=== 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: ed9166ef7988979478d40eb36b21c1c136d249d6 ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, 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 (03.02.2023) # # \ 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 -j${maxtasks} --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/01 job id: 45069474 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/02 job id: 45069475 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/03 job id: 45069476 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/04 job id: 45069477 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/05 job id: 45069479 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/06 job id: 45069480 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/07 job id: 45069481 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/08 job id: 45069482 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/09 job id: 45069483 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/10 job id: 45069484 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/11 job id: 45069485 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/12 job id: 45069488 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/13 job id: 45069490 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/14 job id: 45069491 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/15 job id: 45069492 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/16 job id: 45069493 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/17 job id: 45069494 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/18 job id: 45069495 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/19 job id: 45069496 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/20 job id: 45069497 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/21 job id: 45069498 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/22 job id: 45069499 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/23 job id: 45069500 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/24 job id: 45069501 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/25 job id: 45069502 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/26 job id: 45069503 --- 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/ed9166ef7988979478d40eb36b21c1c136d249d6_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 0.033 135.766 135.766 farming_run 1 2.0 135.247 135.248 135.737 135.740 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452593E+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 231 1098811. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 117.031 117.032 qs_energies 1 2.0 0.000 0.000 116.774 116.788 mp2_main 1 3.0 0.000 0.000 114.651 114.665 mp2_gpw_main 1 4.0 0.020 0.026 113.607 113.621 mp2_ri_gpw_compute_in 1 5.0 0.173 0.183 94.347 94.844 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.590 56.087 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.911 47.377 get_2c_integrals 1 6.0 0.001 0.010 37.591 38.584 integrate_v_rspace 273 8.0 0.434 0.449 25.195 30.347 pw_transfer 6555 10.6 0.368 0.390 27.614 28.432 fft_wrap_pw1pw2 5465 11.4 0.043 0.046 26.223 26.964 grid_integrate_task_list 273 9.0 20.997 26.645 20.997 26.645 fft_wrap_pw1pw2_100 2178 12.4 1.269 1.496 23.746 24.481 compute_2c_integrals 1 7.0 0.003 0.004 19.801 19.802 compute_2c_integrals_loop_lm 1 8.0 0.009 0.010 19.111 19.477 mp2_eri_2c_integrate_gpw 1 9.0 2.348 2.436 19.102 19.468 rpa_ri_compute_en 1 5.0 0.003 0.004 19.152 19.397 cp_fm_cholesky_decompose 12 8.2 17.736 18.759 17.736 18.759 cholesky_decomp 1 7.0 0.000 0.000 16.632 17.651 fft3d_s 5443 13.4 16.248 16.775 16.270 16.796 ao_to_mo_and_store_B_mult_1 272 7.0 10.858 15.590 10.858 15.590 calculate_wavefunction 272 8.0 5.482 5.627 12.629 13.214 rpa_num_int 1 6.0 0.002 0.027 10.674 10.690 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.540 10.598 calc_mat_Q 8 8.0 0.000 0.000 9.394 9.540 contract_S_to_Q 8 9.0 0.000 0.000 8.816 8.960 calc_potential_gpw 544 9.5 0.005 0.005 8.343 8.692 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.359 8.595 potential_pw2rs 545 10.0 0.107 0.109 7.764 8.500 parallel_gemm_fm 14 9.1 0.000 0.000 8.402 8.486 parallel_gemm_fm_cosma 14 10.1 8.402 8.486 8.402 8.486 create_integ_mat 1 6.0 0.006 0.008 7.884 7.884 collocate_single_gaussian 272 10.0 0.039 0.043 7.572 7.769 array2fm 1 7.0 0.000 0.000 6.807 7.356 pw_scatter_s 2720 13.7 4.411 4.650 4.411 4.650 pw_gather_s 2722 13.2 3.881 4.323 3.881 4.323 array2fm_buffer_send 1 8.0 2.941 3.170 2.941 3.170 pw_poisson_solve 545 10.5 1.129 1.171 2.233 2.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.621010, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 0.029 0.040 397.462 397.464 farming_run 1 2.0 396.588 396.606 397.420 397.424 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227690E+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 744 385880. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.043 210.127 210.128 qs_energies 1 2.0 0.003 0.012 209.859 209.877 scf_env_do_scf 1 3.0 0.000 0.000 106.425 106.425 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.532 105.541 rebuild_ks_matrix 4 6.0 0.000 0.000 105.531 105.540 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.067 105.531 105.540 hfx_ks_matrix 4 8.0 0.001 0.001 105.138 105.141 integrate_four_center 4 9.0 0.143 0.458 105.137 105.141 mp2_main 1 3.0 0.006 0.032 103.137 103.148 mp2_gpw_main 1 4.0 0.052 0.125 102.222 102.235 integrate_four_center_main 4 10.0 0.084 0.430 96.708 99.102 integrate_four_center_bin 261 11.0 96.624 99.091 96.624 99.091 init_scf_loop 1 4.0 0.000 0.000 92.214 92.214 mp2_ri_gpw_compute_in 1 5.0 0.067 0.091 74.878 76.032 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.558 55.710 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 42.197 47.489 integrate_v_rspace 95 8.0 0.396 0.563 28.572 33.687 pw_transfer 2240 10.6 0.144 0.157 29.917 30.374 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.946 29.443 grid_integrate_task_list 95 9.0 23.856 29.159 23.856 29.159 mp2_ri_gpw_compute_en 1 5.0 0.063 0.125 27.132 28.901 ao_to_mo_and_store_B_mult_1 91 7.0 10.668 28.281 10.668 28.281 fft_wrap_pw1pw2_100 730 12.4 1.256 1.416 26.646 27.201 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.826 1.879 25.221 25.233 get_2c_integrals 1 6.0 0.001 0.007 20.225 20.257 compute_2c_integrals 1 7.0 0.006 0.012 19.199 19.218 compute_2c_integrals_loop_lm 1 8.0 0.008 0.020 18.833 19.051 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.837 18.825 19.050 fft3d_s 1823 13.4 18.412 18.931 18.425 18.944 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.209 14.209 calculate_wavefunction 91 8.0 2.023 2.057 9.745 9.963 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.585 8.877 9.486 potential_pw2rs 186 10.0 0.033 0.035 8.639 9.034 local_gemm 172 8.0 8.318 8.911 8.318 8.911 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.529 8.135 8.649 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.215 8.500 calc_potential_gpw 182 9.5 0.002 0.002 7.892 8.134 collocate_single_gaussian 91 10.0 0.018 0.055 7.861 8.118 mp_sendrecv_dm3 2068 8.0 6.166 6.685 6.166 6.685 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.456 6.339 6.456 mp_sync 37 10.5 3.095 5.998 3.095 5.998 pw_gather_s 912 13.2 4.947 5.262 4.947 5.262 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.236489, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 451.448832E+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 272. MP_Sync 530 MP_Alltoall 2083 578641. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 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 0.052 0.110 55.286 55.290 qs_mol_dyn_low 1 2.0 0.008 0.018 54.748 54.758 qs_forces 11 3.9 0.011 0.052 54.622 54.624 qs_energies 11 4.9 0.037 0.082 52.991 53.012 scf_env_do_scf 11 5.9 0.005 0.023 46.360 46.361 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 44.148 44.149 dbcsr_multiply_generic 2286 12.5 0.094 0.103 33.770 34.357 qs_scf_new_mos 108 7.5 0.000 0.001 33.837 34.106 qs_scf_loop_do_ot 108 8.5 0.001 0.002 33.836 34.106 ot_scf_mini 108 9.5 0.002 0.003 32.184 32.365 multiply_cannon 2286 13.5 0.188 0.209 26.421 28.220 multiply_cannon_loop 2286 14.5 1.452 1.548 25.703 27.484 velocity_verlet 10 3.0 0.002 0.011 27.030 27.038 ot_mini 108 10.5 0.001 0.001 19.542 19.807 qs_ot_get_derivative 108 11.5 0.001 0.001 16.533 16.703 mp_waitall_1 245248 16.5 8.747 15.280 8.747 15.280 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.200 13.155 multiply_cannon_multrec 54864 15.5 4.264 6.705 7.730 11.237 qs_ot_get_p 119 10.4 0.001 0.002 8.119 8.403 rebuild_ks_matrix 119 8.3 0.000 0.000 8.092 8.280 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 8.092 8.280 multiply_cannon_sync_h2d 54864 15.5 6.001 7.502 6.001 7.502 mp_sum_l 7207 12.9 5.388 7.418 5.388 7.418 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.077 7.247 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.104 6.543 qs_ot_p2m_diag 50 11.0 0.004 0.008 5.152 5.198 qs_ot_get_derivative_diag 49 12.0 0.002 0.010 5.090 5.192 init_scf_run 11 5.9 0.000 0.001 5.111 5.111 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.110 5.111 dbcsr_mm_accdrv_process 76910 16.1 1.113 1.812 3.386 4.792 sum_up_and_integrate 119 10.3 0.012 0.015 4.609 4.639 integrate_v_rspace 119 11.3 0.003 0.003 4.597 4.628 cp_dbcsr_syevd 50 12.0 0.003 0.007 4.444 4.445 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.344 4.431 calculate_rho_elec 119 8.7 0.012 0.017 4.343 4.430 cp_fm_diag_elpa 50 13.0 0.000 0.001 4.229 4.229 cp_fm_redistribute_end 50 14.0 2.156 4.199 2.162 4.202 cp_fm_diag_elpa_base 50 14.0 2.034 4.095 2.038 4.104 multiply_cannon_metrocomm1 54864 15.5 0.054 0.060 1.851 3.185 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.915 3.151 apply_single 119 13.6 0.000 0.000 2.915 3.151 calculate_dm_sparse 119 9.5 0.000 0.001 2.999 3.134 rs_pw_transfer 974 11.9 0.012 0.014 2.823 2.972 jit_kernel_multiply 13 15.8 2.213 2.889 2.213 2.889 