=== 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: e58a972ed010720609b145a32ea323a5cbbe9a85 ################# 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/01 job id: 42642646 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/02 job id: 42642647 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/03 job id: 42642648 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/04 job id: 42642649 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/05 job id: 42642650 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/06 job id: 42642651 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/07 job id: 42642653 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/08 job id: 42642654 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/09 job id: 42642655 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/10 job id: 42642656 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/11 job id: 42642657 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/12 job id: 42642658 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/13 job id: 42642659 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/14 job id: 42642660 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/15 job id: 42642661 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/16 job id: 42642662 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/17 job id: 42642664 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/18 job id: 42642666 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/19 job id: 42642667 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/20 job id: 42642668 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/21 job id: 42642669 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/22 job id: 42642670 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/23 job id: 42642672 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/24 job id: 42642675 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/25 job id: 42642676 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/26 job id: 42642677 --- 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/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.042 133.073 133.074 farming_run 1 2.0 132.368 132.386 133.036 133.040 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459700E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 115.555 115.555 qs_energies 1 2.0 0.000 0.000 115.267 115.269 mp2_main 1 3.0 0.000 0.000 113.139 113.141 mp2_gpw_main 1 4.0 0.019 0.026 112.318 112.320 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.678 93.986 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.548 55.858 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.174 41.873 47.121 get_2c_integrals 1 6.0 0.000 0.000 37.298 37.958 integrate_v_rspace 273 8.0 0.436 0.451 25.203 30.274 pw_transfer 6555 10.6 0.372 0.390 27.468 28.217 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.115 26.736 grid_integrate_task_list 273 9.0 21.002 26.561 21.002 26.561 fft_wrap_pw1pw2_100 2178 12.4 1.206 1.406 23.683 24.300 compute_2c_integrals 1 7.0 0.002 0.002 19.712 19.712 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.962 19.432 mp2_eri_2c_integrate_gpw 1 9.0 2.372 2.440 18.959 19.430 rpa_ri_compute_en 1 5.0 0.000 0.000 18.532 18.794 cp_fm_cholesky_decompose 12 8.2 17.542 18.184 17.542 18.184 cholesky_decomp 1 7.0 0.000 0.000 16.435 17.079 fft3d_s 5443 13.4 16.237 16.692 16.260 16.713 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.558 10.851 15.558 calculate_wavefunction 272 8.0 5.430 5.589 12.582 13.188 rpa_num_int 1 6.0 0.002 0.028 10.563 10.563 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.493 10.498 calc_mat_Q 8 8.0 0.000 0.000 9.335 9.409 contract_S_to_Q 8 9.0 0.000 0.000 8.757 8.830 calc_potential_gpw 544 9.5 0.006 0.006 8.312 8.632 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.269 8.591 potential_pw2rs 545 10.0 0.107 0.109 7.702 8.433 parallel_gemm_fm 14 9.1 0.000 0.000 8.344 8.409 parallel_gemm_fm_cosma 14 10.1 8.344 8.409 8.344 8.409 collocate_single_gaussian 272 10.0 0.041 0.043 7.494 7.749 create_integ_mat 1 6.0 0.022 0.026 7.567 7.567 array2fm 1 7.0 0.000 0.000 6.553 7.034 pw_scatter_s 2720 13.7 4.430 4.601 4.430 4.601 pw_gather_s 2722 13.2 3.874 4.272 3.874 4.272 array2fm_buffer_send 1 8.0 2.867 3.040 2.867 3.040 pw_poisson_solve 545 10.5 1.119 1.188 2.180 2.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.317955, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 0.031 0.045 397.688 397.693 farming_run 1 2.0 396.494 396.516 397.637 397.642 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223954E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.047 208.929 208.930 qs_energies 1 2.0 0.000 0.000 208.550 208.557 scf_env_do_scf 1 3.0 0.000 0.000 105.940 105.941 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.059 105.068 rebuild_ks_matrix 4 6.0 0.000 0.000 105.058 105.066 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.058 105.066 hfx_ks_matrix 4 8.0 0.001 0.001 104.688 104.691 integrate_four_center 4 9.0 0.143 0.471 104.687 104.691 mp2_main 1 3.0 0.000 0.001 102.327 102.334 mp2_gpw_main 1 4.0 0.039 0.410 101.458 101.468 integrate_four_center_main 4 10.0 0.087 0.611 96.688 99.005 integrate_four_center_bin 261 11.0 96.601 98.953 96.601 98.953 init_scf_loop 1 4.0 0.000 0.000 91.846 91.847 mp2_ri_gpw_compute_in 1 5.0 0.072 0.105 74.877 75.896 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.096 54.447 55.452 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.118 47.235 integrate_v_rspace 95 8.0 0.398 0.574 28.483 33.424 pw_transfer 2240 10.6 0.145 0.166 29.924 30.416 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.925 29.316 grid_integrate_task_list 95 9.0 23.781 28.926 23.781 28.926 ao_to_mo_and_store_B_mult_1 91 7.0 10.643 28.796 10.643 28.796 mp2_ri_gpw_compute_en 1 5.0 0.065 0.118 26.416 28.050 fft_wrap_pw1pw2_100 730 12.4 1.295 1.469 26.662 27.063 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.848 1.893 24.685 24.695 get_2c_integrals 1 6.0 0.002 0.009 20.333 20.357 compute_2c_integrals 1 7.0 0.006 0.020 19.320 19.321 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.918 19.182 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.894 18.916 19.181 fft3d_s 1823 13.4 18.412 18.963 18.425 18.977 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.092 14.092 calculate_wavefunction 91 8.0 2.013 2.043 9.742 9.999 potential_pw2rs 186 10.0 0.034 0.035 8.638 9.273 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.579 8.723 9.194 local_gemm 172 8.0 8.167 8.641 8.167 8.641 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.277 8.596 calc_potential_gpw 182 9.5 0.002 0.002 7.937 8.160 mp2_ri_gpw_compute_en_comm 22 7.0 0.510 0.534 7.723 8.145 collocate_single_gaussian 91 10.0 0.017 0.023 7.902 8.144 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.444 6.347 6.444 mp_sendrecv_dm3 2068 8.0 5.730 6.114 5.730 6.114 mp_sync 38 10.4 2.592 5.552 2.592 5.552 pw_gather_s 912 13.2 4.906 5.538 4.906 5.538 pw_scatter_s 910 13.7 3.943 4.219 3.943 4.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.459043, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 451.096576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.102 51.387 51.387 qs_mol_dyn_low 1 2.0 0.005 0.021 50.899 50.915 qs_forces 11 3.9 0.004 0.021 50.786 50.790 qs_energies 11 4.9 0.003 0.011 49.230 49.241 scf_env_do_scf 11 5.9 0.000 0.001 43.005 43.005 scf_env_do_scf_inner_loop 108 6.5 0.005 0.028 40.939 40.940 dbcsr_multiply_generic 2286 12.5 0.093 0.097 32.630 33.114 qs_scf_new_mos 108 7.5 0.000 0.001 31.036 31.332 qs_scf_loop_do_ot 108 8.5 0.001 0.001 31.036 31.332 ot_scf_mini 108 9.5 0.002 0.002 29.415 29.600 multiply_cannon 2286 13.5 0.181 0.187 25.679 27.085 multiply_cannon_loop 2286 14.5 1.515 1.592 25.016 26.451 velocity_verlet 10 3.0 0.002 0.011 25.398 25.400 ot_mini 108 10.5 0.001 0.001 18.650 18.882 qs_ot_get_derivative 108 11.5 0.001 0.001 15.668 15.860 mp_waitall_1 267858 16.1 8.478 14.223 8.478 14.223 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.921 12.470 multiply_cannon_multrec 54864 15.5 4.208 6.569 7.511 10.893 rebuild_ks_matrix 119 8.3 0.000 0.000 7.864 8.000 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 7.863 8.000 multiply_cannon_sync_h2d 54864 15.5 5.876 7.122 5.876 7.122 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.896 7.028 mp_sum_l 7207 12.9 5.105 6.601 5.105 6.601 qs_ot_get_p 119 10.4 0.001 0.002 6.179 6.483 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.377 5.845 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.133 5.229 init_scf_run 11 5.9 0.001 0.007 4.915 4.915 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.914 4.915 dbcsr_mm_accdrv_process 76910 16.1 1.183 1.870 3.225 4.591 sum_up_and_integrate 119 10.3 0.013 0.016 4.510 4.524 integrate_v_rspace 119 11.3 0.002 0.003 4.498 4.513 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.999 4.129 calculate_rho_elec 119 8.7 0.011 0.016 3.998 4.128 qs_ot_p2m_diag 50 11.0 0.004 0.007 3.400 3.435 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.884 3.098 apply_single 119 13.6 0.000 0.000 2.883 3.097 calculate_dm_sparse 119 9.5 0.000 0.000 2.860 2.986 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.745 2.843 rs_pw_transfer 974 11.9 0.012 0.013 2.743 2.832 ot_diis_step 108 11.5 0.006 0.008 2.696 2.697 calculate_first_density_matrix 1 7.0 0.000 0.002 2.687 2.695 jit_kernel_multiply 13 15.8 1.981 2.657 1.981 2.657 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.602 2.602 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.433 2.478 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.400 2.401 cp_fm_redistribute_end 50 14.0 2.180 2.380 2.185 2.382 cp_fm_diag_elpa_base 50 14.0 0.195 2.345 0.196 2.354 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.319 2.323 density_rs2pw 119 9.7 0.004 0.004 2.100 2.198 wfi_extrapolate 11 7.9 0.009 0.070 2.167 2.167 acc_transpose_blocks 54864 15.5 0.230 0.249 1.738 2.165 grid_integrate_task_list 119 12.3 2.025 2.120 2.025 2.120 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.029 2.094 init_scf_loop 11 6.9 0.000 0.001 2.049 2.052 mp_sum_d 4125 12.0 1.438 2.041 1.438 2.041 potential_pw2rs 119 12.3 0.004 0.004 1.875 1.891 pw_transfer 1439 11.6 0.051 0.056 1.615 1.691 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.541 1.618 make_m2s 4572 13.5 0.054 0.056 1.563 1.607 make_images 4572 14.5 0.133 0.138 1.480 1.524 mp_alltoall_d11v 2130 13.8 1.289 1.432 1.289 1.432 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.399 1.424 mp_waitany 12084 13.8 1.267 1.419 1.267 1.419 fft3d_ps 1201 14.6 0.352 0.453 1.317 1.393 grid_collocate_task_list 119 9.7 1.292 1.341 1.292 1.341 fft_wrap_pw1pw2_140 487 13.2 0.081 0.094 1.194 1.271 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.753 1.163 parallel_gemm_fm 81 9.0 0.000 0.000 1.054 1.059 parallel_gemm_fm_cosma 81 10.0 1.054 1.058 1.054 1.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.011 1.034 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.387000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 487.923712E+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 1162195. 