=== 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: a5ffd1932e3d34d3bbecc675223bf506dafc025e ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/01 job id: 43117678 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/02 job id: 43117680 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/03 job id: 43117685 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/04 job id: 43117686 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/05 job id: 43117687 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/06 job id: 43117689 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/07 job id: 43117691 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/08 job id: 43117693 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/09 job id: 43117694 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/10 job id: 43117698 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/11 job id: 43117700 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/12 job id: 43117701 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/13 job id: 43117704 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/14 job id: 43117705 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/15 job id: 43117707 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/16 job id: 43117709 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/17 job id: 43117710 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/18 job id: 43117711 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/19 job id: 43117713 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/20 job id: 43117714 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/21 job id: 43117716 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/22 job id: 43117718 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/23 job id: 43117719 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/24 job id: 43117720 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/25 job id: 43117722 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/26 job id: 43117723 --- 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/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 0.021 0.036 132.535 132.536 farming_run 1 2.0 132.088 132.089 132.508 132.512 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458000E+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 485 2282278. 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 0.028 115.281 115.281 qs_energies 1 2.0 0.000 0.000 115.081 115.084 mp2_main 1 3.0 0.000 0.000 113.157 113.159 mp2_gpw_main 1 4.0 0.019 0.022 112.280 112.282 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.049 93.666 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.360 55.979 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.670 47.245 get_2c_integrals 1 6.0 0.000 0.001 36.919 37.517 integrate_v_rspace 273 8.0 0.439 0.451 25.075 30.316 pw_transfer 6555 10.6 0.374 0.382 27.328 27.937 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.992 26.622 grid_integrate_task_list 273 9.0 20.893 26.607 20.893 26.607 fft_wrap_pw1pw2_100 2178 12.4 1.170 1.225 23.549 24.191 rpa_ri_compute_en 1 5.0 0.000 0.000 19.123 19.313 compute_2c_integrals 1 7.0 0.002 0.003 19.221 19.221 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.849 18.941 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.429 18.846 18.940 cp_fm_cholesky_decompose 12 8.2 17.677 18.298 17.677 18.298 cholesky_decomp 1 7.0 0.000 0.000 16.552 17.167 fft3d_s 5443 13.4 16.136 16.718 16.158 16.740 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.555 10.851 15.555 calculate_wavefunction 272 8.0 5.393 5.571 12.515 13.140 rpa_num_int 1 6.0 0.000 0.001 10.560 10.561 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.537 10.552 calc_mat_Q 8 8.0 0.000 0.000 9.354 9.456 contract_S_to_Q 8 9.0 0.000 0.000 8.774 8.875 calc_potential_gpw 544 9.5 0.005 0.006 8.268 8.602 parallel_gemm_fm 14 9.1 0.000 0.000 8.351 8.452 parallel_gemm_fm_cosma 14 10.1 8.351 8.452 8.351 8.452 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.197 8.398 potential_pw2rs 545 10.0 0.107 0.109 7.646 8.273 create_integ_mat 1 6.0 0.021 0.029 7.850 7.850 collocate_single_gaussian 272 10.0 0.040 0.042 7.452 7.666 array2fm 1 7.0 0.000 0.000 6.904 7.299 pw_scatter_s 2720 13.7 4.451 4.570 4.451 4.570 pw_gather_s 2722 13.2 3.862 4.245 3.862 4.245 array2fm_buffer_send 1 8.0 3.121 3.319 3.121 3.319 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.282448, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 0.038 0.043 397.221 397.225 farming_run 1 2.0 396.474 396.480 397.166 397.172 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223524E+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 1821 23730. MP_Sync 38 MP_Alltoall 77 2393302. 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 0.015 0.052 209.841 209.842 qs_energies 1 2.0 0.000 0.000 209.601 209.610 scf_env_do_scf 1 3.0 0.000 0.000 105.976 105.977 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.037 105.045 rebuild_ks_matrix 4 6.0 0.000 0.000 105.035 105.043 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.035 105.043 hfx_ks_matrix 4 8.0 0.001 0.001 104.629 104.632 integrate_four_center 4 9.0 0.143 0.448 104.628 104.631 mp2_main 1 3.0 0.000 0.000 103.304 103.312 mp2_gpw_main 1 4.0 0.032 0.047 102.395 102.406 integrate_four_center_main 4 10.0 0.106 0.522 96.819 99.155 integrate_four_center_bin 264 11.0 96.713 99.017 96.713 99.017 init_scf_loop 1 4.0 0.000 0.000 91.744 91.744 mp2_ri_gpw_compute_in 1 5.0 0.065 0.074 74.879 75.944 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.503 55.568 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.242 47.380 integrate_v_rspace 95 8.0 0.398 0.565 28.632 33.594 pw_transfer 2240 10.6 0.144 0.162 29.948 30.488 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.958 29.515 ao_to_mo_and_store_B_mult_1 91 7.0 10.583 29.099 10.583 29.099 grid_integrate_task_list 95 9.0 23.894 29.062 23.894 29.062 mp2_ri_gpw_compute_en 1 5.0 0.055 0.065 27.369 29.040 fft_wrap_pw1pw2_100 730 12.4 1.278 1.433 26.678 27.228 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.903 25.660 25.670 get_2c_integrals 1 6.0 0.000 0.003 20.284 20.312 compute_2c_integrals 1 7.0 0.003 0.005 19.263 19.269 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.889 19.139 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.848 18.888 19.137 fft3d_s 1823 13.4 18.426 18.821 18.440 18.835 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.231 14.231 calculate_wavefunction 91 8.0 2.024 2.051 9.751 10.006 mp2_ri_gpw_compute_en_expansio 172 7.0 0.553 0.572 8.811 9.453 potential_pw2rs 186 10.0 0.034 0.035 8.664 9.350 mp2_ri_gpw_compute_en_comm 22 7.0 0.479 0.503 8.622 8.997 local_gemm 172 8.0 8.258 8.882 8.258 8.882 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.264 8.632 calc_potential_gpw 182 9.5 0.002 0.002 7.931 8.177 collocate_single_gaussian 91 10.0 0.017 0.024 7.887 8.100 mp_sendrecv_dm3 2068 8.0 6.718 7.050 6.718 7.050 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.407 6.344 6.407 pw_gather_s 912 13.2 4.926 5.572 4.926 5.572 mp_sync 38 10.4 2.447 4.985 2.447 4.985 pw_scatter_s 910 13.7 3.935 4.422 3.935 4.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.404610, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 451.457024E+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 480344. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 53.470 53.471 qs_mol_dyn_low 1 2.0 0.004 0.007 53.263 53.271 qs_forces 11 3.9 0.002 0.002 53.193 53.194 qs_energies 11 4.9 0.001 0.003 51.723 51.747 scf_env_do_scf 11 5.9 0.000 0.001 45.894 45.894 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.786 43.786 dbcsr_multiply_generic 2286 12.5 0.095 0.098 34.766 35.278 qs_scf_new_mos 108 7.5 0.000 0.001 33.675 33.971 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.675 33.970 ot_scf_mini 108 9.5 0.002 0.002 31.990 32.187 multiply_cannon 2286 13.5 0.185 0.191 26.571 28.214 multiply_cannon_loop 2286 14.5 1.484 1.554 25.856 27.540 velocity_verlet 10 3.0 0.001 0.002 27.015 27.015 ot_mini 108 10.5 0.001 0.001 20.521 20.780 qs_ot_get_derivative 108 11.5 0.001 0.001 17.443 17.648 mp_waitall_1 267858 16.1 9.584 15.401 9.584 15.401 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 6.502 13.458 multiply_cannon_multrec 54864 15.5 4.152 6.501 7.464 10.854 rebuild_ks_matrix 119 8.3 0.000 0.001 7.873 8.015 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.872 8.015 mp_sum_l 7207 12.9 6.176 7.834 6.176 7.834 multiply_cannon_sync_h2d 54864 15.5 5.816 7.158 5.816 7.158 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.958 7.085 qs_ot_get_p 119 10.4 0.001 0.001 6.653 6.993 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.051 6.527 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.703 5.805 init_scf_run 11 5.9 0.000 0.001 4.570 4.570 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.569 4.570 sum_up_and_integrate 119 10.3 0.012 0.014 4.494 4.501 integrate_v_rspace 119 11.3 0.002 0.002 4.482 4.491 dbcsr_mm_accdrv_process 76910 16.1 1.184 1.816 3.234 4.486 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.157 4.296 calculate_rho_elec 119 8.7 0.011 0.016 4.156 4.295 multiply_cannon_metrocomm1 54864 15.5 0.051 0.058 2.169 3.603 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.522 3.568 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.925 3.127 apply_single 119 13.6 0.000 0.000 2.924 3.127 calculate_dm_sparse 119 9.5 0.000 0.000 2.909 3.051 rs_pw_transfer 974 11.9 0.011 0.012 2.795 2.894 ot_diis_step 108 11.5 0.006 0.006 2.770 2.770 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.613 2.705 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.630 2.630 jit_kernel_multiply 13 15.8 1.988 2.586 1.988 2.586 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.401 2.402 calculate_first_density_matrix 1 7.0 0.000 0.001 2.385 2.389 cp_fm_redistribute_end 50 14.0 2.177 2.376 2.181 2.378 cp_fm_diag_elpa_base 50 14.0 0.194 2.333 0.195 2.342 density_rs2pw 119 9.7 0.004 0.004 2.236 2.314 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.003 2.202 2.204 acc_transpose_blocks 54864 15.5 0.231 0.256 1.728 2.133 wfi_extrapolate 11 7.9 0.001 0.001 2.123 2.123 mp_sum_d 4125 12.0 1.485 2.113 1.485 2.113 grid_integrate_task_list 119 12.3 2.019 2.110 2.019 2.110 init_scf_loop 11 6.9 0.000 0.001 2.091 2.092 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.906 1.968 potential_pw2rs 119 12.3 0.004 0.004 1.861 1.876 pw_transfer 1439 11.6 0.051 0.056 1.707 1.781 make_m2s 4572 13.5 0.054 0.057 1.714 1.759 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.632 1.708 make_images 4572 14.5 0.133 0.141 1.632 1.676 mp_alltoall_d11v 2130 13.8 1.381 1.578 1.381 1.578 fft3d_ps 1201 14.6 0.359 0.461 1.411 1.478 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.401 1.436 mp_waitany 12084 13.8 1.242 1.396 1.242 1.396 grid_collocate_task_list 119 9.7 1.282 1.334 1.282 1.334 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.251 1.329 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.814 1.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.471000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.727104E+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 1576612. 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 0.017 0.032 37.848 37.849 qs_mol_dyn_low 1 2.0 0.003 0.004 37.621 37.629 qs_forces 11 3.9 0.002 0.002 37.566 37.567 qs_energies 11 4.9 0.002 0.005 35.888 35.891 scf_env_do_scf 11 5.9 0.001 0.001 30.876 30.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.447 28.449 dbcsr_multiply_generic 2286 12.5 0.102 0.104 21.134 21.488 qs_scf_new_mos 108 7.5 0.001 0.001 19.576 19.820 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.575 19.819 ot_scf_mini 108 9.5 0.003 0.003 18.683 18.856 velocity_verlet 10 3.0 0.001 0.002 17.908 17.909 multiply_cannon 2286 13.5 0.208 0.219 16.232 17.832 multiply_cannon_loop 2286 14.5 0.893 0.960 15.139 16.480 ot_mini 108 10.5 0.001 0.001 11.663 11.899 mp_waitall_1 217478 16.2 5.922 11.093 5.922 11.093 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.133 9.410 qs_ot_get_derivative 108 11.5 0.001 0.001 9.206 9.383 multiply_cannon_multrec 27432 15.5 1.968 4.256 5.851 8.704 rebuild_ks_matrix 119 8.3 0.000 0.000 7.096 7.232 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.096 7.232 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.262 6.386 dbcsr_mm_accdrv_process 47894 16.0 3.033 5.344 3.814 5.629 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.585 4.435 qs_ot_get_p 119 10.4 0.001 0.001 4.151 4.391 sum_up_and_integrate 119 10.3 0.024 0.027 4.130 4.137 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.049 4.121 apply_single 119 13.6 0.000 0.000 3.049 4.121 integrate_v_rspace 119 11.3 0.002 0.002 4.106 4.114 mp_sum_l 7207 12.9 2.013 3.950 2.013 3.950 init_scf_run 11 5.9 0.000 0.001 3.790 3.791 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.790 3.791 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.557 3.585 calculate_rho_elec 119 8.7 0.021 0.024 3.556 3.584 multiply_cannon_sync_h2d 27432 15.5 2.216 2.873 2.216 2.873 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.728 2.747 make_m2s 4572 13.5 0.052 0.054 2.440 2.651 rs_pw_transfer 974 11.9 0.010 0.011 2.496 2.576 make_images 4572 14.5 0.200 0.237 2.351 2.562 init_scf_loop 11 6.9 0.001 0.003 2.407 2.409 ot_diis_step 108 11.5 0.011 0.011 2.408 2.409 calculate_first_density_matrix 1 7.0 0.000 0.002 2.301 2.304 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.301 2.301 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.090 2.181 calculate_dm_sparse 119 9.5 0.000 0.001 2.034 2.111 density_rs2pw 119 9.7 0.004 0.004 1.924 2.003 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.963 1.963 cp_fm_redistribute_end 50 14.0 1.617 1.941 1.621 1.942 grid_integrate_task_list 119 12.3 1.843 1.929 1.843 1.929 cp_fm_diag_elpa_base 50 14.0 0.313 1.891 0.320 1.922 potential_pw2rs 119 12.3 0.006 0.006 1.868 1.