calculate_first_density_matrix 1 7.0 0.003 0.009 2.851 2.868 ot_diis_step 108 11.5 0.006 0.006 2.769 2.770 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.472 2.476 density_rs2pw 119 9.7 0.004 0.006 2.305 2.445 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.349 2.404 acc_transpose_blocks 54864 15.5 0.220 0.247 1.755 2.254 init_scf_loop 11 6.9 0.003 0.010 2.175 2.178 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.127 2.171 wfi_extrapolate 11 7.9 0.003 0.018 2.163 2.163 grid_integrate_task_list 119 12.3 2.012 2.115 2.012 2.115 mp_sum_d 4131 12.0 1.408 2.074 1.408 2.074 pw_transfer 1439 11.6 0.053 0.059 1.941 2.034 potential_pw2rs 119 12.3 0.004 0.004 1.963 1.996 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.865 1.960 make_m2s 4572 13.5 0.054 0.057 1.686 1.725 fft3d_ps 1201 14.6 0.376 0.480 1.633 1.722 make_images 4572 14.5 0.133 0.140 1.603 1.643 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.531 1.573 mp_alltoall_d11v 2130 13.8 1.365 1.558 1.365 1.558 fft_wrap_pw1pw2_140 487 13.2 0.083 0.097 1.430 1.526 mp_waitany 12084 13.8 1.298 1.488 1.298 1.488 grid_collocate_task_list 119 9.7 1.294 1.361 1.294 1.361 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.789 1.249 acc_transpose_blocks_kernels 54864 16.5 0.239 0.375 0.819 1.119 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.290000, 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/ed9166ef7988979478d40eb36b21c1c136d249d6_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 488.099840E+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 54 MP_Alltoall 2060 1031878. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 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 0.023 0.042 39.216 39.218 qs_mol_dyn_low 1 2.0 0.003 0.003 38.958 38.965 qs_forces 11 3.9 0.002 0.003 38.894 38.894 qs_energies 11 4.9 0.003 0.007 37.198 37.201 scf_env_do_scf 11 5.9 0.001 0.001 31.780 31.781 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 29.305 29.307 dbcsr_multiply_generic 2286 12.5 0.100 0.102 21.519 21.877 qs_scf_new_mos 108 7.5 0.001 0.001 20.021 20.272 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.021 20.271 ot_scf_mini 108 9.5 0.003 0.008 19.139 19.312 velocity_verlet 10 3.0 0.001 0.002 18.245 18.246 multiply_cannon 2286 13.5 0.207 0.218 16.576 18.121 multiply_cannon_loop 2286 14.5 0.900 0.977 15.495 16.818 ot_mini 108 10.5 0.001 0.001 11.733 11.970 mp_waitall_1 200699 16.5 5.640 11.028 5.640 11.028 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.204 9.591 qs_ot_get_derivative 108 11.5 0.001 0.001 9.281 9.457 multiply_cannon_multrec 27432 15.5 1.987 4.326 6.129 8.902 rebuild_ks_matrix 119 8.3 0.000 0.000 7.282 7.421 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 7.282 7.421 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.437 6.562 dbcsr_mm_accdrv_process 47894 16.0 3.124 5.317 4.073 5.993 qs_ot_get_p 119 10.4 0.001 0.002 4.549 4.786 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.614 4.454 sum_up_and_integrate 119 10.3 0.025 0.027 4.232 4.255 integrate_v_rspace 119 11.3 0.002 0.003 4.208 4.227 init_scf_run 11 5.9 0.000 0.001 4.165 4.166 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.165 4.166 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.036 4.110 apply_single 119 13.6 0.000 0.000 3.036 4.110 mp_sum_l 7207 12.9 2.058 4.017 2.058 4.017 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.862 3.892 calculate_rho_elec 119 8.7 0.021 0.024 3.861 3.892 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.094 3.112 rs_pw_transfer 974 11.9 0.010 0.012 2.621 3.059 multiply_cannon_sync_h2d 27432 15.5 2.206 2.799 2.206 2.799 make_m2s 4572 13.5 0.052 0.054 2.448 2.662 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.660 2.660 calculate_first_density_matrix 1 7.0 0.001 0.003 2.622 2.625 density_rs2pw 119 9.7 0.004 0.005 2.184 2.623 make_images 4572 14.5 0.199 0.233 2.360 2.573 init_scf_loop 11 6.9 0.001 0.005 2.454 2.457 ot_diis_step 108 11.5 0.011 0.011 2.403 2.404 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.329 2.329 cp_fm_redistribute_end 50 14.0 1.179 2.299 1.183 2.302 calculate_dm_sparse 119 9.5 0.000 0.001 2.203 2.282 cp_fm_diag_elpa_base 50 14.0 1.085 2.197 1.114 2.244 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.120 2.211 pw_transfer 1439 11.6 0.065 0.072 2.041 2.076 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.000 2.002 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.949 1.987 jit_kernel_multiply 9 16.0 0.897 1.980 0.897 1.980 potential_pw2rs 119 12.3 0.006 0.006 1.957 1.967 grid_integrate_task_list 119 12.3 1.835 1.927 1.835 1.927 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.799 1.839 fft3d_ps 1201 14.6 0.517 0.574 1.649 1.683 acc_transpose_blocks 27432 15.5 0.109 0.113 1.237 1.560 fft_wrap_pw1pw2_140 487 13.2 0.080 0.087 1.514 1.553 prepare_preconditioner 11 7.9 0.000 0.001 1.517 1.543 make_preconditioner 11 8.9 0.001 0.005 1.517 1.543 make_images_data 4572 15.5 0.045 0.052 1.123 1.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.416 1.471 wfi_extrapolate 11 7.9 0.001 0.001 1.462 1.462 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.976 1.445 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.430 1.438 grid_collocate_task_list 119 9.7 1.231 1.379 1.231 1.379 mp_alltoall_d11v 2130 13.8 1.172 1.299 1.172 1.299 mp_allgather_i34 2286 14.5 0.522 1.260 0.522 1.260 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.155 1.204 mp_sum_d 4131 12.0 0.595 1.049 0.595 1.049 mp_waitany 5720 13.7 0.529 1.014 0.529 1.014 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.146 0.557 0.985 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.973 0.973 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.706 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.917 mp_alltoall_z22v 1201 16.6 0.714 0.818 0.714 0.818 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.609 0.746 0.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.218000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 523.751424E+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 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 666959. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.015 0.030 32.329 32.330 qs_mol_dyn_low 1 2.0 0.003 0.004 32.038 32.046 qs_forces 11 3.9 0.003 0.003 31.972 31.973 qs_energies 11 4.9 0.001 0.002 30.399 30.402 scf_env_do_scf 11 5.9 0.001 0.002 25.560 25.560 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.057 23.058 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.395 16.470 velocity_verlet 10 3.0 0.001 0.001 15.266 15.268 qs_scf_new_mos 108 7.5 0.001 0.001 14.776 14.796 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.775 14.796 ot_scf_mini 108 9.5 0.002 0.002 14.062 14.074 multiply_cannon 2286 13.5 0.196 0.203 13.182 13.996 multiply_cannon_loop 2286 14.5 0.635 0.665 12.405 13.235 ot_mini 108 10.5 0.001 0.001 8.635 8.650 multiply_cannon_multrec 18288 15.5 1.927 2.895 6.909 7.245 qs_ot_get_derivative 108 11.5 0.001 0.001 7.148 7.161 rebuild_ks_matrix 119 8.3 0.000 0.000 6.425 6.440 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.425 6.440 dbcsr_mm_accdrv_process 38222 16.0 4.254 5.530 4.897 5.735 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.674 5.687 sum_up_and_integrate 119 10.3 0.030 0.031 3.996 4.003 integrate_v_rspace 119 11.3 0.003 0.003 3.966 3.976 init_scf_run 11 5.9 0.000 0.001 3.669 3.669 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.669 3.669 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.514 3.518 calculate_rho_elec 119 8.7 0.031 0.032 3.514 3.518 mp_waitall_1 158411 16.6 2.545 3.449 2.545 3.449 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.699 3.335 qs_ot_get_p 119 10.4 0.001 0.001 3.273 3.297 rs_pw_transfer 974 11.9 0.009 0.010 2.287 2.490 init_scf_loop 11 6.9 0.000 0.000 2.486 2.486 calculate_first_density_matrix 1 7.0 0.000 0.001 2.423 2.423 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.043 2.351 apply_single 119 13.6 0.000 0.000 2.043 2.351 density_rs2pw 119 9.7 0.004 0.004 2.029 2.256 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.216 2.221 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.362 2.090 pw_transfer 1439 11.6 0.065 0.071 1.988 1.999 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.952 1.952 calculate_dm_sparse 119 9.5 0.000 0.000 1.934 1.948 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.895 1.908 grid_integrate_task_list 119 12.3 1.808 1.903 1.808 1.903 make_m2s 4572 13.5 0.044 0.045 1.760 1.898 make_images 4572 14.5 0.189 0.200 1.674 1.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.761 1.763 potential_pw2rs 119 12.3 0.007 0.008 1.753 1.759 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.702 1.712 cp_fm_diag_elpa_base 50 14.0 1.679 1.690 1.700 1.710 prepare_preconditioner 11 7.9 0.000 0.000 1.694 1.697 make_preconditioner 11 8.9 0.000 0.000 1.694 1.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.548 1.633 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.594 1.611 jit_kernel_multiply 8 16.3 0.591 1.608 0.591 1.608 fft3d_ps 1201 14.6 0.527 0.545 1.574 1.587 multiply_cannon_sync_h2d 18288 15.5 1.415 1.547 1.415 1.547 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.526 1.533 fft_wrap_pw1pw2_140 487 13.2 0.091 0.094 1.519 1.531 ot_diis_step 108 11.5 0.011 0.011 1.468 1.469 mp_sum_l 7207 12.9 1.064 1.370 1.064 1.370 grid_collocate_task_list 119 9.7 1.214 1.352 1.214 1.352 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.313 1.319 acc_transpose_blocks 18288 15.5 0.076 0.078 1.283 1.317 wfi_extrapolate 11 7.9 0.001 0.001 1.197 1.197 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.946 0.947 make_images_data 4572 15.5 0.045 0.049 0.789 0.942 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.863 0.887 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.843 0.864 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.674 0.849 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.823 mp_alltoall_d11v 2130 13.8 0.654 0.758 0.654 0.758 mp_alltoall_z22v 1201 16.6 0.652 0.749 0.652 0.749 mp_waitany 9880 13.7 0.514 0.740 0.514 0.740 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.728 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.505 0.708 cp_fm_cholesky_invert 11 10.9 0.703 0.707 0.703 0.