119 13.6 0.000 0.000 1.190 1.225 wfi_extrapolate 11 7.9 0.001 0.001 1.212 1.213 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.153 1.159 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.850 1.103 make_images_data 4572 15.5 0.038 0.042 0.881 1.086 mp_alltoall_d11v 2130 13.8 0.963 1.053 0.963 1.053 cp_fm_cholesky_invert 11 10.9 0.954 0.958 0.954 0.958 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.913 mp_alltoall_z22v 1201 16.6 0.809 0.850 0.809 0.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.836 multiply_cannon_sync_h2d 9144 15.5 0.704 0.773 0.704 0.773 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.759 0.773 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.678 0.726 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.317 0.701 mp_allgather_i34 2286 14.5 0.272 0.695 0.272 0.695 mp_sum_l 7207 12.9 0.550 0.666 0.550 0.666 dbcsr_complete_redistribute 329 12.2 0.163 0.168 0.591 0.620 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.711000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.454545, yerr=5.416534 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 749.285376E+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 0.026 0.045 40.844 40.844 qs_mol_dyn_low 1 2.0 0.003 0.003 40.612 40.619 qs_forces 11 3.9 0.005 0.009 40.359 40.368 qs_energies 11 4.9 0.001 0.001 38.432 38.438 scf_env_do_scf 11 5.9 0.001 0.001 32.384 32.384 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.543 24.543 velocity_verlet 10 3.0 0.002 0.002 22.891 22.913 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.805 18.026 qs_scf_new_mos 108 7.5 0.001 0.001 15.758 15.859 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.757 15.858 multiply_cannon 2286 13.5 0.304 0.313 13.979 14.962 ot_scf_mini 108 9.5 0.002 0.002 14.685 14.784 multiply_cannon_loop 2286 14.5 0.344 0.349 12.725 13.686 multiply_cannon_multrec 9144 15.5 3.377 4.692 8.966 9.075 ot_mini 108 10.5 0.001 0.001 8.766 8.887 init_scf_loop 11 6.9 0.000 0.000 7.815 7.817 rebuild_ks_matrix 119 8.3 0.000 0.000 6.802 6.953 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.801 6.952 prepare_preconditioner 11 7.9 0.000 0.000 6.886 6.901 make_preconditioner 11 8.9 0.000 0.000 6.886 6.901 qs_ot_get_derivative 108 11.5 0.001 0.001 6.720 6.823 dbcsr_mm_accdrv_process 12550 15.8 4.395 5.659 5.464 6.786 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.459 6.773 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.158 6.294 cp_fm_upper_to_full 72 14.2 3.195 4.557 3.195 4.557 init_scf_run 11 5.9 0.000 0.001 3.998 3.998 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.998 3.998 sum_up_and_integrate 119 10.3 0.038 0.038 3.640 3.647 integrate_v_rspace 119 11.3 0.003 0.003 3.602 3.609 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.587 3.595 calculate_rho_elec 119 8.7 0.118 0.121 3.586 3.595 mp_waitall_1 97218 16.6 2.558 3.464 2.558 3.464 qs_ot_get_p 119 10.4 0.001 0.001 3.188 3.329 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.464 2.852 dbcsr_complete_redistribute 329 12.2 0.288 0.291 1.970 2.786 calculate_first_density_matrix 1 7.0 0.001 0.001 2.579 2.579 calculate_dm_sparse 119 9.5 0.000 0.000 2.473 2.490 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.471 apply_single 119 13.6 0.000 0.000 2.168 2.471 make_m2s 4572 13.5 0.038 0.038 2.298 2.471 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.656 2.462 make_images 4572 14.5 0.352 0.381 2.177 2.350 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.422 2.221 mp_alltoall_i22 627 13.8 1.416 2.205 1.416 2.205 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.292 2.160 pw_transfer 1439 11.6 0.067 0.068 2.007 2.018 ot_diis_step 108 11.5 0.014 0.014 2.017 2.018 grid_integrate_task_list 119 12.3 1.997 2.003 1.997 2.003 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.951 1.951 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.911 1.923 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.913 1.914 jit_kernel_multiply 8 15.6 1.041 1.867 1.041 1.867 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.792 1.797 density_rs2pw 119 9.7 0.003 0.003 1.748 1.768 mp_sum_l 7207 12.9 0.990 1.764 0.990 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.722 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.669 1.719 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.636 1.637 fft_wrap_pw1pw2_140 487 13.2 0.089 0.090 1.603 1.613 fft3d_ps 1201 14.6 0.596 0.611 1.544 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.545 1.554 cp_fm_cholesky_invert 11 10.9 1.503 1.507 1.503 1.507 grid_collocate_task_list 119 9.7 1.441 1.462 1.441 1.462 rs_pw_transfer 974 11.9 0.009 0.009 1.412 1.443 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.369 1.369 cp_fm_diag_elpa_base 50 14.0 1.224 1.276 1.367 1.368 wfi_extrapolate 11 7.9 0.001 0.001 1.360 1.360 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.101 1.359 make_images_data 4572 15.5 0.042 0.045 1.121 1.342 potential_pw2rs 119 12.3 0.014 0.014 1.217 1.220 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.141 1.163 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.127 1.143 multiply_cannon_sync_h2d 9144 15.5 1.038 1.042 1.038 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.984 1.028 mp_alltoall_d11v 2130 13.8 1.001 1.012 1.001 1.012 qs_create_task_list 11 7.9 0.001 0.002 0.951 0.975 generate_qs_task_list 11 8.9 0.373 0.393 0.950 0.974 acc_transpose_blocks 9144 15.5 0.039 0.039 0.925 0.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.854 0.869 mp_alltoall_z22v 1201 16.6 0.812 0.823 0.812 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.844000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.090909, yerr=13.317484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 502.153216E+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 1441692. 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 0.022 0.075 84.540 84.547 qs_mol_dyn_low 1 2.0 0.003 0.003 84.231 84.250 qs_forces 11 3.9 0.005 0.010 84.144 84.145 qs_energies 11 4.9 0.001 0.002 81.231 81.244 scf_env_do_scf 11 5.9 0.000 0.001 72.293 72.295 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.518 66.518 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.853 53.236 qs_scf_new_mos 99 7.5 0.000 0.001 48.831 48.946 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.830 48.945 ot_scf_mini 99 9.5 0.002 0.002 46.379 46.507 multiply_cannon 2055 13.4 0.177 0.181 42.845 43.948 velocity_verlet 10 3.0 0.010 0.020 43.325 43.325 multiply_cannon_loop 2055 14.4 1.577 1.624 41.896 43.019 ot_mini 99 10.5 0.001 0.001 28.117 28.215 qs_ot_get_derivative 99 11.5 0.001 0.001 21.258 21.359 multiply_cannon_multrec 49320 15.4 12.051 12.566 17.103 17.919 rebuild_ks_matrix 110 8.3 0.000 0.001 14.658 14.881 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.658 14.881 mp_waitall_1 241148 16.1 12.297 13.185 12.297 13.185 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.831 13.040 multiply_cannon_sync_h2d 49320 15.4 9.885 10.458 9.885 10.458 qs_ot_get_p 110 10.4 0.001 0.001 9.534 9.663 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.176 8.302 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.496 8.016 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.147 7.611 apply_single 110 13.6 0.000 0.001 7.147 7.610 sum_up_and_integrate 110 10.3 0.023 0.028 7.198 7.212 integrate_v_rspace 110 11.3 0.002 0.003 7.175 7.195 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.814 6.930 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.872 6.929 calculate_rho_elec 110 8.6 0.020 0.024 6.813 6.929 init_scf_run 11 5.9 0.000 0.001 6.880 6.881 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.880 6.881 ot_diis_step 99 11.5 0.006 0.006 6.576 6.578 mp_sum_l 6514 12.8 5.618 6.334 5.618 6.334 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.133 6.183 init_scf_loop 11 6.9 0.000 0.000 5.746 5.746 dbcsr_mm_accdrv_process 87628 16.1 2.089 2.215 4.931 5.245 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.096 5.096 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.619 4.621 cp_fm_redistribute_end 48 14.0 4.026 4.596 4.029 4.596 cp_fm_diag_elpa_base 48 14.0 0.561 4.503 0.564 4.527 rs_pw_transfer 902 11.9 0.012 0.014 4.316 4.478 make_m2s 4110 13.4 0.061 0.066 3.965 4.065 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.862 4.064 wfi_extrapolate 11 7.9 0.001 0.001 4.023 4.023 make_images 4110 14.4 0.178 0.192 3.869 3.973 density_rs2pw 110 9.6 0.004 0.004 3.802 3.953 calculate_dm_sparse 110 9.5 0.000 0.001 3.813 3.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.705 3.709 prepare_preconditioner 11 7.9 0.000 0.000 3.583 3.605 make_preconditioner 11 8.9 0.000 0.000 3.583 3.605 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.454 3.523 grid_integrate_task_list 110 12.3 3.252 3.436 3.252 3.436 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.379 3.422 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.277 3.343 pw_transfer 1331 11.6 0.054 0.066 3.072 3.139 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.981 3.047 potential_pw2rs 110 12.3 0.006 0.007 2.803 2.821 calculate_first_density_matrix 1 7.0 0.000 0.001 2.769 2.775 mp_alltoall_d11v 2046 13.8 2.191 2.602 2.191 2.602 jit_kernel_multiply 13 15.9 2.557 2.580 2.557 2.580 fft3d_ps 1111 14.6 0.751 0.839 2.519 2.579 fft_wrap_pw1pw2_140 451 13.1 0.169 0.187 2.411 2.478 acc_transpose_blocks 49320 15.4 0.216 0.228 2.222 2.313 grid_collocate_task_list 110 9.6 2.089 2.240 2.089 2.240 cp_fm_cholesky_invert 11 10.9 2.012 2.016 2.012 2.016 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.968 1.995 mp_sum_d 3883 11.9 1.508 1.968 1.508 1.968 mp_waitany 14300 13.8 1.789 1.932 1.789 1.932 make_images_data 4110 15.4 0.043 0.046 1.805 1.927 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.555 1.