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.868 pw_transfer 1439 11.6 0.063 0.066 1.752 1.787 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.662 1.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.694 jit_kernel_multiply 8 16.1 0.728 1.547 0.728 1.547 make_images_data 4572 15.5 0.045 0.051 1.130 1.537 prepare_preconditioner 11 7.9 0.000 0.000 1.496 1.523 make_preconditioner 11 8.9 0.000 0.001 1.496 1.523 acc_transpose_blocks 27432 15.5 0.109 0.114 1.180 1.469 hybrid_alltoall_any 4725 16.4 0.051 0.110 0.980 1.461 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.402 1.457 wfi_extrapolate 11 7.9 0.001 0.001 1.438 1.438 fft3d_ps 1201 14.6 0.499 0.555 1.370 1.404 fft_wrap_pw1pw2_140 487 13.2 0.076 0.085 1.305 1.340 mp_alltoall_d11v 2130 13.8 1.159 1.318 1.159 1.318 grid_collocate_task_list 119 9.7 1.241 1.311 1.241 1.311 mp_allgather_i34 2286 14.5 0.535 1.272 0.535 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.257 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.163 1.213 mp_sum_d 4125 12.0 0.583 1.039 0.583 1.039 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.951 0.952 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.903 acc_transpose_blocks_kernels 27432 16.5 0.179 0.267 0.651 0.858 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.748 0.788 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.693 0.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=37.849000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 522.051584E+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 0.016 0.031 33.336 33.337 qs_mol_dyn_low 1 2.0 0.003 0.004 33.067 33.074 qs_forces 11 3.9 0.001 0.002 33.011 33.011 qs_energies 11 4.9 0.003 0.007 31.423 31.425 scf_env_do_scf 11 5.9 0.001 0.002 26.586 26.587 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.789 23.790 dbcsr_multiply_generic 2286 12.5 0.096 0.100 16.870 16.969 velocity_verlet 10 3.0 0.001 0.002 15.828 15.830 qs_scf_new_mos 108 7.5 0.001 0.001 15.560 15.581 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.559 15.581 ot_scf_mini 108 9.5 0.003 0.003 14.815 14.832 multiply_cannon 2286 13.5 0.195 0.205 13.290 14.040 multiply_cannon_loop 2286 14.5 0.632 0.666 12.431 13.327 ot_mini 108 10.5 0.001 0.001 9.105 9.121 qs_ot_get_derivative 108 11.5 0.001 0.001 7.546 7.567 multiply_cannon_multrec 18288 15.5 1.899 2.801 6.820 7.108 rebuild_ks_matrix 119 8.3 0.000 0.000 6.499 6.516 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.498 6.515 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.751 5.768 dbcsr_mm_accdrv_process 38222 16.0 3.937 5.573 4.837 5.709 mp_waitall_1 169478 16.3 3.349 4.722 3.349 4.722 sum_up_and_integrate 119 10.3 0.031 0.032 4.100 4.104 integrate_v_rspace 119 11.3 0.002 0.003 4.068 4.077 init_scf_run 11 5.9 0.000 0.001 3.633 3.633 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.633 3.633 qs_ot_get_p 119 10.4 0.001 0.001 3.479 3.502 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.435 3.443 calculate_rho_elec 119 8.7 0.030 0.031 3.434 3.442 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.818 3.435 init_scf_loop 11 6.9 0.001 0.002 2.778 2.779 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.483 2.729 rs_pw_transfer 974 11.9 0.009 0.010 2.531 2.642 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.066 2.379 apply_single 119 13.6 0.000 0.000 2.066 2.379 calculate_first_density_matrix 1 7.0 0.001 0.002 2.315 2.316 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.267 2.273 density_rs2pw 119 9.7 0.004 0.004 2.022 2.150 jit_kernel_multiply 11 15.9 0.848 2.139 0.848 2.139 make_m2s 4572 13.5 0.045 0.046 1.954 2.131 make_images 4572 14.5 0.191 0.203 1.869 2.044 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.989 1.990 prepare_preconditioner 11 7.9 0.000 0.000 1.935 1.939 make_preconditioner 11 8.9 0.000 0.001 1.935 1.939 grid_integrate_task_list 119 12.3 1.811 1.912 1.811 1.912 calculate_dm_sparse 119 9.5 0.000 0.000 1.872 1.884 potential_pw2rs 119 12.3 0.007 0.008 1.865 1.883 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.785 1.865 pw_transfer 1439 11.6 0.063 0.067 1.809 1.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.758 1.760 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.718 1.737 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.715 1.717 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.708 1.714 cp_fm_redistribute_end 50 14.0 1.266 1.683 1.267 1.684 cp_fm_diag_elpa_base 50 14.0 0.399 1.618 0.414 1.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.582 1.591 mp_sum_l 7207 12.9 1.242 1.587 1.242 1.587 multiply_cannon_sync_h2d 18288 15.5 1.401 1.549 1.401 1.549 ot_diis_step 108 11.5 0.011 0.011 1.523 1.524 fft3d_ps 1201 14.6 0.508 0.524 1.405 1.424 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.312 1.330 grid_collocate_task_list 119 9.7 1.203 1.297 1.203 1.297 acc_transpose_blocks 18288 15.5 0.076 0.078 1.261 1.296 wfi_extrapolate 11 7.9 0.001 0.001 1.252 1.252 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.226 make_images_data 4572 15.5 0.045 0.048 0.860 1.059 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.963 0.964 hybrid_alltoall_any 4725 16.4 0.055 0.115 0.732 0.948 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.921 0.940 mp_alltoall_d11v 2130 13.8 0.755 0.907 0.755 0.907 cp_fm_cholesky_invert 11 10.9 0.836 0.844 0.836 0.844 acc_transpose_blocks_kernels 18288 16.5 0.208 0.219 0.823 0.844 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.825 mp_alltoall_z22v 1201 16.6 0.729 0.793 0.729 0.793 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.657 0.736 mp_allgather_i34 2286 14.5 0.318 0.691 0.318 0.691 mp_waitany 9880 13.7 0.537 0.675 0.537 0.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.337000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.641406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.598400E+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 0.019 0.035 35.265 35.266 qs_mol_dyn_low 1 2.0 0.003 0.004 35.018 35.025 qs_forces 11 3.9 0.001 0.002 34.960 34.961 qs_energies 11 4.9 0.003 0.007 33.230 33.235 scf_env_do_scf 11 5.9 0.001 0.002 28.112 28.113 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.599 24.601 dbcsr_multiply_generic 2286 12.5 0.104 0.110 18.280 18.366 velocity_verlet 10 3.0 0.001 0.002 18.028 18.030 qs_scf_new_mos 108 7.5 0.001 0.001 16.301 16.360 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.301 16.359 ot_scf_mini 108 9.5 0.003 0.003 15.354 15.405 multiply_cannon 2286 13.5 0.230 0.268 14.522 15.274 multiply_cannon_loop 2286 14.5 0.933 0.961 13.530 14.153 ot_mini 108 10.5 0.001 0.001 9.453 9.516 multiply_cannon_multrec 27432 15.5 2.334 3.050 8.599 8.941 qs_ot_get_derivative 108 11.5 0.001 0.001 7.609 7.660 dbcsr_mm_accdrv_process 47916 15.9 5.330 7.142 6.171 7.368 rebuild_ks_matrix 119 8.3 0.000 0.000 6.603 6.657 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.603 6.657 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.853 5.899 sum_up_and_integrate 119 10.3 0.035 0.038 3.849 3.859 integrate_v_rspace 119 11.3 0.002 0.003 3.814 3.824 init_scf_run 11 5.9 0.000 0.001 3.714 3.714 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.714 3.714 init_scf_loop 11 6.9 0.001 0.003 3.493 3.505 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.358 3.393 calculate_rho_elec 119 8.7 0.040 0.046 3.357 3.393 qs_ot_get_p 119 10.4 0.001 0.001 3.269 3.344 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.949 3.262 mp_waitall_1 145218 16.4 2.357 2.858 2.357 2.858 prepare_preconditioner 11 7.9 0.000 0.000 2.623 2.630 make_preconditioner 11 8.9 0.000 0.002 2.623 2.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.234 2.555 make_m2s 4572 13.5 0.054 0.057 2.366 2.476 make_images 4572 14.5 0.273 0.334 2.258 2.367 calculate_first_density_matrix 1 7.0 0.001 0.002 2.296 2.298 rs_pw_transfer 974 11.9 0.009 0.009 2.134 2.209 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.985 2.205 apply_single 119 13.6 0.000 0.000 1.985 2.205 calculate_dm_sparse 119 9.5 0.000 0.000 2.036 2.088 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.982 2.009 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.946 1.955 grid_integrate_task_list 119 12.3 1.827 1.939 1.827 1.939 density_rs2pw 119 9.7 0.004 0.004 1.836 1.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.862 1.863 pw_transfer 1439 11.6 0.063 0.066 1.825 1.855 ot_diis_step 108 11.5 0.012 0.012 1.803 1.804 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.735 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.635 1.650 jit_kernel_multiply 9 15.9 0.779 1.650 0.779 1.650 potential_pw2rs 119 12.3 0.008 0.009 1.624 1.630 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.621 1.622 acc_transpose_blocks 27432 15.5 0.111 0.113 1.453 1.495 fft3d_ps 1201 14.6 0.535 0.587 1.416 1.442 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.406 1.441 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.844 1.420 wfi_extrapolate 11 7.9 0.001 0.001 1.369 1.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.364 1.365 cp_fm_redistribute_end 50 14.0 0.898 1.340 0.899 1.341 mp_sum_l 7207 12.9 0.974 1.336 0.974 1.336 grid_collocate_task_list 119 9.7 1.221 1.321 1.221 1.321 cp_fm_diag_elpa_base 50 14.0 0.420 1.278 0.439 1.318 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.219 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.123 1.144 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.140 1.142 dbcsr_complete_redistribute 329 12.2 0.133 0.166 0.825 1.099 cp_fm_upper_to_full 72 13.5 0.791 1.097 0.791 1.097 multiply_cannon_sync_h2d 27432 15.5 0.992 1.069 0.992 1.069 make_images_data 4572 15.5 0.045 0.048 0.929 1.061 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.774 0.972 mp_alltoall_d11v 2130 13.8 0.786 0.924 0.786 0.924 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.832 0.876 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.869 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.599 0.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.845 cp_fm_cholesky_invert 11 10.9 0.815 0.818 0.815 0.818 mp_alltoall_z22v 1201 16.6 0.763 0.789 0.763 0.789 mp_alltoall_i22 627 13.8 0.441 0.729 0.441 0.729 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.266000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.545455, yerr=2.349609 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 602.779648E+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 0.012 0.030 26.906 26.907 qs_mol_dyn_low 1 2.0 0.003 0.004 26.736 26.743 qs_forces 11 3.9 0.002 0.002 26.679 26.680 qs_energies 11 4.9 0.001 0.001 24.997 24.999 scf_env_do_scf 11 5.9 0.000 0.001 20.448 20.448 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.024 18.025 velocity_verlet 10 3.0 0.002 0.002 13.964 13.967 dbcsr_multiply_generic 2286 12.5 0.095 0.098 11.861 11.930 qs_scf_new_mos 108 7.5 0.001 0.001 10.561 10.591 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.561 10.590 ot_scf_mini 108 9.5 0.002 0.002 9.885 9.914 multiply_cannon 2286 13.5 0.230 0.242 9.388 9.828 multiply_cannon_loop 2286 14.5 0.329 0.340 8.482 8.664 multiply_cannon_multrec 9144 15.5 1.569 1.816 5.721 5.942 rebuild_ks_matrix 119 8.3 0.000 0.000 5.712 5.735 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.712 5.735 ot_mini 108 10.5 0.001 0.001 5.503 5.537 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.089 5.111 qs_ot_get_derivative 108 11.5 0.001 0.001 4.206 4.237 dbcsr_mm_accdrv_process 12550 15.8 2.998 4.048 4.051 4.135 sum_up_and_integrate 119 10.3 0.037 0.041 3.469 3.475 integrate_v_rspace 119 11.3 0.003 0.004 3.432 3.439 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.177 3.183 calculate_rho_elec 119 8.7 0.059 0.061 3.177 3.182 init_scf_run 11 5.9 0.000 0.001 3.136 3.136 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.136 3.136 qs_ot_get_p 119 10.4 0.001 0.001 2.618 2.659 init_scf_loop 11 6.9 0.000 0.000 2.404 2.405 calculate_first_density_matrix 1 7.0 0.000 0.000 2.049 2.049 mp_waitall_1 121218 16.5 1.466 1.961 1.466 1.961 grid_integrate_task_list 119 12.3 1.851 1.929 1.851 1.929 make_m2s 4572 13.5 0.034 0.036 1.698 1.868 make_images 4572 14.5 0.267 0.301 1.609 1.776 jit_kernel_multiply 10 15.8 1.015 1.738 1.015 1.738 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.734 1.736 calculate_dm_sparse 119 9.5 0.000 0.000 1.702 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.690 1.694 make_preconditioner 11 8.9 0.000 0.000 1.690 1.694 pw_transfer 1439 11.6 0.063 0.065 1.669 1.676 density_rs2pw 119 9.7 0.003 0.004 1.580 1.653 rs_pw_transfer 974 11.9 0.008 0.008 1.574 1.649 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.579 1.606 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.578 1.586 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.540 1.