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.330000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.181818, yerr=2.724241 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 546.766848E+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 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.047 0.107 35.860 35.893 qs_mol_dyn_low 1 2.0 0.003 0.005 35.465 35.475 qs_forces 11 3.9 0.002 0.003 35.337 35.339 qs_energies 11 4.9 0.003 0.013 33.611 33.616 scf_env_do_scf 11 5.9 0.001 0.004 28.347 28.348 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.956 24.957 dbcsr_multiply_generic 2286 12.5 0.100 0.103 18.243 18.359 velocity_verlet 10 3.0 0.001 0.001 17.802 17.806 qs_scf_new_mos 108 7.5 0.001 0.001 16.360 16.409 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.359 16.409 multiply_cannon 2286 13.5 0.226 0.266 14.710 15.564 ot_scf_mini 108 9.5 0.002 0.003 15.432 15.481 multiply_cannon_loop 2286 14.5 0.934 0.966 13.772 14.527 ot_mini 108 10.5 0.001 0.001 9.290 9.349 multiply_cannon_multrec 27432 15.5 2.396 3.112 8.827 9.237 dbcsr_mm_accdrv_process 47916 15.9 5.213 7.318 6.337 7.603 qs_ot_get_derivative 108 11.5 0.001 0.001 7.486 7.537 rebuild_ks_matrix 119 8.3 0.000 0.000 6.673 6.732 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.673 6.732 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.930 5.983 sum_up_and_integrate 119 10.3 0.035 0.038 3.885 3.892 init_scf_run 11 5.9 0.000 0.001 3.876 3.876 scf_env_initial_rho_setup 11 6.9 0.000 0.002 3.875 3.876 integrate_v_rspace 119 11.3 0.003 0.003 3.850 3.857 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.625 3.654 calculate_rho_elec 119 8.7 0.040 0.046 3.625 3.654 qs_ot_get_p 119 10.4 0.001 0.001 3.534 3.604 init_scf_loop 11 6.9 0.000 0.003 3.369 3.370 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.792 3.161 prepare_preconditioner 11 7.9 0.000 0.000 2.509 2.518 make_preconditioner 11 8.9 0.000 0.001 2.509 2.518 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.998 2.513 apply_single 119 13.6 0.000 0.000 1.998 2.513 calculate_first_density_matrix 1 7.0 0.002 0.007 2.485 2.487 make_full_inverse_cholesky 11 9.9 0.000 0.002 2.126 2.445 mp_waitall_1 137007 16.6 1.823 2.322 1.823 2.322 make_m2s 4572 13.5 0.054 0.056 2.169 2.285 rs_pw_transfer 974 11.9 0.009 0.010 2.073 2.262 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.245 2.254 density_rs2pw 119 9.7 0.004 0.004 2.040 2.227 pw_transfer 1439 11.6 0.066 0.072 2.151 2.191 calculate_dm_sparse 119 9.5 0.000 0.000 2.135 2.190 make_images 4572 14.5 0.269 0.328 2.062 2.178 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.058 2.102 jit_kernel_multiply 10 15.9 1.063 2.026 1.063 2.026 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.915 1.941 grid_integrate_task_list 119 12.3 1.818 1.928 1.818 1.928 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.928 1.928 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.879 1.880 fft3d_ps 1201 14.6 0.562 0.614 1.732 1.767 ot_diis_step 108 11.5 0.012 0.012 1.763 1.763 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.690 1.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.705 1.716 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.664 1.672 cp_fm_diag_elpa_base 50 14.0 1.626 1.642 1.659 1.668 potential_pw2rs 119 12.3 0.008 0.010 1.632 1.637 mp_sum_l 7207 12.9 0.951 1.584 0.951 1.584 acc_transpose_blocks 27432 15.5 0.113 0.114 1.518 1.550 multiply_cannon_metrocomm3 27432 15.5 0.037 0.039 0.814 1.426 grid_collocate_task_list 119 9.7 1.221 1.334 1.221 1.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.333 wfi_extrapolate 11 7.9 0.001 0.002 1.331 1.331 cp_fm_upper_to_full 72 14.2 0.808 1.148 0.808 1.148 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.114 1.115 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.095 1.113 multiply_cannon_sync_h2d 27432 15.5 1.001 1.071 1.001 1.071 dbcsr_complete_redistribute 329 12.2 0.122 0.149 0.757 1.024 make_images_data 4572 15.5 0.045 0.050 0.844 0.972 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.894 0.913 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.727 0.910 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.873 mp_alltoall_d11v 2130 13.8 0.728 0.867 0.728 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.833 0.838 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.822 mp_alltoall_z22v 1201 16.6 0.752 0.791 0.752 0.791 cp_fm_cholesky_invert 11 10.9 0.737 0.740 0.737 0.740 mp_alltoall_i22 627 13.8 0.425 0.722 0.425 0.722 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.893000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.909091, yerr=3.260077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 603.844608E+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 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.057 28.510 28.515 qs_mol_dyn_low 1 2.0 0.003 0.003 28.238 28.266 qs_forces 11 3.9 0.003 0.003 28.170 28.172 qs_energies 11 4.9 0.001 0.001 26.446 26.449 scf_env_do_scf 11 5.9 0.001 0.002 21.644 21.644 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.148 19.149 velocity_verlet 10 3.0 0.001 0.001 14.458 14.462 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.324 12.446 qs_scf_new_mos 108 7.5 0.001 0.001 11.061 11.085 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.061 11.084 ot_scf_mini 108 9.5 0.002 0.002 10.398 10.422 multiply_cannon 2286 13.5 0.227 0.238 9.809 10.328 multiply_cannon_loop 2286 14.5 0.329 0.341 8.918 9.150 multiply_cannon_multrec 9144 15.5 1.647 2.040 5.970 6.326 rebuild_ks_matrix 119 8.3 0.000 0.000 6.036 6.060 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.036 6.059 ot_mini 108 10.5 0.001 0.001 5.902 5.932 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.378 5.399 qs_ot_get_derivative 108 11.5 0.001 0.001 4.630 4.656 dbcsr_mm_accdrv_process 12550 15.8 3.175 4.267 4.221 4.308 sum_up_and_integrate 119 10.3 0.038 0.041 3.704 3.708 integrate_v_rspace 119 11.3 0.003 0.003 3.666 3.670 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.588 3.594 calculate_rho_elec 119 8.7 0.060 0.061 3.588 3.593 init_scf_run 11 5.9 0.000 0.001 3.366 3.366 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.365 3.365 qs_ot_get_p 119 10.4 0.001 0.001 2.743 2.781 init_scf_loop 11 6.9 0.000 0.000 2.473 2.474 calculate_first_density_matrix 1 7.0 0.001 0.001 2.217 2.217 pw_transfer 1439 11.6 0.066 0.069 2.199 2.211 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.105 2.119 density_rs2pw 119 9.7 0.004 0.004 1.937 2.073 mp_waitall_1 115863 16.7 1.497 2.001 1.497 2.001 grid_integrate_task_list 119 12.3 1.848 1.942 1.848 1.942 jit_kernel_multiply 10 15.7 1.008 1.930 1.008 1.930 make_m2s 4572 13.5 0.034 0.035 1.730 1.885 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.812 1.815 rs_pw_transfer 974 11.9 0.008 0.009 1.690 1.807 make_images 4572 14.5 0.267 0.299 1.641 1.794 calculate_dm_sparse 119 9.5 0.000 0.000 1.756 1.776 fft3d_ps 1201 14.6 0.565 0.577 1.764 1.775 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.733 1.745 prepare_preconditioner 11 7.9 0.000 0.000 1.725 1.729 make_preconditioner 11 8.9 0.000 0.000 1.725 1.729 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.612 1.640 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.616 1.616 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.608 1.609 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.525 1.538 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.457 1.465 potential_pw2rs 119 12.3 0.010 0.011 1.433 1.437 grid_collocate_task_list 119 9.7 1.272 1.387 1.272 1.387 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.363 1.371 cp_fm_diag_elpa_base 50 14.0 1.335 1.354 1.361 1.370 ot_diis_step 108 11.5 0.013 0.013 1.259 1.259 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.223 1.225 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.186 1.191 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.133 1.154 apply_single 119 13.6 0.000 0.000 1.132 1.154 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.129 1.141 wfi_extrapolate 11 7.9 0.001 0.001 1.101 1.101 make_images_data 4572 15.5 0.039 0.042 0.798 0.977 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.744 0.976 acc_transpose_blocks 9144 15.5 0.038 0.039 0.918 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.920 cp_fm_cholesky_invert 11 10.9 0.873 0.876 0.873 0.876 mp_alltoall_d11v 2130 13.8 0.712 0.849 0.712 0.849 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.784 multiply_cannon_sync_h2d 9144 15.5 0.710 0.779 0.710 0.779 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.398 0.757 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.677 0.