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 0.040 0.058 40.489 40.490 qs_mol_dyn_low 1 2.0 0.003 0.003 39.942 39.949 qs_forces 11 3.9 0.013 0.018 39.379 39.379 qs_energies 11 4.9 0.004 0.020 37.691 37.706 scf_env_do_scf 11 5.9 0.001 0.001 31.305 31.306 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.793 28.795 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.992 22.341 qs_scf_new_mos 108 7.5 0.001 0.001 19.979 20.218 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.979 20.218 ot_scf_mini 108 9.5 0.003 0.003 19.082 19.260 multiply_cannon 2286 13.5 0.208 0.221 17.082 18.703 velocity_verlet 10 3.0 0.001 0.001 18.402 18.414 multiply_cannon_loop 2286 14.5 0.906 0.977 15.959 17.297 ot_mini 108 10.5 0.001 0.001 11.978 12.218 mp_waitall_1 217478 16.2 6.468 11.868 6.468 11.868 multiply_cannon_metrocomm3 27432 15.5 0.068 0.069 4.679 10.224 qs_ot_get_derivative 108 11.5 0.001 0.001 9.520 9.702 multiply_cannon_multrec 27432 15.5 1.969 4.537 6.041 9.102 rebuild_ks_matrix 119 8.3 0.000 0.000 7.066 7.199 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.066 7.199 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.237 6.356 dbcsr_mm_accdrv_process 47894 16.0 3.146 5.471 4.004 6.146 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.871 4.722 init_scf_run 11 5.9 0.000 0.001 4.510 4.511 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.510 4.511 qs_ot_get_p 119 10.4 0.001 0.001 4.265 4.498 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.045 4.120 apply_single 119 13.6 0.000 0.000 3.045 4.119 sum_up_and_integrate 119 10.3 0.024 0.027 4.089 4.096 integrate_v_rspace 119 11.3 0.002 0.002 4.065 4.074 mp_sum_l 7207 12.9 2.018 3.962 2.018 3.962 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.518 3.552 calculate_rho_elec 119 8.7 0.021 0.024 3.517 3.551 calculate_first_density_matrix 1 7.0 0.000 0.001 3.032 3.034 multiply_cannon_sync_h2d 27432 15.5 2.194 2.798 2.194 2.798 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.752 2.771 make_m2s 4572 13.5 0.053 0.055 2.445 2.659 rs_pw_transfer 974 11.9 0.010 0.011 2.556 2.652 make_images 4572 14.5 0.201 0.238 2.357 2.569 init_scf_loop 11 6.9 0.001 0.002 2.493 2.498 ot_diis_step 108 11.5 0.010 0.011 2.408 2.409 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.378 2.380 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.321 2.321 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.111 2.201 calculate_dm_sparse 119 9.5 0.000 0.000 2.088 2.164 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.111 2.151 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.991 1.992 cp_fm_redistribute_end 50 14.0 1.637 1.965 1.641 1.966 density_rs2pw 119 9.7 0.004 0.004 1.889 1.958 cp_fm_diag_elpa_base 50 14.0 0.316 1.910 0.324 1.943 grid_integrate_task_list 119 12.3 1.834 1.915 1.834 1.915 potential_pw2rs 119 12.3 0.006 0.006 1.836 1.845 pw_transfer 1439 11.6 0.064 0.068 1.742 1.773 jit_kernel_multiply 8 16.2 0.805 1.765 0.805 1.765 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.652 1.684 prepare_preconditioner 11 7.9 0.000 0.000 1.515 1.543 make_preconditioner 11 8.9 0.000 0.001 1.515 1.543 make_images_data 4572 15.5 0.045 0.051 1.131 1.532 acc_transpose_blocks 27432 15.5 0.109 0.113 1.202 1.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.413 1.468 wfi_extrapolate 11 7.9 0.001 0.001 1.427 1.427 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.976 1.416 fft3d_ps 1201 14.6 0.497 0.550 1.359 1.387 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.302 1.333 grid_collocate_task_list 119 9.7 1.239 1.320 1.239 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.298 1.305 mp_allgather_i34 2286 14.5 0.554 1.293 0.554 1.293 mp_alltoall_d11v 2130 13.8 1.175 1.270 1.175 1.270 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.183 1.230 mp_sum_d 4125 12.0 0.638 1.134 0.638 1.134 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.103 1.115 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.885 0.898 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.661 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.490000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 522.514432E+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 0.062 0.110 32.333 32.334 qs_mol_dyn_low 1 2.0 0.003 0.003 31.863 31.870 qs_forces 11 3.9 0.001 0.002 31.807 31.808 qs_energies 11 4.9 0.009 0.062 30.248 30.250 scf_env_do_scf 11 5.9 0.001 0.001 25.076 25.076 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.512 22.512 dbcsr_multiply_generic 2286 12.5 0.093 0.097 16.267 16.353 velocity_verlet 10 3.0 0.001 0.002 14.975 14.976 qs_scf_new_mos 108 7.5 0.001 0.001 14.679 14.702 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.679 14.701 ot_scf_mini 108 9.5 0.002 0.003 13.965 13.984 multiply_cannon 2286 13.5 0.197 0.205 13.015 13.848 multiply_cannon_loop 2286 14.5 0.639 0.666 12.205 13.067 ot_mini 108 10.5 0.001 0.001 8.661 8.675 qs_ot_get_derivative 108 11.5 0.001 0.001 7.176 7.193 multiply_cannon_multrec 18288 15.5 1.942 2.835 6.783 7.014 rebuild_ks_matrix 119 8.3 0.000 0.000 6.270 6.284 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.270 6.283 dbcsr_mm_accdrv_process 38222 16.0 4.022 5.495 4.757 5.586 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.545 5.558 sum_up_and_integrate 119 10.3 0.030 0.031 3.877 3.882 integrate_v_rspace 119 11.3 0.002 0.003 3.847 3.855 init_scf_run 11 5.9 0.000 0.001 3.846 3.848 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.846 3.848 mp_waitall_1 169478 16.3 2.788 3.636 2.788 3.636 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.510 3.514 calculate_rho_elec 119 8.7 0.030 0.031 3.509 3.514 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.684 3.327 qs_ot_get_p 119 10.4 0.001 0.001 3.130 3.164 init_scf_loop 11 6.9 0.001 0.003 2.545 2.546 rs_pw_transfer 974 11.9 0.009 0.010 2.233 2.373 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.044 2.348 apply_single 119 13.6 0.000 0.000 2.043 2.348 calculate_first_density_matrix 1 7.0 0.000 0.003 2.310 2.313 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.066 2.071 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.372 2.007 make_m2s 4572 13.5 0.045 0.045 1.759 1.933 density_rs2pw 119 9.7 0.004 0.004 1.781 1.909 grid_integrate_task_list 119 12.3 1.798 1.902 1.798 1.902 jit_kernel_multiply 10 16.2 0.684 1.860 0.684 1.860 calculate_dm_sparse 119 9.5 0.000 0.000 1.849 1.858 make_images 4572 14.5 0.192 0.204 1.675 1.847 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.802 1.802 prepare_preconditioner 11 7.9 0.000 0.000 1.732 1.734 make_preconditioner 11 8.9 0.000 0.001 1.732 1.734 potential_pw2rs 119 12.3 0.007 0.008 1.669 1.674 pw_transfer 1439 11.6 0.064 0.066 1.661 1.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.593 1.668 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.644 1.645 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.570 1.581 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.570 1.570 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.551 1.560 multiply_cannon_sync_h2d 18288 15.5 1.386 1.555 1.386 1.555 cp_fm_redistribute_end 50 14.0 1.163 1.548 1.164 1.549 cp_fm_diag_elpa_base 50 14.0 0.368 1.490 0.382 1.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.504 1.510 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 mp_sum_l 7207 12.9 1.130 1.445 1.130 1.445 grid_collocate_task_list 119 9.7 1.205 1.284 1.205 1.284 fft3d_ps 1201 14.6 0.502 0.517 1.255 1.266 acc_transpose_blocks 18288 15.5 0.075 0.077 1.230 1.253 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.233 1.244 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.228 1.232 wfi_extrapolate 11 7.9 0.001 0.001 1.156 1.157 mp_alltoall_d11v 2130 13.8 0.967 1.087 0.967 1.087 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.967 0.968 make_images_data 4572 15.5 0.045 0.049 0.776 0.960 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.875 0.897 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.674 0.879 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.795 0.796 acc_transpose_blocks_kernels 18288 16.5 0.210 0.221 0.783 0.795 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.654 0.744 cp_fm_cholesky_invert 11 10.9 0.720 0.724 0.720 0.724 mp_waitany 9880 13.7 0.551 0.700 0.551 0.700 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.537 0.670 mp_alltoall_z22v 1201 16.6 0.596 0.653 0.596 0.653 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.334000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=2.346089 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 553.500672E+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 0.199 0.577 35.922 35.922 qs_mol_dyn_low 1 2.0 0.003 0.003 35.023 35.344 qs_forces 11 3.9 0.002 0.002 34.694 34.694 qs_energies 11 4.9 0.021 0.065 32.980 32.985 scf_env_do_scf 11 5.9 0.001 0.001 27.684 27.685 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.008 24.009 dbcsr_multiply_generic 2286 12.5 0.099 0.102 17.947 18.063 velocity_verlet 10 3.0 0.001 0.002 17.687 17.698 qs_scf_new_mos 108 7.5 0.001 0.001 15.973 16.022 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.973 16.021 ot_scf_mini 108 9.5 0.002 0.003 15.037 15.092 multiply_cannon 2286 13.5 0.230 0.273 14.494 14.959 multiply_cannon_loop 2286 14.5 0.948 0.973 13.551 14.041 ot_mini 108 10.5 0.001 0.001 9.134 9.208 multiply_cannon_multrec 27432 15.5 2.332 2.986 8.610 8.975 qs_ot_get_derivative 108 11.5 0.001 0.001 7.334 7.391 dbcsr_mm_accdrv_process 47916 15.9 5.356 7.295 6.184 7.350 rebuild_ks_matrix 119 8.3 0.000 0.000 6.446 6.497 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.445 6.497 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.713 5.755 init_scf_run 11 5.9 0.004 0.016 3.811 3.814 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.807 3.813 sum_up_and_integrate 119 10.3 0.035 0.038 3.719 3.725 integrate_v_rspace 119 11.3 0.002 0.003 3.684 3.691 init_scf_loop 11 6.9 0.093 0.249 3.656 3.658 qs_ot_get_p 119 10.4 0.001 0.001 3.309 3.389 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.232 3.267 calculate_rho_elec 119 8.7 0.040 0.046 3.231 3.267 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.822 3.140 prepare_preconditioner 11 7.9 0.000 0.000 2.553 2.563 make_preconditioner 11 8.9 0.000 0.001 2.553 2.563 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.151 2.489 mp_waitall_1 145218 16.4 2.021 2.484 2.021 2.484 calculate_first_density_matrix 1 7.0 0.009 0.023 2.436 2.445 make_m2s 4572 13.5 0.054 0.056 2.165 2.283 make_images 4572 14.5 0.274 0.333 2.057 2.174 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.943 2.150 apply_single 119 13.6 0.000 0.000 1.943 2.150 calculate_dm_sparse 119 9.5 0.000 0.000 2.042 2.096 rs_pw_transfer 974 11.9 0.009 0.009 1.939 2.053 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.024 2.032 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.909 1.937 grid_integrate_task_list 119 12.3 1.835 1.923 1.835 1.923 acc_transpose_blocks 27432 15.5 0.111 0.114 1.505 1.857 jit_kernel_multiply 10 15.9 0.767 1.855 0.767 1.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.849 1.850 density_rs2pw 119 9.7 0.004 0.004 1.722 1.824 pw_transfer 1439 11.6 0.063 0.067 1.755 1.785 ot_diis_step 108 11.5 0.012 0.012 1.760 1.761 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.707 1.707 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.665 1.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.653 1.666 potential_pw2rs 119 12.3 0.008 0.009 1.502 1.509 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.469 1.469 cp_fm_redistribute_end 50 14.0 0.967 1.445 0.968 1.446 cp_fm_diag_elpa_base 50 14.0 0.455 1.386 0.476 1.428 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.358 1.394 fft3d_ps 1201 14.6 0.532 0.585 1.346 1.369 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.777 1.327 dbcsr_complete_redistribute 329 12.2 0.123 0.142 0.962 1.323 wfi_extrapolate 11 7.9 0.001 0.001 1.322 1.322 grid_collocate_task_list 119 9.7 1.221 1.302 1.221 1.302 mp_sum_l 7207 12.9 0.896 1.262 0.896 1.262 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.224 1.235 acc_transpose_blocks_kernels 27432 16.5 0.269 0.282 0.871 1.223 qs_energies_init_hamiltonians 11 5.9 0.009 0.013 1.158 1.158 cp_fm_upper_to_full 72 13.5 0.820 1.138 0.820 1.138 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.097 1.118 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.753 1.112 multiply_cannon_sync_h2d 27432 15.5 0.990 1.097 0.990 1.097 mp_alltoall_i22 627 13.8 0.587 0.975 0.587 0.975 make_images_data 4572 15.5 0.045 0.049 0.836 0.961 jit_kernel_transpose 5 15.6 0.601 0.958 0.601 0.958 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.718 0.906 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.872 mp_alltoall_d11v 2130 13.8 0.733 0.845 0.733 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.829 mp_sum_d 4123 12.0 0.589 0.769 0.589 0.769 cp_fm_cholesky_invert 11 10.9 0.738 0.741 0.738 0.741 mp_alltoall_z22v 1201 16.6 0.695 0.722 0.695 0.722 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.922000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.363636, yerr=3.820346 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 600.211456E+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 0.061 0.101 29.141 29.141 qs_mol_dyn_low 1 2.0 0.003 0.003 28.742 28.749 qs_forces 11 3.9 0.002 0.002 28.530 28.530 qs_energies 11 4.9 0.001 0.001 26.832 26.835 scf_env_do_scf 11 5.9 0.000 0.001 21.635 21.635 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.191 19.192 velocity_verlet 10 3.0 0.012 0.015 14.093 14.105 dbcsr_multiply_generic 2286 12.5 0.091 0.094 13.373 13.445 qs_scf_new_mos 108 7.5 0.001 0.001 11.781 11.800 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.780 11.799 multiply_cannon 2286 13.5 0.228 0.234 10.822 11.285 ot_scf_mini 108 9.5 0.002 0.002 11.111 11.139 multiply_cannon_loop 2286 14.5 0.332 0.342 9.921 10.110 multiply_cannon_multrec 9144 15.5 2.060 2.259 7.070 7.326 ot_mini 108 10.5 0.001 0.001 6.598 6.632 rebuild_ks_matrix 119 8.3 0.000 0.000 5.727 5.749 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.727 5.748 qs_ot_get_derivative 108 11.5 0.001 0.001 5.311 5.340 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.094 5.114 dbcsr_mm_accdrv_process 12550 15.8 3.487 4.952 4.910 5.009 init_scf_run 11 5.9 0.000 0.001 3.746 3.746 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.746 3.746 sum_up_and_integrate 119 10.3 0.037 0.040 3.475 3.483 integrate_v_rspace 119 11.3 0.003 0.003 3.438 3.447 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.123 3.130 calculate_rho_elec 119 8.7 0.059 0.061 3.123 3.129 qs_ot_get_p 119 10.4 0.001 0.001 2.748 2.790 calculate_first_density_matrix 1 7.0 0.000 0.000 2.649 2.650 init_scf_loop 11 6.9 0.000 0.000 2.424 2.424 jit_kernel_multiply 9 15.7 1.385 2.313 1.385 2.313 mp_waitall_1 121218 16.5 1.613 2.060 1.613 2.060 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.989 2.004 grid_integrate_task_list 119 12.3 1.844 1.907 1.844 1.907 make_m2s 4572 13.5 0.035 0.036 1.763 1.893 calculate_dm_sparse 119 9.5 0.000 0.000 1.850 1.869 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.823 1.825 make_images 4572 14.5 0.269 0.302 1.674 1.