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.547000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.937363 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.280384E+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 2695002. 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 0.045 0.118 70.281 70.282 qs_mol_dyn_low 1 2.0 0.003 0.003 69.903 69.924 qs_forces 11 3.9 0.002 0.002 69.786 69.787 qs_energies 11 4.9 0.007 0.037 66.439 66.443 scf_env_do_scf 11 5.9 0.001 0.001 57.222 57.225 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.644 49.645 dbcsr_multiply_generic 2055 12.4 0.112 0.115 38.223 38.404 velocity_verlet 10 3.0 0.001 0.002 36.107 36.109 multiply_cannon 2055 13.4 0.221 0.245 31.544 32.917 qs_scf_new_mos 99 7.5 0.001 0.001 32.645 32.787 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.645 32.787 multiply_cannon_loop 2055 14.4 0.937 0.959 30.227 31.258 ot_scf_mini 99 9.5 0.003 0.003 30.990 31.113 ot_mini 99 10.5 0.001 0.001 18.407 18.538 multiply_cannon_multrec 24660 15.4 7.641 9.535 14.214 15.898 rebuild_ks_matrix 110 8.3 0.000 0.001 13.824 13.934 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.823 13.933 qs_ot_get_derivative 99 11.5 0.001 0.001 12.637 12.769 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.164 12.265 mp_waitall_1 186928 16.3 7.949 10.707 7.949 10.707 multiply_cannon_sync_h2d 24660 15.4 7.026 8.015 7.026 8.015 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.207 7.800 init_scf_loop 11 6.9 0.000 0.002 7.543 7.546 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.534 7.254 apply_single 110 13.6 0.000 0.001 6.534 7.254 init_scf_run 11 5.9 0.000 0.001 6.741 6.744 scf_env_initial_rho_setup 11 6.9 0.004 0.005 6.740 6.743 dbcsr_mm_accdrv_process 52282 16.1 4.940 5.664 6.412 6.696 sum_up_and_integrate 110 10.3 0.028 0.033 6.668 6.682 integrate_v_rspace 110 11.3 0.002 0.003 6.640 6.651 qs_ot_get_p 110 10.4 0.001 0.001 5.968 6.111 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.914 5.924 calculate_rho_elec 110 8.6 0.039 0.047 5.913 5.924 ot_diis_step 99 11.5 0.010 0.011 5.723 5.723 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.780 5.545 prepare_preconditioner 11 7.9 0.000 0.000 5.497 5.517 make_preconditioner 11 8.9 0.000 0.000 5.497 5.517 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.065 5.231 make_m2s 4110 13.4 0.057 0.060 4.221 4.719 make_images 4110 14.4 0.400 0.447 4.112 4.608 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.089 4.110 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.661 3.661 pw_transfer 1331 11.6 0.067 0.072 3.385 3.532 wfi_extrapolate 11 7.9 0.001 0.001 3.526 3.526 density_rs2pw 110 9.6 0.004 0.004 3.213 3.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.455 3.456 rs_pw_transfer 902 11.9 0.012 0.013 3.243 3.450 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.274 3.423 grid_integrate_task_list 110 12.3 3.137 3.315 3.137 3.315 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.148 3.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.109 3.173 calculate_dm_sparse 110 9.5 0.001 0.001 3.127 3.158 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.130 3.132 calculate_first_density_matrix 1 7.0 0.004 0.009 3.126 3.131 cp_fm_redistribute_end 48 14.0 2.336 3.106 2.338 3.106 cp_fm_diag_elpa_base 48 14.0 0.734 2.978 0.765 3.073 fft_wrap_pw1pw2_140 451 13.1 0.202 0.218 2.736 2.882 make_images_data 4110 15.4 0.047 0.051 2.362 2.876 hybrid_alltoall_any 4261 16.3 0.102 0.440 2.071 2.812 fft3d_ps 1111 14.6 1.101 1.302 2.630 2.768 cp_fm_cholesky_invert 11 10.9 2.552 2.561 2.552 2.561 potential_pw2rs 110 12.3 0.009 0.009 2.530 2.544 mp_sum_l 6514 12.8 1.902 2.479 1.902 2.479 jit_kernel_multiply 11 16.3 1.119 2.420 1.119 2.420 grid_collocate_task_list 110 9.6 2.117 2.287 2.117 2.287 mp_alltoall_d11v 2046 13.8 1.820 2.195 1.820 2.195 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.941 1.963 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.930 1.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.823 1.838 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.713 1.724 mp_allgather_i34 2055 14.4 0.646 1.673 0.646 1.673 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.790 1.667 acc_transpose_blocks 24660 15.4 0.108 0.113 1.606 1.665 mp_irecv_dv 57340 16.2 0.666 1.552 0.666 1.552 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.475 dbcsr_complete_redistribute 325 12.2 0.243 0.304 1.152 1.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.282000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.000000, yerr=7.920055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 657.297408E+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 0.015 0.031 60.812 60.812 qs_mol_dyn_low 1 2.0 0.003 0.003 60.509 60.520 qs_forces 11 3.9 0.002 0.002 60.436 60.437 qs_energies 11 4.9 0.007 0.024 57.098 57.103 scf_env_do_scf 11 5.9 0.000 0.001 49.072 49.072 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.224 40.224 velocity_verlet 10 3.0 0.002 0.002 33.052 33.054 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.686 28.950 qs_scf_new_mos 99 7.5 0.001 0.001 25.314 25.445 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.314 25.444 ot_scf_mini 99 9.5 0.002 0.003 24.042 24.193 multiply_cannon 2055 13.4 0.211 0.218 22.178 23.350 multiply_cannon_loop 2055 14.4 0.622 0.636 20.986 22.203 ot_mini 99 10.5 0.001 0.001 13.771 13.920 rebuild_ks_matrix 110 8.3 0.000 0.000 12.473 12.619 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.473 12.619 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.003 11.132 multiply_cannon_multrec 16440 15.4 3.976 5.119 9.757 10.884 mp_waitall_1 146766 16.3 7.344 10.567 7.344 10.567 qs_ot_get_derivative 99 11.5 0.001 0.001 9.327 9.478 init_scf_loop 11 6.9 0.000 0.000 8.807 8.808 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.209 7.277 prepare_preconditioner 11 7.9 0.000 0.000 7.026 7.044 make_preconditioner 11 8.9 0.000 0.000 7.026 7.044 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.368 6.723 sum_up_and_integrate 110 10.3 0.033 0.034 6.701 6.714 integrate_v_rspace 110 11.3 0.002 0.003 6.668 6.682 dbcsr_mm_accdrv_process 34862 16.1 4.979 5.336 5.634 5.774 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.720 5.727 calculate_rho_elec 110 8.6 0.058 0.058 5.719 5.726 qs_ot_get_p 110 10.4 0.001 0.001 5.413 5.556 init_scf_run 11 5.9 0.000 0.001 5.523 5.523 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.523 5.523 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.933 5.354 apply_single 110 13.6 0.000 0.000 4.933 5.354 make_m2s 4110 13.4 0.050 0.052 4.346 4.681 make_images 4110 14.4 0.391 0.509 4.230 4.563 multiply_cannon_sync_h2d 16440 15.4 3.699 4.416 3.699 4.416 ot_diis_step 99 11.5 0.011 0.011 4.415 4.415 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.107 3.736 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.715 3.719 grid_integrate_task_list 110 12.3 3.192 3.393 3.192 3.393 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.371 3.371 pw_transfer 1331 11.6 0.066 0.073 3.293 3.304 rs_pw_transfer 902 11.9 0.010 0.011 3.016 3.241 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.182 3.195 density_rs2pw 110 9.6 0.004 0.004 2.993 3.164 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 make_images_data 4110 15.4 0.043 0.047 2.454 2.894 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.801 2.816 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.157 2.791 cp_fm_redistribute_end 48 14.0 1.755 2.785 1.759 2.787 cp_fm_cholesky_invert 11 10.9 2.758 2.764 2.758 2.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.758 2.759 cp_fm_diag_elpa_base 48 14.0 0.952 2.590 1.009 2.696 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 2.681 2.695 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.518 2.595 calculate_dm_sparse 110 9.5 0.001 0.001 2.561 2.591 fft3d_ps 1111 14.6 1.062 1.074 2.494 2.506 calculate_first_density_matrix 1 7.0 0.001 0.001 2.494 2.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.444 2.491 potential_pw2rs 110 12.3 0.011 0.011 2.456 2.474 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.864 2.429 grid_collocate_task_list 110 9.6 2.176 2.369 2.176 2.369 mp_irecv_dv 48980 15.7 0.795 2.302 0.795 2.302 mp_alltoall_d11v 2046 13.8 1.888 2.294 1.888 2.294 mp_sum_l 6514 12.8 1.547 2.192 1.547 2.192 qs_energies_init_hamiltonians 11 5.9 0.008 0.026 2.036 2.036 dbcsr_complete_redistribute 325 12.2 0.318 0.357 1.424 1.877 cp_fm_upper_to_full 70 13.6 1.385 1.850 1.385 1.850 cp_fm_cholesky_decompose 22 10.9 1.762 1.785 1.762 1.