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.712 0.721 mp_alltoall_z22v 1201 16.6 0.645 0.701 0.645 0.701 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.678 0.681 yz_to_x 606 15.1 0.265 0.277 0.604 0.621 mp_allgather_i34 2286 14.5 0.223 0.619 0.223 0.619 x_to_yz 595 16.2 0.276 0.290 0.582 0.588 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.515000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.000000, yerr=6.701425 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 747.945984E+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_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 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 0.041 0.071 43.600 43.605 qs_mol_dyn_low 1 2.0 0.003 0.004 43.294 43.306 qs_forces 11 3.9 0.002 0.003 43.178 43.181 qs_energies 11 4.9 0.001 0.002 41.197 41.199 scf_env_do_scf 11 5.9 0.002 0.002 33.536 33.536 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.656 25.657 velocity_verlet 10 3.0 0.001 0.001 23.423 23.428 dbcsr_multiply_generic 2286 12.5 0.100 0.101 19.395 19.585 multiply_cannon 2286 13.5 0.300 0.312 15.640 16.549 qs_scf_new_mos 108 7.5 0.001 0.001 15.803 15.900 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.802 15.899 multiply_cannon_loop 2286 14.5 0.343 0.349 14.398 15.356 ot_scf_mini 108 9.5 0.002 0.002 14.723 14.826 ot_mini 108 10.5 0.001 0.001 8.867 8.986 multiply_cannon_multrec 9144 15.5 3.481 4.914 8.755 8.879 init_scf_loop 11 6.9 0.000 0.000 7.852 7.853 rebuild_ks_matrix 119 8.3 0.000 0.001 7.356 7.504 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.356 7.504 qs_ot_get_derivative 108 11.5 0.001 0.001 6.859 6.962 prepare_preconditioner 11 7.9 0.000 0.000 6.873 6.886 make_preconditioner 11 8.9 0.000 0.000 6.873 6.886 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.653 6.787 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.416 6.762 dbcsr_mm_accdrv_process 12550 15.8 4.096 5.640 5.151 6.473 init_scf_run 11 5.9 0.000 0.001 5.589 5.589 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.589 5.589 cp_fm_upper_to_full 72 14.2 3.211 4.607 3.211 4.607 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.274 4.280 calculate_rho_elec 119 8.7 0.118 0.121 4.273 4.280 calculate_first_density_matrix 1 7.0 0.001 0.001 4.143 4.145 sum_up_and_integrate 119 10.3 0.065 0.066 4.046 4.053 integrate_v_rspace 119 11.3 0.004 0.004 3.981 3.988 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.786 3.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.486 3.500 mp_waitall_1 94719 16.7 2.302 3.351 2.302 3.351 qs_ot_get_p 119 10.4 0.001 0.001 3.139 3.282 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.491 2.913 pw_transfer 1439 11.6 0.068 0.068 2.827 2.834 dbcsr_complete_redistribute 329 12.2 0.287 0.291 1.993 2.811 acc_transpose_blocks 9144 15.5 0.038 0.038 2.780 2.799 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.729 2.735 acc_transpose_blocks_kernels 9144 16.5 0.117 0.118 2.532 2.549 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.505 jit_kernel_transpose 5 15.6 2.415 2.432 2.415 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.409 apply_single 119 13.6 0.000 0.000 2.167 2.409 fft3d_ps 1201 14.6 0.601 0.610 2.352 2.357 make_m2s 4572 13.5 0.038 0.038 2.192 2.353 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.305 2.315 mp_alltoall_i22 627 13.8 1.484 2.295 1.484 2.295 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.321 2.278 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.452 2.265 calculate_dm_sparse 119 9.5 0.000 0.000 2.214 2.232 make_images 4572 14.5 0.349 0.377 2.070 2.229 density_rs2pw 119 9.7 0.004 0.004 2.204 2.218 grid_integrate_task_list 119 12.3 2.059 2.071 2.059 2.071 ot_diis_step 108 11.5 0.014 0.014 1.981 1.981 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.895 1.896 jit_kernel_multiply 10 15.7 1.028 1.859 1.028 1.859 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.794 1.797 mp_sum_l 7207 12.9 1.028 1.750 1.028 1.750 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.664 1.717 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.622 1.622 grid_collocate_task_list 119 9.7 1.491 1.506 1.491 1.506 rs_pw_transfer 974 11.9 0.009 0.010 1.438 1.463 cp_fm_cholesky_invert 11 10.9 1.449 1.452 1.449 1.452 potential_pw2rs 119 12.3 0.014 0.014 1.445 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.380 1.380 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.339 1.339 cp_fm_diag_elpa_base 50 14.0 1.193 1.245 1.337 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.281 1.289 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.065 1.262 make_images_data 4572 15.5 0.043 0.047 1.013 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.136 1.156 qs_env_update_s_mstruct 11 6.9 0.011 0.019 1.128 1.147 mp_alltoall_d11v 2130 13.8 1.042 1.063 1.042 1.063 multiply_cannon_sync_h2d 9144 15.5 1.046 1.048 1.046 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.982 1.022 qs_create_task_list 11 7.9 0.008 0.019 0.955 0.969 generate_qs_task_list 11 8.9 0.371 0.390 0.947 0.968 yz_to_x 606 15.1 0.458 0.470 0.943 0.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.605000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=697.727273, yerr=14.955029 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.509568E+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 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.030 0.040 81.044 81.045 qs_mol_dyn_low 1 2.0 0.003 0.003 80.699 80.710 qs_forces 11 3.9 0.003 0.005 80.593 80.594 qs_energies 11 4.9 0.002 0.004 77.720 77.733 scf_env_do_scf 11 5.9 0.001 0.002 68.734 68.736 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.167 63.168 dbcsr_multiply_generic 2055 12.4 0.104 0.109 49.664 49.994 qs_scf_new_mos 99 7.5 0.000 0.001 45.682 45.789 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.681 45.788 ot_scf_mini 99 9.5 0.002 0.002 43.381 43.473 multiply_cannon 2055 13.4 0.195 0.210 41.883 42.624 velocity_verlet 10 3.0 0.001 0.001 42.224 42.232 multiply_cannon_loop 2055 14.4 1.489 1.526 40.882 41.541 ot_mini 99 10.5 0.001 0.001 25.504 25.585 qs_ot_get_derivative 99 11.5 0.001 0.001 18.725 18.828 multiply_cannon_multrec 49320 15.4 12.594 13.270 17.570 18.138 rebuild_ks_matrix 110 8.3 0.000 0.001 14.712 14.842 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.711 14.841 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.916 13.024 multiply_cannon_sync_h2d 49320 15.4 10.478 11.114 10.478 11.114 mp_waitall_1 220248 16.4 10.080 10.765 10.080 10.765 qs_ot_get_p 110 10.4 0.001 0.001 9.614 9.707 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.204 7.665 apply_single 110 13.6 0.000 0.000 7.204 7.664 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.940 7.493 sum_up_and_integrate 110 10.3 0.036 0.042 7.155 7.167 integrate_v_rspace 110 11.3 0.003 0.004 7.119 7.140 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.093 7.018 init_scf_run 11 5.9 0.000 0.001 6.903 6.904 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.903 6.903 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.654 6.761 calculate_rho_elec 110 8.6 0.021 0.026 6.653 6.761 ot_diis_step 99 11.5 0.006 0.006 6.543 6.543 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.519 6.541 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.730 5.731 init_scf_loop 11 6.9 0.000 0.000 5.533 5.534 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.153 5.206 dbcsr_mm_accdrv_process 87628 16.1 1.869 2.003 4.850 5.175 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.148 5.173 cp_fm_diag_elpa_base 48 14.0 5.133 5.159 5.146 5.171 mp_sum_l 6514 12.8 3.445 4.369 3.445 4.369 rs_pw_transfer 902 11.9 0.012 0.013 3.718 4.281 wfi_extrapolate 11 7.9 0.001 0.001 4.053 4.053 density_rs2pw 110 9.6 0.004 0.005 3.495 4.013 make_m2s 4110 13.4 0.060 0.064 3.873 3.996 make_images 4110 14.4 0.177 0.190 3.777 3.903 calculate_dm_sparse 110 9.5 0.001 0.001 3.758 3.839 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.591 3.596 grid_integrate_task_list 110 12.3 3.235 3.382 3.235 3.382 pw_transfer 1331 11.6 0.054 0.068 3.316 3.380 prepare_preconditioner 11 7.9 0.000 0.000 3.334 3.357 make_preconditioner 11 8.9 0.000 0.000 3.333 3.357 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.256 3.311 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.229 3.295 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.112 3.270 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.073 3.119 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.980 3.027 fft3d_ps 1111 14.6 0.780 0.880 2.755 2.812 fft_wrap_pw1pw2_140 451 13.1 0.171 0.191 2.737 2.806 calculate_first_density_matrix 1 7.0 0.001 0.001 2.758 2.763 jit_kernel_multiply 13 15.9 2.706 2.733 2.706 2.733 potential_pw2rs 110 12.3 0.006 0.007 2.690 2.714 mp_alltoall_d11v 2046 13.8 2.087 2.539 2.087 2.539 mp_waitany 14300 13.8 1.796 2.437 1.796 2.437 grid_collocate_task_list 110 9.6 2.082 2.346 2.082 2.346 acc_transpose_blocks 49320 15.4 0.204 0.210 2.046 2.127 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.966 make_images_data 4110 15.4 0.042 0.046 1.770 1.917 mp_sum_d 3885 11.9 1.351 1.826 1.351 1.826 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.538 1.825 cp_fm_cholesky_invert 11 10.9 1.806 1.810 1.806 1.810 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.745 1.771 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.045000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.757409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 588.132352E+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 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2855968. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.064 0.208 70.190 70.192 qs_mol_dyn_low 1 2.0 0.003 0.003 69.668 69.675 qs_forces 11 3.9 0.003 0.005 69.548 69.549 qs_energies 11 4.9 0.007 0.045 66.200 66.203 scf_env_do_scf 11 5.9 0.001 0.002 57.367 57.370 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.764 49.765 dbcsr_multiply_generic 2055 12.4 0.113 0.119 38.122 38.259 velocity_verlet 10 3.0 0.002 0.012 36.198 36.205 qs_scf_new_mos 99 7.5 0.001 0.001 33.311 33.