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.505 1.506 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.482 1.505 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.348 1.356 grid_collocate_task_list 119 9.7 1.271 1.326 1.271 1.326 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.298 1.306 ot_diis_step 108 11.5 0.013 0.013 1.284 1.284 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.259 1.260 potential_pw2rs 119 12.3 0.010 0.010 1.255 1.259 fft3d_ps 1201 14.6 0.540 0.553 1.245 1.254 cp_fm_redistribute_end 50 14.0 0.625 1.242 0.626 1.243 cp_fm_diag_elpa_base 50 14.0 0.573 1.176 0.615 1.231 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.208 1.208 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.164 1.189 apply_single 119 13.6 0.000 0.000 1.164 1.189 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.118 1.123 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.096 1.111 wfi_extrapolate 11 7.9 0.001 0.001 1.044 1.044 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.766 1.020 make_images_data 4572 15.5 0.039 0.042 0.788 0.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.923 mp_alltoall_d11v 2130 13.8 0.794 0.906 0.794 0.906 acc_transpose_blocks 9144 15.5 0.038 0.039 0.892 0.904 cp_fm_cholesky_invert 11 10.9 0.851 0.854 0.851 0.854 multiply_cannon_sync_h2d 9144 15.5 0.708 0.798 0.708 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.744 0.746 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.714 0.723 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.665 0.713 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.652 0.661 mp_allgather_i34 2286 14.5 0.240 0.639 0.240 0.639 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.311 0.629 mp_alltoall_z22v 1201 16.6 0.579 0.611 0.579 0.611 qs_create_task_list 11 7.9 0.002 0.002 0.545 0.566 generate_qs_task_list 11 8.9 0.189 0.212 0.543 0.565 jit_kernel_transpose 5 15.6 0.535 0.545 0.535 0.545 mp_waitany 5200 13.7 0.458 0.543 0.458 0.543 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=26.907000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.545455, yerr=6.213821 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 753.463296E+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 0.013 0.032 39.572 39.572 qs_mol_dyn_low 1 2.0 0.003 0.003 39.385 39.391 qs_forces 11 3.9 0.001 0.002 39.328 39.328 qs_energies 11 4.9 0.001 0.001 37.405 37.408 scf_env_do_scf 11 5.9 0.001 0.001 31.785 31.785 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.041 24.042 velocity_verlet 10 3.0 0.002 0.002 22.530 22.535 dbcsr_multiply_generic 2286 12.5 0.106 0.108 17.197 17.439 qs_scf_new_mos 108 7.5 0.001 0.001 15.356 15.454 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.356 15.454 multiply_cannon 2286 13.5 0.301 0.309 13.565 14.491 ot_scf_mini 108 9.5 0.002 0.002 14.285 14.385 multiply_cannon_loop 2286 14.5 0.341 0.347 12.347 13.306 multiply_cannon_multrec 9144 15.5 3.311 4.629 8.637 8.784 ot_mini 108 10.5 0.001 0.001 8.565 8.678 init_scf_loop 11 6.9 0.000 0.000 7.718 7.719 prepare_preconditioner 11 7.9 0.000 0.000 6.818 6.832 make_preconditioner 11 8.9 0.000 0.000 6.818 6.832 rebuild_ks_matrix 119 8.3 0.000 0.001 6.676 6.815 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.675 6.815 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.332 6.714 qs_ot_get_derivative 108 11.5 0.001 0.001 6.564 6.662 dbcsr_mm_accdrv_process 12550 15.8 4.279 5.722 5.201 6.528 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.039 6.165 cp_fm_upper_to_full 72 14.2 3.210 4.644 3.210 4.644 init_scf_run 11 5.9 0.000 0.001 3.616 3.616 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.615 3.615 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.591 3.596 calculate_rho_elec 119 8.7 0.118 0.121 3.591 3.595 sum_up_and_integrate 119 10.3 0.064 0.067 3.566 3.573 integrate_v_rspace 119 11.3 0.003 0.004 3.502 3.510 mp_waitall_1 97218 16.6 2.313 3.251 2.313 3.251 qs_ot_get_p 119 10.4 0.001 0.001 3.018 3.152 dbcsr_complete_redistribute 329 12.2 0.289 0.293 2.008 2.868 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.429 2.811 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.712 2.572 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.120 2.420 apply_single 119 13.6 0.000 0.000 2.120 2.420 mp_alltoall_i22 627 13.8 1.461 2.361 1.461 2.361 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.480 2.330 make_m2s 4572 13.5 0.038 0.038 2.138 2.290 calculate_dm_sparse 119 9.5 0.000 0.000 2.230 2.249 calculate_first_density_matrix 1 7.0 0.000 0.000 2.231 2.231 make_images 4572 14.5 0.353 0.389 2.018 2.169 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.232 2.118 grid_integrate_task_list 119 12.3 1.990 2.000 1.990 2.000 ot_diis_step 108 11.5 0.014 0.015 1.976 1.976 pw_transfer 1439 11.6 0.066 0.067 1.961 1.963 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.866 1.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.857 1.857 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.802 1.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.742 1.743 density_rs2pw 119 9.7 0.003 0.004 1.713 1.726 mp_sum_l 7207 12.9 0.948 1.685 0.948 1.685 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.629 1.678 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.606 1.651 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.573 1.576 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.537 1.537 fft3d_ps 1201 14.6 0.566 0.575 1.500 1.502 grid_collocate_task_list 119 9.7 1.435 1.458 1.435 1.458 cp_fm_cholesky_invert 11 10.9 1.382 1.386 1.382 1.386 rs_pw_transfer 974 11.9 0.009 0.009 1.305 1.333 wfi_extrapolate 11 7.9 0.001 0.001 1.328 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.304 1.311 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.256 1.256 cp_fm_diag_elpa_base 50 14.0 1.111 1.162 1.254 1.254 hybrid_alltoall_any 4725 16.4 0.086 0.146 1.055 1.252 make_images_data 4572 15.5 0.042 0.046 0.976 1.162 potential_pw2rs 119 12.3 0.014 0.014 1.140 1.142 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.126 1.142 mp_alltoall_d11v 2130 13.8 1.102 1.111 1.102 1.111 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.082 1.099 jit_kernel_multiply 6 15.6 0.895 1.096 0.895 1.096 multiply_cannon_sync_h2d 9144 15.5 1.041 1.045 1.041 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.991 1.035 qs_create_task_list 11 7.9 0.000 0.000 0.939 0.951 generate_qs_task_list 11 8.9 0.372 0.392 0.938 0.951 acc_transpose_blocks 9144 15.5 0.038 0.038 0.922 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.847 0.860 mp_alltoall_z22v 1201 16.6 0.798 0.813 0.798 0.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=39.572000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.454545, yerr=18.386574 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.554048E+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 2493559. 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 0.015 0.032 82.691 82.693 qs_mol_dyn_low 1 2.0 0.004 0.006 82.381 82.391 qs_forces 11 3.9 0.002 0.005 82.149 82.150 qs_energies 11 4.9 0.008 0.017 79.334 79.353 scf_env_do_scf 11 5.9 0.001 0.002 70.434 70.437 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.962 64.963 dbcsr_multiply_generic 2055 12.4 0.109 0.117 52.601 52.948 qs_scf_new_mos 99 7.5 0.000 0.001 48.354 48.480 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.354 48.479 ot_scf_mini 99 9.5 0.002 0.003 45.956 46.066 multiply_cannon 2055 13.4 0.179 0.185 42.817 43.885 multiply_cannon_loop 2055 14.4 1.538 1.577 41.903 42.993 velocity_verlet 10 3.0 0.001 0.002 41.546 41.549 ot_mini 99 10.5 0.001 0.001 28.021 28.144 qs_ot_get_derivative 99 11.5 0.001 0.001 21.171 21.273 multiply_cannon_multrec 49320 15.4 12.118 12.807 17.144 17.887 rebuild_ks_matrix 110 8.3 0.000 0.001 14.264 14.438 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.264 14.438 mp_waitall_1 241148 16.1 11.959 12.991 11.959 12.991 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.527 12.680 multiply_cannon_sync_h2d 49320 15.4 9.929 10.541 9.929 10.541 qs_ot_get_p 110 10.4 0.001 0.005 9.279 9.439 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.238 8.406 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.171 7.675 apply_single 110 13.6 0.000 0.001 7.171 7.675 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.880 7.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.005 7.049 7.156 sum_up_and_integrate 110 10.3 0.036 0.044 6.893 6.907 integrate_v_rspace 110 11.3 0.002 0.003 6.857 6.878 init_scf_run 11 5.9 0.000 0.001 6.783 6.783 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.782 6.783 ot_diis_step 99 11.5 0.006 0.008 6.549 6.549 mp_sum_l 6514 12.8 5.524 6.269 5.524 6.269 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.025 6.175 calculate_rho_elec 110 8.6 0.020 0.024 6.025 6.175 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.076 6.117 init_scf_loop 11 6.9 0.001 0.003 5.446 5.449 dbcsr_mm_accdrv_process 87628 16.1 2.086 2.169 4.905 5.190 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.038 5.039 cp_fm_diag_elpa 48 13.0 0.000 0.001 4.578 4.580 cp_fm_redistribute_end 48 14.0 3.986 4.556 3.989 4.557 cp_fm_diag_elpa_base 48 14.0 0.562 4.501 0.565 4.525 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 2.798 3.967 wfi_extrapolate 11 7.9 0.001 0.003 3.949 3.949 make_m2s 4110 13.4 0.061 0.065 3.806 3.891 calculate_dm_sparse 110 9.5 0.000 0.001 3.773 3.884 make_images 4110 14.4 0.178 0.190 3.711 3.799 rs_pw_transfer 902 11.9 0.011 0.013 3.585 3.766 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.578 3.582 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.397 3.448 grid_integrate_task_list 110 12.3 3.247 3.396 3.247 3.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.235 3.306 prepare_preconditioner 11 7.9 0.000 0.000 3.277 3.297 make_preconditioner 11 8.9 0.000 0.001 3.277 3.297 density_rs2pw 110 9.6 0.004 0.005 3.107 3.244 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.051 3.096 pw_transfer 1331 11.6 0.054 0.066 2.760 2.824 calculate_first_density_matrix 1 7.0 0.000 0.003 2.733 2.743 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.672 2.735 mp_alltoall_d11v 2046 13.8 2.094 2.588 2.094 2.588 jit_kernel_multiply 13 15.9 2.534 2.551 2.534 2.551 potential_pw2rs 110 12.3 0.006 0.007 2.506 2.525 fft_wrap_pw1pw2_140 451 13.1 0.170 0.190 2.270 2.330 fft3d_ps 1111 14.6 0.752 0.835 2.209 2.268 acc_transpose_blocks 49320 15.4 0.220 0.230 2.191 2.262 grid_collocate_task_list 110 9.6 2.087 2.187 2.087 2.187 mp_sum_d 3879 11.9 1.565 2.079 1.565 2.079 mp_waitany 14300 13.8 1.826 2.020 1.826 2.020 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.877 1.901 make_images_data 4110 15.4 0.042 0.045 1.746 1.901 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.531 1.829 cp_fm_cholesky_invert 11 10.9 1.773 1.777 1.773 1.777 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.645 1.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.693000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.678349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 587.304960E+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 2294321. 