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.669 1.686 mp_allgather_i34 2055 14.4 0.515 1.524 0.515 1.524 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.494 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.470 1.483 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.946 1.395 rs_gather_matrices 110 12.3 0.137 0.150 0.964 1.358 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.301 1.310 mp_waitany 17072 13.8 1.106 1.268 1.106 1.268 mp_alltoall_z22v 1111 16.6 1.219 1.242 1.219 1.242 acc_transpose_blocks 16440 15.4 0.073 0.075 1.224 1.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.812000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.545455, yerr=7.655819 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 741.634048E+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 0.090 0.234 65.974 65.976 qs_mol_dyn_low 1 2.0 0.007 0.037 65.284 65.293 qs_forces 11 3.9 0.039 0.113 65.206 65.215 qs_energies 11 4.9 0.012 0.032 61.804 61.828 scf_env_do_scf 11 5.9 0.001 0.001 52.304 52.307 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.552 40.553 velocity_verlet 10 3.0 0.001 0.002 36.777 36.779 dbcsr_multiply_generic 2055 12.4 0.115 0.122 29.552 29.754 qs_scf_new_mos 99 7.5 0.001 0.001 25.985 26.094 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.984 26.094 ot_scf_mini 99 9.5 0.003 0.003 24.364 24.474 multiply_cannon 2055 13.4 0.247 0.276 22.417 23.773 multiply_cannon_loop 2055 14.4 0.886 0.906 21.054 21.633 ot_mini 99 10.5 0.001 0.001 13.898 14.027 multiply_cannon_multrec 24660 15.4 4.207 6.754 12.627 13.781 rebuild_ks_matrix 110 8.3 0.000 0.000 12.034 12.139 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.033 12.139 init_scf_loop 11 6.9 0.000 0.001 11.711 11.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.658 10.751 prepare_preconditioner 11 7.9 0.000 0.000 9.915 9.929 make_preconditioner 11 8.9 0.000 0.001 9.915 9.929 qs_ot_get_derivative 99 11.5 0.001 0.001 9.755 9.868 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.097 9.580 dbcsr_mm_accdrv_process 52304 16.0 7.004 8.662 8.272 9.200 mp_waitall_1 126806 16.4 4.563 6.578 4.563 6.578 sum_up_and_integrate 110 10.3 0.038 0.040 6.364 6.378 integrate_v_rspace 110 11.3 0.003 0.003 6.327 6.339 init_scf_run 11 5.9 0.000 0.001 5.969 5.970 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.969 5.970 make_m2s 4110 13.4 0.059 0.061 5.552 5.858 make_images 4110 14.4 0.578 0.700 5.411 5.712 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.699 5.709 calculate_rho_elec 110 8.6 0.077 0.081 5.698 5.708 qs_ot_get_p 110 10.4 0.001 0.001 5.474 5.619 cp_fm_upper_to_full 70 13.8 3.341 4.759 3.341 4.759 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.021 4.124 apply_single 110 13.6 0.000 0.000 4.020 4.123 ot_diis_step 99 11.5 0.011 0.011 4.106 4.107 dbcsr_complete_redistribute 325 12.2 0.454 0.517 2.841 4.007 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.692 3.709 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.306 3.463 grid_integrate_task_list 110 12.3 3.275 3.392 3.275 3.392 multiply_cannon_sync_h2d 24660 15.4 3.191 3.350 3.191 3.350 rs_pw_transfer 902 11.9 0.010 0.011 3.008 3.336 pw_transfer 1331 11.6 0.066 0.073 3.305 3.328 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.240 3.301 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.388 3.228 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.195 3.220 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.427 3.201 make_images_data 4110 15.4 0.046 0.049 2.862 3.192 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.175 3.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.046 3.047 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.973 3.032 qs_energies_init_hamiltonians 11 5.9 0.023 0.027 2.957 2.979 density_rs2pw 110 9.6 0.004 0.004 2.791 2.970 wfi_extrapolate 11 7.9 0.001 0.001 2.954 2.954 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.809 2.945 calculate_dm_sparse 110 9.5 0.001 0.001 2.906 2.937 calculate_first_density_matrix 1 7.0 0.001 0.001 2.890 2.893 mp_alltoall_i22 605 13.7 1.674 2.891 1.674 2.891 fft_wrap_pw1pw2_140 451 13.1 0.225 0.258 2.720 2.747 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.622 2.662 cp_fm_cholesky_invert 11 10.9 2.630 2.639 2.630 2.639 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.588 2.589 cp_fm_redistribute_end 48 14.0 1.288 2.559 1.289 2.560 cp_fm_diag_elpa_base 48 14.0 1.188 2.430 1.268 2.540 fft3d_ps 1111 14.6 1.065 1.103 2.475 2.508 grid_collocate_task_list 110 9.6 2.227 2.352 2.227 2.352 mp_alltoall_d11v 2046 13.8 2.008 2.177 2.008 2.177 potential_pw2rs 110 12.3 0.013 0.013 2.154 2.162 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.773 1.981 mp_waitany 13376 13.8 1.495 1.892 1.495 1.892 jit_kernel_multiply 10 15.7 0.939 1.846 0.939 1.846 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.690 1.723 mp_allgather_i34 2055 14.4 0.462 1.711 0.462 1.711 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.602 1.703 cp_fm_cholesky_decompose 22 10.9 1.647 1.694 1.647 1.694 rs_pw_transfer_RS2PW_140 121 11.5 0.243 0.251 1.328 1.653 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.632 acc_transpose_blocks 24660 15.4 0.106 0.108 1.562 1.596 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.555 1.574 mp_sum_l 6514 12.8 0.958 1.572 0.958 1.572 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.855 1.455 mp_irecv_dv 62702 16.1 0.756 1.376 0.756 1.376 build_core_hamiltonian_matrix 11 6.9 0.007 0.045 0.945 1.365 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.976000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.090909, yerr=13.681139 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 826.982400E+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 0.022 0.058 55.679 55.680 qs_mol_dyn_low 1 2.0 0.003 0.003 55.365 55.397 qs_forces 11 3.9 0.002 0.002 55.282 55.282 qs_energies 11 4.9 0.001 0.002 51.581 51.585 scf_env_do_scf 11 5.9 0.000 0.001 43.239 43.239 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.407 35.408 velocity_verlet 10 3.0 0.010 0.016 31.097 31.100 dbcsr_multiply_generic 2055 12.4 0.112 0.120 23.050 23.241 qs_scf_new_mos 99 7.5 0.001 0.001 20.933 20.982 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.932 20.982 ot_scf_mini 99 9.5 0.002 0.002 19.688 19.722 multiply_cannon 2055 13.4 0.246 0.259 17.333 18.862 multiply_cannon_loop 2055 14.4 0.324 0.337 15.963 16.373 rebuild_ks_matrix 110 8.3 0.000 0.000 11.801 11.830 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.800 11.829 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.507 10.532 ot_mini 99 10.5 0.001 0.001 10.459 10.486 multiply_cannon_multrec 8220 15.4 3.194 4.388 7.482 8.479 mp_waitall_1 106626 16.5 6.271 8.022 6.271 8.022 init_scf_loop 11 6.9 0.000 0.000 7.785 7.786 qs_ot_get_derivative 99 11.5 0.001 0.001 6.688 6.721 sum_up_and_integrate 110 10.3 0.048 0.049 6.433 6.446 integrate_v_rspace 110 11.3 0.003 0.003 6.385 6.399 prepare_preconditioner 11 7.9 0.000 0.000 6.141 6.145 make_preconditioner 11 8.9 0.000 0.000 6.141 6.145 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.731 5.801 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.702 5.715 calculate_rho_elec 110 8.6 0.115 0.116 5.701 5.714 qs_ot_get_p 110 10.4 0.001 0.001 5.247 5.273 init_scf_run 11 5.9 0.000 0.001 5.224 5.224 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.224 5.224 dbcsr_mm_accdrv_process 17442 15.9 2.915 3.979 4.158 5.065 make_m2s 4110 13.4 0.039 0.041 4.367 4.636 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.014 4.563 make_images 4110 14.4 0.635 0.687 4.236 4.505 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.751 3.754 ot_diis_step 99 11.5 0.012 0.012 3.750 3.751 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.673 3.699 apply_single 110 13.6 0.000 0.000 3.673 3.698 grid_integrate_task_list 110 12.3 3.338 3.479 3.338 3.479 pw_transfer 1331 11.6 0.066 0.071 3.454 3.467 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.447 3.447 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.343 3.360 multiply_cannon_sync_h2d 8220 15.4 2.915 3.048 2.915 3.048 cp_fm_cholesky_invert 11 10.9 3.000 3.005 3.000 3.005 make_images_data 4110 15.4 0.038 0.043 2.486 2.945 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.896 2.897 density_rs2pw 110 9.6 0.004 0.004 2.758 2.877 cp_fm_redistribute_end 48 14.0 0.734 2.860 0.739 2.862 hybrid_alltoall_any 4261 16.3 0.200 0.868 2.399 2.841 cp_fm_diag_elpa_base 48 14.0 1.939 2.658 2.114 2.824 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 2.785 2.805 wfi_extrapolate 11 7.9 0.001 0.001 2.731 2.731 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.664 2.664 fft3d_ps 1111 14.6 1.114 1.135 2.600 2.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.571 2.572 grid_collocate_task_list 110 9.6 2.325 2.542 2.325 2.542 calculate_dm_sparse 110 9.5 0.001 0.001 2.494 2.536 rs_pw_transfer 902 11.9 0.010 0.011 2.361 2.511 calculate_first_density_matrix 1 7.0 0.001 0.001 2.402 2.404 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.220 2.231 potential_pw2rs 110 12.3 0.015 0.015 2.178 2.193 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 2.019 mp_alltoall_d11v 2046 13.8 1.723 1.988 1.723 1.988 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.944 1.958 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.803 1.820 cp_fm_cholesky_decompose 22 10.9 1.676 1.707 1.676 1.707 mp_allgather_i34 2055 14.4 0.554 1.684 0.554 1.684 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.633 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.509 1.622 dbcsr_complete_redistribute 325 12.2 0.554 0.574 1.462 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.452 1.465 qs_create_task_list 11 7.9 0.001 0.001 1.228 1.325 generate_qs_task_list 11 8.9 0.379 0.446 1.227 1.324 mp_alltoall_z22v 1111 16.6 1.253 1.271 1.253 1.271 mp_waitany 9240 13.8 1.048 1.212 1.048 1.212 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.173 1.195 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.479 1.167 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.774 1.166 mp_irecv_dv 24056 15.7 0.455 1.124 0.455 1.