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.311 33.445 multiply_cannon 2055 13.4 0.221 0.241 31.490 32.729 ot_scf_mini 99 9.5 0.003 0.004 31.662 31.809 multiply_cannon_loop 2055 14.4 0.927 0.951 30.253 31.132 ot_mini 99 10.5 0.001 0.001 18.600 18.747 multiply_cannon_multrec 24660 15.4 7.674 9.355 14.217 15.929 rebuild_ks_matrix 110 8.3 0.000 0.000 13.869 13.980 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.868 13.980 qs_ot_get_derivative 99 11.5 0.001 0.001 12.824 12.976 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.212 12.302 mp_waitall_1 176588 16.5 7.721 10.251 7.721 10.251 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.277 8.168 multiply_cannon_sync_h2d 24660 15.4 7.022 8.072 7.022 8.072 init_scf_loop 11 6.9 0.001 0.006 7.567 7.568 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.650 7.344 apply_single 110 13.6 0.000 0.001 6.650 7.344 dbcsr_mm_accdrv_process 52282 16.1 4.876 5.990 6.382 6.726 sum_up_and_integrate 110 10.3 0.053 0.059 6.699 6.711 integrate_v_rspace 110 11.3 0.003 0.003 6.646 6.659 qs_ot_get_p 110 10.4 0.001 0.001 6.474 6.633 init_scf_run 11 5.9 0.000 0.001 6.332 6.332 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.331 6.332 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.245 6.255 calculate_rho_elec 110 8.6 0.040 0.048 6.245 6.254 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.858 5.742 ot_diis_step 99 11.5 0.010 0.010 5.729 5.730 prepare_preconditioner 11 7.9 0.000 0.001 5.530 5.552 make_preconditioner 11 8.9 0.000 0.002 5.530 5.552 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.117 5.274 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.596 4.616 make_m2s 4110 13.4 0.057 0.060 4.110 4.517 make_images 4110 14.4 0.397 0.444 4.001 4.404 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.170 4.170 pw_transfer 1331 11.6 0.066 0.074 3.778 3.927 density_rs2pw 110 9.6 0.004 0.004 3.367 3.914 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.671 3.824 rs_pw_transfer 902 11.9 0.012 0.014 3.027 3.636 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.598 3.609 cp_fm_diag_elpa_base 48 14.0 3.550 3.565 3.595 3.606 wfi_extrapolate 11 7.9 0.001 0.001 3.508 3.508 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.333 3.335 grid_integrate_task_list 110 12.3 3.154 3.305 3.154 3.305 fft_wrap_pw1pw2_140 451 13.1 0.204 0.220 3.129 3.286 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.141 3.209 fft3d_ps 1111 14.6 1.103 1.320 3.022 3.164 calculate_dm_sparse 110 9.5 0.001 0.001 3.060 3.090 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.000 3.036 calculate_first_density_matrix 1 7.0 0.001 0.003 2.735 2.736 make_images_data 4110 15.4 0.047 0.050 2.312 2.729 hybrid_alltoall_any 4261 16.3 0.103 0.446 2.037 2.711 mp_sum_l 6514 12.8 2.012 2.688 2.012 2.688 cp_fm_cholesky_invert 11 10.9 2.559 2.565 2.559 2.565 potential_pw2rs 110 12.3 0.008 0.008 2.482 2.495 grid_collocate_task_list 110 9.6 2.048 2.481 2.048 2.481 jit_kernel_multiply 10 16.4 1.156 2.232 1.156 2.232 mp_alltoall_d11v 2046 13.8 1.759 2.105 1.759 2.105 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.935 1.956 mp_waitany 10164 13.8 1.259 1.932 1.259 1.932 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.884 1.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.821 1.840 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.787 1.799 mp_irecv_dv 57340 16.2 0.662 1.680 0.662 1.680 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.642 1.654 acc_transpose_blocks 24660 15.4 0.112 0.115 1.584 1.618 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.216 0.988 1.576 mp_allgather_i34 2055 14.4 0.564 1.566 0.564 1.566 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.192000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.636364, yerr=6.892324 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.664320E+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 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.017 0.033 60.614 60.616 qs_mol_dyn_low 1 2.0 0.003 0.004 60.315 60.328 qs_forces 11 3.9 0.003 0.004 60.194 60.195 qs_energies 11 4.9 0.001 0.002 57.000 57.004 scf_env_do_scf 11 5.9 0.001 0.002 48.819 48.820 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.232 40.233 velocity_verlet 10 3.0 0.001 0.001 32.744 32.748 dbcsr_multiply_generic 2055 12.4 0.107 0.111 28.689 28.933 qs_scf_new_mos 99 7.5 0.001 0.001 25.000 25.099 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.999 25.098 ot_scf_mini 99 9.5 0.002 0.003 23.783 23.894 multiply_cannon 2055 13.4 0.210 0.221 22.412 23.682 multiply_cannon_loop 2055 14.4 0.613 0.629 21.248 22.382 ot_mini 99 10.5 0.001 0.001 13.539 13.644 rebuild_ks_matrix 110 8.3 0.000 0.000 12.532 12.678 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.018 12.532 12.677 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.032 11.167 multiply_cannon_multrec 16440 15.4 3.906 4.828 10.001 10.876 mp_waitall_1 139946 16.5 7.034 10.072 7.034 10.072 qs_ot_get_derivative 99 11.5 0.001 0.001 9.135 9.243 init_scf_loop 11 6.9 0.000 0.000 8.551 8.552 multiply_cannon_metrocomm3 16440 15.4 0.044 0.046 4.351 7.203 prepare_preconditioner 11 7.9 0.000 0.000 6.759 6.779 make_preconditioner 11 8.9 0.000 0.000 6.759 6.779 sum_up_and_integrate 110 10.3 0.061 0.062 6.650 6.666 integrate_v_rspace 110 11.3 0.003 0.003 6.589 6.605 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.104 6.454 dbcsr_mm_accdrv_process 34862 16.1 4.842 5.589 5.950 6.117 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.070 6.081 calculate_rho_elec 110 8.6 0.059 0.059 6.070 6.081 init_scf_run 11 5.9 0.000 0.001 5.807 5.807 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.807 5.807 qs_ot_get_p 110 10.4 0.001 0.001 5.443 5.583 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.917 5.368 apply_single 110 13.6 0.000 0.000 4.917 5.368 make_m2s 4110 13.4 0.049 0.051 4.182 4.539 make_images 4110 14.4 0.388 0.507 4.068 4.424 density_rs2pw 110 9.6 0.004 0.004 3.170 4.393 ot_diis_step 99 11.5 0.011 0.011 4.376 4.376 multiply_cannon_sync_h2d 16440 15.4 3.657 3.997 3.657 3.997 rs_pw_transfer 902 11.9 0.011 0.011 2.715 3.935 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.826 3.829 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.145 3.791 pw_transfer 1331 11.6 0.065 0.073 3.767 3.776 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.661 3.672 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.491 3.492 grid_integrate_task_list 110 12.3 3.172 3.391 3.172 3.391 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.150 3.166 wfi_extrapolate 11 7.9 0.001 0.001 2.995 2.995 fft3d_ps 1111 14.6 1.091 1.107 2.960 2.974 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.938 2.950 cp_fm_diag_elpa_base 48 14.0 2.872 2.904 2.937 2.949 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.925 2.927 make_images_data 4110 15.4 0.044 0.048 2.444 2.912 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.142 2.817 calculate_first_density_matrix 1 7.0 0.000 0.001 2.728 2.728 calculate_dm_sparse 110 9.5 0.001 0.001 2.661 2.694 cp_fm_cholesky_invert 11 10.9 2.656 2.662 2.656 2.662 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.611 2.660 grid_collocate_task_list 110 9.6 2.084 2.466 2.084 2.466 mp_waitany 17072 13.8 1.166 2.452 1.166 2.452 potential_pw2rs 110 12.3 0.011 0.011 2.371 2.380 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.873 2.370 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.307 2.366 mp_irecv_dv 48980 15.7 0.802 2.241 0.802 2.241 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.872 2.083 mp_sum_l 6514 12.8 1.478 2.079 1.478 2.079 mp_alltoall_d11v 2046 13.8 1.744 2.004 1.744 2.004 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.944 1.946 dbcsr_complete_redistribute 325 12.2 0.322 0.346 1.449 1.902 cp_fm_upper_to_full 70 14.2 1.384 1.750 1.384 1.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.699 1.714 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.627 1.641 cp_fm_cholesky_decompose 22 10.9 1.598 1.618 1.598 1.618 mp_allgather_i34 2055 14.4 0.491 1.602 0.491 1.602 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.362 1.447 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.945 1.388 jit_kernel_multiply 6 16.2 0.720 1.367 0.720 1.367 acc_transpose_blocks 16440 15.4 0.072 0.075 1.277 1.336 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.233 1.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.616000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.909091, yerr=7.982419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 739.262464E+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 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.023 0.044 66.972 66.973 qs_mol_dyn_low 1 2.0 0.003 0.004 66.671 66.682 qs_forces 11 3.9 0.003 0.003 66.596 66.597 qs_energies 11 4.9 0.003 0.007 63.112 63.117 scf_env_do_scf 11 5.9 0.001 0.003 54.723 54.726 scf_env_do_scf_inner_loop 99 6.5 0.004 0.008 42.739 42.742 velocity_verlet 10 3.0 0.001 0.002 37.696 37.698 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.537 29.784 qs_scf_new_mos 99 7.5 0.001 0.001 27.289 27.381 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.288 27.380 ot_scf_mini 99 9.5 0.003 0.004 25.652 25.758 multiply_cannon 2055 13.4 0.242 0.262 22.423 23.434 multiply_cannon_loop 2055 14.4 0.884 0.904 20.977 21.640 ot_mini 99 10.5 0.001 0.001 14.226 14.343 multiply_cannon_multrec 24660 15.4 4.200 6.694 12.592 13.834 rebuild_ks_matrix 110 8.3 0.000 0.000 12.452 12.550 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.451 12.550 init_scf_loop 11 6.9 0.001 0.005 11.940 11.946 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.984 11.072 prepare_preconditioner 11 7.9 0.000 0.001 10.152 10.167 make_preconditioner 11 8.9 0.001 0.002 10.152 10.167 qs_ot_get_derivative 99 11.5 0.001 0.001 10.025 10.129 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.390 9.839 dbcsr_mm_accdrv_process 52304 16.0 6.795 7.930 8.246 9.199 sum_up_and_integrate 110 10.3 0.067 0.071 6.703 6.717 integrate_v_rspace 110 11.3 0.003 0.003 6.636 6.650 qs_ot_get_p 110 10.4 0.001 0.002 6.424 6.575 mp_waitall_1 121746 16.5 4.401 6.439 4.401 6.439 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.359 6.371 calculate_rho_elec 110 8.6 0.077 0.081 6.359 6.371 make_m2s 4110 13.4 0.060 0.064 5.460 5.822 make_images 4110 14.4 0.571 0.692 5.321 5.679 init_scf_run 11 5.9 0.000 0.001 5.591 5.592 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.591 5.592 cp_fm_upper_to_full 70 14.2 3.333 4.832 3.333 4.832 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.533 4.549 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.164 apply_single 110 13.6 0.000 0.000 4.056 4.164 ot_diis_step 99 11.5 0.011 0.011 4.162 4.162 pw_transfer 1331 11.6 0.065 0.074 4.047 4.075 cp_dbcsr_syevd 48 12.0 0.003 0.005 4.044 4.045 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.941 3.972 dbcsr_complete_redistribute 325 12.2 0.418 0.459 2.707 3.855 density_rs2pw 110 9.6 0.004 0.005 3.261 3.775 grid_integrate_task_list 110 12.3 3.270 3.453 3.270 3.453 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.417 3.426 cp_fm_diag_elpa_base 48 14.0 3.259 3.322 3.413 3.422 fft_wrap_pw1pw2_140 451 13.1 0.203 0.214 3.368 3.401 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.324 3.380 multiply_cannon_sync_h2d 24660 15.4 3.175 3.341 3.175 3.341 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.494 3.338 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.135 3.268 fft3d_ps 1111 14.6 1.095 1.134 3.231 3.256 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.072 3.120 make_images_data 4110 15.4 0.047 0.051 2.742 3.098 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.450 3.083 rs_pw_transfer 902 11.9 0.010 0.011 2.543 3.068 wfi_extrapolate 11 7.9 0.001 0.006 3.032 3.032 calculate_dm_sparse 110 9.5 0.001 0.001 2.938 2.968 cp_fm_cholesky_invert 11 10.9 2.897 2.907 2.897 2.907 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.753 2.869 mp_alltoall_i22 605 13.7 1.649 2.811 1.649 2.