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 0.023 0.034 68.925 68.926 qs_mol_dyn_low 1 2.0 0.004 0.005 68.613 68.623 qs_forces 11 3.9 0.002 0.002 68.527 68.527 qs_energies 11 4.9 0.003 0.008 65.219 65.223 scf_env_do_scf 11 5.9 0.001 0.001 56.717 56.720 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 48.947 48.948 dbcsr_multiply_generic 2055 12.4 0.116 0.121 37.844 37.984 velocity_verlet 10 3.0 0.001 0.002 36.246 36.247 qs_scf_new_mos 99 7.5 0.001 0.001 32.993 33.133 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.992 33.133 multiply_cannon 2055 13.4 0.223 0.250 31.086 32.177 ot_scf_mini 99 9.5 0.003 0.004 31.314 31.455 multiply_cannon_loop 2055 14.4 0.919 0.946 29.801 30.769 ot_mini 99 10.5 0.001 0.001 18.344 18.475 multiply_cannon_multrec 24660 15.4 7.633 9.945 13.818 15.938 rebuild_ks_matrix 110 8.3 0.000 0.001 13.682 13.773 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 13.682 13.772 qs_ot_get_derivative 99 11.5 0.001 0.001 12.488 12.622 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.059 12.144 mp_waitall_1 186928 16.3 8.020 10.792 8.020 10.792 multiply_cannon_sync_h2d 24660 15.4 7.099 8.206 7.099 8.206 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.256 8.057 init_scf_loop 11 6.9 0.001 0.006 7.732 7.733 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.593 7.347 apply_single 110 13.6 0.000 0.001 6.592 7.347 sum_up_and_integrate 110 10.3 0.052 0.059 6.558 6.570 integrate_v_rspace 110 11.3 0.002 0.003 6.506 6.521 qs_ot_get_p 110 10.4 0.001 0.002 6.287 6.435 dbcsr_mm_accdrv_process 52282 16.1 4.658 5.446 6.026 6.338 init_scf_run 11 5.9 0.000 0.001 6.068 6.069 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.068 6.069 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.815 5.823 calculate_rho_elec 110 8.6 0.039 0.047 5.814 5.822 ot_diis_step 99 11.5 0.010 0.010 5.806 5.806 prepare_preconditioner 11 7.9 0.000 0.000 5.719 5.737 make_preconditioner 11 8.9 0.000 0.001 5.719 5.737 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.309 5.469 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.689 5.438 make_m2s 4110 13.4 0.057 0.061 4.278 4.712 make_images 4110 14.4 0.399 0.442 4.169 4.598 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.360 4.382 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.919 3.919 wfi_extrapolate 11 7.9 0.001 0.001 3.509 3.509 pw_transfer 1331 11.6 0.066 0.071 3.342 3.477 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.235 3.372 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.355 3.357 cp_fm_redistribute_end 48 14.0 2.495 3.326 2.498 3.326 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.250 3.318 cp_fm_diag_elpa_base 48 14.0 0.792 3.216 0.825 3.302 grid_integrate_task_list 110 12.3 3.151 3.296 3.151 3.296 density_rs2pw 110 9.6 0.004 0.004 3.099 3.290 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.246 3.248 rs_pw_transfer 902 11.9 0.012 0.014 3.022 3.227 calculate_dm_sparse 110 9.5 0.001 0.001 2.977 3.015 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.890 2.929 make_images_data 4110 15.4 0.047 0.050 2.398 2.843 fft_wrap_pw1pw2_140 451 13.1 0.200 0.219 2.701 2.842 hybrid_alltoall_any 4261 16.3 0.102 0.448 2.092 2.738 fft3d_ps 1111 14.6 1.071 1.269 2.594 2.723 cp_fm_cholesky_invert 11 10.9 2.653 2.660 2.653 2.660 calculate_first_density_matrix 1 7.0 0.000 0.002 2.474 2.477 mp_sum_l 6514 12.8 1.856 2.458 1.856 2.458 potential_pw2rs 110 12.3 0.008 0.009 2.414 2.434 grid_collocate_task_list 110 9.6 2.117 2.277 2.117 2.277 jit_kernel_multiply 11 16.4 1.016 2.113 1.016 2.113 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.006 2.028 mp_alltoall_d11v 2046 13.8 1.815 1.990 1.815 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.894 1.896 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.764 1.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.801 mp_allgather_i34 2055 14.4 0.611 1.640 0.611 1.640 mp_irecv_dv 57340 16.2 0.641 1.598 0.641 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.544 acc_transpose_blocks 24660 15.4 0.110 0.112 1.510 1.536 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.478 mp_waitany 10164 13.8 1.221 1.463 1.221 1.463 cp_fm_cholesky_decompose 22 10.9 1.424 1.429 1.424 1.429 dbcsr_complete_redistribute 325 12.2 0.249 0.313 1.162 1.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.926000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.636364, yerr=6.852640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.689472E+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 0.018 0.033 59.770 59.771 qs_mol_dyn_low 1 2.0 0.003 0.005 59.449 59.458 qs_forces 11 3.9 0.002 0.003 59.155 59.156 qs_energies 11 4.9 0.002 0.007 56.020 56.023 scf_env_do_scf 11 5.9 0.001 0.002 48.222 48.222 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.577 39.577 velocity_verlet 10 3.0 0.001 0.002 32.562 32.571 dbcsr_multiply_generic 2055 12.4 0.111 0.116 28.420 28.664 qs_scf_new_mos 99 7.5 0.001 0.001 25.212 25.320 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.212 25.319 ot_scf_mini 99 9.5 0.003 0.004 23.986 24.114 multiply_cannon 2055 13.4 0.211 0.219 22.025 23.345 multiply_cannon_loop 2055 14.4 0.611 0.629 20.826 22.019 ot_mini 99 10.5 0.001 0.001 13.698 13.826 rebuild_ks_matrix 110 8.3 0.000 0.001 12.074 12.219 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.074 12.219 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.660 10.793 multiply_cannon_multrec 16440 15.4 3.861 4.784 9.601 10.554 mp_waitall_1 146766 16.3 7.119 10.209 7.119 10.209 qs_ot_get_derivative 99 11.5 0.001 0.002 9.176 9.299 init_scf_loop 11 6.9 0.001 0.003 8.610 8.611 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.336 7.304 prepare_preconditioner 11 7.9 0.000 0.000 6.854 6.870 make_preconditioner 11 8.9 0.000 0.001 6.854 6.870 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.184 6.548 sum_up_and_integrate 110 10.3 0.060 0.062 6.382 6.394 integrate_v_rspace 110 11.3 0.002 0.003 6.321 6.334 dbcsr_mm_accdrv_process 34862 16.1 4.637 4.925 5.595 5.680 qs_ot_get_p 110 10.4 0.001 0.001 5.504 5.654 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.484 5.498 calculate_rho_elec 110 8.6 0.058 0.058 5.483 5.497 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 5.036 5.469 apply_single 110 13.6 0.000 0.000 5.036 5.468 init_scf_run 11 5.9 0.000 0.001 5.403 5.403 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.403 5.403 make_m2s 4110 13.4 0.050 0.052 4.177 4.563 ot_diis_step 99 11.5 0.011 0.013 4.490 4.491 make_images 4110 14.4 0.394 0.514 4.062 4.449 multiply_cannon_sync_h2d 16440 15.4 3.688 4.214 3.688 4.214 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.861 3.867 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.130 3.764 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.522 3.522 grid_integrate_task_list 110 12.3 3.198 3.399 3.198 3.399 pw_transfer 1331 11.6 0.064 0.070 3.137 3.143 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.031 3.040 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.003 3.004 density_rs2pw 110 9.6 0.004 0.007 2.752 2.984 cp_fm_redistribute_end 48 14.0 1.864 2.976 1.866 2.977 cp_fm_diag_elpa_base 48 14.0 1.048 2.845 1.107 2.954 wfi_extrapolate 11 7.9 0.001 0.003 2.934 2.934 make_images_data 4110 15.4 0.043 0.046 2.407 2.876 rs_pw_transfer 902 11.9 0.011 0.012 2.621 2.840 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.182 2.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.755 2.757 cp_fm_cholesky_invert 11 10.9 2.689 2.695 2.689 2.695 fft_wrap_pw1pw2_140 451 13.1 0.210 0.216 2.581 2.592 calculate_dm_sparse 110 9.5 0.001 0.001 2.506 2.536 multiply_cannon_metrocomm4 14385 15.4 0.043 0.047 0.891 2.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.417 2.471 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.397 2.464 calculate_first_density_matrix 1 7.0 0.001 0.002 2.379 2.380 mp_irecv_dv 48980 15.7 0.823 2.372 0.823 2.372 fft3d_ps 1111 14.6 1.064 1.073 2.346 2.354 grid_collocate_task_list 110 9.6 2.179 2.331 2.179 2.331 mp_sum_l 6514 12.8 1.609 2.251 1.609 2.251 potential_pw2rs 110 12.3 0.010 0.010 2.200 2.207 mp_alltoall_d11v 2046 13.8 1.805 2.077 1.805 2.077 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.956 1.956 dbcsr_complete_redistribute 325 12.2 0.326 0.371 1.459 1.930 cp_fm_upper_to_full 70 13.6 1.401 1.855 1.401 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.613 1.629 cp_fm_cholesky_decompose 22 10.9 1.594 1.616 1.594 1.616 mp_allgather_i34 2055 14.4 0.523 1.588 0.523 1.588 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.473 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.967 1.428 mp_waitany 17072 13.8 1.134 1.327 1.134 1.327 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.266 1.274 acc_transpose_blocks 16440 15.4 0.071 0.072 1.202 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.771000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.090909, yerr=9.548484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 724.795392E+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 0.018 0.030 64.376 64.377 qs_mol_dyn_low 1 2.0 0.003 0.004 64.134 64.144 qs_forces 11 3.9 0.002 0.002 64.062 64.063 qs_energies 11 4.9 0.004 0.010 60.699 60.703 scf_env_do_scf 11 5.9 0.001 0.001 52.390 52.393 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.665 40.666 velocity_verlet 10 3.0 0.001 0.002 36.615 36.617 dbcsr_multiply_generic 2055 12.4 0.117 0.120 29.545 29.759 qs_scf_new_mos 99 7.5 0.001 0.001 26.423 26.528 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.422 26.528 ot_scf_mini 99 9.5 0.003 0.004 24.786 24.873 multiply_cannon 2055 13.4 0.245 0.263 22.429 23.642 multiply_cannon_loop 2055 14.4 0.882 0.905 21.020 21.608 ot_mini 99 10.5 0.001 0.001 14.205 14.320 multiply_cannon_multrec 24660 15.4 4.181 6.829 12.602 13.666 rebuild_ks_matrix 110 8.3 0.000 0.000 11.836 11.946 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 11.836 11.946 init_scf_loop 11 6.9 0.002 0.007 11.678 11.679 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.476 10.579 qs_ot_get_derivative 99 11.5 0.001 0.001 10.022 10.118 prepare_preconditioner 11 7.9 0.000 0.000 9.940 9.959 make_preconditioner 11 8.9 0.000 0.002 9.940 9.959 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.184 9.621 dbcsr_mm_accdrv_process 52304 16.0 6.641 7.798 8.275 9.172 mp_waitall_1 126806 16.4 4.543 6.731 4.543 6.731 sum_up_and_integrate 110 10.3 0.068 0.071 6.277 6.291 integrate_v_rspace 110 11.3 0.002 0.003 6.209 6.223 make_m2s 4110 13.4 0.059 0.061 5.435 5.774 qs_ot_get_p 110 10.4 0.001 0.001 5.561 5.681 make_images 4110 14.4 0.574 0.698 5.295 5.629 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.553 5.577 calculate_rho_elec 110 8.6 0.077 0.080 5.553 5.577 init_scf_run 11 5.9 0.000 0.001 5.563 5.564 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.563 5.564 cp_fm_upper_to_full 70 13.8 3.248 4.612 3.248 4.612 ot_diis_step 99 11.5 0.011 0.012 4.145 4.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.041 4.140 apply_single 110 13.6 0.000 0.000 4.041 4.140 dbcsr_complete_redistribute 325 12.2 0.419 0.461 2.675 3.798 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.769 3.784 grid_integrate_task_list 110 12.3 3.276 3.458 3.276 3.458 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.530 3.411 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.323 3.376 multiply_cannon_sync_h2d 24660 15.4 3.182 3.370 3.182 3.370 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.281 3.282 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.130 3.245 pw_transfer 1331 11.6 0.064 0.072 3.175 3.200 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.070 3.098 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.037 3.088 make_images_data 4110 15.4 0.046 0.050 2.734 3.073 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.317 3.028 calculate_dm_sparse 110 9.5 0.001 0.001 2.960 2.992 wfi_extrapolate 11 7.9 0.001 0.001 2.938 2.938 density_rs2pw 110 9.6 0.004 0.004 2.722 2.904 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.747 2.849 cp_fm_cholesky_invert 11 10.9 2.784 2.792 2.784 2.792 mp_alltoall_i22 605 13.7 1.639 2.785 1.639 2.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.776 2.778 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.720 2.722 cp_fm_redistribute_end 48 14.0 1.352 2.690 1.353 2.691 cp_fm_diag_elpa_base 48 14.0 1.257 2.561 1.335 2.671 fft_wrap_pw1pw2_140 451 13.1 0.202 0.216 2.600 2.630 rs_pw_transfer 902 11.9 0.010 0.011 2.391 2.567 calculate_first_density_matrix 1 7.0 0.001 0.002 2.515 2.519 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.407 2.445 fft3d_ps 1111 14.6 1.061 1.100 2.372 2.397 grid_collocate_task_list 110 9.6 2.220 2.369 2.220 2.369 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.239 2.240 jit_kernel_multiply 12 15.8 1.304 2.178 1.304 2.178 potential_pw2rs 110 12.3 0.012 0.013 2.063 2.073 mp_alltoall_d11v 2046 13.8 1.807 1.969 1.807 1.969 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.711 1.747 cp_fm_cholesky_decompose 22 10.9 1.691 1.733 1.691 1.733 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.604 1.702 mp_allgather_i34 2055 14.4 0.512 1.627 0.512 1.627 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.608 1.620 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.564 1.582 mp_sum_l 6514 12.8 0.998 1.558 0.998 1.558 acc_transpose_blocks 24660 15.4 0.104 0.105 1.490 1.516 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.862 1.497 mp_irecv_dv 62702 16.1 0.764 1.421 0.764 1.