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.791 1.792 prepare_preconditioner 11 7.9 0.000 0.000 1.712 1.718 make_preconditioner 11 8.9 0.000 0.000 1.712 1.718 rs_pw_transfer 974 11.9 0.008 0.008 1.595 1.659 pw_transfer 1439 11.6 0.063 0.065 1.646 1.654 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.607 1.632 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.630 1.630 density_rs2pw 119 9.7 0.003 0.004 1.557 1.619 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.607 1.616 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.556 1.563 grid_collocate_task_list 119 9.7 1.272 1.353 1.272 1.353 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.345 1.345 cp_fm_redistribute_end 50 14.0 0.666 1.323 0.667 1.324 cp_fm_diag_elpa_base 50 14.0 0.611 1.252 0.655 1.311 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.271 1.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.273 1.278 ot_diis_step 108 11.5 0.012 0.013 1.275 1.275 potential_pw2rs 119 12.3 0.010 0.011 1.271 1.274 qs_energies_init_hamiltonians 11 5.9 0.006 0.009 1.243 1.243 fft3d_ps 1201 14.6 0.535 0.546 1.224 1.231 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.149 1.171 apply_single 119 13.6 0.000 0.000 1.149 1.171 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.119 1.135 wfi_extrapolate 11 7.9 0.001 0.001 1.051 1.051 hybrid_alltoall_any 4725 16.4 0.062 0.173 0.806 1.012 make_images_data 4572 15.5 0.039 0.042 0.842 0.995 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.919 acc_transpose_blocks 9144 15.5 0.038 0.039 0.903 0.911 cp_fm_cholesky_invert 11 10.9 0.857 0.859 0.857 0.859 mp_alltoall_d11v 2130 13.8 0.746 0.813 0.746 0.813 multiply_cannon_sync_h2d 9144 15.5 0.705 0.792 0.705 0.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.757 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.688 0.733 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.717 0.725 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.347 0.711 mp_allgather_i34 2286 14.5 0.235 0.691 0.235 0.691 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.659 0.665 mp_alltoall_z22v 1201 16.6 0.563 0.595 0.563 0.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.141000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.272727, yerr=5.241286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 761.700352E+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 0.025 0.042 41.425 41.426 qs_mol_dyn_low 1 2.0 0.003 0.003 41.151 41.158 qs_forces 11 3.9 0.014 0.021 40.695 40.696 qs_energies 11 4.9 0.001 0.001 38.754 38.757 scf_env_do_scf 11 5.9 0.001 0.001 32.395 32.395 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.608 24.609 velocity_verlet 10 3.0 0.002 0.002 23.141 23.196 dbcsr_multiply_generic 2286 12.5 0.099 0.100 17.816 17.903 qs_scf_new_mos 108 7.5 0.001 0.001 15.781 15.894 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.780 15.893 ot_scf_mini 108 9.5 0.002 0.002 14.698 14.810 multiply_cannon 2286 13.5 0.309 0.317 13.948 14.742 multiply_cannon_loop 2286 14.5 0.345 0.351 12.630 13.427 multiply_cannon_multrec 9144 15.5 3.428 5.161 8.882 9.072 ot_mini 108 10.5 0.001 0.001 8.769 8.893 init_scf_loop 11 6.9 0.000 0.000 7.760 7.764 rebuild_ks_matrix 119 8.3 0.000 0.000 6.784 6.930 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.784 6.930 qs_ot_get_derivative 108 11.5 0.001 0.001 6.751 6.860 prepare_preconditioner 11 7.9 0.000 0.000 6.833 6.847 make_preconditioner 11 8.9 0.000 0.000 6.833 6.847 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.374 6.722 dbcsr_mm_accdrv_process 12550 15.8 3.942 5.559 5.334 6.618 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.134 6.266 cp_fm_upper_to_full 72 14.2 3.166 4.565 3.166 4.565 init_scf_run 11 5.9 0.000 0.001 4.320 4.321 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.320 4.321 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.742 3.807 calculate_rho_elec 119 8.7 0.118 0.121 3.742 3.806 sum_up_and_integrate 119 10.3 0.064 0.066 3.637 3.643 integrate_v_rspace 119 11.3 0.003 0.003 3.573 3.580 mp_waitall_1 97218 16.6 2.528 3.401 2.528 3.401 qs_ot_get_p 119 10.4 0.001 0.001 3.201 3.342 calculate_first_density_matrix 1 7.0 0.000 0.000 2.816 2.883 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.477 2.859 dbcsr_complete_redistribute 329 12.2 0.290 0.292 2.023 2.830 make_m2s 4572 13.5 0.038 0.038 2.321 2.518 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.707 2.515 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.149 2.401 apply_single 119 13.6 0.000 0.000 2.149 2.400 make_images 4572 14.5 0.354 0.388 2.201 2.397 mp_alltoall_i22 627 13.8 1.515 2.359 1.515 2.359 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.339 2.340 calculate_dm_sparse 119 9.5 0.000 0.000 2.210 2.274 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.454 2.257 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.318 2.132 pw_transfer 1439 11.6 0.066 0.067 2.035 2.040 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.989 2.036 jit_kernel_multiply 10 15.6 1.365 2.025 1.365 2.025 grid_integrate_task_list 119 12.3 2.000 2.019 2.000 2.019 ot_diis_step 108 11.5 0.014 0.014 1.996 1.996 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.940 1.944 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.936 1.937 density_rs2pw 119 9.7 0.004 0.004 1.782 1.789 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.765 1.768 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.701 1.756 mp_sum_l 7207 12.9 1.006 1.686 1.006 1.686 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.658 1.659 fft_wrap_pw1pw2_140 487 13.2 0.087 0.088 1.621 1.627 fft3d_ps 1201 14.6 0.569 0.581 1.573 1.578 grid_collocate_task_list 119 9.7 1.443 1.458 1.443 1.458 cp_fm_cholesky_invert 11 10.9 1.454 1.457 1.454 1.457 rs_pw_transfer 974 11.9 0.009 0.009 1.366 1.392 wfi_extrapolate 11 7.9 0.001 0.001 1.380 1.380 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.123 1.370 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.352 1.352 cp_fm_diag_elpa_base 50 14.0 1.205 1.258 1.350 1.350 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.276 1.339 make_images_data 4572 15.5 0.043 0.046 1.084 1.322 mp_alltoall_d11v 2130 13.8 1.228 1.303 1.228 1.303 potential_pw2rs 119 12.3 0.014 0.015 1.186 1.189 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.130 1.145 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.106 1.120 multiply_cannon_sync_h2d 9144 15.5 1.037 1.040 1.037 1.040 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.984 1.018 qs_create_task_list 11 7.9 0.000 0.000 0.941 0.952 generate_qs_task_list 11 8.9 0.373 0.392 0.941 0.952 acc_transpose_blocks 9144 15.5 0.038 0.038 0.897 0.907 mp_alltoall_z22v 1201 16.6 0.868 0.881 0.868 0.881 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.862 0.876 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.788 0.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.426000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.181818, yerr=13.326789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.850688E+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 1851502. 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 0.053 0.071 83.807 83.808 qs_mol_dyn_low 1 2.0 0.003 0.003 83.387 83.398 qs_forces 11 3.9 0.018 0.026 82.993 82.995 qs_energies 11 4.9 0.007 0.027 80.055 80.083 scf_env_do_scf 11 5.9 0.000 0.001 70.238 70.240 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.789 64.790 dbcsr_multiply_generic 2055 12.4 0.103 0.109 52.547 52.942 qs_scf_new_mos 99 7.5 0.000 0.001 47.987 48.114 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.987 48.113 ot_scf_mini 99 9.5 0.002 0.002 45.583 45.708 multiply_cannon 2055 13.4 0.176 0.185 43.019 44.160 multiply_cannon_loop 2055 14.4 1.581 1.619 42.079 43.238 velocity_verlet 10 3.0 0.007 0.016 42.659 42.663 ot_mini 99 10.5 0.001 0.001 27.540 27.659 qs_ot_get_derivative 99 11.5 0.001 0.001 20.613 20.740 multiply_cannon_multrec 49320 15.4 12.097 12.913 17.498 18.188 rebuild_ks_matrix 110 8.3 0.000 0.001 14.348 14.611 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.348 14.611 mp_waitall_1 241148 16.1 11.764 12.911 11.764 12.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.593 12.827 multiply_cannon_sync_h2d 49320 15.4 9.934 10.725 9.934 10.725 qs_ot_get_p 110 10.4 0.001 0.001 9.522 9.680 multiply_cannon_metrocomm3 49320 15.4 0.080 0.086 7.009 8.417 init_scf_run 11 5.9 0.000 0.001 7.775 7.776 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.775 7.776 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.209 7.669 apply_single 110 13.6 0.000 0.001 7.209 7.669 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.017 7.548 sum_up_and_integrate 110 10.3 0.037 0.044 6.944 6.960 integrate_v_rspace 110 11.3 0.002 0.003 6.907 6.931 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.624 6.695 ot_diis_step 99 11.5 0.005 0.005 6.589 6.589 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.153 6.302 calculate_rho_elec 110 8.6 0.020 0.024 6.153 6.302 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.122 6.158 mp_sum_l 6514 12.8 5.181 6.008 5.181 6.008 dbcsr_mm_accdrv_process 87628 16.1 2.112 2.225 5.280 5.667 init_scf_loop 11 6.9 0.001 0.003 5.421 5.422 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.155 5.156 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.678 4.680 cp_fm_redistribute_end 48 14.0 4.072 4.655 4.076 4.656 cp_fm_diag_elpa_base 48 14.0 0.573 4.594 0.577 4.619 wfi_extrapolate 11 7.9 0.001 0.001 4.536 4.536 calculate_dm_sparse 110 9.5 0.000 0.001 3.982 4.119 make_m2s 4110 13.4 0.060 0.065 3.898 3.989 rs_pw_transfer 902 11.9 0.012 0.013 3.706 3.922 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.893 3.899 make_images 4110 14.4 0.177 0.191 3.802 3.898 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.712 3.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.465 3.532 density_rs2pw 110 9.6 0.004 0.004 3.229 3.413 grid_integrate_task_list 110 12.3 3.219 3.389 3.219 3.389 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.299 3.357 prepare_preconditioner 11 7.9 0.000 0.000 3.223 3.242 make_preconditioner 11 8.9 0.000 0.001 3.223 3.241 calculate_first_density_matrix 1 7.0 0.000 0.002 3.154 3.160 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.999 3.048 jit_kernel_multiply 13 15.9 2.882 2.962 2.882 2.962 pw_transfer 1331 11.6 0.054 0.069 2.797 2.862 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.710 2.778 potential_pw2rs 110 12.3 0.006 0.007 2.559 2.583 mp_alltoall_d11v 2046 13.8 2.154 2.524 2.154 2.524 fft_wrap_pw1pw2_140 451 13.1 0.170 0.188 2.292 2.363 acc_transpose_blocks 49320 15.4 0.224 0.236 2.278 2.359 fft3d_ps 1111 14.6 0.743 0.824 2.247 2.307 mp_sum_d 3879 11.9 1.635 2.185 1.635 2.185 grid_collocate_task_list 110 9.6 2.088 2.179 2.088 2.179 mp_waitany 14300 13.8 1.831 2.025 1.831 2.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.930 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.863 1.896 make_images_data 4110 15.4 0.043 0.046 1.772 1.891 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.541 1.793 parallel_gemm_fm 81 9.0 0.000 0.000 1.772 1.779 parallel_gemm_fm_cosma 81 10.0 1.772 1.779 1.772 1.779 cp_fm_cholesky_invert 11 10.9 1.762 1.766 1.762 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.808000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.636364, yerr=2.100767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 587.698176E+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 1372800. 