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.802 2.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.419 2.459 calculate_first_density_matrix 1 7.0 0.001 0.004 2.439 2.442 grid_collocate_task_list 110 9.6 2.182 2.438 2.182 2.438 potential_pw2rs 110 12.3 0.012 0.013 2.327 2.339 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.286 2.287 jit_kernel_multiply 12 15.8 1.122 2.194 1.122 2.194 mp_alltoall_d11v 2046 13.8 1.879 2.182 1.879 2.182 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.724 1.754 cp_fm_cholesky_decompose 22 10.9 1.699 1.739 1.699 1.739 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.714 1.724 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.708 mp_waitany 13376 13.8 1.086 1.570 1.086 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.559 acc_transpose_blocks 24660 15.4 0.107 0.109 1.505 1.534 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.856 1.486 mp_sum_l 6514 12.8 0.977 1.465 0.977 1.465 mp_irecv_dv 62702 16.1 0.756 1.411 0.756 1.411 mp_alltoall_z22v 1111 16.6 1.272 1.383 1.272 1.383 mp_allgather_i34 2055 14.4 0.547 1.378 0.547 1.378 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.176 0.825 1.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.973000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.363636, yerr=12.382339 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 827.207680E+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 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.016 0.034 57.059 57.060 qs_mol_dyn_low 1 2.0 0.003 0.003 56.783 56.794 qs_forces 11 3.9 0.003 0.003 56.642 56.643 qs_energies 11 4.9 0.001 0.002 52.884 52.888 scf_env_do_scf 11 5.9 0.001 0.002 44.528 44.528 scf_env_do_scf_inner_loop 99 6.5 0.006 0.010 36.496 36.497 velocity_verlet 10 3.0 0.001 0.001 31.931 31.939 dbcsr_multiply_generic 2055 12.4 0.105 0.108 23.100 23.242 qs_scf_new_mos 99 7.5 0.001 0.001 21.060 21.124 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.059 21.123 ot_scf_mini 99 9.5 0.002 0.002 19.809 19.863 multiply_cannon 2055 13.4 0.248 0.259 17.580 18.766 multiply_cannon_loop 2055 14.4 0.320 0.332 16.227 16.490 rebuild_ks_matrix 110 8.3 0.000 0.000 12.143 12.189 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.142 12.189 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.775 10.817 ot_mini 99 10.5 0.001 0.001 10.624 10.680 multiply_cannon_multrec 8220 15.4 3.218 4.854 7.500 8.692 init_scf_loop 11 6.9 0.000 0.000 7.983 7.984 mp_waitall_1 103326 16.6 6.229 7.877 6.229 7.877 qs_ot_get_derivative 99 11.5 0.001 0.001 6.760 6.814 sum_up_and_integrate 110 10.3 0.080 0.081 6.669 6.679 integrate_v_rspace 110 11.3 0.003 0.003 6.589 6.599 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.519 6.534 calculate_rho_elec 110 8.6 0.116 0.116 6.519 6.533 prepare_preconditioner 11 7.9 0.000 0.000 6.284 6.289 make_preconditioner 11 8.9 0.000 0.000 6.284 6.289 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.852 5.923 qs_ot_get_p 110 10.4 0.001 0.001 5.245 5.312 init_scf_run 11 5.9 0.000 0.001 5.203 5.203 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.203 5.203 dbcsr_mm_accdrv_process 17442 15.9 2.804 3.960 4.151 5.047 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.204 4.527 make_m2s 4110 13.4 0.039 0.040 4.181 4.405 pw_transfer 1331 11.6 0.066 0.071 4.265 4.282 make_images 4110 14.4 0.633 0.688 4.051 4.274 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.157 4.178 ot_diis_step 99 11.5 0.012 0.012 3.841 3.842 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.740 3.812 apply_single 110 13.6 0.000 0.000 3.740 3.812 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.803 3.807 density_rs2pw 110 9.6 0.004 0.004 3.242 3.617 fft_wrap_pw1pw2_140 451 13.1 0.217 0.220 3.580 3.608 grid_integrate_task_list 110 12.3 3.336 3.571 3.336 3.571 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.496 3.496 fft3d_ps 1111 14.6 1.149 1.173 3.400 3.414 cp_fm_cholesky_invert 11 10.9 3.069 3.073 3.069 3.073 multiply_cannon_sync_h2d 8220 15.4 2.905 3.017 2.905 3.017 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.941 2.949 cp_fm_diag_elpa_base 48 14.0 2.886 2.908 2.940 2.947 wfi_extrapolate 11 7.9 0.001 0.001 2.792 2.792 make_images_data 4110 15.4 0.038 0.044 2.364 2.749 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.691 2.692 rs_pw_transfer 902 11.9 0.010 0.010 2.262 2.661 hybrid_alltoall_any 4261 16.3 0.199 0.863 2.304 2.657 calculate_dm_sparse 110 9.5 0.001 0.001 2.528 2.568 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.544 2.546 grid_collocate_task_list 110 9.6 2.279 2.541 2.279 2.541 calculate_first_density_matrix 1 7.0 0.001 0.001 2.312 2.312 potential_pw2rs 110 12.3 0.015 0.015 2.199 2.206 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.179 2.191 mp_alltoall_d11v 2046 13.8 1.738 2.138 1.738 2.138 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 2.008 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.941 1.965 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.846 1.880 jit_kernel_multiply 10 15.8 1.041 1.819 1.041 1.819 cp_fm_cholesky_decompose 22 10.9 1.726 1.744 1.726 1.744 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.702 1.707 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.525 1.650 mp_allgather_i34 2055 14.4 0.535 1.600 0.535 1.600 dbcsr_complete_redistribute 325 12.2 0.559 0.571 1.500 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.482 1.493 mp_waitany 9240 13.8 1.074 1.482 1.074 1.482 rs_gather_matrices 110 12.3 0.325 0.364 0.963 1.342 qs_create_task_list 11 7.9 0.000 0.001 1.232 1.337 generate_qs_task_list 11 8.9 0.379 0.447 1.231 1.336 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.323 mp_irecv_dv 24056 15.7 0.457 1.281 0.457 1.281 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.194 1.227 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.802 1.201 mp_alltoall_z22v 1111 16.6 1.140 1.164 1.140 1.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.060000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.545455, yerr=10.543103 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 1.352806E+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_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 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 0.020 0.042 89.461 89.463 qs_mol_dyn_low 1 2.0 0.003 0.003 89.051 89.061 qs_forces 11 3.9 0.003 0.003 88.975 88.977 qs_energies 11 4.9 0.002 0.002 84.742 84.744 scf_env_do_scf 11 5.9 0.004 0.008 74.535 74.536 velocity_verlet 10 3.0 0.001 0.001 57.037 57.044 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.338 45.339 init_scf_loop 11 6.9 0.000 0.000 29.119 29.121 dbcsr_multiply_generic 2055 12.4 0.121 0.124 28.790 28.953 prepare_preconditioner 11 7.9 0.000 0.000 27.042 27.049 make_preconditioner 11 8.9 0.000 0.000 27.042 27.049 qs_scf_new_mos 99 7.5 0.001 0.001 26.640 26.696 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.639 26.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.108 26.521 ot_scf_mini 99 9.5 0.002 0.002 24.875 24.919 multiply_cannon 2055 13.4 0.350 0.388 21.645 22.483 multiply_cannon_loop 2055 14.4 0.341 0.347 19.803 20.236 cp_fm_upper_to_full 70 14.2 12.989 18.657 12.989 18.657 rebuild_ks_matrix 110 8.3 0.001 0.001 14.402 14.482 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.016 14.402 14.481 ot_mini 99 10.5 0.001 0.001 13.753 13.794 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.053 13.124 dbcsr_complete_redistribute 325 12.2 1.013 1.043 7.559 10.909 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.497 9.841 multiply_cannon_multrec 8220 15.4 4.393 4.575 9.698 9.834 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.919 9.224 qs_ot_get_derivative 99 11.5 0.001 0.001 9.149 9.197 mp_alltoall_i22 605 13.7 5.569 8.928 5.569 8.928 mp_waitall_1 84994 16.7 7.514 8.482 7.514 8.482 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.936 7.976 calculate_rho_elec 110 8.6 0.227 0.227 7.935 7.975 sum_up_and_integrate 110 10.3 0.150 0.152 7.542 7.554 integrate_v_rspace 110 11.3 0.004 0.004 7.393 7.405 qs_ot_get_p 110 10.4 0.001 0.001 5.918 5.970 init_scf_run 11 5.9 0.000 0.001 5.910 5.910 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.910 5.910 make_m2s 4110 13.4 0.043 0.044 5.295 5.821 make_images 4110 14.4 0.878 0.936 5.106 5.631 cp_fm_cholesky_invert 11 10.9 5.517 5.522 5.517 5.522 dbcsr_mm_accdrv_process 11614 15.7 3.365 3.710 5.162 5.447 pw_transfer 1331 11.6 0.075 0.075 5.343 5.354 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.226 5.237 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.667 5.205 apply_single 110 13.6 0.000 0.000 4.666 5.204 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.731 5.046 ot_diis_step 99 11.5 0.015 0.016 4.579 4.580 fft_wrap_pw1pw2_140 451 13.1 0.222 0.225 4.521 4.539 fft3d_ps 1111 14.6 1.298 1.308 4.419 4.435 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.284 4.290 density_rs2pw 110 9.6 0.004 0.004 3.916 3.969 multiply_cannon_sync_h2d 8220 15.4 3.953 3.955 3.953 3.955 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.836 3.837 grid_integrate_task_list 110 12.3 3.718 3.815 3.718 3.815 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 3.690 3.691 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.866 3.691 make_images_data 4110 15.4 0.042 0.045 2.860 3.576 wfi_extrapolate 11 7.9 0.001 0.001 3.392 3.392 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.907 3.354 calculate_dm_sparse 110 9.5 0.001 0.001 3.232 3.254 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.216 3.216 cp_fm_diag_elpa_base 48 14.0 2.663 2.871 3.214 3.214 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.922 2.924 grid_collocate_task_list 110 9.6 2.678 2.768 2.678 2.768 potential_pw2rs 110 12.3 0.021 0.021 2.567 2.572 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.457 2.485 rs_pw_transfer 902 11.9 0.011 0.011 2.377 2.