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.377000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.181818, yerr=7.529556 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 833.294336E+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 0.014 0.037 53.574 53.574 qs_mol_dyn_low 1 2.0 0.003 0.003 53.336 53.346 qs_forces 11 3.9 0.002 0.002 53.274 53.275 qs_energies 11 4.9 0.001 0.002 49.679 49.685 scf_env_do_scf 11 5.9 0.000 0.001 41.641 41.641 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.000 34.001 velocity_verlet 10 3.0 0.002 0.002 30.248 30.251 dbcsr_multiply_generic 2055 12.4 0.106 0.108 22.465 22.633 qs_scf_new_mos 99 7.5 0.001 0.001 20.047 20.086 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.046 20.086 ot_scf_mini 99 9.5 0.002 0.002 18.818 18.841 multiply_cannon 2055 13.4 0.247 0.261 17.110 18.350 multiply_cannon_loop 2055 14.4 0.320 0.334 15.764 16.000 rebuild_ks_matrix 110 8.3 0.000 0.000 11.374 11.414 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.374 11.414 ot_mini 99 10.5 0.001 0.001 10.175 10.193 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.134 10.171 multiply_cannon_multrec 8220 15.4 3.196 4.500 7.403 8.351 mp_waitall_1 106626 16.5 5.962 7.707 5.962 7.707 init_scf_loop 11 6.9 0.000 0.000 7.594 7.595 qs_ot_get_derivative 99 11.5 0.001 0.001 6.389 6.412 sum_up_and_integrate 110 10.3 0.079 0.081 6.119 6.132 integrate_v_rspace 110 11.3 0.003 0.003 6.040 6.054 prepare_preconditioner 11 7.9 0.000 0.000 5.980 5.984 make_preconditioner 11 8.9 0.000 0.000 5.980 5.984 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.578 5.648 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.516 5.528 calculate_rho_elec 110 8.6 0.115 0.115 5.516 5.527 dbcsr_mm_accdrv_process 17442 15.9 2.811 3.929 4.079 4.978 init_scf_run 11 5.9 0.000 0.001 4.954 4.954 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.954 4.954 qs_ot_get_p 110 10.4 0.001 0.001 4.741 4.763 make_m2s 4110 13.4 0.038 0.040 4.146 4.422 make_images 4110 14.4 0.636 0.688 4.017 4.291 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.979 4.278 ot_diis_step 99 11.5 0.012 0.012 3.766 3.766 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.680 3.721 apply_single 110 13.6 0.000 0.000 3.680 3.720 grid_integrate_task_list 110 12.3 3.359 3.530 3.359 3.530 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.449 3.452 pw_transfer 1331 11.6 0.065 0.071 3.201 3.223 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.153 3.153 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.094 3.118 multiply_cannon_sync_h2d 8220 15.4 2.922 3.029 2.922 3.029 cp_fm_cholesky_invert 11 10.9 2.905 2.908 2.905 2.908 make_images_data 4110 15.4 0.038 0.043 2.389 2.814 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.315 2.736 density_rs2pw 110 9.6 0.004 0.004 2.580 2.707 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.667 2.702 wfi_extrapolate 11 7.9 0.001 0.001 2.660 2.660 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.644 2.644 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.640 2.641 cp_fm_redistribute_end 48 14.0 0.667 2.621 0.671 2.622 cp_fm_diag_elpa_base 48 14.0 1.772 2.432 1.944 2.595 calculate_dm_sparse 110 9.5 0.001 0.001 2.449 2.493 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.482 2.484 grid_collocate_task_list 110 9.6 2.328 2.459 2.328 2.459 fft3d_ps 1111 14.6 1.118 1.161 2.354 2.374 rs_pw_transfer 902 11.9 0.010 0.011 2.097 2.239 calculate_first_density_matrix 1 7.0 0.000 0.000 2.209 2.210 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.129 2.148 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.997 mp_alltoall_d11v 2046 13.8 1.582 1.904 1.582 1.904 potential_pw2rs 110 12.3 0.015 0.015 1.877 1.882 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.851 1.865 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.754 1.767 jit_kernel_multiply 10 15.7 0.959 1.734 0.959 1.734 cp_fm_cholesky_decompose 22 10.9 1.650 1.669 1.650 1.669 mp_allgather_i34 2055 14.4 0.529 1.633 0.529 1.633 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.488 1.605 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.571 1.574 dbcsr_complete_redistribute 325 12.2 0.570 0.592 1.408 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.432 qs_create_task_list 11 7.9 0.001 0.001 1.218 1.315 generate_qs_task_list 11 8.9 0.378 0.445 1.218 1.315 mp_waitany 9240 13.8 1.038 1.184 1.038 1.184 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.111 1.127 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.481 1.106 rs_gather_matrices 110 12.3 0.149 0.162 0.753 1.080 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.574000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.818182, yerr=11.899719 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 1.384886E+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 0.017 0.038 86.773 86.773 qs_mol_dyn_low 1 2.0 0.004 0.004 86.437 86.446 qs_forces 11 3.9 0.002 0.003 86.371 86.372 qs_energies 11 4.9 0.001 0.001 82.303 82.306 scf_env_do_scf 11 5.9 0.000 0.001 72.394 72.394 velocity_verlet 10 3.0 0.002 0.002 55.624 55.630 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.470 43.471 dbcsr_multiply_generic 2055 12.4 0.121 0.124 28.737 28.853 init_scf_loop 11 6.9 0.000 0.000 28.846 28.848 qs_scf_new_mos 99 7.5 0.001 0.001 26.923 26.962 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.922 26.961 prepare_preconditioner 11 7.9 0.000 0.000 26.909 26.918 make_preconditioner 11 8.9 0.000 0.000 26.909 26.918 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.039 26.385 ot_scf_mini 99 9.5 0.002 0.002 25.153 25.190 multiply_cannon 2055 13.4 0.349 0.385 21.355 22.161 multiply_cannon_loop 2055 14.4 0.339 0.342 19.551 19.844 cp_fm_upper_to_full 70 14.2 12.747 18.340 12.747 18.340 ot_mini 99 10.5 0.001 0.001 13.789 13.827 rebuild_ks_matrix 110 8.3 0.001 0.001 13.242 13.277 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.242 13.276 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.051 12.085 dbcsr_complete_redistribute 325 12.2 1.028 1.057 7.540 10.735 multiply_cannon_multrec 8220 15.4 4.366 4.577 9.545 9.702 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.457 9.650 qs_ot_get_derivative 99 11.5 0.001 0.001 9.172 9.207 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.854 9.025 mp_alltoall_i22 605 13.7 5.498 8.723 5.498 8.723 mp_waitall_1 87304 16.6 7.924 8.644 7.924 8.644 sum_up_and_integrate 110 10.3 0.151 0.152 6.726 6.743 integrate_v_rspace 110 11.3 0.003 0.003 6.575 6.591 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.545 6.585 calculate_rho_elec 110 8.6 0.227 0.227 6.544 6.585 qs_ot_get_p 110 10.4 0.001 0.001 6.137 6.172 make_m2s 4110 13.4 0.043 0.043 5.508 6.006 make_images 4110 14.4 0.875 0.920 5.320 5.817 init_scf_run 11 5.9 0.000 0.001 5.668 5.668 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.668 5.668 cp_fm_cholesky_invert 11 10.9 5.656 5.660 5.656 5.660 dbcsr_mm_accdrv_process 11614 15.7 3.350 3.711 5.037 5.326 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.668 5.130 apply_single 110 13.6 0.000 0.000 4.668 5.130 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.635 4.951 ot_diis_step 99 11.5 0.020 0.054 4.591 4.592 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.516 4.523 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.037 4.068 multiply_cannon_sync_h2d 8220 15.4 3.953 3.958 3.953 3.958 hybrid_alltoall_any 4261 16.3 0.256 0.550 3.041 3.859 pw_transfer 1331 11.6 0.073 0.074 3.828 3.836 make_images_data 4110 15.4 0.041 0.045 3.072 3.746 grid_integrate_task_list 110 12.3 3.664 3.721 3.664 3.721 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.712 3.720 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.632 3.632 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.389 3.390 cp_fm_diag_elpa_base 48 14.0 2.851 3.050 3.387 3.387 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.896 3.350 wfi_extrapolate 11 7.9 0.001 0.001 3.289 3.289 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.180 3.188 calculate_dm_sparse 110 9.5 0.001 0.001 3.140 3.167 density_rs2pw 110 9.6 0.004 0.004 3.027 3.046 fft3d_ps 1111 14.6 1.266 1.277 2.921 2.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.901 2.904 grid_collocate_task_list 110 9.6 2.632 2.655 2.632 2.655 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.466 2.486 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.378 2.388 calculate_first_density_matrix 1 7.0 0.000 0.000 2.271 2.274 rs_pw_transfer 902 11.9 0.010 0.011 2.193 2.227 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.170 2.226 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.101 2.190 mp_alltoall_d11v 2046 13.8 2.128 2.184 2.128 2.184 potential_pw2rs 110 12.3 0.021 0.022 2.132 2.136 cp_fm_cholesky_decompose 22 10.9 2.104 2.127 2.104 2.127 qs_create_task_list 11 7.9 0.000 0.001 1.896 1.943 generate_qs_task_list 11 8.9 0.737 0.789 1.896 1.942 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.849 1.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.796 1.798 jit_kernel_multiply 10 15.2 1.485 1.740 1.485 1.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.773000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1237.181818, yerr=70.183938 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 1.103378E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 12049536 0.0% 0.0% 100.0% average stack size 0.0 0.0 564.9 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 631.644160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10439712 MPI messages size (bytes): total size 4.526311E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.566688E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3611304 96569262080 32768 < size <= 131072 1307392 74803314688 131072 < size <= 4194304 5190114 3200149557592 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56191. MP_Allreduce 11358 781. MP_Sync 172 MP_Alltoall 2242 2032940. MP_ISendRecv 49020 18752. MP_Wait 67312 MP_comm_split 84 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 205.243 205.245 qs_mol_dyn_low 1 2.0 0.004 0.007 204.829 204.842 qs_forces 11 3.9 0.003 0.003 204.742 204.745 qs_energies 11 4.9 0.004 0.021 199.212 199.224 scf_env_do_scf 11 5.9 0.001 0.001 182.774 182.778 scf_env_do_scf_inner_loop 118 6.6 0.003 0.007 161.847 161.849 dbcsr_multiply_generic 2529 12.6 0.182 0.188 125.000 125.619 velocity_verlet 10 3.0 0.001 0.002 123.726 123.727 qs_scf_new_mos 118 7.6 0.001 0.001 122.455 122.601 qs_scf_loop_do_ot 118 8.6 0.001 0.001 122.454 122.600 ot_scf_mini 118 9.6 0.003 0.003 115.773 115.931 multiply_cannon 2529 13.6 0.243 0.252 101.166 102.840 multiply_cannon_loop 2529 14.6 2.126 2.185 98.926 100.048 ot_mini 118 10.6 0.001 0.001 66.341 66.499 multiply_cannon_multrec 60696 15.6 33.302 34.596 41.701 43.256 qs_ot_get_derivative 118 11.6 0.001 0.001 41.329 41.499 rebuild_ks_matrix 129 8.3 0.001 0.001 33.639 33.801 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 33.638 33.800 mp_waitall_1 293956 16.2 29.010 32.463 29.010 32.463 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.273 30.440 multiply_cannon_sync_h2d 60696 15.6 27.414 28.886 27.414 28.886 qs_ot_get_p 129 10.4 0.001 0.001 27.470 27.686 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.462 25.403 apply_single 129 13.6 0.001 0.001 24.462 25.403 ot_diis_step 118 11.6 0.008 0.008 24.646 24.647 init_scf_loop 11 6.9 0.000 0.000 20.848 20.850 qs_ot_p2m_diag 84 11.4 0.078 0.093 20.700 20.803 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 19.324 19.444 cp_dbcsr_syevd 84 12.4 0.004 0.005 18.053 18.055 multiply_cannon_metrocomm3 60696 15.6 0.116 0.122 15.623 17.427 prepare_preconditioner 11 7.9 0.000 0.000 16.255 16.286 make_preconditioner 11 8.9 0.000 0.000 16.255 16.286 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.447 15.619 cp_fm_diag_elpa 84 13.4 0.000 0.001 15.025 15.032 cp_fm_redistribute_end 84 14.4 11.841 14.949 11.853 14.952 cp_fm_diag_elpa_base 84 14.4 3.056 14.708 3.088 14.832 make_m2s 5058 13.6 0.104 0.111 14.149 14.692 make_images 5058 14.6 0.405 0.423 13.967 14.522 sum_up_and_integrate 129 10.3 0.090 0.108 14.266 14.279 integrate_v_rspace 129 11.3 0.003 0.004 14.176 14.194 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.970 13.079 calculate_rho_elec 129 8.7 0.045 0.064 12.969 13.078 init_scf_run 11 5.9 0.000 0.001 12.256 12.257 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.256 12.256 mp_sum_l 7936 13.1 8.412 9.756 8.412 9.756 cp_fm_cholesky_invert 11 10.9 9.269 9.277 9.269 9.277 wfi_extrapolate 11 7.9 0.001 0.001 9.035 9.035 calculate_dm_sparse 129 9.5 0.001 0.001 8.485 8.568 dbcsr_mm_accdrv_process 125516 16.2 3.347 3.463 7.962 8.409 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.892 7.987 make_images_data 5058 15.6 0.067 0.073 6.970 7.967 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.867 7.946 multiply_cannon_metrocomm1 60696 15.6 0.086 0.091 6.077 7.655 grid_integrate_task_list 129 12.3 7.148 7.532 7.148 7.532 hybrid_alltoall_any 5245 16.5 0.294 2.269 6.073 7.531 pw_transfer 1559 11.6 0.075 0.106 7.094 7.369 density_rs2pw 129 9.7 0.006 0.007 6.676 7.207 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 6.889 7.138 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.006 6.666 6.675 rs_pw_transfer 1054 12.0 0.016 0.019 5.863 6.430 fft_wrap_pw1pw2_140 527 13.2 0.448 0.512 5.845 6.016 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.768 5.865 fft3d_ps 1301 14.7 2.109 2.579 5.671 5.858 mp_alltoall_d11v 2429 14.1 4.404 5.558 4.404 5.558 grid_collocate_task_list 129 9.7 4.758 5.172 4.758 5.172 cp_fm_cholesky_decompose 22 10.9 4.745 4.759 4.745 4.759 potential_pw2rs 129 12.3 0.009 0.011 4.653 4.677 mp_sum_d 4499 12.2 3.573 4.248 3.573 4.248 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.245000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.818182, yerr=6.952020 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 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 827.346944E+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 1463834560144 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 5413965. 