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 0.042 0.095 68.622 68.623 qs_mol_dyn_low 1 2.0 0.003 0.003 68.096 68.105 qs_forces 11 3.9 0.002 0.002 68.027 68.028 qs_energies 11 4.9 0.008 0.031 64.690 64.694 scf_env_do_scf 11 5.9 0.001 0.001 56.155 56.159 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.554 48.563 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.679 37.876 velocity_verlet 10 3.0 0.012 0.014 36.147 36.148 qs_scf_new_mos 99 7.5 0.001 0.001 32.891 33.039 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.891 33.039 multiply_cannon 2055 13.4 0.222 0.246 31.035 32.071 ot_scf_mini 99 9.5 0.003 0.003 31.246 31.376 multiply_cannon_loop 2055 14.4 0.930 0.948 29.738 30.450 ot_mini 99 10.5 0.001 0.001 18.688 18.831 multiply_cannon_multrec 24660 15.4 7.607 9.436 13.784 15.507 rebuild_ks_matrix 110 8.3 0.000 0.001 13.535 13.647 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.534 13.647 qs_ot_get_derivative 99 11.5 0.001 0.001 12.909 13.042 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.889 12.001 mp_waitall_1 186928 16.3 7.880 10.231 7.880 10.231 multiply_cannon_sync_h2d 24660 15.4 6.927 8.346 6.927 8.346 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.238 8.105 init_scf_loop 11 6.9 0.001 0.003 7.566 7.590 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.473 7.107 apply_single 110 13.6 0.000 0.001 6.473 7.107 sum_up_and_integrate 110 10.3 0.052 0.059 6.505 6.534 integrate_v_rspace 110 11.3 0.002 0.002 6.452 6.479 dbcsr_mm_accdrv_process 52282 16.1 4.597 5.463 6.018 6.315 qs_ot_get_p 110 10.4 0.001 0.001 6.044 6.243 init_scf_run 11 5.9 0.000 0.001 6.112 6.112 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.111 6.112 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.231 5.962 ot_diis_step 99 11.5 0.010 0.010 5.730 5.730 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.667 5.674 calculate_rho_elec 110 8.6 0.039 0.047 5.666 5.674 prepare_preconditioner 11 7.9 0.000 0.000 5.509 5.530 make_preconditioner 11 8.9 0.000 0.001 5.509 5.530 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.100 5.263 make_m2s 4110 13.4 0.057 0.060 4.156 4.617 make_images 4110 14.4 0.402 0.447 4.051 4.508 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.125 4.146 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.694 3.695 wfi_extrapolate 11 7.9 0.001 0.001 3.441 3.441 pw_transfer 1331 11.6 0.066 0.071 3.281 3.432 grid_integrate_task_list 110 12.3 3.154 3.371 3.154 3.371 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.174 3.326 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.168 3.236 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.212 3.214 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.169 3.171 density_rs2pw 110 9.6 0.004 0.004 2.967 3.154 cp_fm_redistribute_end 48 14.0 2.361 3.147 2.363 3.147 cp_fm_diag_elpa_base 48 14.0 0.748 3.035 0.780 3.123 rs_pw_transfer 902 11.9 0.012 0.014 2.870 3.063 calculate_dm_sparse 110 9.5 0.001 0.001 2.938 2.967 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.882 2.935 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.704 2.858 make_images_data 4110 15.4 0.047 0.051 2.314 2.777 hybrid_alltoall_any 4261 16.3 0.104 0.446 2.035 2.729 fft3d_ps 1111 14.6 1.071 1.266 2.529 2.668 calculate_first_density_matrix 1 7.0 0.001 0.002 2.584 2.588 cp_fm_cholesky_invert 11 10.9 2.570 2.577 2.570 2.577 mp_sum_l 6514 12.8 1.889 2.535 1.889 2.535 potential_pw2rs 110 12.3 0.008 0.009 2.368 2.397 grid_collocate_task_list 110 9.6 2.119 2.261 2.119 2.261 jit_kernel_multiply 12 16.2 1.073 2.244 1.073 2.244 mp_alltoall_d11v 2046 13.8 1.875 2.098 1.875 2.098 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.939 1.958 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.860 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.809 1.817 multiply_cannon_metrocomm4 22605 15.4 0.075 0.078 0.814 1.737 mp_irecv_dv 57340 16.2 0.690 1.618 0.690 1.618 mp_allgather_i34 2055 14.4 0.626 1.574 0.626 1.574 acc_transpose_blocks 24660 15.4 0.110 0.114 1.530 1.553 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.538 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.487 mp_waitany 10164 13.8 1.171 1.386 1.171 1.386 dbcsr_complete_redistribute 325 12.2 0.241 0.298 1.112 1.386 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.623000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.000000, yerr=7.483315 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 659.427328E+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 0.020 0.038 60.312 60.313 qs_mol_dyn_low 1 2.0 0.003 0.003 59.991 60.000 qs_forces 11 3.9 0.008 0.014 59.926 59.927 qs_energies 11 4.9 0.002 0.003 56.791 56.801 scf_env_do_scf 11 5.9 0.001 0.001 48.498 48.498 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.805 39.819 velocity_verlet 10 3.0 0.002 0.002 32.588 32.589 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.835 29.159 qs_scf_new_mos 99 7.5 0.001 0.001 25.216 25.351 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.215 25.351 ot_scf_mini 99 9.5 0.002 0.003 24.001 24.141 multiply_cannon 2055 13.4 0.211 0.223 22.645 23.674 multiply_cannon_loop 2055 14.4 0.619 0.640 21.516 22.669 ot_mini 99 10.5 0.001 0.001 13.848 13.988 rebuild_ks_matrix 110 8.3 0.000 0.000 12.071 12.256 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.070 12.255 multiply_cannon_multrec 16440 15.4 3.763 4.596 10.109 10.916 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.659 10.829 mp_waitall_1 146766 16.3 7.364 10.490 7.364 10.490 qs_ot_get_derivative 99 11.5 0.001 0.001 9.359 9.500 init_scf_loop 11 6.9 0.000 0.001 8.418 8.419 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.538 7.475 prepare_preconditioner 11 7.9 0.000 0.000 6.669 6.686 make_preconditioner 11 8.9 0.000 0.000 6.669 6.686 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.045 6.391 dbcsr_mm_accdrv_process 34862 16.1 4.679 5.991 6.201 6.377 sum_up_and_integrate 110 10.3 0.060 0.060 6.319 6.335 integrate_v_rspace 110 11.3 0.002 0.003 6.259 6.275 init_scf_run 11 5.9 0.001 0.005 5.811 5.811 scf_env_initial_rho_setup 11 6.9 0.003 0.017 5.810 5.811 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.694 5.708 calculate_rho_elec 110 8.6 0.058 0.058 5.694 5.708 qs_ot_get_p 110 10.4 0.001 0.001 5.368 5.546 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.987 5.451 apply_single 110 13.6 0.000 0.000 4.987 5.451 ot_diis_step 99 11.5 0.010 0.011 4.458 4.458 make_m2s 4110 13.4 0.050 0.051 4.107 4.448 make_images 4110 14.4 0.394 0.508 3.995 4.336 multiply_cannon_sync_h2d 16440 15.4 3.636 4.046 3.636 4.046 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.780 3.784 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.084 3.736 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.447 3.448 grid_integrate_task_list 110 12.3 3.198 3.363 3.198 3.363 pw_transfer 1331 11.6 0.065 0.071 3.244 3.264 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.137 3.158 density_rs2pw 110 9.6 0.004 0.004 2.978 3.129 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.939 2.940 wfi_extrapolate 11 7.9 0.001 0.001 2.925 2.925 cp_fm_redistribute_end 48 14.0 1.825 2.915 1.828 2.915 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.230 2.905 jit_kernel_multiply 12 16.3 1.138 2.897 1.138 2.897 cp_fm_diag_elpa_base 48 14.0 1.024 2.780 1.083 2.892 rs_pw_transfer 902 11.9 0.011 0.011 2.704 2.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.814 2.815 make_images_data 4110 15.4 0.044 0.048 2.389 2.815 calculate_first_density_matrix 1 7.0 0.000 0.000 2.801 2.805 calculate_dm_sparse 110 9.5 0.001 0.001 2.680 2.710 fft_wrap_pw1pw2_140 451 13.1 0.209 0.212 2.674 2.706 cp_fm_cholesky_invert 11 10.9 2.633 2.639 2.633 2.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.465 2.540 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.915 2.508 fft3d_ps 1111 14.6 1.065 1.123 2.450 2.468 mp_irecv_dv 48980 15.7 0.846 2.384 0.846 2.384 grid_collocate_task_list 110 9.6 2.177 2.368 2.177 2.368 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.291 2.362 potential_pw2rs 110 12.3 0.011 0.011 2.160 2.167 mp_alltoall_d11v 2046 13.8 1.859 2.161 1.859 2.161 mp_sum_l 6514 12.8 1.468 2.078 1.468 2.078 dbcsr_complete_redistribute 325 12.2 0.327 0.372 1.572 2.024 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.935 1.941 cp_fm_upper_to_full 70 13.6 1.386 1.838 1.386 1.838 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.650 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.612 1.627 cp_fm_cholesky_decompose 22 10.9 1.547 1.575 1.547 1.575 mp_allgather_i34 2055 14.4 0.453 1.501 0.453 1.501 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.356 1.456 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.928 1.372 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.245 1.330 mp_waitany 17072 13.8 1.123 1.270 1.123 1.270 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.247 1.254 acc_transpose_blocks 16440 15.4 0.074 0.078 1.225 1.238 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.313000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.363636, yerr=9.098179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 727.183360E+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 0.446 0.551 65.811 65.812 qs_mol_dyn_low 1 2.0 0.003 0.003 64.732 64.745 qs_forces 11 3.9 0.002 0.002 64.658 64.658 qs_energies 11 4.9 0.002 0.005 61.261 61.264 scf_env_do_scf 11 5.9 0.001 0.001 52.697 52.700 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.031 41.032 velocity_verlet 10 3.0 0.001 0.002 36.835 36.837 dbcsr_multiply_generic 2055 12.4 0.114 0.116 29.985 30.214 qs_scf_new_mos 99 7.5 0.001 0.001 26.821 26.941 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.821 26.940 ot_scf_mini 99 9.5 0.003 0.003 24.928 25.065 multiply_cannon 2055 13.4 0.242 0.264 22.902 24.191 multiply_cannon_loop 2055 14.4 0.893 0.908 21.535 22.168 ot_mini 99 10.5 0.001 0.001 14.255 14.360 multiply_cannon_multrec 24660 15.4 4.213 6.651 12.695 13.921 rebuild_ks_matrix 110 8.3 0.000 0.000 11.798 11.917 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 11.798 11.916 init_scf_loop 11 6.9 0.000 0.002 11.624 11.626 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.399 10.505 qs_ot_get_derivative 99 11.5 0.001 0.001 10.094 10.194 prepare_preconditioner 11 7.9 0.000 0.000 9.911 9.925 make_preconditioner 11 8.9 0.000 0.001 9.911 9.925 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.126 9.605 dbcsr_mm_accdrv_process 52304 16.0 6.707 8.069 8.337 9.240 mp_waitall_1 126806 16.4 4.651 6.680 4.651 6.680 sum_up_and_integrate 110 10.3 0.067 0.070 6.268 6.288 integrate_v_rspace 110 11.3 0.003 0.003 6.200 6.220 qs_ot_get_p 110 10.4 0.001 0.001 5.633 5.766 make_m2s 4110 13.4 0.059 0.060 5.440 5.710 make_images 4110 14.4 0.580 0.695 5.300 5.573 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.542 5.551 calculate_rho_elec 110 8.6 0.077 0.081 5.542 5.550 init_scf_run 11 5.9 0.000 0.001 5.541 5.541 scf_env_initial_rho_setup 11 6.9 0.004 0.013 5.540 5.541 cp_fm_upper_to_full 70 13.8 3.356 4.765 3.356 4.765 ot_diis_step 99 11.5 0.011 0.011 4.126 4.126 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.034 4.123 apply_single 110 13.6 0.000 0.000 4.033 4.123 dbcsr_complete_redistribute 325 12.2 0.429 0.483 2.709 3.865 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.772 3.787 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.585 3.641 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.600 3.521 grid_integrate_task_list 110 12.3 3.274 3.474 3.274 3.474 multiply_cannon_sync_h2d 24660 15.4 3.168 3.337 3.168 3.337 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.296 3.296 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.153 3.295 pw_transfer 1331 11.6 0.065 0.071 3.219 3.249 calculate_dm_sparse 110 9.5 0.001 0.001 3.187 3.230 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.113 3.145 make_images_data 4110 15.4 0.047 0.051 2.741 3.059 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.972 3.026 hybrid_alltoall_any 4261 16.3 0.120 0.462 2.390 2.921 wfi_extrapolate 11 7.9 0.001 0.001 2.912 2.912 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.776 2.901 mp_alltoall_i22 605 13.7 1.661 2.863 1.661 2.