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.680000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=777.000000, yerr=11.661904 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 1.196876E+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 0.031 0.056 87.826 87.826 qs_mol_dyn_low 1 2.0 0.003 0.003 87.402 87.424 qs_forces 11 3.9 0.018 0.021 87.331 87.332 qs_energies 11 4.9 0.002 0.002 83.289 83.291 scf_env_do_scf 11 5.9 0.001 0.001 72.556 72.556 velocity_verlet 10 3.0 0.002 0.002 55.366 55.372 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.589 43.589 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.756 29.848 init_scf_loop 11 6.9 0.000 0.000 28.890 28.891 qs_scf_new_mos 99 7.5 0.001 0.001 27.100 27.149 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.099 27.148 prepare_preconditioner 11 7.9 0.000 0.000 26.875 26.883 make_preconditioner 11 8.9 0.000 0.000 26.875 26.883 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.958 26.294 ot_scf_mini 99 9.5 0.002 0.002 25.348 25.387 multiply_cannon 2055 13.4 0.345 0.372 22.453 23.208 multiply_cannon_loop 2055 14.4 0.343 0.344 20.603 20.901 cp_fm_upper_to_full 70 14.2 12.768 18.355 12.768 18.355 ot_mini 99 10.5 0.001 0.001 14.077 14.119 rebuild_ks_matrix 110 8.3 0.001 0.001 13.305 13.358 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.304 13.358 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.118 12.167 dbcsr_complete_redistribute 325 12.2 1.024 1.054 7.534 10.773 multiply_cannon_multrec 8220 15.4 4.435 4.614 10.185 10.325 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.489 9.721 qs_ot_get_derivative 99 11.5 0.001 0.001 9.422 9.457 mp_waitall_1 87304 16.6 8.175 9.191 8.175 9.191 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.901 9.111 mp_alltoall_i22 605 13.7 5.483 8.722 5.483 8.722 sum_up_and_integrate 110 10.3 0.090 0.091 6.616 6.629 integrate_v_rspace 110 11.3 0.003 0.004 6.526 6.540 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.492 6.528 calculate_rho_elec 110 8.6 0.226 0.227 6.491 6.528 init_scf_run 11 5.9 0.000 0.001 6.464 6.464 scf_env_initial_rho_setup 11 6.9 0.002 0.002 6.464 6.464 qs_ot_get_p 110 10.4 0.001 0.001 5.979 6.032 make_m2s 4110 13.4 0.043 0.043 5.449 5.988 dbcsr_mm_accdrv_process 11614 15.7 3.224 3.347 5.606 5.852 make_images 4110 14.4 0.880 0.928 5.261 5.799 cp_fm_cholesky_invert 11 10.9 5.482 5.486 5.482 5.486 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.040 5.393 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.710 5.189 apply_single 110 13.6 0.000 0.000 4.710 5.189 ot_diis_step 99 11.5 0.015 0.016 4.624 4.625 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.335 4.343 multiply_cannon_sync_h2d 8220 15.4 3.946 3.951 3.946 3.951 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.885 3.885 pw_transfer 1331 11.6 0.073 0.073 3.876 3.879 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.751 3.755 grid_integrate_task_list 110 12.3 3.657 3.707 3.657 3.707 make_images_data 4110 15.4 0.041 0.044 2.975 3.695 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 3.665 3.666 hybrid_alltoall_any 4261 16.3 0.257 0.553 2.895 3.627 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.943 3.398 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.365 3.368 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.322 3.323 cp_fm_diag_elpa_base 48 14.0 2.765 2.967 3.320 3.321 calculate_dm_sparse 110 9.5 0.001 0.001 3.285 3.306 wfi_extrapolate 11 7.9 0.001 0.001 3.278 3.278 fft_wrap_pw1pw2_140 451 13.1 0.220 0.224 3.232 3.234 calculate_first_density_matrix 1 7.0 0.001 0.001 3.079 3.080 density_rs2pw 110 9.6 0.004 0.004 3.052 3.071 fft3d_ps 1111 14.6 1.259 1.271 2.957 2.965 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.864 2.873 grid_collocate_task_list 110 9.6 2.627 2.648 2.627 2.648 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.564 2.582 jit_kernel_multiply 10 15.4 2.181 2.335 2.181 2.335 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.209 2.267 cp_fm_cholesky_decompose 22 10.9 2.230 2.246 2.230 2.246 rs_pw_transfer 902 11.9 0.011 0.011 2.196 2.245 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.178 potential_pw2rs 110 12.3 0.021 0.021 2.117 2.122 mp_alltoall_d11v 2046 13.8 1.939 1.987 1.939 1.987 qs_create_task_list 11 7.9 0.003 0.003 1.929 1.976 generate_qs_task_list 11 8.9 0.738 0.791 1.926 1.974 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.833 1.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.797 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.826000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1122.090909, yerr=25.798007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.310592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955597216 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2919571. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.049 206.375 206.377 qs_mol_dyn_low 1 2.0 0.003 0.005 205.993 206.012 qs_forces 11 3.9 0.006 0.022 205.864 205.868 qs_energies 11 4.9 0.002 0.005 200.220 200.237 scf_env_do_scf 11 5.9 0.001 0.001 183.219 183.223 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 162.347 162.349 dbcsr_multiply_generic 2507 12.6 0.175 0.179 124.851 125.408 velocity_verlet 10 3.0 0.003 0.013 124.002 124.005 qs_scf_new_mos 117 7.6 0.001 0.001 122.539 122.715 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.539 122.714 ot_scf_mini 117 9.6 0.003 0.004 115.898 116.065 multiply_cannon 2507 13.6 0.237 0.247 100.981 102.739 multiply_cannon_loop 2507 14.6 2.134 2.205 98.770 100.273 ot_mini 117 10.6 0.001 0.002 66.506 66.674 multiply_cannon_multrec 60168 15.6 32.942 35.656 41.332 43.091 qs_ot_get_derivative 117 11.6 0.001 0.002 41.577 41.733 rebuild_ks_matrix 128 8.3 0.001 0.001 34.002 34.278 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.001 34.278 mp_waitall_1 291448 16.2 29.279 32.201 29.279 32.201 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.537 30.773 multiply_cannon_sync_h2d 60168 15.6 27.274 29.416 27.274 29.416 qs_ot_get_p 128 10.4 0.001 0.001 27.529 27.758 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.236 25.067 apply_single 128 13.6 0.001 0.001 24.235 25.066 ot_diis_step 117 11.6 0.008 0.011 24.549 24.550 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.842 20.964 init_scf_loop 11 6.9 0.000 0.000 20.796 20.797 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.514 19.657 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.266 18.267 multiply_cannon_metrocomm3 60168 15.6 0.111 0.118 15.729 17.468 prepare_preconditioner 11 7.9 0.000 0.000 16.181 16.234 make_preconditioner 11 8.9 0.000 0.000 16.181 16.234 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.407 15.554 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.361 15.366 cp_fm_redistribute_end 83 14.4 12.126 15.283 12.140 15.289 cp_fm_diag_elpa_base 83 14.4 3.103 14.981 3.135 15.103 sum_up_and_integrate 128 10.3 0.052 0.063 14.527 14.545 integrate_v_rspace 128 11.3 0.003 0.004 14.474 14.494 make_m2s 5014 13.6 0.103 0.111 14.017 14.411 make_images 5014 14.6 0.402 0.419 13.836 14.240 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.184 13.293 calculate_rho_elec 128 8.7 0.045 0.063 13.183 13.292 init_scf_run 11 5.9 0.000 0.001 12.620 12.620 scf_env_initial_rho_setup 11 6.9 0.002 0.004 12.620 12.620 mp_sum_l 7870 13.0 8.611 9.795 8.611 9.795 cp_fm_cholesky_invert 11 10.9 9.332 9.340 9.332 9.340 wfi_extrapolate 11 7.9 0.001 0.001 9.126 9.126 multiply_cannon_metrocomm1 60168 15.6 0.086 0.090 6.350 8.845 dbcsr_mm_accdrv_process 124484 16.2 3.318 3.471 7.951 8.617 calculate_dm_sparse 128 9.5 0.001 0.001 8.467 8.545 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.960 8.100 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.830 7.994 make_images_data 5014 15.6 0.069 0.075 6.874 7.883 density_rs2pw 128 9.7 0.006 0.007 6.991 7.831 grid_integrate_task_list 128 12.3 7.105 7.610 7.105 7.610 rs_pw_transfer 1046 11.9 0.017 0.019 6.702 7.606 pw_transfer 1547 11.6 0.076 0.104 7.046 7.338 hybrid_alltoall_any 5200 16.5 0.291 2.265 5.973 7.226 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.831 7.097 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.742 6.749 fft_wrap_pw1pw2_140 523 13.2 0.458 0.521 5.817 6.012 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.841 5.998 fft3d_ps 1291 14.7 2.091 2.595 5.609 5.847 mp_alltoall_d11v 2415 14.1 4.272 5.315 4.272 5.315 grid_collocate_task_list 128 9.7 4.725 5.179 4.725 5.179 potential_pw2rs 128 12.3 0.009 0.011 4.997 5.029 cp_fm_cholesky_decompose 22 10.9 4.644 4.658 4.644 4.658 mp_sum_d 4473 12.2 3.764 4.516 3.764 4.516 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.377000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.000000, yerr=6.324555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 830.324736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5061807. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.067 191.128 191.129 qs_mol_dyn_low 1 2.0 0.003 0.003 190.636 190.659 qs_forces 11 3.9 0.003 0.004 190.493 190.494 qs_energies 11 4.9 0.007 0.008 183.657 183.667 scf_env_do_scf 11 5.9 0.001 0.001 167.015 167.025 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.695 133.697 velocity_verlet 10 3.0 0.001 0.002 120.217 120.220 dbcsr_multiply_generic 2507 12.6 0.185 0.190 97.714 98.878 qs_scf_new_mos 117 7.6 0.001 0.001 94.601 95.080 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.601 95.079 ot_scf_mini 117 9.6 0.004 0.004 89.806 90.291 multiply_cannon 2507 13.6 0.478 0.532 77.617 82.394 multiply_cannon_loop 2507 14.6 1.266 1.308 74.262 76.586 ot_mini 117 10.6 0.001 0.001 49.824 50.323 mp_waitall_1 226760 16.4 25.035 40.022 25.035 40.022 multiply_cannon_multrec 30084 15.6 22.155 26.723 31.803 36.576 rebuild_ks_matrix 128 8.3 0.001 0.001 33.170 33.706 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.170 33.706 init_scf_loop 11 6.9 0.000 0.000 33.232 33.233 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.830 30.331 prepare_preconditioner 11 7.9 0.000 0.000 28.822 28.890 make_preconditioner 11 8.9 0.000 0.000 28.822 28.890 multiply_cannon_metrocomm3 30084 15.6 0.096 0.099 15.641 28.645 qs_ot_get_derivative 117 11.6 0.001 0.002 27.953 28.434 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.474 28.003 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.942 23.106 apply_single 128 13.6 0.001 0.001 21.942 23.106 qs_ot_get_p 128 10.4 0.001 0.001 21.574 22.267 multiply_cannon_sync_h2d 30084 15.6 19.352 21.900 19.352 21.900 ot_diis_step 117 11.6 0.014 0.015 21.696 21.699 cp_fm_cholesky_invert 11 10.9 16.966 16.978 16.966 16.978 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.676 16.708 make_m2s 5014 13.6 0.090 0.095 14.415 16.064 make_images 5014 14.6 1.152 1.342 14.204 15.854 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.535 15.536 sum_up_and_integrate 128 10.3 0.061 0.071 14.501 14.530 integrate_v_rspace 128 11.3 0.003 0.003 14.440 14.470 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.257 13.292 calculate_rho_elec 128 8.7 0.088 0.105 13.256 13.291 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.423 12.435 cp_fm_redistribute_end 83 14.4 7.272 12.357 7.288 12.367 cp_fm_diag_elpa_base 83 14.4 4.844 11.879 5.057 12.240 init_scf_run 11 5.9 0.000 0.001 11.801 11.803 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.801 11.802 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.983 11.311 make_images_data 5014 15.6 0.066 0.073 8.781 10.875 multiply_cannon_metrocomm4 27577 15.6 0.097 0.114 3.699 10.834 mp_irecv_dv 69486 16.3 3.505 10.445 3.505 10.445 hybrid_alltoall_any 5200 16.5 0.341 1.499 7.480 10.099 dbcsr_mm_accdrv_process 62242 16.2 4.615 5.297 9.099 9.690 wfi_extrapolate 11 7.9 0.001 0.001 8.508 8.508 pw_transfer 1547 11.6 0.083 0.103 7.852 7.917 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.623 7.687 grid_integrate_task_list 128 12.3 7.121 7.609 7.121 7.609 density_rs2pw 128 9.7 0.006 0.006 6.986 7.485 cp_fm_cholesky_decompose 22 10.9 6.915 6.993 6.915 6.993 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.199 6.935 fft_wrap_pw1pw2_140 523 13.2 0.461 0.495 6.676 6.751 calculate_dm_sparse 128 9.5 0.001 0.001 6.470 6.589 rs_pw_transfer 1046 11.9 0.014 0.015 6.031 6.502 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.253 6.264 fft3d_ps 1291 14.7 2.775 2.937 6.001 6.068 mp_sum_l 7870 13.0 3.987 5.827 3.987 5.827 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.317 5.492 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.320 5.383 grid_collocate_task_list 128 9.7 4.910 5.360 4.910 5.360 mp_alltoall_d11v 2415 14.1 4.371 5.271 4.371 5.271 mp_allgather_i34 2507 14.6 1.931 5.014 1.931 5.014 potential_pw2rs 128 12.3 0.016 0.018 4.928 4.942 dbcsr_complete_redistribute 395 12.7 0.774 0.882 3.267 4.126 mp_sum_d 4459 12.1 2.654 3.861 2.654 3.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.129000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.727273, yerr=5.785419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 939.200512E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.253 181.438 181.440 qs_mol_dyn_low 1 2.0 0.006 0.023 180.410 180.425 qs_forces 11 3.9 0.003 0.007 180.208 180.215 qs_energies 11 4.9 0.014 0.049 173.646 173.660 scf_env_do_scf 11 5.9 0.001 0.004 156.929 156.930 scf_env_do_scf_inner_loop 117 6.6 0.011 0.051 120.890 120.893 velocity_verlet 10 3.0 0.008 0.011 115.284 115.294 dbcsr_multiply_generic 2507 12.6 0.185 0.207 82.934 84.226 qs_scf_new_mos 117 7.6 0.001 0.001 83.205 83.510 qs_scf_loop_do_ot 117 8.6 0.001 0.002 83.204 83.509 ot_scf_mini 117 9.6 0.004 0.008 79.017 79.425 multiply_cannon 2507 13.6 0.500 0.533 62.445 66.685 multiply_cannon_loop 2507 14.6 0.865 0.893 59.302 62.132 ot_mini 117 10.6 0.001 0.001 42.845 43.243 init_scf_loop 11 6.9 0.001 0.004 35.936 35.943 mp_waitall_1 178456 16.5 25.993 35.620 25.993 35.620 prepare_preconditioner 11 7.9 0.000 0.000 31.796 31.855 make_preconditioner 11 8.9 0.002 0.009 31.795 31.855 rebuild_ks_matrix 128 8.3 0.001 0.001 30.767 31.254 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.043 30.766 31.253 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.372 30.770 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.833 28.273 multiply_cannon_multrec 20056 15.6 13.532 16.435 22.644 25.507 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.209 24.473 qs_ot_get_derivative 117 11.6 0.002 0.002 23.059 23.459 qs_ot_get_p 128 10.4 0.001 0.001 21.505 21.998 apply_preconditioner_dbcsr 128 12.6 0.001 0.002 19.858 20.869 apply_single 128 13.6 0.001 0.001 19.858 20.868 ot_diis_step 117 11.6 0.020 0.026 19.679 19.681 qs_ot_p2m_diag 83 11.4 0.266 0.272 16.948 16.954 make_m2s 5014 13.6 0.081 0.088 15.302 16.282 make_images 5014 14.6 1.179 1.279 15.068 16.035 multiply_cannon_sync_h2d 20056 15.6 14.307 15.957 14.307 15.957 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.935 15.936 cp_fm_cholesky_invert 11 10.9 14.921 14.930 14.921 14.930 sum_up_and_integrate 128 10.3 0.071 0.081 14.383 14.409 integrate_v_rspace 128 11.3 0.003 0.003 14.312 14.339 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.613 13.694 calculate_rho_elec 128 8.7 0.132 0.146 13.612 13.694 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.758 12.759 cp_fm_redistribute_end 83 14.4 4.819 12.676 4.835 12.678 cp_fm_diag_elpa_base 83 14.4 7.407 12.045 7.822 12.570 init_scf_run 11 5.9 0.000 0.001 11.356 11.357 scf_env_initial_rho_setup 11 6.9 0.007 0.011 11.356 11.357 make_images_data 5014 15.6 0.060 0.069 9.487 10.889 hybrid_alltoall_any 5200 16.5 0.431 1.967 8.241 9.950 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.043 9.344 multiply_cannon_metrocomm4 17549 15.6 0.063 0.075 3.420 9.062 mp_irecv_dv 50230 16.2 3.296 8.818 3.296 8.818 dbcsr_mm_accdrv_process 41502 16.2 4.396 5.736 8.565 8.679 pw_transfer 1547 11.6 0.083 0.104 8.031 8.140 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.805 7.920 grid_integrate_task_list 128 12.3 7.287 7.742 7.287 7.742 cp_fm_cholesky_decompose 22 10.9 7.575 7.649 7.575 7.649 cp_fm_upper_to_full 105 14.5 5.830 7.573 5.830 7.573 wfi_extrapolate 11 7.9 0.007 0.051 7.557 7.558 density_rs2pw 128 9.7 0.006 0.006 6.878 7.152 fft_wrap_pw1pw2_140 523 13.2 0.470 0.513 6.707 6.822 dbcsr_complete_redistribute 395 12.7 1.168 1.212 4.792 6.588 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.211 6.218 fft3d_ps 1291 14.7 2.698 2.914 6.082 6.173 rs_pw_transfer 1046 11.9 0.013 0.014 5.554 5.935 calculate_dm_sparse 128 9.5 0.001 0.002 5.782 5.898 grid_collocate_task_list 128 9.7 5.072 5.472 5.072 5.472 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.586 5.381 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.200 5.317 mp_alltoall_d11v 2415 14.1 4.665 5.266 4.665 5.266 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.579 5.264 mp_allgather_i34 2507 14.6 1.668 5.104 1.668 5.104 mp_sum_l 7870 13.0 3.448 5.007 3.448 5.007 potential_pw2rs 128 12.3 0.021 0.023 4.773 4.789 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.401 4.176 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.006 4.034 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 4.005 4.008 mp_alltoall_i22 716 14.1 2.004 3.882 2.004 3.882 mp_sum_d 4457 12.1 2.467 3.829 2.467 3.