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 0.014 0.033 187.580 187.580 qs_mol_dyn_low 1 2.0 0.003 0.004 187.260 187.274 qs_forces 11 3.9 0.003 0.003 187.180 187.182 qs_energies 11 4.9 0.001 0.002 180.498 180.508 scf_env_do_scf 11 5.9 0.001 0.001 164.413 164.423 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.835 131.838 velocity_verlet 10 3.0 0.001 0.002 118.140 118.142 dbcsr_multiply_generic 2507 12.6 0.188 0.195 97.028 98.140 qs_scf_new_mos 117 7.6 0.001 0.001 93.558 94.167 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.557 94.166 ot_scf_mini 117 9.6 0.004 0.004 88.777 89.457 multiply_cannon 2507 13.6 0.481 0.538 77.377 81.937 multiply_cannon_loop 2507 14.6 1.241 1.276 74.035 76.182 ot_mini 117 10.6 0.001 0.001 49.421 49.999 mp_waitall_1 226760 16.4 24.468 38.601 24.468 38.601 multiply_cannon_multrec 30084 15.6 22.029 26.618 31.685 36.483 rebuild_ks_matrix 128 8.3 0.001 0.001 32.554 33.068 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.554 33.067 init_scf_loop 11 6.9 0.000 0.000 32.489 32.490 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.339 29.815 multiply_cannon_metrocomm3 30084 15.6 0.092 0.097 15.601 28.264 qs_ot_get_derivative 117 11.6 0.001 0.002 27.597 28.246 prepare_preconditioner 11 7.9 0.000 0.000 28.098 28.178 make_preconditioner 11 8.9 0.000 0.000 28.098 28.178 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.777 27.310 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.985 23.093 apply_single 128 13.6 0.001 0.001 21.984 23.092 qs_ot_get_p 128 10.4 0.001 0.001 21.169 21.889 ot_diis_step 117 11.6 0.014 0.015 21.649 21.651 multiply_cannon_sync_h2d 30084 15.6 19.316 21.536 19.316 21.536 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.430 16.463 cp_fm_cholesky_invert 11 10.9 16.414 16.426 16.414 16.426 make_m2s 5014 13.6 0.090 0.096 14.013 15.699 make_images 5014 14.6 1.154 1.340 13.803 15.493 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.307 15.308 sum_up_and_integrate 128 10.3 0.117 0.132 14.181 14.207 integrate_v_rspace 128 11.3 0.003 0.003 14.064 14.095 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.761 12.790 calculate_rho_elec 128 8.7 0.087 0.105 12.760 12.789 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.278 12.285 cp_fm_redistribute_end 83 14.4 7.141 12.223 7.152 12.224 cp_fm_diag_elpa_base 83 14.4 4.842 11.775 5.056 12.143 init_scf_run 11 5.9 0.000 0.001 11.365 11.367 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.365 11.367 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.781 11.276 multiply_cannon_metrocomm4 27577 15.6 0.097 0.111 3.751 10.758 make_images_data 5014 15.6 0.065 0.071 8.532 10.535 mp_irecv_dv 69486 16.3 3.556 10.369 3.556 10.369 hybrid_alltoall_any 5200 16.5 0.342 1.517 7.235 9.725 dbcsr_mm_accdrv_process 62242 16.2 4.471 5.058 9.113 9.681 wfi_extrapolate 11 7.9 0.001 0.001 8.292 8.292 pw_transfer 1547 11.6 0.085 0.104 7.650 7.715 grid_integrate_task_list 128 12.3 7.180 7.494 7.180 7.494 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.425 7.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.201 6.996 density_rs2pw 128 9.7 0.006 0.006 6.560 6.899 cp_fm_cholesky_decompose 22 10.9 6.798 6.869 6.798 6.869 fft_wrap_pw1pw2_140 523 13.2 0.468 0.516 6.536 6.603 calculate_dm_sparse 128 9.5 0.001 0.001 6.455 6.584 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.126 6.135 fft3d_ps 1291 14.7 2.774 2.921 5.797 5.834 mp_sum_l 7870 13.0 3.850 5.801 3.850 5.801 rs_pw_transfer 1046 11.9 0.014 0.017 5.336 5.704 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.266 5.425 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.290 5.363 grid_collocate_task_list 128 9.7 4.909 5.188 4.909 5.188 mp_allgather_i34 2507 14.6 1.905 5.010 1.905 5.010 potential_pw2rs 128 12.3 0.014 0.016 4.633 4.649 mp_alltoall_d11v 2415 14.1 4.080 4.580 4.080 4.580 dbcsr_complete_redistribute 395 12.7 0.780 0.885 3.139 3.980 mp_sum_d 4455 12.2 2.563 3.880 2.563 3.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.580000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.272727, yerr=1.600620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 938.078208E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57167. MP_Allreduce 11256 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 172.826 172.829 qs_mol_dyn_low 1 2.0 0.003 0.003 172.388 172.401 qs_forces 11 3.9 0.003 0.003 172.290 172.294 qs_energies 11 4.9 0.019 0.029 165.829 165.841 scf_env_do_scf 11 5.9 0.001 0.002 150.557 150.559 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.817 115.818 velocity_verlet 10 3.0 0.002 0.002 110.496 110.499 dbcsr_multiply_generic 2485 12.5 0.182 0.187 80.698 81.830 qs_scf_new_mos 116 7.6 0.001 0.001 79.936 80.288 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.935 80.288 ot_scf_mini 116 9.6 0.003 0.004 75.836 76.241 multiply_cannon 2485 13.5 0.496 0.532 61.130 66.010 multiply_cannon_loop 2485 14.5 0.843 0.869 58.114 60.789 ot_mini 116 10.6 0.001 0.001 42.143 42.545 mp_waitall_1 176908 16.5 24.816 34.838 24.816 34.838 init_scf_loop 11 6.9 0.001 0.004 34.644 34.646 prepare_preconditioner 11 7.9 0.000 0.000 30.689 30.745 make_preconditioner 11 8.9 0.000 0.001 30.688 30.745 rebuild_ks_matrix 127 8.3 0.001 0.001 29.647 30.206 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.646 30.206 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.314 29.739 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.735 27.253 multiply_cannon_multrec 19880 15.5 13.499 16.738 22.086 25.331 multiply_cannon_metrocomm3 19880 15.5 0.060 0.062 14.936 24.900 qs_ot_get_derivative 116 11.6 0.001 0.002 22.581 22.989 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.667 20.719 apply_single 127 13.6 0.001 0.001 19.667 20.719 qs_ot_get_p 127 10.4 0.001 0.001 19.282 19.816 ot_diis_step 116 11.6 0.018 0.018 19.460 19.460 multiply_cannon_sync_h2d 19880 15.5 14.233 15.719 14.233 15.719 make_m2s 4970 13.5 0.080 0.086 14.539 15.467 make_images 4970 14.5 1.178 1.261 14.306 15.234 qs_ot_p2m_diag 82 11.4 0.262 0.269 15.047 15.054 cp_fm_cholesky_invert 11 10.9 14.188 14.197 14.188 14.197 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.074 14.075 sum_up_and_integrate 127 10.3 0.133 0.145 13.897 13.921 integrate_v_rspace 127 11.3 0.003 0.004 13.764 13.791 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.735 12.764 calculate_rho_elec 127 8.7 0.131 0.145 12.734 12.763 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.075 11.075 cp_fm_redistribute_end 82 14.4 4.157 11.025 4.170 11.027 cp_fm_diag_elpa_base 82 14.4 6.435 10.487 6.838 10.952 make_images_data 4970 15.5 0.059 0.065 9.016 10.380 init_scf_run 11 5.9 0.000 0.001 10.274 10.274 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.273 10.274 hybrid_alltoall_any 5155 16.4 0.426 1.950 7.812 9.889 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.807 9.102 multiply_cannon_metrocomm4 17395 15.5 0.059 0.068 3.363 8.888 mp_irecv_dv 49801 16.2 3.246 8.653 3.246 8.653 dbcsr_mm_accdrv_process 41158 16.2 4.314 5.260 8.041 8.231 grid_integrate_task_list 127 12.3 7.247 7.815 7.247 7.815 pw_transfer 1535 11.6 0.084 0.105 7.591 7.690 cp_fm_upper_to_full 104 14.5 5.803 7.515 5.803 7.515 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.370 7.474 cp_fm_cholesky_decompose 22 10.9 7.261 7.299 7.261 7.299 wfi_extrapolate 11 7.9 0.001 0.001 7.282 7.282 density_rs2pw 127 9.7 0.005 0.006 6.260 6.860 fft_wrap_pw1pw2_140 519 13.2 0.476 0.525 6.525 6.631 dbcsr_complete_redistribute 393 12.7 1.184 1.220 4.546 6.300 calculate_dm_sparse 127 9.5 0.001 0.001 5.688 5.802 fft3d_ps 1281 14.7 2.673 2.868 5.648 5.710 grid_collocate_task_list 127 9.7 5.034 5.569 5.034 5.569 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.385 5.398 rs_pw_transfer 1038 11.9 0.013 0.014 4.771 5.377 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.544 5.235 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.406 5.155 mp_alltoall_d11v 2401 14.1 4.234 5.147 4.234 5.147 mp_sum_l 7804 13.0 3.227 4.926 3.227 4.926 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.493 4.635 mp_allgather_i34 2485 14.5 1.545 4.539 1.545 4.539 potential_pw2rs 127 12.3 0.020 0.022 4.320 4.331 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.355 4.065 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.939 3.981 mp_alltoall_i22 712 14.1 1.937 3.837 1.937 3.837 qs_energies_init_hamiltonians 11 5.9 0.006 0.015 3.706 3.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.417 3.458 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=172.829000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.272727, yerr=9.185868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.150603E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57136. MP_Allreduce 11261 1067. MP_Sync 168 MP_Alltoall 1700 12496371. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.033 184.167 184.168 qs_mol_dyn_low 1 2.0 0.003 0.004 183.802 183.814 qs_forces 11 3.9 0.003 0.006 183.701 183.709 qs_energies 11 4.9 0.003 0.008 176.732 176.740 scf_env_do_scf 11 5.9 0.001 0.002 160.382 160.393 velocity_verlet 10 3.0 0.002 0.002 121.962 121.965 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.098 114.099 qs_scf_new_mos 116 7.6 0.001 0.001 79.236 79.483 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.236 79.482 dbcsr_multiply_generic 2485 12.5 0.191 0.196 78.243 78.856 ot_scf_mini 116 9.6 0.003 0.004 74.767 75.026 multiply_cannon 2485 13.5 0.553 0.599 54.161 56.623 multiply_cannon_loop 2485 14.5 1.170 1.205 50.492 51.730 init_scf_loop 11 6.9 0.001 0.004 46.160 46.161 prepare_preconditioner 11 7.9 0.000 0.000 42.145 42.172 make_preconditioner 11 8.9 0.000 0.003 42.145 42.172 ot_mini 116 10.6 0.001 0.001 41.692 41.933 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.798 40.724 multiply_cannon_multrec 29820 15.5 14.164 19.048 25.988 30.308 rebuild_ks_matrix 127 8.3 0.001 0.001 28.487 28.660 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 28.486 28.660 mp_waitall_1 152434 16.5 16.919 26.759 16.919 26.759 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.742 25.905 qs_ot_get_derivative 116 11.6 0.001 0.002 22.339 22.594 make_m2s 4970 13.5 0.094 0.098 19.812 20.892 make_images 4970 14.5 1.937 2.117 19.508 20.589 qs_ot_get_p 127 10.4 0.001 0.001 19.537 19.799 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.825 19.329 apply_single 127 13.6 0.001 0.001 18.824 19.328 ot_diis_step 116 11.6 0.017 0.018 19.227 19.229 cp_fm_cholesky_invert 11 10.9 16.261 16.270 16.261 16.270 cp_fm_upper_to_full 104 14.7 10.900 16.136 10.900 16.136 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.502 15.553 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.172 14.691 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.242 14.243 sum_up_and_integrate 127 10.3 0.140 0.151 13.807 13.836 integrate_v_rspace 127 11.3 0.003 0.004 13.667 13.702 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.895 12.936 calculate_rho_elec 127 8.7 0.173 0.189 12.895 12.935 dbcsr_complete_redistribute 393 12.7 1.505 1.618 8.916 12.654 multiply_cannon_sync_h2d 29820 15.5 11.657 12.307 11.657 12.307 make_images_data 4970 15.5 0.063 0.066 10.557 12.000 dbcsr_mm_accdrv_process 61748 16.2 7.260 8.221 11.401 11.912 hybrid_alltoall_any 5155 16.4 0.515 2.160 9.649 11.366 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.564 11.279 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.210 11.213 cp_fm_redistribute_end 82 14.4 1.912 11.137 1.925 11.142 cp_fm_diag_elpa_base 82 14.4 8.618 10.552 9.185 11.031 init_scf_run 11 5.9 0.000 0.001 10.467 10.468 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.467 10.468 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.323 9.972 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.198 9.381 mp_alltoall_i22 712 14.1 5.463 9.166 5.463 9.166 grid_integrate_task_list 127 12.3 7.473 7.830 7.473 7.830 pw_transfer 1535 11.6 0.084 0.100 7.640 7.718 cp_fm_cholesky_decompose 22 10.9 7.525 7.623 7.525 7.623 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.417 7.502 wfi_extrapolate 11 7.9 0.001 0.001 7.411 7.411 multiply_cannon_metrocomm4 24850 15.5 0.074 0.084 2.692 6.926 mp_irecv_dv 75445 16.2 2.550 6.667 2.550 6.667 fft_wrap_pw1pw2_140 519 13.2 0.478 0.490 6.546 6.648 density_rs2pw 127 9.7 0.005 0.006 6.084 6.305 calculate_dm_sparse 127 9.5 0.001 0.001 6.127 6.224 fft3d_ps 1281 14.7 2.777 2.881 5.663 5.726 grid_collocate_task_list 127 9.7 5.161 5.633 5.161 5.633 mp_alltoall_d11v 2401 14.1 4.775 5.504 4.775 5.504 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.354 5.390 rs_pw_transfer 1038 11.9 0.013 0.014 4.468 4.728 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.342 4.416 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.411 4.