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.855 2.857 density_rs2pw 110 9.6 0.004 0.004 2.688 2.854 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.769 2.771 cp_fm_redistribute_end 48 14.0 1.377 2.742 1.379 2.743 cp_fm_diag_elpa_base 48 14.0 1.281 2.613 1.362 2.725 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.664 2.699 cp_fm_cholesky_invert 11 10.9 2.642 2.650 2.642 2.650 calculate_first_density_matrix 1 7.0 0.000 0.000 2.541 2.546 rs_pw_transfer 902 11.9 0.010 0.011 2.324 2.534 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.511 fft3d_ps 1111 14.6 1.062 1.093 2.413 2.436 grid_collocate_task_list 110 9.6 2.221 2.384 2.221 2.384 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.238 2.239 potential_pw2rs 110 12.3 0.012 0.013 2.072 2.085 jit_kernel_multiply 11 15.5 1.297 2.069 1.297 2.069 mp_alltoall_d11v 2046 13.8 1.771 1.912 1.771 1.912 acc_transpose_blocks 24660 15.4 0.104 0.107 1.750 1.872 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.782 1.872 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.711 cp_fm_cholesky_decompose 22 10.9 1.630 1.675 1.630 1.675 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.645 1.655 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.541 mp_allgather_i34 2055 14.4 0.469 1.511 0.469 1.511 multiply_cannon_metrocomm4 20550 15.4 0.057 0.061 0.850 1.501 mp_sum_l 6514 12.8 0.981 1.478 0.981 1.478 mp_irecv_dv 62702 16.1 0.751 1.426 0.751 1.426 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.812000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.818182, yerr=9.123578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 827.092992E+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 0.253 0.263 55.847 55.848 qs_mol_dyn_low 1 2.0 0.003 0.003 55.213 55.227 qs_forces 11 3.9 0.007 0.007 55.140 55.141 qs_energies 11 4.9 0.001 0.002 51.534 51.540 scf_env_do_scf 11 5.9 0.000 0.001 42.550 42.550 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.660 34.661 velocity_verlet 10 3.0 0.011 0.012 30.730 30.733 dbcsr_multiply_generic 2055 12.4 0.103 0.105 23.600 23.763 qs_scf_new_mos 99 7.5 0.001 0.001 20.775 20.827 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.775 20.827 ot_scf_mini 99 9.5 0.002 0.002 19.555 19.578 multiply_cannon 2055 13.4 0.246 0.261 17.919 19.228 multiply_cannon_loop 2055 14.4 0.324 0.337 16.469 16.734 rebuild_ks_matrix 110 8.3 0.000 0.000 11.457 11.511 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.456 11.510 ot_mini 99 10.5 0.001 0.001 10.767 10.786 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.217 10.264 multiply_cannon_multrec 8220 15.4 3.237 4.507 7.785 8.881 mp_waitall_1 106626 16.5 6.344 8.245 6.344 8.245 init_scf_loop 11 6.9 0.000 0.000 7.843 7.845 qs_ot_get_derivative 99 11.5 0.001 0.001 6.990 7.014 sum_up_and_integrate 110 10.3 0.080 0.081 6.089 6.103 prepare_preconditioner 11 7.9 0.000 0.000 6.056 6.060 make_preconditioner 11 8.9 0.000 0.000 6.056 6.060 integrate_v_rspace 110 11.3 0.003 0.005 6.009 6.022 init_scf_run 11 5.9 0.000 0.001 5.870 5.871 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.870 5.870 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.587 5.659 dbcsr_mm_accdrv_process 17442 15.9 2.900 4.161 4.419 5.513 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.489 5.503 calculate_rho_elec 110 8.6 0.115 0.115 5.489 5.502 qs_ot_get_p 110 10.4 0.001 0.001 4.921 4.962 make_m2s 4110 13.4 0.039 0.040 4.387 4.711 make_images 4110 14.4 0.642 0.702 4.259 4.585 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.069 4.387 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.797 3.857 apply_single 110 13.6 0.000 0.000 3.797 3.857 ot_diis_step 99 11.5 0.012 0.012 3.758 3.758 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.584 3.588 grid_integrate_task_list 110 12.3 3.374 3.469 3.374 3.469 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.286 3.286 pw_transfer 1331 11.6 0.065 0.072 3.145 3.168 calculate_first_density_matrix 1 7.0 0.000 0.000 3.139 3.139 multiply_cannon_sync_h2d 8220 15.4 2.916 3.103 2.916 3.103 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.038 3.063 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.027 3.028 make_images_data 4110 15.4 0.039 0.044 2.497 2.958 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.370 2.910 cp_fm_cholesky_invert 11 10.9 2.883 2.887 2.883 2.887 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.762 2.762 cp_fm_redistribute_end 48 14.0 0.696 2.733 0.701 2.733 cp_fm_diag_elpa_base 48 14.0 1.850 2.540 2.025 2.704 density_rs2pw 110 9.6 0.004 0.004 2.568 2.674 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.671 2.672 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.644 2.666 wfi_extrapolate 11 7.9 0.001 0.001 2.646 2.646 fft_wrap_pw1pw2_140 451 13.1 0.211 0.216 2.607 2.639 calculate_dm_sparse 110 9.5 0.001 0.001 2.564 2.611 grid_collocate_task_list 110 9.6 2.330 2.525 2.330 2.525 fft3d_ps 1111 14.6 1.112 1.169 2.295 2.328 rs_pw_transfer 902 11.9 0.010 0.011 2.111 2.269 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 2.000 jit_kernel_multiply 10 15.8 1.212 1.994 1.212 1.994 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.950 1.968 potential_pw2rs 110 12.3 0.015 0.016 1.864 1.868 mp_allgather_i34 2055 14.4 0.612 1.834 0.612 1.834 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.753 1.764 mp_alltoall_d11v 2046 13.8 1.594 1.737 1.594 1.737 cp_fm_cholesky_decompose 22 10.9 1.662 1.677 1.662 1.677 dbcsr_complete_redistribute 325 12.2 0.578 0.622 1.559 1.659 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.511 1.630 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.571 1.578 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.556 qs_create_task_list 11 7.9 0.000 0.000 1.221 1.319 generate_qs_task_list 11 8.9 0.380 0.448 1.221 1.319 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.484 1.212 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.176 1.208 mp_waitany 9240 13.8 1.036 1.180 1.036 1.180 mp_irecv_dv 24056 15.7 0.460 1.168 0.460 1.168 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.868 1.122 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.848000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.636364, yerr=8.803643 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 1.353900E+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 0.024 0.049 88.195 88.196 qs_mol_dyn_low 1 2.0 0.003 0.003 87.465 87.775 qs_forces 11 3.9 0.002 0.002 87.355 87.356 qs_energies 11 4.9 0.001 0.002 83.319 83.321 scf_env_do_scf 11 5.9 0.001 0.001 71.272 71.273 velocity_verlet 10 3.0 0.002 0.002 57.066 57.072 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.930 42.932 dbcsr_multiply_generic 2055 12.4 0.118 0.121 29.210 29.290 init_scf_loop 11 6.9 0.000 0.000 28.269 28.272 qs_scf_new_mos 99 7.5 0.001 0.001 26.383 26.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.383 26.445 prepare_preconditioner 11 7.9 0.000 0.000 26.335 26.342 make_preconditioner 11 8.9 0.000 0.000 26.335 26.342 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.610 25.813 ot_scf_mini 99 9.5 0.002 0.002 24.537 24.569 multiply_cannon 2055 13.4 0.346 0.367 22.043 22.908 multiply_cannon_loop 2055 14.4 0.343 0.347 20.202 20.495 cp_fm_upper_to_full 70 14.2 12.753 18.206 12.753 18.206 ot_mini 99 10.5 0.001 0.001 13.751 13.786 rebuild_ks_matrix 110 8.3 0.000 0.001 13.090 13.130 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.089 13.129 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.912 11.948 dbcsr_complete_redistribute 325 12.2 1.029 1.044 7.562 10.669 multiply_cannon_multrec 8220 15.4 4.382 4.582 9.692 9.893 mp_waitall_1 87304 16.6 8.429 9.489 8.429 9.489 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.362 9.468 qs_ot_get_derivative 99 11.5 0.001 0.001 9.193 9.226 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.709 8.794 mp_alltoall_i22 605 13.7 5.366 8.480 5.366 8.480 init_scf_run 11 5.9 0.000 0.001 7.826 7.827 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.826 7.826 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.830 6.853 calculate_rho_elec 110 8.6 0.227 0.227 6.829 6.852 sum_up_and_integrate 110 10.3 0.152 0.153 6.597 6.610 integrate_v_rspace 110 11.3 0.004 0.004 6.445 6.460 make_m2s 4110 13.4 0.043 0.044 5.406 5.965 make_images 4110 14.4 0.887 0.943 5.219 5.779 qs_ot_get_p 110 10.4 0.001 0.001 5.572 5.617 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.147 5.569 dbcsr_mm_accdrv_process 11614 15.7 3.407 3.873 5.169 5.514 wfi_extrapolate 11 7.9 0.001 0.001 5.288 5.289 cp_fm_cholesky_invert 11 10.9 5.256 5.261 5.256 5.261 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.666 5.138 apply_single 110 13.6 0.000 0.000 4.665 5.137 ot_diis_step 99 11.5 0.015 0.015 4.534 4.534 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.035 4.042 multiply_cannon_sync_h2d 8220 15.4 3.943 3.952 3.943 3.952 pw_transfer 1331 11.6 0.073 0.074 3.774 3.781 grid_integrate_task_list 110 12.3 3.660 3.712 3.660 3.712 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.982 3.705 make_images_data 4110 15.4 0.042 0.045 2.972 3.701 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.658 3.664 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.624 3.625 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.601 3.601 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.517 3.520 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.057 3.505 calculate_dm_sparse 110 9.5 0.001 0.001 3.302 3.318 density_rs2pw 110 9.6 0.004 0.004 3.205 3.225 fft_wrap_pw1pw2_140 451 13.1 0.215 0.215 3.172 3.178 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.004 3.004 cp_fm_diag_elpa_base 48 14.0 2.466 2.667 3.002 3.002 fft3d_ps 1111 14.6 1.267 1.278 2.866 2.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.725 2.738 grid_collocate_task_list 110 9.6 2.630 2.647 2.630 2.647 calculate_first_density_matrix 1 7.0 0.000 0.000 2.432 2.435 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.378 2.398 rs_pw_transfer 902 11.9 0.010 0.010 2.322 2.360 cp_fm_cholesky_decompose 22 10.9 2.326 2.344 2.326 2.344 mp_alltoall_d11v 2046 13.8 2.287 2.333 2.287 2.333 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.169 2.214 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.076 2.209 potential_pw2rs 110 12.3 0.021 0.021 2.031 2.033 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.975 2.020 jit_kernel_multiply 10 15.2 1.560 1.970 1.560 1.970 qs_create_task_list 11 7.9 0.000 0.000 1.897 1.941 generate_qs_task_list 11 8.9 0.741 0.794 1.897 1.941 make_basis_sm 11 9.8 0.000 0.000 1.919 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.782 1.786 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.752 1.771 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.196000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1210.727273, yerr=64.320350 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.908608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2489541. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.046 209.469 209.471 qs_mol_dyn_low 1 2.0 0.003 0.003 208.897 208.910 qs_forces 11 3.9 0.003 0.003 208.796 208.797 qs_energies 11 4.9 0.002 0.006 202.997 203.015 scf_env_do_scf 11 5.9 0.001 0.002 183.741 183.745 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.874 162.881 dbcsr_multiply_generic 2507 12.6 0.177 0.182 127.494 128.192 velocity_verlet 10 3.0 0.002 0.004 125.117 125.118 qs_scf_new_mos 117 7.6 0.001 0.001 123.740 124.155 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.739 124.154 ot_scf_mini 117 9.6 0.003 0.004 117.081 117.487 multiply_cannon 2507 13.6 0.235 0.241 102.928 105.203 multiply_cannon_loop 2507 14.6 2.140 2.206 100.578 102.520 ot_mini 117 10.6 0.001 0.001 67.920 68.261 multiply_cannon_multrec 60168 15.6 33.212 35.741 42.123 44.609 qs_ot_get_derivative 117 11.6 0.001 0.001 42.868 43.258 rebuild_ks_matrix 128 8.3 0.001 0.001 34.037 34.463 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 34.036 34.462 mp_waitall_1 291448 16.2 30.703 33.009 30.703 33.009 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.422 30.826 multiply_cannon_sync_h2d 60168 15.6 27.070 29.119 27.070 29.119 qs_ot_get_p 128 10.4 0.001 0.001 27.101 27.443 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.407 25.050 apply_single 128 13.6 0.001 0.001 24.407 25.050 ot_diis_step 117 11.6 0.007 0.008 24.669 24.670 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.737 21.007 init_scf_loop 11 6.9 0.001 0.003 20.662 20.696 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.310 20.419 multiply_cannon_metrocomm3 60168 15.6 0.117 0.125 16.125 18.177 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.596 17.598 prepare_preconditioner 11 7.9 0.000 0.000 16.023 16.076 make_preconditioner 11 8.9 0.000 0.001 16.023 16.076 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.273 15.470 make_m2s 5014 13.6 0.103 0.111 14.355 14.846 init_scf_run 11 5.9 0.000 0.001 14.798 14.799 scf_env_initial_rho_setup 11 6.9 0.000 0.001 14.798 14.799 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.778 14.783 cp_fm_redistribute_end 83 14.4 11.648 14.708 11.659 14.711 make_images 5014 14.6 0.406 0.425 14.180 14.681 cp_fm_diag_elpa_base 83 14.4 3.009 14.481 3.040 14.603 sum_up_and_integrate 128 10.3 0.090 0.108 14.441 14.495 integrate_v_rspace 128 11.3 0.003 0.004 14.350 14.406 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.580 12.707 calculate_rho_elec 128 8.7 0.045 0.063 12.579 12.707 wfi_extrapolate 11 7.9 0.001 0.001 10.856 10.856 mp_sum_l 7870 13.0 8.952 10.546 8.952 10.546 multiply_cannon_metrocomm1 60168 15.6 0.088 0.092 7.182 9.763 cp_fm_cholesky_invert 11 10.9 9.217 9.225 9.217 9.225 dbcsr_mm_accdrv_process 124484 16.2 3.287 3.461 8.480 9.097 calculate_dm_sparse 128 9.5 0.001 0.001 8.946 9.051 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.954 8.135 make_images_data 5014 15.6 0.070 0.076 7.060 8.116 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.872 7.994 grid_integrate_task_list 128 12.3 7.015 7.507 7.015 7.507 hybrid_alltoall_any 5200 16.5 0.294 2.252 6.201 7.486 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.325 7.340 pw_transfer 1547 11.6 0.075 0.106 7.008 7.308 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.804 7.073 density_rs2pw 128 9.7 0.006 0.007 6.404 6.972 rs_pw_transfer 1046 11.9 0.017 0.019 5.887 6.