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.398 2.424 calculate_first_density_matrix 1 7.0 0.001 0.001 2.393 2.394 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.206 2.264 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.085 2.166 mp_alltoall_d11v 2046 13.8 2.030 2.141 2.030 2.141 cp_fm_cholesky_decompose 22 10.9 2.081 2.106 2.081 2.106 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.955 1.963 qs_create_task_list 11 7.9 0.001 0.001 1.903 1.958 generate_qs_task_list 11 8.9 0.736 0.790 1.902 1.957 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.832 1.870 jit_kernel_multiply 10 15.2 1.597 1.865 1.597 1.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.463000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1218.545455, yerr=58.244568 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.960704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2523700. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 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 0.015 0.033 207.386 207.387 qs_mol_dyn_low 1 2.0 0.003 0.004 207.003 207.019 qs_forces 11 3.9 0.005 0.006 206.911 206.912 qs_energies 11 4.9 0.004 0.006 201.210 201.229 scf_env_do_scf 11 5.9 0.001 0.001 184.602 184.607 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 163.688 163.690 velocity_verlet 10 3.0 0.001 0.001 124.707 124.708 dbcsr_multiply_generic 2507 12.6 0.177 0.186 123.822 124.624 qs_scf_new_mos 117 7.6 0.001 0.001 123.234 123.553 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.233 123.552 ot_scf_mini 117 9.6 0.003 0.003 116.589 116.878 multiply_cannon 2507 13.6 0.238 0.247 100.640 102.976 multiply_cannon_loop 2507 14.6 2.032 2.094 98.412 100.585 ot_mini 117 10.6 0.001 0.001 65.119 65.433 multiply_cannon_multrec 60168 15.6 33.336 35.611 41.567 43.913 qs_ot_get_derivative 117 11.6 0.001 0.001 40.437 40.707 rebuild_ks_matrix 128 8.3 0.001 0.001 33.941 34.351 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 33.940 34.351 mp_waitall_1 267128 16.5 27.845 31.306 27.845 31.306 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.509 30.889 qs_ot_get_p 128 10.4 0.001 0.001 29.503 29.838 multiply_cannon_sync_h2d 60168 15.6 27.792 29.436 27.792 29.436 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.273 25.288 apply_single 128 13.6 0.001 0.001 24.272 25.288 ot_diis_step 117 11.6 0.007 0.008 24.423 24.424 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.892 22.987 init_scf_loop 11 6.9 0.000 0.000 20.836 20.837 cp_dbcsr_syevd 83 12.4 0.004 0.005 20.406 20.406 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.959 19.174 multiply_cannon_metrocomm3 60168 15.6 0.112 0.118 15.403 17.856 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.334 17.373 cp_fm_diag_elpa_base 83 14.4 17.264 17.299 17.330 17.370 prepare_preconditioner 11 7.9 0.000 0.000 16.181 16.224 make_preconditioner 11 8.9 0.000 0.000 16.181 16.224 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.429 15.592 sum_up_and_integrate 128 10.3 0.089 0.106 14.413 14.428 integrate_v_rspace 128 11.3 0.004 0.005 14.323 14.342 make_m2s 5014 13.6 0.105 0.114 13.885 14.211 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.942 14.095 calculate_rho_elec 128 8.7 0.046 0.065 13.942 14.095 make_images 5014 14.6 0.401 0.420 13.701 14.036 init_scf_run 11 5.9 0.000 0.001 12.359 12.359 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.359 12.359 density_rs2pw 128 9.7 0.006 0.007 7.268 10.493 cp_fm_cholesky_invert 11 10.9 9.376 9.384 9.376 9.384 mp_sum_l 7870 13.0 8.021 9.177 8.021 9.177 wfi_extrapolate 11 7.9 0.001 0.001 9.132 9.132 rs_pw_transfer 1046 11.9 0.017 0.019 5.764 8.994 calculate_dm_sparse 128 9.5 0.001 0.001 8.493 8.610 pw_transfer 1547 11.6 0.076 0.100 8.192 8.437 dbcsr_mm_accdrv_process 124484 16.2 3.079 3.243 7.791 8.261 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.986 8.236 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.036 8.170 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.975 8.096 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 5.848 7.810 make_images_data 5014 15.6 0.066 0.071 6.838 7.780 grid_integrate_task_list 128 12.3 7.094 7.660 7.094 7.660 hybrid_alltoall_any 5200 16.5 0.291 2.253 5.982 7.471 fft_wrap_pw1pw2_140 523 13.2 0.446 0.514 6.921 7.212 fft3d_ps 1291 14.7 2.110 2.901 6.770 6.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.622 6.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.779 5.919 mp_waitany 16020 13.9 2.640 5.913 2.640 5.913 grid_collocate_task_list 128 9.7 4.564 5.772 4.564 5.772 mp_alltoall_d11v 2415 14.1 4.207 5.754 4.207 5.754 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.290 2.115 5.371 potential_pw2rs 128 12.3 0.009 0.011 4.788 4.817 cp_fm_cholesky_decompose 22 10.9 4.654 4.668 4.654 4.668 mp_sum_d 4461 12.1 3.564 4.312 3.564 4.312 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.387000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.272727, yerr=6.823634 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 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 826.548224E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463835059104 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57964. MP_Allreduce 11048 965. MP_Sync 87 MP_Alltoall 1969 6422314. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 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 0.031 0.048 192.941 192.943 qs_mol_dyn_low 1 2.0 0.003 0.005 192.586 192.602 qs_forces 11 3.9 0.005 0.005 192.478 192.480 qs_energies 11 4.9 0.002 0.003 185.617 185.626 scf_env_do_scf 11 5.9 0.001 0.002 169.113 169.123 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.978 135.981 velocity_verlet 10 3.0 0.001 0.001 121.036 121.039 dbcsr_multiply_generic 2507 12.6 0.186 0.193 97.512 98.695 qs_scf_new_mos 117 7.6 0.001 0.001 95.321 95.760 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.321 95.759 ot_scf_mini 117 9.6 0.004 0.004 90.506 91.014 multiply_cannon 2507 13.6 0.475 0.528 77.419 81.584 multiply_cannon_loop 2507 14.6 1.256 1.298 74.387 76.971 ot_mini 117 10.6 0.001 0.001 49.984 50.419 mp_waitall_1 214728 16.6 24.710 37.144 24.710 37.144 multiply_cannon_multrec 30084 15.6 21.957 25.610 31.778 35.878 rebuild_ks_matrix 128 8.3 0.001 0.001 33.678 34.203 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.678 34.202 init_scf_loop 11 6.9 0.000 0.000 33.044 33.045 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.299 30.771 prepare_preconditioner 11 7.9 0.000 0.000 28.610 28.669 make_preconditioner 11 8.9 0.000 0.000 28.610 28.669 qs_ot_get_derivative 117 11.6 0.001 0.002 28.004 28.502 multiply_cannon_metrocomm3 30084 15.6 0.092 0.098 15.753 28.026 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.268 27.821 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.050 23.207 apply_single 128 13.6 0.001 0.001 22.050 23.207 qs_ot_get_p 128 10.4 0.001 0.001 22.335 22.934 ot_diis_step 117 11.6 0.014 0.015 21.807 21.810 multiply_cannon_sync_h2d 30084 15.6 19.336 21.391 19.336 21.391 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.535 17.571 cp_fm_cholesky_invert 11 10.9 16.792 16.805 16.792 16.805 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.375 16.376 make_m2s 5014 13.6 0.089 0.095 14.415 15.659 make_images 5014 14.6 1.146 1.337 14.205 15.452 sum_up_and_integrate 128 10.3 0.117 0.135 15.036 15.068 integrate_v_rspace 128 11.3 0.004 0.004 14.919 14.955 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.390 14.426 calculate_rho_elec 128 8.7 0.088 0.106 14.390 14.425 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.138 13.166 cp_fm_diag_elpa_base 83 14.4 12.878 12.946 13.132 13.161 init_scf_run 11 5.9 0.000 0.001 11.701 11.702 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.700 11.702 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.993 11.413 multiply_cannon_metrocomm4 27577 15.6 0.097 0.111 3.793 10.518 make_images_data 5014 15.6 0.064 0.072 8.783 10.321 mp_irecv_dv 69486 16.3 3.596 10.124 3.596 10.124 density_rs2pw 128 9.7 0.006 0.007 7.736 10.097 dbcsr_mm_accdrv_process 62242 16.2 4.504 5.834 9.273 9.867 hybrid_alltoall_any 5200 16.5 0.342 1.511 7.436 9.685 pw_transfer 1547 11.6 0.087 0.113 9.222 9.319 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 8.994 9.095 wfi_extrapolate 11 7.9 0.001 0.001 8.460 8.460 rs_pw_transfer 1046 11.9 0.014 0.017 5.909 8.282 fft_wrap_pw1pw2_140 523 13.2 0.473 0.530 7.897 8.020 grid_integrate_task_list 128 12.3 7.179 7.732 7.179 7.732 fft3d_ps 1291 14.7 2.754 2.948 7.363 7.417 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.216 7.016 cp_fm_cholesky_decompose 22 10.9 6.935 7.000 6.935 7.000 calculate_dm_sparse 128 9.5 0.001 0.001 6.540 6.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.174 6.184 mp_sum_l 7870 13.0 3.886 6.008 3.886 6.008 grid_collocate_task_list 128 9.7 4.708 5.874 4.708 5.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.301 5.438 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.354 5.410 potential_pw2rs 128 12.3 0.015 0.017 5.262 5.293 mp_alltoall_d11v 2415 14.1 4.210 4.947 4.210 4.947 mp_waitany 11748 13.9 2.481 4.907 2.481 4.907 mp_allgather_i34 2507 14.6 1.603 4.651 1.603 4.651 rs_pw_transfer_RS2PW_140 139 11.5 0.350 0.386 2.070 4.437 dbcsr_complete_redistribute 395 12.7 0.770 0.842 3.171 4.010 mp_sum_d 4475 12.1 2.654 3.985 2.654 3.