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.332 4.398 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.132 4.185 potential_pw2rs 127 12.3 0.022 0.023 4.139 4.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=184.168000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.000000, yerr=21.349899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.489658E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828158. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 166.685 166.686 qs_mol_dyn_low 1 2.0 0.003 0.003 166.316 166.328 qs_forces 11 3.9 0.003 0.003 166.207 166.209 qs_energies 11 4.9 0.002 0.002 158.951 158.959 scf_env_do_scf 11 5.9 0.001 0.001 142.121 142.139 velocity_verlet 10 3.0 0.002 0.002 109.117 109.121 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.164 107.165 dbcsr_multiply_generic 2485 12.5 0.183 0.188 73.075 73.391 qs_scf_new_mos 116 7.6 0.001 0.001 72.229 72.321 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.228 72.321 ot_scf_mini 116 9.6 0.003 0.004 67.871 67.935 multiply_cannon 2485 13.5 0.578 0.600 53.841 58.049 multiply_cannon_loop 2485 14.5 0.441 0.451 49.344 50.089 ot_mini 116 10.6 0.001 0.001 38.874 38.944 init_scf_loop 11 6.9 0.000 0.000 34.810 34.811 mp_waitall_1 128490 16.6 25.653 33.001 25.653 33.001 prepare_preconditioner 11 7.9 0.000 0.000 30.996 31.029 make_preconditioner 11 8.9 0.000 0.000 30.996 31.029 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.868 29.176 rebuild_ks_matrix 127 8.3 0.001 0.001 28.112 28.198 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.111 28.197 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.663 25.743 multiply_cannon_multrec 9940 15.5 10.292 13.975 17.520 20.476 ot_diis_step 116 11.6 0.019 0.022 19.546 19.546 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.190 19.413 apply_single 127 13.6 0.001 0.001 19.190 19.413 qs_ot_get_derivative 116 11.6 0.002 0.002 19.258 19.327 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.248 19.101 make_m2s 4970 13.5 0.065 0.069 15.764 17.720 cp_fm_cholesky_invert 11 10.9 17.600 17.607 17.600 17.607 make_images 4970 14.5 2.269 2.654 15.460 17.413 qs_ot_get_p 127 10.4 0.001 0.001 16.255 16.331 sum_up_and_integrate 127 10.3 0.179 0.190 13.796 13.845 integrate_v_rspace 127 11.3 0.004 0.004 13.617 13.674 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.868 12.905 calculate_rho_elec 127 8.7 0.256 0.266 12.867 12.904 qs_ot_p2m_diag 82 11.4 0.489 0.494 12.654 12.671 multiply_cannon_sync_h2d 9940 15.5 11.585 12.001 11.585 12.001 make_images_data 4970 15.5 0.052 0.059 9.634 11.969 hybrid_alltoall_any 5155 16.4 0.833 3.766 9.495 11.914 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.555 11.558 init_scf_run 11 5.9 0.000 0.001 10.075 10.076 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.074 10.075 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.657 8.658 cp_fm_diag_elpa_base 82 14.4 8.425 8.492 8.654 8.654 grid_integrate_task_list 127 12.3 7.745 7.985 7.745 7.985 cp_fm_cholesky_decompose 22 10.9 7.742 7.829 7.742 7.829 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.705 7.755 pw_transfer 1535 11.6 0.083 0.090 7.580 7.591 dbcsr_mm_accdrv_process 20590 16.1 2.654 3.431 6.861 7.470 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.361 7.377 wfi_extrapolate 11 7.9 0.001 0.001 7.170 7.170 mp_allgather_i34 2485 14.5 2.709 7.163 2.709 7.163 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.407 7.095 fft_wrap_pw1pw2_140 519 13.2 0.500 0.519 6.514 6.532 density_rs2pw 127 9.7 0.005 0.006 5.783 6.150 calculate_dm_sparse 127 9.5 0.001 0.001 6.011 6.100 grid_collocate_task_list 127 9.7 5.436 5.741 5.436 5.741 fft3d_ps 1281 14.7 2.698 2.771 5.535 5.566 dbcsr_complete_redistribute 393 12.7 2.114 2.192 5.162 5.552 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.222 5.224 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.094 5.104 mp_alltoall_d11v 2401 14.1 4.536 4.909 4.536 4.909 rs_pw_transfer 1038 11.9 0.012 0.013 3.928 4.318 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.029 4.050 potential_pw2rs 127 12.3 0.026 0.027 3.918 3.930 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.868 multiply_cannon_metrocomm4 7455 15.5 0.022 0.025 1.830 3.867 mp_irecv_dv 28618 15.9 1.794 3.799 1.794 3.799 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.369 3.699 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.641 3.665 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.494 3.569 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.383 3.403 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.348 3.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=166.686000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1402.545455, yerr=43.139022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 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 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.928513E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.041 291.738 291.738 qs_mol_dyn_low 1 2.0 0.003 0.004 291.226 291.238 qs_forces 11 3.9 0.003 0.003 291.131 291.135 qs_energies 11 4.9 0.002 0.002 282.589 282.599 scf_env_do_scf 11 5.9 0.001 0.001 260.886 260.898 velocity_verlet 10 3.0 0.002 0.002 211.425 211.432 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.528 135.530 init_scf_loop 11 6.9 0.000 0.000 125.105 125.106 prepare_preconditioner 11 7.9 0.000 0.000 120.381 120.418 make_preconditioner 11 8.9 0.000 0.000 120.381 120.418 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.018 117.526 qs_scf_new_mos 118 7.6 0.001 0.001 93.118 93.202 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.117 93.201 ot_scf_mini 118 9.6 0.004 0.004 88.281 88.337 dbcsr_multiply_generic 2535 12.6 0.219 0.227 84.937 85.650 cp_fm_upper_to_full 106 14.8 53.283 76.635 53.283 76.635 multiply_cannon 2535 13.6 0.710 0.745 59.459 60.442 multiply_cannon_loop 2535 14.6 0.476 0.494 55.509 57.230 ot_mini 118 10.6 0.001 0.001 45.502 45.552 dbcsr_complete_redistribute 397 12.7 3.982 4.037 30.016 43.238 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.627 39.865 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.317 37.489 mp_alltoall_i22 720 14.1 22.136 35.209 22.136 35.209 rebuild_ks_matrix 129 8.3 0.001 0.001 33.843 33.940 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 33.843 33.939 cp_fm_cholesky_invert 11 10.9 33.600 33.606 33.600 33.606 mp_waitall_1 106626 16.7 29.345 33.278 29.345 33.278 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.641 31.731 qs_ot_get_p 129 10.4 0.001 0.001 26.863 26.925 qs_ot_get_derivative 118 11.6 0.002 0.002 24.562 24.620 qs_ot_p2m_diag 84 11.4 0.890 0.895 22.667 22.696 make_m2s 5070 13.6 0.077 0.080 20.646 21.872 make_images 5070 14.6 3.809 3.909 20.164 21.391 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.905 20.907 ot_diis_step 118 11.6 0.022 0.022 20.901 20.901 multiply_cannon_metrocomm3 10140 15.6 0.024 0.024 19.166 20.664 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.051 20.318 apply_single 129 13.6 0.001 0.001 20.051 20.317 multiply_cannon_multrec 10140 15.6 10.570 12.310 18.000 18.067 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.655 17.656 cp_fm_diag_elpa_base 84 14.4 13.222 14.823 17.651 17.651 sum_up_and_integrate 129 10.3 0.324 0.326 15.793 15.891 multiply_cannon_sync_h2d 10140 15.6 15.778 15.791 15.778 15.791 integrate_v_rspace 129 11.3 0.004 0.004 15.469 15.565 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.062 15.068 calculate_rho_elec 129 8.7 0.487 0.487 15.061 15.067 hybrid_alltoall_any 5257 16.5 1.308 3.053 11.326 13.380 make_images_data 5070 15.6 0.060 0.063 11.269 13.350 init_scf_run 11 5.9 0.000 0.001 11.870 11.870 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.869 11.870 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.663 9.709 dbcsr_mm_accdrv_process 20958 16.1 3.919 5.762 7.193 9.022 cp_fm_cholesky_decompose 22 10.9 8.931 8.957 8.931 8.957 wfi_extrapolate 11 7.9 0.001 0.001 8.845 8.845 grid_integrate_task_list 129 12.3 8.601 8.782 8.601 8.782 pw_transfer 1559 11.6 0.091 0.092 8.415 8.427 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.179 8.192 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.873 7.873 fft_wrap_pw1pw2_140 527 13.2 0.544 0.546 7.234 7.250 mp_alltoall_d11v 2429 14.1 6.950 7.084 6.950 7.084 calculate_dm_sparse 129 9.5 0.001 0.001 6.682 6.731 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.371 6.432 grid_collocate_task_list 129 9.7 6.388 6.420 6.388 6.420 fft3d_ps 1301 14.7 2.782 2.793 6.206 6.218 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.071 6.117 density_rs2pw 129 9.7 0.005 0.005 6.063 6.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.738000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2584.727273, yerr=182.802472 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 1.261728E+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 14436293. MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.040 85.079 85.080 qs_energies 1 2.0 0.000 0.001 84.608 84.622 ls_scf 1 3.0 0.000 0.000 83.712 83.726 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.897 73.125 multiply_cannon 111 7.7 0.018 0.021 56.039 57.050 multiply_cannon_loop 111 8.7 0.207 0.220 52.569 53.841 ls_scf_main 1 4.0 0.000 0.000 52.401 52.401 density_matrix_trs4 2 5.0 0.002 0.003 46.933 47.021 ls_scf_init_scf 1 4.0 0.000 0.001 28.285 28.286 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.215 27.263 mp_waitall_1 11316 10.9 22.454 25.937 22.454 25.937 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.106 25.131 multiply_cannon_multrec 2664 9.7 8.218 9.087 15.517 17.378 multiply_cannon_sync_h2d 2664 9.7 13.730 16.077 13.730 16.077 make_m2s 222 7.7 0.008 0.010 13.036 13.514 make_images 222 8.7 0.099 0.110 13.014 13.493 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.745 12.854 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.415 8.755 make_images_data 222 9.7 0.004 0.005 7.609 8.258 hybrid_alltoall_any 227 10.6 0.215 1.837 6.454 8.110 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.616 6.918 7.877 dbcsr_mm_accdrv_process_sort 4760 11.4 6.211 7.105 6.211 7.105 calculate_norms 4752 9.8 5.523 6.166 5.523 6.166 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.070 5.192 mp_sum_l 807 5.4 3.295 4.772 3.295 4.772 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.461 3.655 make_images_sizes 222 9.7 0.000 0.000 0.822 3.611 mp_alltoall_i44 222 10.7 0.822 3.611 0.822 3.611 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.046 3.561 mp_irecv_dv 6231 10.9 2.029 3.532 2.029 3.532 arnoldi_extremal 4 6.8 0.000 0.000 3.253 3.285 arnoldi_normal_ev 4 7.8 0.001 0.003 3.253 3.285 build_subspace 16 8.4 0.009 0.012 3.147 3.150 ls_scf_post 1 4.0 0.000 0.000 3.027 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.889 dbcsr_special_finalize 555 9.7 0.005 0.005 2.250 2.674 dbcsr_merge_single_wm 555 10.7 0.463 0.617 2.243 2.666 make_images_pack 222 9.7 2.209 2.632 2.211 2.633 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.340 2.590 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.466 2.072 2.468 dbcsr_sort_data 658 11.4 2.036 2.444 2.036 2.444 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.198 2.285 buffer_matrices_ensure_size 222 8.7 1.775 2.168 1.775 2.168 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.734 1.736 rebuild_ks_matrix 3 7.3 0.000 0.000 1.725 1.726 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.725 1.