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.007 6.172 fft_wrap_pw1pw2_140 523 13.2 0.446 0.511 5.835 6.032 mp_alltoall_d11v 2415 14.1 4.654 5.979 4.654 5.979 fft3d_ps 1291 14.7 2.087 2.559 5.592 5.803 grid_collocate_task_list 128 9.7 4.723 5.085 4.723 5.085 potential_pw2rs 128 12.3 0.009 0.011 4.940 5.016 cp_fm_cholesky_decompose 22 10.9 4.650 4.664 4.650 4.664 mp_sum_d 4455 12.2 3.775 4.628 3.775 4.628 parallel_gemm_fm 81 9.0 0.000 0.000 4.436 4.461 parallel_gemm_fm_cosma 81 10.0 4.436 4.461 4.436 4.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.471000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.818182, yerr=6.520153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 832.880640E+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 6187102. 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 0.044 0.073 195.790 195.791 qs_mol_dyn_low 1 2.0 0.003 0.003 195.230 195.242 qs_forces 11 3.9 0.003 0.003 195.143 195.146 qs_energies 11 4.9 0.003 0.014 188.407 188.419 scf_env_do_scf 11 5.9 0.001 0.002 171.539 171.549 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 138.163 138.170 velocity_verlet 10 3.0 0.008 0.012 121.518 121.519 dbcsr_multiply_generic 2507 12.6 0.184 0.189 101.339 103.005 qs_scf_new_mos 117 7.6 0.001 0.001 98.387 98.803 qs_scf_loop_do_ot 117 8.6 0.001 0.002 98.386 98.802 ot_scf_mini 117 9.6 0.004 0.005 92.790 93.275 multiply_cannon 2507 13.6 0.479 0.534 79.384 83.382 multiply_cannon_loop 2507 14.6 1.252 1.297 75.783 79.599 ot_mini 117 10.6 0.001 0.001 52.639 53.100 mp_waitall_1 226760 16.4 26.852 40.109 26.852 40.109 multiply_cannon_multrec 30084 15.6 21.925 26.705 31.912 37.206 rebuild_ks_matrix 128 8.3 0.001 0.001 32.783 33.346 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.782 33.346 init_scf_loop 11 6.9 0.001 0.003 33.288 33.305 qs_ot_get_derivative 117 11.6 0.001 0.002 30.422 30.911 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.533 30.041 multiply_cannon_metrocomm3 30084 15.6 0.093 0.098 16.713 29.364 prepare_preconditioner 11 7.9 0.000 0.000 28.716 28.775 make_preconditioner 11 8.9 0.000 0.002 28.716 28.775 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.296 27.952 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 23.169 24.640 apply_single 128 13.6 0.001 0.001 23.169 24.640 qs_ot_get_p 128 10.4 0.001 0.002 21.827 22.405 ot_diis_step 117 11.6 0.014 0.015 22.024 22.026 multiply_cannon_sync_h2d 30084 15.6 19.208 21.729 19.208 21.729 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.812 16.846 cp_fm_cholesky_invert 11 10.9 16.599 16.612 16.599 16.612 make_m2s 5014 13.6 0.092 0.098 14.778 16.288 make_images 5014 14.6 1.169 1.354 14.577 16.085 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.675 15.676 sum_up_and_integrate 128 10.3 0.116 0.134 14.389 14.421 integrate_v_rspace 128 11.3 0.004 0.004 14.273 14.309 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.088 14.180 calculate_rho_elec 128 8.7 0.087 0.105 14.087 14.179 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.571 12.577 cp_fm_redistribute_end 83 14.4 7.304 12.496 7.325 12.507 cp_fm_diag_elpa_base 83 14.4 4.952 12.037 5.167 12.409 init_scf_run 11 5.9 0.000 0.001 12.036 12.037 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.035 12.037 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.950 11.318 make_images_data 5014 15.6 0.067 0.074 8.810 10.736 multiply_cannon_metrocomm4 27577 15.6 0.099 0.114 3.762 10.491 mp_irecv_dv 69486 16.3 3.567 10.095 3.567 10.095 dbcsr_mm_accdrv_process 62242 16.2 4.611 5.812 9.450 10.055 hybrid_alltoall_any 5200 16.5 0.344 1.509 7.471 9.973 wfi_extrapolate 11 7.9 0.001 0.001 8.302 8.303 density_rs2pw 128 9.7 0.006 0.007 7.400 7.921 calculate_dm_sparse 128 9.5 0.001 0.001 7.658 7.863 pw_transfer 1547 11.6 0.085 0.103 7.771 7.815 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.546 7.591 grid_integrate_task_list 128 12.3 7.160 7.590 7.160 7.590 mp_sum_l 7870 13.0 5.395 7.413 5.395 7.413 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.463 7.254 cp_fm_cholesky_decompose 22 10.9 6.905 6.980 6.905 6.980 rs_pw_transfer 1046 11.9 0.014 0.016 6.282 6.820 fft_wrap_pw1pw2_140 523 13.2 0.469 0.520 6.608 6.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.158 6.166 fft3d_ps 1291 14.7 2.780 2.909 5.911 5.946 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.656 5.915 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.293 5.476 grid_collocate_task_list 128 9.7 4.914 5.272 4.914 5.272 mp_alltoall_d11v 2415 14.1 4.651 5.094 4.651 5.094 mp_allgather_i34 2507 14.6 2.173 5.045 2.173 5.045 potential_pw2rs 128 12.3 0.015 0.017 4.817 4.846 dbcsr_complete_redistribute 395 12.7 0.770 0.867 3.278 4.339 mp_sum_d 4455 12.2 2.853 4.316 2.853 4.316 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.791000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.363636, yerr=3.471251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 939.999232E+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 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.137 178.192 178.193 qs_mol_dyn_low 1 2.0 0.003 0.003 177.442 177.455 qs_forces 11 3.9 0.003 0.003 177.354 177.358 qs_energies 11 4.9 0.008 0.037 170.868 170.881 scf_env_do_scf 11 5.9 0.001 0.001 152.562 152.566 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.967 117.969 velocity_verlet 10 3.0 0.002 0.002 113.378 113.381 dbcsr_multiply_generic 2485 12.5 0.177 0.182 81.547 82.809 qs_scf_new_mos 116 7.6 0.001 0.001 81.763 82.081 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.762 82.080 ot_scf_mini 116 9.6 0.004 0.004 77.660 78.025 multiply_cannon 2485 13.5 0.496 0.520 61.647 66.569 multiply_cannon_loop 2485 14.5 0.852 0.879 58.455 61.265 ot_mini 116 10.6 0.001 0.001 42.285 42.665 init_scf_loop 11 6.9 0.006 0.043 34.496 34.499 mp_waitall_1 176908 16.5 25.166 33.919 25.166 33.919 prepare_preconditioner 11 7.9 0.000 0.000 30.386 30.437 make_preconditioner 11 8.9 0.001 0.002 30.386 30.437 rebuild_ks_matrix 127 8.3 0.001 0.001 29.855 30.316 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 29.854 30.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.033 29.435 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.969 27.392 multiply_cannon_multrec 19880 15.5 13.439 16.492 22.105 25.126 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 15.158 24.337 qs_ot_get_derivative 116 11.6 0.001 0.002 22.769 23.120 qs_ot_get_p 127 10.4 0.001 0.001 20.910 21.354 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.619 20.596 apply_single 127 13.6 0.001 0.001 19.619 20.595 ot_diis_step 116 11.6 0.017 0.018 19.416 19.418 multiply_cannon_sync_h2d 19880 15.5 14.144 15.824 14.144 15.824 make_m2s 4970 13.5 0.082 0.086 14.617 15.748 qs_ot_p2m_diag 82 11.4 0.262 0.269 15.691 15.699 make_images 4970 14.5 1.183 1.286 14.391 15.523 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.698 14.704 cp_fm_cholesky_invert 11 10.9 14.161 14.170 14.161 14.170 sum_up_and_integrate 127 10.3 0.132 0.143 14.042 14.070 integrate_v_rspace 127 11.3 0.004 0.004 13.910 13.942 init_scf_run 11 5.9 0.000 0.001 13.190 13.191 scf_env_initial_rho_setup 11 6.9 0.005 0.028 13.190 13.190 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.964 12.990 calculate_rho_elec 127 8.7 0.130 0.145 12.963 12.989 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.599 11.600 cp_fm_redistribute_end 82 14.4 4.353 11.539 4.367 11.541 cp_fm_diag_elpa_base 82 14.4 6.749 10.969 7.151 11.456 make_images_data 4970 15.5 0.061 0.068 8.904 10.570 hybrid_alltoall_any 5155 16.4 0.425 1.926 7.810 9.764 wfi_extrapolate 11 7.9 0.001 0.001 9.612 9.612 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.841 9.110 multiply_cannon_metrocomm4 17395 15.5 0.060 0.070 3.395 8.983 mp_irecv_dv 49801 16.2 3.277 8.745 3.277 8.745 cp_fm_cholesky_decompose 22 10.9 8.481 8.519 8.481 8.519 dbcsr_mm_accdrv_process 41158 16.2 4.417 5.199 8.123 8.229 pw_transfer 1535 11.6 0.084 0.104 7.753 7.858 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.532 7.644 grid_integrate_task_list 127 12.3 7.240 7.626 7.240 7.626 cp_fm_upper_to_full 104 14.5 5.667 7.355 5.667 7.355 density_rs2pw 127 9.7 0.005 0.006 6.443 6.873 fft_wrap_pw1pw2_140 519 13.2 0.474 0.531 6.621 6.745 dbcsr_complete_redistribute 393 12.7 1.176 1.209 4.712 6.444 calculate_dm_sparse 127 9.5 0.001 0.001 6.108 6.214 fft3d_ps 1281 14.7 2.674 2.881 5.806 5.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.596 5.602 rs_pw_transfer 1038 11.9 0.013 0.014 4.976 5.449 grid_collocate_task_list 127 9.7 5.048 5.361 5.048 5.361 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.660 5.324 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.487 5.223 mp_sum_l 7804 13.0 3.484 5.132 3.484 5.132 mp_alltoall_d11v 2401 14.1 4.372 4.994 4.372 4.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.562 4.702 mp_allgather_i34 2485 14.5 1.724 4.580 1.724 4.580 potential_pw2rs 127 12.3 0.020 0.022 4.460 4.477 make_basis_sm 11 9.8 0.000 0.000 4.408 4.416 reorthogonalize_vectors 10 9.0 0.000 0.000 4.040 4.047 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.334 4.047 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.944 3.978 mp_alltoall_i22 712 14.1 1.981 3.825 1.981 3.825 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.742 3.745 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.193000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.636364, yerr=9.177767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.161204E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4063 57495. MP_Allreduce 11196 1069. MP_Sync 168 MP_Alltoall 1700 12496361. 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 0.072 0.132 189.096 189.099 qs_mol_dyn_low 1 2.0 0.003 0.003 188.486 188.500 qs_forces 11 3.9 0.003 0.003 188.387 188.395 qs_energies 11 4.9 0.002 0.003 180.963 180.975 scf_env_do_scf 11 5.9 0.001 0.001 164.070 164.083 velocity_verlet 10 3.0 0.002 0.002 124.706 124.711 scf_env_do_scf_inner_loop 116 6.6 0.029 0.211 115.150 115.180 dbcsr_multiply_generic 2485 12.5 0.185 0.190 79.530 80.110 qs_scf_new_mos 116 7.6 0.001 0.001 79.408 79.640 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.408 79.639 ot_scf_mini 116 9.6 0.003 0.004 74.940 75.207 multiply_cannon 2485 13.5 0.552 0.583 54.976 58.108 multiply_cannon_loop 2485 14.5 1.181 1.209 51.028 52.246 init_scf_loop 11 6.9 0.000 0.001 48.797 48.819 prepare_preconditioner 11 7.9 0.000 0.000 43.831 43.864 make_preconditioner 11 8.9 0.000 0.001 43.831 43.864 ot_mini 116 10.6 0.001 0.001 42.090 42.339 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.886 41.129 multiply_cannon_multrec 29820 15.5 14.298 19.536 26.354 30.999 rebuild_ks_matrix 127 8.3 0.001 0.001 28.823 28.963 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.823 28.963 mp_waitall_1 152434 16.5 17.747 27.263 17.747 27.263 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.986 26.116 qs_ot_get_derivative 116 11.6 0.001 0.002 22.750 23.006 make_m2s 4970 13.5 0.096 0.098 20.050 21.306 make_images 4970 14.5 1.925 2.186 19.749 21.003 qs_ot_get_p 127 10.4 0.001 0.002 19.576 19.891 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.773 19.223 apply_single 127 13.6 0.001 0.001 18.772 19.223 ot_diis_step 116 11.6 0.017 0.018 19.207 19.209 cp_fm_upper_to_full 104 14.7 11.343 16.912 11.343 16.912 cp_fm_cholesky_invert 11 10.9 16.026 16.035 16.026 16.035 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.205 15.257 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.263 15.148 dbcsr_complete_redistribute 393 12.7 1.520 1.666 10.898 14.823 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.955 13.956 sum_up_and_integrate 127 10.3 0.141 0.154 13.921 13.956 integrate_v_rspace 127 11.3 0.004 0.004 13.780 13.820 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.547 13.602 calculate_rho_elec 127 8.7 0.174 0.189 13.547 13.601 copy_fm_to_dbcsr 208 11.6 0.002 0.002 9.517 13.458 make_images_data 4970 15.5 0.065 0.068 10.706 12.622 multiply_cannon_sync_h2d 29820 15.5 11.635 12.442 11.635 12.442 dbcsr_mm_accdrv_process 61748 16.2 7.326 8.433 11.638 12.147 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 7.921 11.759 hybrid_alltoall_any 5155 16.4 0.514 2.163 9.443 11.525 init_scf_run 11 5.9 0.000 0.001 10.964 10.965 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.963 10.965 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.917 10.919 cp_fm_redistribute_end 82 14.4 1.861 10.849 1.877 10.854 cp_fm_diag_elpa_base 82 14.4 8.359 10.256 8.946 10.745 mp_alltoall_i22 712 14.1 5.830 9.798 5.830 9.798 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.307 9.494 pw_transfer 1535 11.6 0.084 0.097 7.881 7.975 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.657 7.757 grid_integrate_task_list 127 12.3 7.460 7.697 7.460 7.697 cp_fm_cholesky_decompose 22 10.9 7.465 7.529 7.465 7.529 wfi_extrapolate 11 7.9 0.001 0.001 7.489 7.489 multiply_cannon_metrocomm4 24850 15.5 0.075 0.086 2.637 7.272 density_rs2pw 127 9.7 0.005 0.006 6.565 7.118 mp_irecv_dv 75445 16.2 2.498 7.010 2.498 7.010 fft_wrap_pw1pw2_140 519 13.2 0.478 0.489 6.792 6.904 mp_alltoall_d11v 2401 14.1 6.254 6.681 6.254 6.681 calculate_dm_sparse 127 9.5 0.001 0.001 6.262 6.351 fft3d_ps 1281 14.7 2.776 2.893 5.898 5.974 grid_collocate_task_list 127 9.7 5.166 5.617 5.166 5.