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=181.440000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.727273, yerr=11.817482 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.130394E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.128 0.405 188.966 188.967 qs_mol_dyn_low 1 2.0 0.003 0.006 188.002 188.016 qs_forces 11 3.9 0.034 0.041 187.791 187.795 qs_energies 11 4.9 0.002 0.005 180.682 180.715 scf_env_do_scf 11 5.9 0.001 0.001 163.744 163.756 velocity_verlet 10 3.0 0.002 0.003 125.148 125.153 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.520 116.523 qs_scf_new_mos 117 7.6 0.001 0.001 80.667 80.925 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.666 80.925 dbcsr_multiply_generic 2507 12.6 0.190 0.195 79.365 80.104 ot_scf_mini 117 9.6 0.003 0.004 76.184 76.439 multiply_cannon 2507 13.6 0.560 0.586 54.767 57.842 multiply_cannon_loop 2507 14.6 1.185 1.224 50.958 52.793 init_scf_loop 11 6.9 0.001 0.005 47.099 47.102 prepare_preconditioner 11 7.9 0.000 0.000 42.980 43.027 make_preconditioner 11 8.9 0.000 0.002 42.980 43.026 ot_mini 117 10.6 0.001 0.002 42.429 42.673 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.457 41.596 multiply_cannon_multrec 30084 15.6 14.133 19.276 26.074 30.667 rebuild_ks_matrix 128 8.3 0.001 0.001 29.126 29.396 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.126 29.395 mp_waitall_1 153770 16.5 17.420 27.723 17.420 27.723 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.279 26.528 qs_ot_get_derivative 117 11.6 0.002 0.002 22.789 23.048 make_m2s 5014 13.6 0.095 0.100 20.179 21.151 make_images 5014 14.6 1.972 2.278 19.873 20.845 qs_ot_get_p 128 10.4 0.001 0.001 20.025 20.277 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.995 19.585 apply_single 128 13.6 0.001 0.001 18.994 19.585 ot_diis_step 117 11.6 0.018 0.020 19.521 19.522 cp_fm_upper_to_full 105 14.7 11.315 16.815 11.315 16.815 cp_fm_cholesky_invert 11 10.9 16.400 16.408 16.400 16.408 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.773 15.827 multiply_cannon_metrocomm3 30084 15.6 0.046 0.048 6.397 15.012 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.478 14.481 sum_up_and_integrate 128 10.3 0.079 0.087 14.107 14.138 integrate_v_rspace 128 11.3 0.003 0.004 14.027 14.060 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.560 13.604 calculate_rho_elec 128 8.7 0.174 0.190 13.560 13.604 dbcsr_complete_redistribute 395 12.7 1.506 1.622 9.135 12.996 multiply_cannon_sync_h2d 30084 15.6 11.690 12.529 11.690 12.529 make_images_data 5014 15.6 0.063 0.067 10.766 12.430 dbcsr_mm_accdrv_process 62264 16.2 7.377 8.417 11.513 12.055 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.774 11.627 hybrid_alltoall_any 5200 16.5 0.526 2.211 9.760 11.483 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.384 11.387 cp_fm_redistribute_end 83 14.4 1.968 11.313 1.987 11.318 cp_fm_diag_elpa_base 83 14.4 8.702 10.687 9.298 11.172 init_scf_run 11 5.9 0.000 0.001 10.878 10.879 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.877 10.879 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.500 10.254 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.503 9.691 mp_alltoall_i22 716 14.1 5.744 9.561 5.744 9.561 pw_transfer 1547 11.6 0.084 0.096 7.999 8.077 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.771 7.854 grid_integrate_task_list 128 12.3 7.523 7.837 7.523 7.837 cp_fm_cholesky_decompose 22 10.9 7.656 7.762 7.656 7.762 wfi_extrapolate 11 7.9 0.001 0.001 7.677 7.677 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.769 7.399 mp_irecv_dv 76098 16.2 2.624 7.129 2.624 7.129 density_rs2pw 128 9.7 0.005 0.006 6.564 6.948 fft_wrap_pw1pw2_140 523 13.2 0.476 0.485 6.828 6.939 calculate_dm_sparse 128 9.5 0.001 0.002 6.174 6.261 fft3d_ps 1291 14.7 2.803 2.864 6.003 6.058 mp_alltoall_d11v 2415 14.1 4.898 5.704 4.898 5.704 grid_collocate_task_list 128 9.7 5.210 5.656 5.210 5.656 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.487 5.542 rs_pw_transfer 1046 11.9 0.013 0.014 5.010 5.299 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.561 4.593 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.426 4.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.345 4.425 potential_pw2rs 128 12.3 0.024 0.024 4.374 4.386 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.195 4.256 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.967000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1065.272727, yerr=21.037157 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 1.522377E+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 4045 57602. MP_Allreduce 11109 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 0.020 0.037 174.055 174.055 qs_mol_dyn_low 1 2.0 0.003 0.003 173.621 173.634 qs_forces 11 3.9 0.013 0.023 173.520 173.523 qs_energies 11 4.9 0.014 0.014 166.141 166.156 scf_env_do_scf 11 5.9 0.001 0.001 148.887 148.899 velocity_verlet 10 3.0 0.002 0.002 114.447 114.451 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 112.050 112.051 qs_scf_new_mos 118 7.6 0.001 0.001 75.414 75.529 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.413 75.528 dbcsr_multiply_generic 2529 12.6 0.182 0.189 74.959 75.444 ot_scf_mini 118 9.6 0.004 0.004 70.872 70.945 multiply_cannon 2529 13.6 0.588 0.626 54.797 58.982 multiply_cannon_loop 2529 14.6 0.454 0.467 50.140 51.029 ot_mini 118 10.6 0.001 0.001 39.853 39.919 init_scf_loop 11 6.9 0.000 0.000 36.682 36.683 mp_waitall_1 130746 16.6 26.574 33.792 26.574 33.792 prepare_preconditioner 11 7.9 0.000 0.000 32.773 32.812 make_preconditioner 11 8.9 0.000 0.000 32.773 32.812 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.545 30.855 rebuild_ks_matrix 129 8.3 0.001 0.001 29.219 29.312 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.219 29.312 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.706 26.784 multiply_cannon_multrec 10116 15.6 10.404 14.612 17.831 20.895 ot_diis_step 118 11.6 0.020 0.021 20.047 20.048 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.512 19.823 apply_single 129 13.6 0.001 0.001 19.512 19.823 qs_ot_get_derivative 118 11.6 0.002 0.002 19.735 19.799 multiply_cannon_metrocomm3 10116 15.6 0.022 0.024 12.336 19.659 cp_fm_cholesky_invert 11 10.9 18.818 18.827 18.818 18.827 make_m2s 5058 13.6 0.067 0.073 16.495 18.769 make_images 5058 14.6 2.320 2.881 16.185 18.453 qs_ot_get_p 129 10.4 0.001 0.001 17.898 17.957 sum_up_and_integrate 129 10.3 0.104 0.111 14.217 14.266 integrate_v_rspace 129 11.3 0.004 0.004 14.113 14.167 qs_ot_p2m_diag 84 11.4 0.502 0.508 14.136 14.156 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.583 13.619 calculate_rho_elec 129 8.7 0.259 0.270 13.583 13.618 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.983 12.986 make_images_data 5058 15.6 0.052 0.061 10.114 12.815 hybrid_alltoall_any 5245 16.5 0.846 3.821 9.889 12.207 multiply_cannon_sync_h2d 10116 15.6 11.715 12.120 11.715 12.120 init_scf_run 11 5.9 0.000 0.001 10.447 10.447 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.447 10.447 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.800 9.804 cp_fm_diag_elpa_base 84 14.4 9.537 9.627 9.791 9.795 cp_fm_cholesky_decompose 22 10.9 8.143 8.234 8.143 8.234 grid_integrate_task_list 129 12.3 7.823 8.215 7.823 8.215 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.039 8.093 pw_transfer 1559 11.6 0.083 0.091 8.058 8.086 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.618 8.005 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.829 7.864 dbcsr_mm_accdrv_process 20934 16.1 2.605 3.427 7.050 7.683 wfi_extrapolate 11 7.9 0.001 0.001 7.350 7.350 mp_allgather_i34 2529 14.6 2.847 7.120 2.847 7.120 fft_wrap_pw1pw2_140 527 13.2 0.497 0.515 6.926 6.956 density_rs2pw 129 9.7 0.005 0.006 6.239 6.519 calculate_dm_sparse 129 9.5 0.001 0.001 6.262 6.331 fft3d_ps 1301 14.7 2.750 2.830 5.986 6.011 grid_collocate_task_list 129 9.7 5.534 5.878 5.534 5.878 dbcsr_complete_redistribute 397 12.7 2.125 2.190 5.409 5.838 mp_alltoall_d11v 2429 14.1 5.010 5.833 5.010 5.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.265 5.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.218 5.228 rs_pw_transfer 1054 12.0 0.013 0.013 4.367 4.661 potential_pw2rs 129 12.3 0.027 0.028 4.212 4.225 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.112 4.131 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.899 4.079 mp_irecv_dv 29102 15.9 1.861 4.007 1.861 4.007 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.549 3.929 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.864 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.772 3.804 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.650 3.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.055000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1418.636364, yerr=58.914843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 2.260341E+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 0.029 0.044 291.109 291.110 qs_mol_dyn_low 1 2.0 0.003 0.003 290.509 290.520 qs_forces 11 3.9 0.003 0.004 290.415 290.419 qs_energies 11 4.9 0.002 0.002 281.790 281.800 scf_env_do_scf 11 5.9 0.001 0.001 259.173 259.189 velocity_verlet 10 3.0 0.002 0.002 209.767 209.775 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.403 133.405 init_scf_loop 11 6.9 0.000 0.000 125.518 125.522 prepare_preconditioner 11 7.9 0.000 0.000 120.652 120.694 make_preconditioner 11 8.9 0.000 0.000 120.652 120.