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.080000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1130.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 2.104414E+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 0.026 0.042 90.358 90.360 qs_energies 1 2.0 0.000 0.000 89.896 89.901 ls_scf 1 3.0 0.000 0.000 88.594 88.598 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.801 75.152 multiply_cannon 111 7.7 0.029 0.037 53.212 56.729 ls_scf_main 1 4.0 0.000 0.000 54.421 54.429 multiply_cannon_loop 111 8.7 0.116 0.123 49.917 52.925 density_matrix_trs4 2 5.0 0.002 0.003 48.810 49.013 ls_scf_init_scf 1 4.0 0.000 0.000 30.610 30.611 mp_waitall_1 9246 10.9 20.704 30.530 20.704 30.530 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.492 29.584 multiply_cannon_multrec 1332 9.7 13.150 16.944 22.357 27.266 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.067 27.079 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.659 21.546 make_m2s 222 7.7 0.006 0.008 14.978 15.616 make_images 222 8.7 1.577 1.935 14.948 15.587 dbcsr_mm_accdrv_process 4041 10.4 0.279 0.445 8.805 10.357 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.912 8.399 9.912 make_images_data 222 9.7 0.004 0.004 8.426 9.370 mp_sum_l 807 5.4 5.319 8.816 5.319 8.816 hybrid_alltoall_any 227 10.6 0.520 2.440 7.867 8.639 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.215 7.798 mp_irecv_dv 3311 11.0 3.196 7.745 3.196 7.745 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.085 7.015 calculate_norms 2376 9.8 6.019 6.783 6.019 6.783 multiply_cannon_sync_h2d 1332 9.7 4.853 6.057 4.853 6.057 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.040 5.270 arnoldi_extremal 4 6.8 0.000 0.000 4.641 4.659 arnoldi_normal_ev 4 7.8 0.001 0.004 4.641 4.659 build_subspace 16 8.4 0.014 0.021 4.388 4.391 ls_scf_post 1 4.0 0.000 0.000 3.563 3.568 ls_scf_store_result 1 5.0 0.000 0.000 3.260 3.388 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.147 3.375 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.224 2.744 3.226 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.245 2.881 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.512 2.604 make_images_pack 222 9.7 2.026 2.378 2.028 2.380 mp_allgather_i34 111 8.7 0.991 2.223 0.991 2.223 dbcsr_sort_data 436 11.2 1.818 2.068 1.818 2.068 dbcsr_data_new 4174 10.1 1.612 1.846 1.612 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.360000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1706.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 2.705175E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.060 94.517 94.519 qs_energies 1 2.0 0.000 0.001 93.925 93.928 ls_scf 1 3.0 0.000 0.001 92.525 92.529 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.457 77.724 ls_scf_main 1 4.0 0.000 0.000 58.342 58.347 multiply_cannon 111 7.7 0.044 0.093 53.189 57.764 multiply_cannon_loop 111 8.7 0.099 0.106 49.637 53.853 density_matrix_trs4 2 5.0 0.002 0.003 52.523 52.711 mp_waitall_1 7374 11.0 24.455 34.716 24.455 34.716 ls_scf_init_scf 1 4.0 0.000 0.002 30.595 30.597 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.477 29.533 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.101 27.114 multiply_cannon_multrec 888 9.7 12.579 15.045 21.079 24.111 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.599 23.927 make_m2s 222 7.7 0.006 0.007 17.195 18.488 make_images 222 8.7 1.979 2.291 17.157 18.449 hybrid_alltoall_any 227 10.6 0.620 2.863 9.561 10.967 make_images_data 222 9.7 0.003 0.004 9.917 10.884 mp_sum_l 807 5.4 5.521 9.749 5.521 9.749 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.411 8.036 9.225 dbcsr_mm_accdrv_process_sort 3754 11.4 7.667 8.814 7.667 8.814 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.346 7.797 multiply_cannon_sync_h2d 888 9.7 6.054 7.381 6.054 7.381 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.559 6.948 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.466 6.814 mp_irecv_dv 2335 11.1 2.451 6.774 2.451 6.774 arnoldi_extremal 4 6.8 0.000 0.000 5.045 5.063 arnoldi_normal_ev 4 7.8 0.001 0.005 5.045 5.063 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.847 5.058 build_subspace 16 8.4 0.014 0.020 4.739 4.745 calculate_norms 1584 9.8 4.264 4.586 4.264 4.586 mp_allgather_i34 111 8.7 1.388 3.925 1.388 3.925 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.418 3.752 dbcsr_matrix_vector_mult_local 304 10.0 3.023 3.603 3.025 3.606 ls_scf_post 1 4.0 0.001 0.002 3.588 3.592 ls_scf_store_result 1 5.0 0.000 0.000 3.297 3.414 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.732 2.840 dbcsr_sort_data 325 11.1 1.911 2.146 1.911 2.146 make_images_pack 222 9.7 1.816 2.132 1.819 2.135 make_images_sizes 222 9.7 0.000 0.000 0.922 2.009 mp_alltoall_i44 222 10.7 0.921 2.009 0.921 2.009 dbcsr_data_release 9322 10.9 1.300 1.971 1.300 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.519000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2213.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 3.314110E+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 0.032 0.050 96.420 96.421 qs_energies 1 2.0 0.000 0.000 95.823 95.831 ls_scf 1 3.0 0.000 0.000 94.174 94.180 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.010 78.230 ls_scf_main 1 4.0 0.000 0.000 58.392 58.393 multiply_cannon 111 7.7 0.053 0.098 51.409 55.743 density_matrix_trs4 2 5.0 0.002 0.003 52.356 52.458 multiply_cannon_loop 111 8.7 0.115 0.140 46.348 49.262 ls_scf_init_scf 1 4.0 0.000 0.000 32.525 32.527 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.383 31.443 mp_waitall_1 6438 11.0 22.388 29.528 22.388 29.528 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.891 28.903 multiply_cannon_multrec 1332 9.7 14.244 17.580 22.090 25.334 make_m2s 222 7.7 0.007 0.008 21.026 22.478 make_images 222 8.7 3.137 3.592 20.976 22.430 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.133 16.892 make_images_data 222 9.7 0.004 0.004 11.697 13.368 hybrid_alltoall_any 227 10.6 0.799 3.827 11.017 12.867 dbcsr_mm_accdrv_process 3641 10.4 0.228 0.405 7.482 9.020 dbcsr_mm_accdrv_process_sort 3641 11.4 7.106 8.595 7.106 8.595 mp_sum_l 807 5.4 4.001 7.220 4.001 7.220 multiply_cannon_sync_h2d 1332 9.7 5.555 6.186 5.555 6.186 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.054 6.036 mp_irecv_dv 3229 10.9 2.032 5.960 2.032 5.960 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.075 5.563 arnoldi_extremal 4 6.8 0.000 0.000 5.303 5.322 arnoldi_normal_ev 4 7.8 0.007 0.019 5.303 5.322 build_subspace 16 8.4 0.014 0.021 4.957 4.962 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.599 4.815 mp_allgather_i34 111 8.7 2.192 4.587 2.192 4.587 calculate_norms 2376 9.8 4.194 4.548 4.194 4.548 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.383 4.278 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.607 3.929 dbcsr_matrix_vector_mult_local 304 10.0 3.198 3.702 3.200 3.704 dbcsr_sort_data 658 11.4 3.060 3.335 3.060 3.335 ls_scf_post 1 4.0 0.000 0.000 3.257 3.264 dbcsr_special_finalize 555 9.7 0.006 0.007 2.805 3.121 dbcsr_merge_single_wm 555 10.7 0.535 0.668 2.797 3.112 ls_scf_store_result 1 5.0 0.000 0.000 3.004 3.059 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.901 2.968 dbcsr_data_release 10477 10.7 1.580 2.501 1.580 2.501 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.421000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 4.623520E+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 0.045 0.062 97.668 97.668 qs_energies 1 2.0 0.000 0.000 96.899 96.905 ls_scf 1 3.0 0.000 0.000 94.975 94.980 dbcsr_multiply_generic 111 6.7 0.017 0.018 76.838 77.081 ls_scf_main 1 4.0 0.000 0.000 61.144 61.145 multiply_cannon 111 7.7 0.088 0.232 54.728 60.247 density_matrix_trs4 2 5.0 0.002 0.002 54.206 54.287 multiply_cannon_loop 111 8.7 0.069 0.076 50.150 52.120 mp_waitall_1 5481 11.0 25.611 31.222 25.611 31.222 ls_scf_init_scf 1 4.0 0.000 0.000 30.249 30.252 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.080 29.115 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.926 26.934 multiply_cannon_multrec 444 9.7 13.997 16.336 21.012 23.195 make_m2s 222 7.7 0.004 0.005 17.389 19.966 make_images 222 8.7 3.730 4.416 17.327 19.905 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.666 16.002 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.110 14.956 make_images_data 222 9.7 0.003 0.004 9.634 12.158 hybrid_alltoall_any 227 10.6 0.790 3.758 9.358 11.894 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.338 6.717 7.835 multiply_cannon_sync_h2d 444 9.7 6.529 7.775 6.529 7.775 dbcsr_mm_accdrv_process_sort 3003 11.4 6.403 7.490 6.403 7.490 mp_allgather_i34 111 8.7 2.811 6.989 2.811 6.989 arnoldi_extremal 4 6.8 0.000 0.000 5.814 5.824 arnoldi_normal_ev 4 7.8 0.002 0.005 5.814 5.824 build_subspace 16 8.4 0.015 0.020 5.422 5.434 mp_sum_l 807 5.4 2.776 4.856 2.776 4.856 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.589 4.734 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.181 4.371 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.173 3.699 4.176 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.655 3.945 mp_irecv_dv 1241 11.2 1.636 3.921 1.636 3.921 calculate_norms 792 9.8 3.545 3.719 3.545 3.719 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.909 3.699 ls_scf_post 1 4.0 0.000 0.000 3.582 3.588 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.343 3.420 ls_scf_store_result 1 5.0 0.000 0.000 3.371 3.411 make_images_sizes 222 9.7 0.000 0.000 1.107 3.264 mp_alltoall_i44 222 10.7 1.106 3.264 1.106 3.264 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.261 dbcsr_merge_all 275 8.9 0.473 0.519 2.048 2.087 dbcsr_data_release 10123 10.8 1.321 1.961 1.321 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=97.668000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3650.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5ffd1932e3d34d3bbecc675223bf506dafc025e_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 8.740762E+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 0.076 0.097 106.867 106.867 qs_energies 1 2.0 0.000 0.000 105.463 105.478 ls_scf 1 3.0 0.000 0.000 102.550 102.564 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.786 76.884 ls_scf_main 1 4.0 0.000 0.000 65.024 65.024 density_matrix_trs4 2 5.0 0.002 0.003 56.201 56.259 multiply_cannon 111 7.7 0.154 0.192 49.346 51.310 multiply_cannon_loop 111 8.7 0.068 0.071 45.864 46.681 ls_scf_init_scf 1 4.0 0.000 0.000 33.863 33.864 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.541 32.554 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.716 29.727 mp_waitall_1 4569 11.1 21.643 25.231 21.643 25.231 make_m2s 222 7.7 0.005 0.005 23.907 24.911 make_images 222 8.7 4.594 4.976 23.801 24.804 multiply_cannon_multrec 444 9.7 17.885 18.683 22.491 23.104 hybrid_alltoall_any 227 10.6 1.664 3.652 12.855 15.892 make_images_data 222 9.7 0.003 0.003 13.133 15.577 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.928 10.408 multiply_cannon_sync_h2d 444 9.7 8.847 8.884 8.847 8.884 arnoldi_extremal 4 6.8 0.000 0.000 7.393 7.404 arnoldi_normal_ev 4 7.8 0.003 0.008 7.393 7.404 build_subspace 16 8.4 0.026 0.037 6.845 6.857 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.482 5.635 dbcsr_matrix_vector_mult_local 304 10.0 5.050 5.361 5.053 5.363 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.016 5.259 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.773 4.861 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.320 4.420 4.534 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.250 4.122 4.250 ls_scf_post 1 4.0 0.000 0.000 3.663 3.677 make_images_sizes 222 9.7 0.000 0.000 1.435 3.519 mp_alltoall_i44 222 10.7 1.435 3.519 1.435 3.519 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.426 mp_allgather_i34 111 8.7 1.108 3.368 1.108 3.368 calculate_norms 792 9.8 3.241 3.282 3.241 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.089 3.131 dbcsr_merge_all 275 8.9 0.889 0.909 2.876 2.912 dbcsr_complete_redistribute 5 7.6 1.432 1.456 2.773 2.890 qs_energies_init_hamiltonians 1 3.0 0.002 0.005 2.883 2.883 dbcsr_data_release 12724 10.6 2.330 2.852 2.330 2.852 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.538 dbcsr_sort_data 325 11.1 2.446 2.508 2.446 2.508 dbcsr_new_transposed 4 7.5 0.251 0.265 2.288 2.303 dbcsr_frobenius_norm 74 6.6 2.056 2.137 2.205 2.248 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.200 dbcsr_add_anytype 103 7.2 0.860 0.892 2.130 2.199 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.183 2.185 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.867000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6848.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a5ffd1932e3d34d3bbecc675223bf506dafc025e Summary: empty Status: OK