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.533 5.586 rs_pw_transfer 1038 11.9 0.013 0.015 4.877 5.410 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.526 4.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.449 4.503 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.466 4.468 potential_pw2rs 127 12.3 0.022 0.023 4.237 4.260 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.135 4.193 mp_allgather_i34 2485 14.5 2.066 4.080 2.066 4.080 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.099000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1094.454545, yerr=20.557519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.514750E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.060 173.169 173.170 qs_mol_dyn_low 1 2.0 0.003 0.003 172.178 172.190 qs_forces 11 3.9 0.014 0.033 171.966 171.971 qs_energies 11 4.9 0.001 0.002 164.677 164.695 scf_env_do_scf 11 5.9 0.001 0.001 147.760 147.773 velocity_verlet 10 3.0 0.002 0.002 113.770 113.781 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 111.890 111.891 dbcsr_multiply_generic 2507 12.6 0.190 0.204 76.128 76.575 qs_scf_new_mos 117 7.6 0.001 0.001 76.026 76.112 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.025 76.111 ot_scf_mini 117 9.6 0.004 0.004 71.625 71.707 multiply_cannon 2507 13.6 0.587 0.622 55.340 59.421 multiply_cannon_loop 2507 14.6 0.450 0.463 50.426 51.370 ot_mini 117 10.6 0.001 0.001 41.837 41.932 init_scf_loop 11 6.9 0.000 0.000 35.720 35.722 mp_waitall_1 129618 16.6 27.035 34.427 27.035 34.427 prepare_preconditioner 11 7.9 0.000 0.000 31.682 31.708 make_preconditioner 11 8.9 0.000 0.000 31.682 31.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.510 29.767 rebuild_ks_matrix 128 8.3 0.001 0.001 29.163 29.248 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.163 29.247 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.711 26.793 qs_ot_get_derivative 117 11.6 0.002 0.002 21.897 21.986 multiply_cannon_multrec 10028 15.6 10.535 14.685 17.860 20.945 ot_diis_step 117 11.6 0.020 0.021 19.873 19.873 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.431 19.736 apply_single 128 13.6 0.001 0.001 19.431 19.736 make_m2s 5014 13.6 0.068 0.071 16.916 19.215 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.295 19.183 make_images 5014 14.6 2.346 2.884 16.612 18.912 cp_fm_cholesky_invert 11 10.9 18.087 18.093 18.087 18.093 qs_ot_get_p 128 10.4 0.001 0.001 16.954 17.048 sum_up_and_integrate 128 10.3 0.181 0.192 14.398 14.450 integrate_v_rspace 128 11.3 0.004 0.005 14.216 14.272 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.333 13.349 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.941 12.979 calculate_rho_elec 128 8.7 0.259 0.270 12.940 12.979 make_images_data 5014 15.6 0.055 0.061 10.250 12.784 hybrid_alltoall_any 5200 16.5 0.812 3.583 9.870 12.278 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.217 12.218 multiply_cannon_sync_h2d 10028 15.6 11.617 12.064 11.617 12.064 init_scf_run 11 5.9 0.000 0.001 10.139 10.139 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.138 10.139 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.878 9.945 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.249 9.250 cp_fm_diag_elpa_base 83 14.4 9.004 9.080 9.242 9.242 grid_integrate_task_list 128 12.3 7.731 8.147 7.731 8.147 cp_fm_cholesky_decompose 22 10.9 7.884 7.978 7.884 7.978 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.985 7.802 dbcsr_mm_accdrv_process 20762 16.1 2.709 3.455 6.955 7.669 mp_allgather_i34 2507 14.6 3.119 7.656 3.119 7.656 pw_transfer 1547 11.6 0.083 0.091 7.618 7.633 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.397 7.418 wfi_extrapolate 11 7.9 0.001 0.001 7.166 7.167 fft_wrap_pw1pw2_140 523 13.2 0.503 0.532 6.527 6.553 calculate_dm_sparse 128 9.5 0.001 0.001 6.077 6.150 density_rs2pw 128 9.7 0.005 0.006 5.765 6.043 mp_alltoall_d11v 2415 14.1 5.093 5.736 5.093 5.736 grid_collocate_task_list 128 9.7 5.491 5.718 5.491 5.718 dbcsr_complete_redistribute 395 12.7 2.154 2.243 5.226 5.619 fft3d_ps 1291 14.7 2.724 2.800 5.551 5.576 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.228 5.240 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.156 5.163 rs_pw_transfer 1046 11.9 0.013 0.013 4.098 4.403 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.868 4.298 mp_irecv_dv 28860 15.9 1.833 4.234 1.833 4.234 potential_pw2rs 128 12.3 0.026 0.027 4.122 4.148 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.095 4.120 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.886 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.799 3.823 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.447 3.760 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.704 3.733 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.506 3.581 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.170000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.636364, yerr=51.802549 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.897555E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59299. MP_Allreduce 10975 1515. MP_Sync 86 MP_Alltoall 1700 36954339. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.293 0.525 293.707 293.707 qs_mol_dyn_low 1 2.0 0.015 0.098 292.428 292.456 qs_forces 11 3.9 0.026 0.027 292.104 292.106 qs_energies 11 4.9 0.005 0.027 283.488 283.500 scf_env_do_scf 11 5.9 0.001 0.001 261.424 261.437 velocity_verlet 10 3.0 0.014 0.059 209.730 209.744 scf_env_do_scf_inner_loop 116 6.6 0.014 0.087 135.282 135.284 init_scf_loop 11 6.9 0.000 0.000 125.869 125.873 prepare_preconditioner 11 7.9 0.000 0.000 121.206 121.237 make_preconditioner 11 8.9 0.000 0.000 121.206 121.237 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.561 118.305 qs_scf_new_mos 116 7.6 0.001 0.001 93.510 93.628 qs_scf_loop_do_ot 116 8.6 0.003 0.019 93.509 93.628 ot_scf_mini 116 9.6 0.004 0.004 88.722 88.748 dbcsr_multiply_generic 2485 12.5 0.209 0.214 85.090 85.634 cp_fm_upper_to_full 104 14.8 53.567 77.084 53.567 77.084 multiply_cannon 2485 13.5 0.702 0.761 59.563 60.431 multiply_cannon_loop 2485 14.5 0.476 0.510 55.878 57.184 ot_mini 116 10.6 0.002 0.010 46.518 46.568 dbcsr_complete_redistribute 393 12.7 4.007 4.070 30.370 43.663 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.918 40.228 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.599 37.814 mp_alltoall_i22 712 14.1 22.524 35.975 22.524 35.975 cp_fm_cholesky_invert 11 10.9 33.550 33.556 33.550 33.556 rebuild_ks_matrix 127 8.3 0.001 0.004 33.239 33.292 qs_ks_build_kohn_sham_matrix 127 9.3 0.024 0.072 33.238 33.291 mp_waitall_1 104546 16.7 29.546 32.832 29.546 32.832 qs_ks_update_qs_env 127 7.6 0.004 0.025 31.118 31.185 qs_ot_get_p 127 10.4 0.003 0.013 26.702 26.795 qs_ot_get_derivative 116 11.6 0.002 0.002 26.545 26.579 qs_ot_p2m_diag 82 11.4 0.869 0.878 22.632 22.661 make_m2s 4970 13.5 0.077 0.081 20.332 21.242 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.927 20.930 make_images 4970 14.5 3.738 3.885 19.861 20.772 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 19.464 20.487 ot_diis_step 116 11.6 0.022 0.024 19.925 19.925 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.142 19.356 apply_single 127 13.6 0.001 0.001 19.142 19.355 multiply_cannon_multrec 9940 15.5 10.543 12.345 18.136 18.244 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.762 17.764 cp_fm_diag_elpa_base 82 14.4 13.439 15.003 17.758 17.758 sum_up_and_integrate 127 10.3 0.319 0.320 15.534 15.620 multiply_cannon_sync_h2d 9940 15.5 15.513 15.526 15.513 15.526 integrate_v_rspace 127 11.3 0.004 0.007 15.215 15.301 qs_rho_update_rho_low 127 7.7 0.003 0.020 14.839 14.866 calculate_rho_elec 127 8.7 0.483 0.505 14.836 14.866 make_images_data 4970 15.5 0.060 0.063 11.016 12.930 hybrid_alltoall_any 5155 16.4 1.290 3.017 10.963 12.704 init_scf_run 11 5.9 0.000 0.001 12.037 12.037 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.037 12.037 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 11.066 11.131 cp_fm_cholesky_decompose 22 10.9 9.149 9.166 9.149 9.166 dbcsr_mm_accdrv_process 20590 16.0 3.692 5.538 7.359 9.122 wfi_extrapolate 11 7.9 0.001 0.001 8.696 8.696 grid_integrate_task_list 127 12.3 8.480 8.668 8.480 8.668 pw_transfer 1535 11.6 0.089 0.090 8.230 8.239 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.998 8.007 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.878 7.881 mp_alltoall_d11v 2401 14.1 6.971 7.099 6.971 7.099 fft_wrap_pw1pw2_140 519 13.2 0.534 0.539 7.080 7.089 calculate_dm_sparse 127 9.5 0.001 0.001 6.755 6.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.427 6.515 grid_collocate_task_list 127 9.7 6.293 6.330 6.293 6.330 copy_dbcsr_to_fm 185 11.7 0.006 0.017 6.130 6.195 fft3d_ps 1281 14.7 2.737 2.748 6.055 6.064 density_rs2pw 127 9.7 0.005 0.006 5.995 6.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.707000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2578.181818, yerr=172.640361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 1.261507E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.128 86.163 86.173 qs_energies 1 2.0 0.014 0.112 85.456 85.584 ls_scf 1 3.0 0.000 0.001 84.540 84.571 dbcsr_multiply_generic 111 6.7 0.016 0.016 73.514 73.680 multiply_cannon 111 7.7 0.017 0.020 56.161 57.485 multiply_cannon_loop 111 8.7 0.212 0.225 52.678 54.165 ls_scf_main 1 4.0 0.000 0.017 53.064 53.065 density_matrix_trs4 2 5.0 0.003 0.017 47.580 47.658 ls_scf_init_scf 1 4.0 0.001 0.002 28.238 28.240 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.165 27.221 mp_waitall_1 11316 10.9 23.138 25.794 23.138 25.794 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.045 25.071 multiply_cannon_multrec 2664 9.7 8.185 8.879 15.620 17.302 multiply_cannon_sync_h2d 2664 9.7 13.469 15.125 13.469 15.125 make_m2s 222 7.7 0.009 0.012 13.553 14.193 make_images 222 8.7 0.102 0.111 13.531 14.173 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.936 12.974 make_images_data 222 9.7 0.005 0.006 8.065 8.593 hybrid_alltoall_any 227 10.6 0.217 1.821 6.775 8.194 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.478 8.156 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.608 7.054 8.001 dbcsr_mm_accdrv_process_sort 4760 11.4 6.341 7.207 6.341 7.207 calculate_norms 4752 9.8 5.501 6.132 5.501 6.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.119 5.245 mp_sum_l 807 5.4 3.289 4.772 3.289 4.772 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.441 3.537 make_images_sizes 222 9.7 0.000 0.000 0.842 3.522 mp_alltoall_i44 222 10.7 0.842 3.522 0.842 3.522 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.081 3.397 mp_irecv_dv 6231 10.9 2.068 3.371 2.068 3.371 arnoldi_extremal 4 6.8 0.000 0.000 3.254 3.292 arnoldi_normal_ev 4 7.8 0.002 0.014 3.254 3.292 ls_scf_post 1 4.0 0.003 0.021 3.237 3.268 build_subspace 16 8.4 0.009 0.012 3.117 3.119 ls_scf_store_result 1 5.0 0.000 0.000 2.894 2.931 dbcsr_special_finalize 555 9.7 0.005 0.006 2.299 2.759 dbcsr_merge_single_wm 555 10.7 0.450 0.580 2.291 2.751 make_images_pack 222 9.7 2.198 2.621 2.200 2.623 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.308 2.570 dbcsr_sort_data 658 11.4 2.100 2.494 2.100 2.494 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.482 2.071 2.483 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.189 2.298 buffer_matrices_ensure_size 222 8.7 1.750 2.085 1.750 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.733 1.735 rebuild_ks_matrix 3 7.3 0.000 0.000 1.724 1.726 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.724 1.