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.779 117.777 qs_scf_new_mos 116 7.6 0.001 0.001 91.163 91.268 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.162 91.267 ot_scf_mini 116 9.6 0.004 0.004 86.433 86.476 dbcsr_multiply_generic 2485 12.5 0.212 0.218 83.584 84.151 cp_fm_upper_to_full 104 14.8 52.980 75.810 52.980 75.810 multiply_cannon 2485 13.5 0.690 0.732 59.302 59.767 multiply_cannon_loop 2485 14.5 0.472 0.481 55.659 56.951 ot_mini 116 10.6 0.001 0.001 44.365 44.410 dbcsr_complete_redistribute 393 12.7 3.993 4.057 29.592 42.674 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.180 39.248 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.829 36.737 mp_alltoall_i22 712 14.1 21.710 34.891 21.710 34.891 cp_fm_cholesky_invert 11 10.9 34.404 34.411 34.404 34.411 rebuild_ks_matrix 127 8.3 0.001 0.001 33.857 33.889 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.857 33.889 mp_waitall_1 104546 16.7 29.148 32.922 29.148 32.922 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.620 31.657 qs_ot_get_p 127 10.4 0.001 0.001 26.416 26.444 qs_ot_get_derivative 116 11.6 0.002 0.002 24.273 24.317 qs_ot_p2m_diag 82 11.4 0.868 0.874 22.381 22.410 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 19.791 20.906 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.600 20.601 make_m2s 4970 13.5 0.076 0.078 19.607 20.554 make_images 4970 14.5 3.740 3.876 19.136 20.085 ot_diis_step 116 11.6 0.022 0.022 20.058 20.058 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.361 19.566 apply_single 127 13.6 0.001 0.001 19.360 19.566 multiply_cannon_multrec 9940 15.5 10.358 12.077 17.735 17.828 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.447 17.450 cp_fm_diag_elpa_base 82 14.4 13.032 14.694 17.444 17.446 sum_up_and_integrate 127 10.3 0.192 0.196 15.628 15.715 multiply_cannon_sync_h2d 9940 15.5 15.534 15.543 15.534 15.543 integrate_v_rspace 127 11.3 0.004 0.004 15.435 15.522 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.006 15.038 calculate_rho_elec 127 8.7 0.478 0.480 15.005 15.038 init_scf_run 11 5.9 0.000 0.001 12.564 12.565 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.564 12.565 make_images_data 4970 15.5 0.059 0.064 10.431 12.184 hybrid_alltoall_any 5155 16.4 1.293 3.030 10.478 12.078 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.425 9.474 cp_fm_cholesky_decompose 22 10.9 9.136 9.154 9.136 9.154 dbcsr_mm_accdrv_process 20590 16.0 3.769 5.563 7.142 8.904 wfi_extrapolate 11 7.9 0.001 0.001 8.802 8.802 grid_integrate_task_list 127 12.3 8.480 8.659 8.480 8.659 pw_transfer 1535 11.6 0.092 0.094 8.443 8.453 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.206 8.217 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.013 8.014 fft_wrap_pw1pw2_140 519 13.2 0.539 0.553 7.215 7.229 mp_alltoall_d11v 2401 14.1 7.005 7.144 7.005 7.144 calculate_dm_sparse 127 9.5 0.001 0.001 6.622 6.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.513 6.594 grid_collocate_task_list 127 9.7 6.288 6.319 6.288 6.319 fft3d_ps 1281 14.7 2.742 2.752 6.261 6.284 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.107 6.192 density_rs2pw 127 9.7 0.005 0.005 6.086 6.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.110000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2124.363636, yerr=49.485118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 1.260282E+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 0.043 0.108 88.392 88.393 qs_energies 1 2.0 0.000 0.000 87.770 87.815 ls_scf 1 3.0 0.001 0.010 86.788 86.804 dbcsr_multiply_generic 111 6.7 0.014 0.015 75.800 75.961 multiply_cannon 111 7.7 0.017 0.020 58.820 60.397 multiply_cannon_loop 111 8.7 0.209 0.224 55.420 57.138 ls_scf_main 1 4.0 0.000 0.002 52.169 52.170 density_matrix_trs4 2 5.0 0.002 0.018 46.731 46.820 ls_scf_init_scf 1 4.0 0.000 0.001 31.585 31.586 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.490 30.538 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.348 28.369 mp_waitall_1 11316 10.9 23.606 27.096 23.606 27.096 multiply_cannon_multrec 2664 9.7 8.177 8.945 15.478 17.218 multiply_cannon_sync_h2d 2664 9.7 13.728 15.203 13.728 15.203 make_m2s 222 7.7 0.008 0.012 13.217 13.725 make_images 222 8.7 0.098 0.108 13.195 13.706 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 10.328 13.234 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.844 10.146 make_images_data 222 9.7 0.004 0.005 7.757 8.365 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.657 6.919 7.836 hybrid_alltoall_any 227 10.6 0.215 1.828 6.692 7.789 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.096 6.212 7.096 calculate_norms 4752 9.8 5.540 6.122 5.540 6.122 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.974 5.081 mp_sum_l 807 5.4 3.233 4.606 3.233 4.606 compute_matrix_preconditioner 1 6.0 0.000 0.001 4.306 4.310 make_images_sizes 222 9.7 0.000 0.000 0.723 3.457 mp_alltoall_i44 222 10.7 0.723 3.457 0.723 3.457 acc_transpose_blocks 2664 9.7 0.018 0.021 2.155 3.433 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.424 3.408 acc_transpose_blocks_kernels 2664 10.7 0.036 0.039 2.051 3.318 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.052 3.302 jit_kernel_transpose 1 13.0 2.015 3.282 2.015 3.282 mp_irecv_dv 6231 10.9 2.035 3.271 2.035 3.271 arnoldi_extremal 4 6.8 0.000 0.000 3.213 3.237 arnoldi_normal_ev 4 7.8 0.001 0.003 3.213 3.237 build_subspace 16 8.4 0.009 0.012 3.119 3.121 ls_scf_post 1 4.0 0.002 0.019 3.033 3.041 ls_scf_store_result 1 5.0 0.000 0.000 2.843 2.882 dbcsr_special_finalize 555 9.7 0.005 0.006 2.385 2.749 dbcsr_merge_single_wm 555 10.7 0.464 0.580 2.377 2.741 make_images_pack 222 9.7 2.207 2.620 2.208 2.622 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.312 2.562 dbcsr_sort_data 658 11.4 2.172 2.490 2.172 2.490 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.463 2.066 2.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.230 2.312 buffer_matrices_ensure_size 222 8.7 1.780 2.048 1.780 2.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=88.393000, 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/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 2.109096E+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 0.082 0.225 93.030 93.031 qs_energies 1 2.0 0.000 0.000 91.965 92.030 ls_scf 1 3.0 0.001 0.005 90.585 90.645 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.239 76.597 multiply_cannon 111 7.7 0.027 0.038 53.588 58.024 ls_scf_main 1 4.0 0.000 0.008 55.824 55.828 multiply_cannon_loop 111 8.7 0.116 0.124 50.319 53.871 density_matrix_trs4 2 5.0 0.002 0.003 50.165 50.359 ls_scf_init_scf 1 4.0 0.000 0.001 31.282 31.283 mp_waitall_1 9246 10.9 21.844 30.376 21.844 30.376 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.951 30.061 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.528 27.543 multiply_cannon_multrec 1332 9.7 13.080 16.412 22.293 26.760 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.195 21.134 make_m2s 222 7.7 0.006 0.008 15.478 16.075 make_images 222 8.7 1.577 1.922 15.448 16.047 dbcsr_mm_accdrv_process 4041 10.4 0.279 0.463 8.811 10.425 dbcsr_mm_accdrv_process_sort 4041 11.4 8.397 9.962 8.397 9.962 make_images_data 222 9.7 0.004 0.004 8.979 9.926 mp_sum_l 807 5.4 5.978 9.549 5.978 9.549 hybrid_alltoall_any 227 10.6 0.522 2.471 8.314 9.378 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.222 7.829 mp_irecv_dv 3311 11.0 3.203 7.782 3.203 7.782 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.623 7.618 calculate_norms 2376 9.8 6.004 6.608 6.004 6.608 multiply_cannon_sync_h2d 1332 9.7 4.727 5.930 4.727 5.930 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.089 5.298 arnoldi_extremal 4 6.8 0.000 0.001 4.802 4.833 arnoldi_normal_ev 4 7.8 0.004 0.015 4.802 4.833 build_subspace 16 8.4 0.014 0.022 4.461 4.467 ls_scf_post 1 4.0 0.008 0.065 3.479 3.536 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.193 3.393 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.289 dbcsr_matrix_vector_mult_local 304 10.0 2.751 3.222 2.753 3.224 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.299 3.027 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.527 2.644 mp_allgather_i34 111 8.7 0.975 2.453 0.975 2.453 make_images_pack 222 9.7 2.022 2.432 2.024 2.434 dbcsr_sort_data 436 11.2 1.775 1.985 1.775 1.985 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.930 1.932 rebuild_ks_matrix 3 7.3 0.000 0.000 1.918 1.920 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.018 1.917 1.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.031000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 2.715501E+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 0.088 0.174 96.228 96.229 qs_energies 1 2.0 0.000 0.000 95.318 95.328 ls_scf 1 3.0 0.000 0.001 93.808 93.812 dbcsr_multiply_generic 111 6.7 0.016 0.020 77.928 78.176 ls_scf_main 1 4.0 0.000 0.001 58.683 58.689 multiply_cannon 111 7.7 0.035 0.060 53.388 57.448 multiply_cannon_loop 111 8.7 0.101 0.114 49.802 53.621 density_matrix_trs4 2 5.0 0.002 0.004 52.742 52.951 mp_waitall_1 7374 11.0 25.147 35.148 25.147 35.148 ls_scf_init_scf 1 4.0 0.000 0.001 31.524 31.528 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.145 30.212 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.689 27.705 multiply_cannon_multrec 888 9.7 12.456 14.869 20.965 24.014 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.771 23.686 make_m2s 222 7.7 0.006 0.007 17.706 19.023 make_images 222 8.7 1.968 2.281 17.668 18.983 hybrid_alltoall_any 227 10.6 0.621 2.857 9.983 11.460 make_images_data 222 9.7 0.003 0.004 10.389 11.417 dbcsr_mm_accdrv_process 3754 10.4 0.244 0.432 8.043 9.266 mp_sum_l 807 5.4 5.303 9.091 5.303 9.091 dbcsr_mm_accdrv_process_sort 3754 11.4 7.665 8.834 7.665 8.834 multiply_cannon_sync_h2d 888 9.7 6.033 7.306 6.033 7.306 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.505 7.075 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.013 7.073 mp_irecv_dv 2335 11.1 2.491 7.025 2.491 7.025 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.605 6.553 arnoldi_extremal 4 6.8 0.000 0.002 5.420 5.473 arnoldi_normal_ev 4 7.8 0.002 0.005 5.420 5.473 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.985 5.197 build_subspace 16 8.4 0.014 0.020 4.956 4.966 calculate_norms 1584 9.8 4.259 4.613 4.259 4.613 mp_allgather_i34 111 8.7 1.415 3.948 1.415 3.948 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.558 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.612 3.039 3.614 ls_scf_post 1 4.0 0.002 0.035 3.601 3.611 ls_scf_store_result 1 5.0 0.000 0.000 3.352 3.423 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.784 2.895 make_images_sizes 222 9.7 0.000 0.000 1.006 2.151 mp_alltoall_i44 222 10.7 1.006 2.151 1.006 2.151 dbcsr_sort_data 325 11.1 1.869 2.146 1.869 2.146 make_images_pack 222 9.7 1.810 2.097 1.813 2.100 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.963 1.964 dbcsr_data_release 9322 10.9 1.302 1.947 1.302 1.947 rebuild_ks_matrix 3 7.3 0.000 0.000 1.945 1.947 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.945 1.947 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.229000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2210.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 3.347337E+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 0.066 0.110 97.608 97.609 qs_energies 1 2.0 0.000 0.000 96.660 96.668 ls_scf 1 3.0 0.000 0.000 94.929 94.935 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.425 78.709 ls_scf_main 1 4.0 0.000 0.000 59.127 59.128 multiply_cannon 111 7.7 0.052 0.141 51.686 56.442 density_matrix_trs4 2 5.0 0.002 0.003 53.000 53.130 multiply_cannon_loop 111 8.7 0.115 0.125 46.588 49.061 ls_scf_init_scf 1 4.0 0.000 0.000 32.559 32.560 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.381 31.451 mp_waitall_1 6438 11.0 22.793 29.342 22.793 29.342 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.818 28.828 multiply_cannon_multrec 1332 9.7 14.218 17.671 22.049 24.433 make_m2s 222 7.7 0.006 0.008 21.083 22.576 make_images 222 8.7 3.129 3.586 21.033 22.529 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.305 16.952 make_images_data 222 9.7 0.004 0.004 11.810 13.444 hybrid_alltoall_any 227 10.6 0.794 3.784 11.258 13.088 dbcsr_mm_accdrv_process 3641 10.4 0.210 0.408 7.471 8.991 dbcsr_mm_accdrv_process_sort 3641 11.4 7.089 8.584 7.089 8.584 mp_sum_l 807 5.4 4.103 7.548 4.103 7.548 multiply_cannon_sync_h2d 1332 9.7 5.512 6.183 5.512 6.183 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.076 5.949 mp_irecv_dv 3229 10.9 2.053 5.905 2.053 5.905 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.101 5.892 arnoldi_extremal 4 6.8 0.000 0.000 5.447 5.463 arnoldi_normal_ev 4 7.8 0.009 0.020 5.447 5.463 build_subspace 16 8.4 0.014 0.021 5.097 5.104 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.589 4.808 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.500 4.735 mp_allgather_i34 111 8.7 2.220 4.708 2.220 4.708 calculate_norms 2376 9.8 4.193 4.534 4.193 4.534 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.686 4.028 dbcsr_matrix_vector_mult_local 304 10.0 3.221 3.716 3.223 3.718 dbcsr_sort_data 658 11.4 3.104 3.389 3.104 3.389 ls_scf_post 1 4.0 0.000 0.000 3.243 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.843 3.137 dbcsr_merge_single_wm 555 10.7 0.536 0.642 2.835 3.129 ls_scf_store_result 1 5.0 0.000 0.000 2.983 3.041 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.981 3.038 dbcsr_data_release 10477 10.7 1.588 2.460 1.588 2.460 dbcsr_finalize 304 7.8 0.049 0.061 1.806 2.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.609000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 4.627808E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.057 99.078 99.079 qs_energies 1 2.0 0.000 0.000 98.294 98.300 ls_scf 1 3.0 0.000 0.000 96.359 96.364 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.961 78.188 ls_scf_main 1 4.0 0.000 0.000 62.125 62.126 multiply_cannon 111 7.7 0.093 0.170 55.282 60.796 density_matrix_trs4 2 5.0 0.002 0.003 55.086 55.169 multiply_cannon_loop 111 8.7 0.069 0.076 50.920 52.566 mp_waitall_1 5481 11.0 26.904 32.150 26.904 32.150 ls_scf_init_scf 1 4.0 0.000 0.000 30.660 30.664 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.427 29.468 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.261 27.272 multiply_cannon_multrec 444 9.7 13.898 16.153 20.913 23.083 make_m2s 222 7.7 0.004 0.005 17.913 20.279 make_images 222 8.7 3.707 4.390 17.851 20.219 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.211 16.909 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.360 14.491 make_images_data 222 9.7 0.003 0.004 10.133 12.487 hybrid_alltoall_any 227 10.6 0.788 3.749 9.978 12.307 dbcsr_mm_accdrv_process 3003 10.4 0.128 0.162 6.718 7.866 dbcsr_mm_accdrv_process_sort 3003 11.4 6.406 7.516 6.406 7.516 multiply_cannon_sync_h2d 444 9.7 6.586 7.502 6.586 7.502 mp_allgather_i34 111 8.7 2.601 6.986 2.601 6.986 arnoldi_extremal 4 6.8 0.000 0.000 5.877 5.889 arnoldi_normal_ev 4 7.8 0.021 0.039 5.877 5.889 build_subspace 16 8.4 0.015 0.020 5.436 5.445 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.582 4.727 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.671 4.551 mp_sum_l 807 5.4 2.811 4.530 2.811 4.530 mp_irecv_dv 1241 11.2 1.652 4.516 1.652 4.516 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.175 4.378 dbcsr_matrix_vector_mult_local 304 10.0 3.715 4.175 3.717 4.177 calculate_norms 792 9.8 3.543 3.643 3.543 3.643 make_images_sizes 222 9.7 0.000 0.000 1.164 3.592 mp_alltoall_i44 222 10.7 1.163 3.592 1.163 3.592 ls_scf_post 1 4.0 0.000 0.000 3.574 3.580 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.437 3.537 ls_scf_store_result 1 5.0 0.000 0.000 3.359 3.393 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.912 3.317 dbcsr_finalize 304 7.8 0.062 0.078 2.194 2.247 dbcsr_merge_all 275 8.9 0.472 0.521 2.040 2.086 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.080 2.082 rebuild_ks_matrix 3 7.3 0.000 0.000 2.047 2.048 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.047 2.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.079000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3576.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594_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 8.774996E+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 0.100 0.200 108.699 108.699 qs_energies 1 2.0 0.000 0.000 107.152 107.167 ls_scf 1 3.0 0.000 0.000 104.133 104.148 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.238 78.381 ls_scf_main 1 4.0 0.000 0.000 66.375 66.376 density_matrix_trs4 2 5.0 0.002 0.004 57.374 57.438 multiply_cannon 111 7.7 0.117 0.191 50.102 52.140 multiply_cannon_loop 111 8.7 0.067 0.070 46.562 47.768 ls_scf_init_scf 1 4.0 0.000 0.000 34.062 34.063 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.674 32.690 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.850 29.866 mp_waitall_1 4569 11.1 22.783 26.852 22.783 26.852 make_m2s 222 7.7 0.005 0.005 24.200 25.249 make_images 222 8.7 4.574 4.949 24.095 25.141 multiply_cannon_multrec 444 9.7 17.750 18.565 22.331 22.935 hybrid_alltoall_any 227 10.6 1.659 3.619 13.085 16.041 make_images_data 222 9.7 0.003 0.003 13.412 15.956 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.719 11.278 multiply_cannon_sync_h2d 444 9.7 8.849 8.898 8.849 8.898 arnoldi_extremal 4 6.8 0.000 0.000 7.353 7.373 arnoldi_normal_ev 4 7.8 0.003 0.011 7.353 7.373 build_subspace 16 8.4 0.026 0.036 6.804 6.816 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.477 5.590 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.150 5.412 dbcsr_matrix_vector_mult_local 304 10.0 5.063 5.325 5.066 5.328 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.836 4.927 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.314 4.413 4.540 dbcsr_mm_accdrv_process_sort 1814 11.4 4.115 4.246 4.115 4.246 make_images_sizes 222 9.7 0.000 0.000 1.487 3.729 mp_alltoall_i44 222 10.7 1.487 3.729 1.487 3.729 mp_allgather_i34 111 8.7 1.186 3.724 1.186 3.724 ls_scf_post 1 4.0 0.000 0.000 3.696 3.710 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.441 calculate_norms 792 9.8 3.240 3.278 3.240 3.278 dbcsr_finalize 304 7.8 0.082 0.089 3.092 3.139 qs_energies_init_hamiltonians 1 3.0 0.014 0.026 2.988 2.988 dbcsr_merge_all 275 8.9 0.891 0.919 2.863 2.909 dbcsr_complete_redistribute 5 7.6 1.437 1.472 2.767 2.902 dbcsr_data_release 12724 10.6 2.331 2.830 2.331 2.830 mp_sum_l 807 5.4 1.676 2.690 1.676 2.690 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.572 dbcsr_sort_data 325 11.1 2.437 2.496 2.437 2.496 dbcsr_new_transposed 4 7.5 0.243 0.251 2.274 2.291 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.268 2.269 dbcsr_frobenius_norm 74 6.6 2.059 2.131 2.204 2.246 dbcsr_add_d 103 6.2 0.000 0.000 2.150 2.219 dbcsr_add_anytype 103 7.2 0.860 0.895 2.149 2.219 rebuild_ks_matrix 3 7.3 0.000 0.000 2.201 2.202 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.201 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.699000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6878.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e2bd27a0fbe7d28ec5009d2fd2a3e163d28b4594 Summary: empty Status: OK