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.173000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 2.110747E+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 0.087 0.241 92.713 92.770 qs_energies 1 2.0 0.002 0.011 91.875 91.892 ls_scf 1 3.0 0.000 0.001 90.558 90.567 dbcsr_multiply_generic 111 6.7 0.015 0.017 76.420 76.851 multiply_cannon 111 7.7 0.028 0.039 53.715 57.676 ls_scf_main 1 4.0 0.000 0.008 56.043 56.044 multiply_cannon_loop 111 8.7 0.118 0.125 50.358 53.549 density_matrix_trs4 2 5.0 0.003 0.011 50.443 50.639 ls_scf_init_scf 1 4.0 0.000 0.001 31.027 31.028 mp_waitall_1 9246 10.9 21.312 30.371 21.312 30.371 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.639 29.739 multiply_cannon_multrec 1332 9.7 13.177 17.237 22.549 27.780 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.256 27.271 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.937 20.974 make_m2s 222 7.7 0.008 0.009 15.363 16.052 make_images 222 8.7 1.588 1.945 15.332 16.024 dbcsr_mm_accdrv_process 4041 10.4 0.254 0.471 8.980 10.554 dbcsr_mm_accdrv_process_sort 4041 11.4 8.571 10.110 8.571 10.110 make_images_data 222 9.7 0.004 0.005 8.750 9.712 mp_sum_l 807 5.4 6.107 9.683 6.107 9.683 hybrid_alltoall_any 227 10.6 0.526 2.467 8.080 9.213 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.212 7.869 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.853 7.839 mp_irecv_dv 3311 11.0 3.192 7.810 3.192 7.810 calculate_norms 2376 9.8 6.028 6.782 6.028 6.782 multiply_cannon_sync_h2d 1332 9.7 4.790 6.082 4.790 6.082 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.970 5.155 arnoldi_extremal 4 6.8 0.000 0.000 4.642 4.666 arnoldi_normal_ev 4 7.8 0.001 0.004 4.642 4.666 build_subspace 16 8.4 0.014 0.021 4.381 4.384 ls_scf_post 1 4.0 0.003 0.015 3.488 3.497 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.122 3.354 ls_scf_store_result 1 5.0 0.000 0.000 3.193 3.302 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.222 2.743 3.223 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.508 2.593 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.260 2.573 mp_allgather_i34 111 8.7 1.031 2.492 1.031 2.492 make_images_pack 222 9.7 2.025 2.441 2.027 2.443 dbcsr_sort_data 436 11.2 1.843 2.095 1.843 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.770000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1734.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 2.715996E+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 0.050 0.101 95.949 95.950 qs_energies 1 2.0 0.000 0.000 95.284 95.289 ls_scf 1 3.0 0.000 0.001 93.536 93.541 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.805 78.118 ls_scf_main 1 4.0 0.000 0.002 58.432 58.436 multiply_cannon 111 7.7 0.044 0.075 53.397 57.951 multiply_cannon_loop 111 8.7 0.101 0.107 49.808 53.833 density_matrix_trs4 2 5.0 0.003 0.009 52.501 52.745 mp_waitall_1 7374 11.0 24.559 34.274 24.559 34.274 ls_scf_init_scf 1 4.0 0.000 0.002 31.494 31.496 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.769 29.859 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.352 27.368 multiply_cannon_multrec 888 9.7 12.628 15.180 21.295 24.443 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.486 23.029 make_m2s 222 7.7 0.007 0.008 17.550 18.820 make_images 222 8.7 1.985 2.315 17.511 18.780 hybrid_alltoall_any 227 10.6 0.623 2.888 9.623 11.153 make_images_data 222 9.7 0.004 0.005 10.063 11.083 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.410 8.196 9.413 mp_sum_l 807 5.4 5.313 9.014 5.313 9.014 dbcsr_mm_accdrv_process_sort 3754 11.4 7.827 9.003 7.827 9.003 multiply_cannon_sync_h2d 888 9.7 5.996 7.814 5.996 7.814 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.485 7.268 mp_irecv_dv 2335 11.1 2.472 7.225 2.472 7.225 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.627 6.919 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.031 6.900 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.970 5.165 arnoldi_extremal 4 6.8 0.000 0.000 5.077 5.098 arnoldi_normal_ev 4 7.8 0.001 0.005 5.077 5.098 build_subspace 16 8.4 0.014 0.021 4.772 4.780 calculate_norms 1584 9.8 4.298 4.662 4.298 4.662 mp_allgather_i34 111 8.7 1.408 3.906 1.408 3.906 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.444 3.778 ls_scf_post 1 4.0 0.000 0.000 3.610 3.614 dbcsr_matrix_vector_mult_local 304 10.0 3.025 3.601 3.027 3.603 ls_scf_store_result 1 5.0 0.000 0.000 3.346 3.437 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.758 2.847 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.385 2.387 rebuild_ks_matrix 3 7.3 0.000 0.000 2.367 2.369 qs_ks_build_kohn_sham_matrix 3 8.3 0.053 0.425 2.367 2.369 make_images_sizes 222 9.7 0.000 0.000 1.124 2.259 mp_alltoall_i44 222 10.7 1.123 2.258 1.123 2.258 dbcsr_sort_data 325 11.1 1.907 2.140 1.907 2.140 make_images_pack 222 9.7 1.811 2.121 1.814 2.124 dbcsr_data_release 9322 10.9 1.321 1.973 1.321 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.950000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2174.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 3.336970E+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 0.046 0.074 99.551 99.553 qs_energies 1 2.0 0.002 0.016 98.825 98.837 ls_scf 1 3.0 0.000 0.001 96.757 96.760 dbcsr_multiply_generic 111 6.7 0.017 0.019 80.669 80.913 ls_scf_main 1 4.0 0.000 0.002 60.879 60.881 multiply_cannon 111 7.7 0.054 0.129 52.512 58.225 density_matrix_trs4 2 5.0 0.003 0.004 54.812 54.979 multiply_cannon_loop 111 8.7 0.116 0.128 46.571 49.232 ls_scf_init_scf 1 4.0 0.000 0.001 32.636 32.637 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.488 31.556 mp_waitall_1 6438 11.0 23.753 30.579 23.753 30.579 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.026 29.041 multiply_cannon_multrec 1332 9.7 14.264 17.082 22.286 25.006 make_m2s 222 7.7 0.008 0.009 22.355 24.374 make_images 222 8.7 3.146 3.600 22.304 24.323 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.186 17.152 make_images_data 222 9.7 0.004 0.005 13.030 15.295 hybrid_alltoall_any 227 10.6 0.800 3.756 11.525 14.827 dbcsr_mm_accdrv_process 3641 10.4 0.222 0.409 7.658 9.199 dbcsr_mm_accdrv_process_sort 3641 11.4 7.283 8.782 7.283 8.782 mp_sum_l 807 5.4 4.222 8.011 4.222 8.011 mp_allgather_i34 111 8.7 3.036 6.417 3.036 6.417 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.256 6.413 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.079 6.126 multiply_cannon_sync_h2d 1332 9.7 5.511 6.080 5.511 6.080 mp_irecv_dv 3229 10.9 2.061 6.047 2.061 6.047 arnoldi_extremal 4 6.8 0.000 0.000 5.204 5.217 arnoldi_normal_ev 4 7.8 0.001 0.005 5.204 5.217 build_subspace 16 8.4 0.015 0.021 4.870 4.879 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.563 4.738 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.370 4.523 calculate_norms 2376 9.8 4.189 4.506 4.189 4.506 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.587 3.893 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.706 3.204 3.708 dbcsr_sort_data 658 11.4 3.060 3.452 3.060 3.452 ls_scf_post 1 4.0 0.000 0.001 3.242 3.246 dbcsr_special_finalize 555 9.7 0.006 0.006 2.805 3.219 dbcsr_merge_single_wm 555 10.7 0.531 0.657 2.797 3.211 ls_scf_store_result 1 5.0 0.000 0.000 2.984 3.047 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.919 2.959 dbcsr_data_release 10477 10.7 1.591 2.372 1.591 2.372 qs_energies_init_hamiltonians 1 3.0 0.000 0.002 2.055 2.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.553000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2704.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 4.644532E+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 0.121 0.146 100.687 100.688 qs_energies 1 2.0 0.000 0.000 99.793 99.798 ls_scf 1 3.0 0.000 0.000 97.876 97.880 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.650 79.856 ls_scf_main 1 4.0 0.000 0.000 63.580 63.580 multiply_cannon 111 7.7 0.084 0.174 55.999 61.825 density_matrix_trs4 2 5.0 0.002 0.003 56.614 56.707 multiply_cannon_loop 111 8.7 0.070 0.077 51.243 53.006 mp_waitall_1 5481 11.0 27.472 34.083 27.472 34.083 ls_scf_init_scf 1 4.0 0.000 0.001 30.696 30.702 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.521 29.567 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.347 27.362 multiply_cannon_multrec 444 9.7 14.145 16.910 21.284 23.341 make_m2s 222 7.7 0.006 0.007 18.747 21.421 make_images 222 8.7 3.726 4.412 18.681 21.360 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.193 16.472 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.382 15.424 make_images_data 222 9.7 0.003 0.004 10.698 13.268 hybrid_alltoall_any 227 10.6 0.791 3.771 10.231 12.728 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.340 6.849 8.011 multiply_cannon_sync_h2d 444 9.7 6.516 7.916 6.516 7.916 dbcsr_mm_accdrv_process_sort 3003 11.4 6.530 7.646 6.530 7.646 mp_allgather_i34 111 8.7 2.971 7.458 2.971 7.458 arnoldi_extremal 4 6.8 0.000 0.000 5.792 5.808 arnoldi_normal_ev 4 7.8 0.002 0.005 5.792 5.808 build_subspace 16 8.4 0.015 0.020 5.401 5.416 mp_sum_l 807 5.4 2.953 5.175 2.953 5.175 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.594 4.754 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.187 4.379 make_images_sizes 222 9.7 0.000 0.000 1.379 4.221 mp_alltoall_i44 222 10.7 1.378 4.221 1.378 4.221 dbcsr_matrix_vector_mult_local 304 10.0 3.726 4.194 3.728 4.196 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.037 3.973 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.686 3.865 mp_irecv_dv 1241 11.2 1.668 3.844 1.668 3.844 calculate_norms 792 9.8 3.544 3.663 3.544 3.663 ls_scf_post 1 4.0 0.000 0.000 3.600 3.605 ls_scf_store_result 1 5.0 0.000 0.000 3.381 3.429 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.351 3.427 dbcsr_finalize 304 7.8 0.062 0.078 2.216 2.290 dbcsr_merge_all 275 8.9 0.475 0.526 2.059 2.098 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.688000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3592.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e58a972ed010720609b145a32ea323a5cbbe9a85_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 8.736195E+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 0.099 0.154 110.439 110.449 qs_energies 1 2.0 0.000 0.000 108.664 108.677 ls_scf 1 3.0 0.000 0.000 105.739 105.752 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.400 77.514 ls_scf_main 1 4.0 0.000 0.000 67.640 67.641 density_matrix_trs4 2 5.0 0.002 0.003 58.808 58.861 multiply_cannon 111 7.7 0.177 0.301 49.607 50.836 multiply_cannon_loop 111 8.7 0.068 0.071 46.298 47.415 ls_scf_init_scf 1 4.0 0.000 0.001 34.405 34.405 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.858 32.871 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.995 30.005 mp_waitall_1 4569 11.1 21.867 25.327 21.867 25.327 make_m2s 222 7.7 0.007 0.007 23.861 24.634 make_images 222 8.7 4.583 4.946 23.751 24.522 multiply_cannon_multrec 444 9.7 17.937 18.768 22.628 23.225 hybrid_alltoall_any 227 10.6 1.661 3.623 15.201 17.724 make_images_data 222 9.7 0.004 0.004 13.080 15.289 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.196 11.081 multiply_cannon_sync_h2d 444 9.7 8.847 8.893 8.847 8.893 arnoldi_extremal 4 6.8 0.000 0.000 7.460 7.475 arnoldi_normal_ev 4 7.8 0.003 0.009 7.459 7.474 build_subspace 16 8.4 0.026 0.036 6.899 6.910 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.539 5.702 dbcsr_matrix_vector_mult_local 304 10.0 5.106 5.415 5.109 5.417 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.068 5.315 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.792 4.888 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.321 4.521 4.643 dbcsr_mm_accdrv_process_sort 1814 11.4 4.216 4.346 4.216 4.346 mp_alltoall_d11v 48 9.2 4.263 4.340 4.263 4.340 ls_scf_post 1 4.0 0.000 0.000 3.694 3.707 dbcsr_desymmetrize_deep 5 6.8 0.523 0.536 3.561 3.577 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.448 make_images_sizes 222 9.7 0.000 0.000 1.437 3.444 mp_alltoall_i44 222 10.7 1.437 3.444 1.437 3.444 calculate_norms 792 9.8 3.238 3.271 3.238 3.271 dbcsr_finalize 304 7.8 0.083 0.090 3.092 3.145 dbcsr_merge_all 275 8.9 0.895 0.921 2.877 2.935 dbcsr_complete_redistribute 5 7.6 1.451 1.492 2.783 2.908 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.894 2.894 dbcsr_data_release 12724 10.6 2.327 2.833 2.327 2.833 mp_allgather_i34 111 8.7 0.922 2.685 0.922 2.685 matrix_ls_to_qs 2 6.0 0.000 0.000 2.418 2.559 dbcsr_sort_data 325 11.1 2.437 2.494 2.437 2.494 mp_sum_l 807 5.4 1.613 2.358 1.613 2.358 dbcsr_new_transposed 4 7.5 0.243 0.249 2.291 2.307 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.240 2.242 dbcsr_frobenius_norm 74 6.6 2.056 2.126 2.186 2.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.449000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6861.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e58a972ed010720609b145a32ea323a5cbbe9a85 Summary: empty Status: OK