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.943000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.363636, yerr=2.100767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 936.304640E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11002 1006. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.085 178.984 178.986 qs_mol_dyn_low 1 2.0 0.003 0.004 178.497 178.518 qs_forces 11 3.9 0.004 0.006 178.373 178.380 qs_energies 11 4.9 0.002 0.010 171.702 171.710 scf_env_do_scf 11 5.9 0.001 0.003 156.141 156.144 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.789 120.790 velocity_verlet 10 3.0 0.001 0.001 114.400 114.403 dbcsr_multiply_generic 2507 12.6 0.184 0.188 81.739 82.710 qs_scf_new_mos 117 7.6 0.001 0.001 81.787 82.178 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.786 82.177 ot_scf_mini 117 9.6 0.004 0.005 77.643 78.086 multiply_cannon 2507 13.6 0.499 0.517 61.710 66.521 multiply_cannon_loop 2507 14.6 0.852 0.887 58.776 61.132 ot_mini 117 10.6 0.001 0.001 42.693 43.135 mp_waitall_1 170520 16.6 25.684 35.551 25.684 35.551 init_scf_loop 11 6.9 0.001 0.006 35.249 35.250 rebuild_ks_matrix 128 8.3 0.001 0.001 31.259 31.687 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.259 31.686 prepare_preconditioner 11 7.9 0.000 0.001 31.143 31.187 make_preconditioner 11 8.9 0.001 0.010 31.142 31.187 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.814 30.201 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.170 28.562 multiply_cannon_metrocomm3 20056 15.6 0.060 0.067 15.641 25.419 multiply_cannon_multrec 20056 15.6 13.388 16.839 22.018 25.380 qs_ot_get_derivative 117 11.6 0.001 0.002 22.815 23.268 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.020 20.951 apply_single 128 13.6 0.001 0.001 20.019 20.951 qs_ot_get_p 128 10.4 0.001 0.001 20.236 20.697 ot_diis_step 117 11.6 0.018 0.018 19.775 19.776 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.984 15.994 make_m2s 5014 13.6 0.080 0.085 15.095 15.858 multiply_cannon_sync_h2d 20056 15.6 14.270 15.808 14.270 15.808 make_images 5014 14.6 1.159 1.238 14.861 15.620 sum_up_and_integrate 128 10.3 0.134 0.146 15.004 15.032 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.962 14.963 integrate_v_rspace 128 11.3 0.004 0.004 14.869 14.899 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.715 14.739 calculate_rho_elec 128 8.7 0.133 0.148 14.714 14.739 cp_fm_cholesky_invert 11 10.9 14.586 14.596 14.586 14.596 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.700 11.722 cp_fm_diag_elpa_base 83 14.4 11.287 11.445 11.697 11.718 make_images_data 5014 15.6 0.060 0.069 9.439 10.569 init_scf_run 11 5.9 0.000 0.001 10.504 10.504 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.504 10.504 density_rs2pw 128 9.7 0.006 0.007 7.559 9.621 hybrid_alltoall_any 5200 16.5 0.435 1.992 8.251 9.619 pw_transfer 1547 11.6 0.086 0.109 9.403 9.524 multiply_cannon_metrocomm4 17549 15.6 0.061 0.072 3.481 9.441 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.178 9.305 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.954 9.278 mp_irecv_dv 50230 16.2 3.359 9.183 3.359 9.183 fft_wrap_pw1pw2_140 523 13.2 0.478 0.526 8.071 8.207 dbcsr_mm_accdrv_process 41502 16.2 4.498 5.313 8.083 8.192 grid_integrate_task_list 128 12.3 7.310 7.931 7.310 7.931 fft3d_ps 1291 14.7 2.671 2.910 7.444 7.522 wfi_extrapolate 11 7.9 0.001 0.001 7.493 7.493 cp_fm_cholesky_decompose 22 10.9 7.466 7.492 7.466 7.492 rs_pw_transfer 1046 11.9 0.014 0.015 5.391 7.468 cp_fm_upper_to_full 105 14.8 5.711 7.189 5.711 7.189 dbcsr_complete_redistribute 395 12.7 1.167 1.196 4.610 6.340 calculate_dm_sparse 128 9.5 0.001 0.001 5.753 5.839 grid_collocate_task_list 128 9.7 4.897 5.744 4.897 5.744 mp_alltoall_d11v 2415 14.1 4.453 5.487 4.453 5.487 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.418 5.425 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.540 5.210 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.399 5.126 potential_pw2rs 128 12.3 0.020 0.022 5.041 5.063 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.510 4.642 mp_waitany 11748 13.9 2.394 4.597 2.394 4.597 mp_allgather_i34 2507 14.6 1.456 4.591 1.456 4.591 mp_sum_l 7870 13.0 3.142 4.416 3.142 4.416 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.062 4.097 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.308 4.026 rs_pw_transfer_RS2PW_140 139 11.5 0.330 0.347 1.922 4.003 mp_alltoall_i22 716 14.1 1.917 3.817 1.917 3.817 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.775 3.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.603 3.645 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.986000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.363636, yerr=9.518143 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.139712E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57899. MP_Allreduce 11057 1088. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.032 190.895 190.896 qs_mol_dyn_low 1 2.0 0.003 0.004 190.506 190.521 qs_forces 11 3.9 0.004 0.005 190.394 190.401 qs_energies 11 4.9 0.002 0.002 183.143 183.153 scf_env_do_scf 11 5.9 0.001 0.002 166.344 166.352 velocity_verlet 10 3.0 0.001 0.001 126.952 126.955 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.540 119.541 qs_scf_new_mos 118 7.6 0.001 0.001 81.613 81.936 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.612 81.935 dbcsr_multiply_generic 2527 12.6 0.191 0.196 79.660 80.391 ot_scf_mini 118 9.6 0.003 0.003 77.107 77.429 multiply_cannon 2527 13.6 0.557 0.598 55.276 58.792 multiply_cannon_loop 2527 14.6 1.190 1.219 51.346 53.215 init_scf_loop 11 6.9 0.000 0.000 46.680 46.682 ot_mini 118 10.6 0.001 0.001 42.685 42.984 prepare_preconditioner 11 7.9 0.000 0.000 42.556 42.584 make_preconditioner 11 8.9 0.000 0.000 42.556 42.584 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.191 41.225 multiply_cannon_multrec 30324 15.6 14.235 19.373 26.357 31.330 rebuild_ks_matrix 129 8.3 0.001 0.001 30.168 30.487 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.020 30.167 30.486 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.169 27.450 mp_waitall_1 149060 16.7 16.882 27.032 16.882 27.032 qs_ot_get_derivative 118 11.6 0.001 0.002 22.859 23.173 make_m2s 5054 13.6 0.095 0.098 19.891 21.106 qs_ot_get_p 129 10.4 0.001 0.001 20.652 20.935 make_images 5054 14.6 1.932 2.256 19.590 20.806 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.165 19.825 apply_single 129 13.6 0.001 0.001 19.165 19.825 ot_diis_step 118 11.6 0.018 0.019 19.690 19.691 cp_fm_upper_to_full 105 14.8 11.450 16.898 11.450 16.898 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.426 16.478 cp_fm_cholesky_invert 11 10.9 16.323 16.332 16.323 16.332 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.135 15.138 sum_up_and_integrate 129 10.3 0.142 0.153 14.972 14.999 multiply_cannon_metrocomm3 30324 15.6 0.046 0.049 6.437 14.914 integrate_v_rspace 129 11.3 0.004 0.004 14.830 14.862 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.788 14.830 calculate_rho_elec 129 8.7 0.177 0.192 14.787 14.829 multiply_cannon_sync_h2d 30324 15.6 11.793 12.854 11.793 12.854 dbcsr_complete_redistribute 395 12.7 1.507 1.637 8.955 12.697 make_images_data 5054 15.6 0.065 0.070 10.554 12.497 dbcsr_mm_accdrv_process 62756 16.2 7.554 8.417 11.692 12.194 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.951 11.965 cp_fm_diag_elpa_base 83 14.4 10.961 11.287 11.945 11.958 hybrid_alltoall_any 5240 16.5 0.530 2.227 9.475 11.532 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.557 11.314 init_scf_run 11 5.9 0.000 0.001 10.906 10.908 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.906 10.907 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.343 10.026 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.329 9.547 pw_transfer 1559 11.6 0.088 0.106 9.386 9.448 mp_alltoall_i22 718 14.1 5.584 9.321 5.584 9.321 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.157 9.228 density_rs2pw 129 9.7 0.006 0.006 7.205 8.717 fft_wrap_pw1pw2_140 527 13.2 0.487 0.495 8.078 8.165 grid_integrate_task_list 129 12.3 7.526 8.103 7.526 8.103 cp_fm_cholesky_decompose 22 10.9 7.729 7.817 7.729 7.817 wfi_extrapolate 11 7.9 0.001 0.001 7.683 7.683 fft3d_ps 1301 14.7 2.799 2.866 7.373 7.434 multiply_cannon_metrocomm4 25270 15.6 0.078 0.087 2.839 7.191 mp_irecv_dv 76685 16.2 2.690 6.906 2.690 6.906 rs_pw_transfer 1054 12.0 0.013 0.014 4.900 6.447 calculate_dm_sparse 129 9.5 0.001 0.001 6.239 6.321 grid_collocate_task_list 129 9.7 5.102 5.829 5.102 5.829 mp_alltoall_d11v 2423 14.1 4.812 5.653 4.812 5.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.471 5.518 potential_pw2rs 129 12.3 0.023 0.023 4.763 4.781 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.565 4.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.421 4.485 qs_energies_init_hamiltonians 11 5.9 0.003 0.009 4.460 4.461 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.214 4.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.896000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.363636, yerr=18.927002 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.546023E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108623056 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828148. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.038 174.235 174.236 qs_mol_dyn_low 1 2.0 0.003 0.004 173.790 173.803 qs_forces 11 3.9 0.004 0.004 173.649 173.654 qs_energies 11 4.9 0.017 0.036 166.027 166.035 scf_env_do_scf 11 5.9 0.002 0.003 148.678 148.683 velocity_verlet 10 3.0 0.001 0.001 113.453 113.471 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 112.867 112.868 qs_scf_new_mos 116 7.6 0.001 0.001 74.169 74.272 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.168 74.271 dbcsr_multiply_generic 2485 12.5 0.181 0.189 73.479 73.936 ot_scf_mini 116 9.6 0.003 0.004 69.777 69.843 multiply_cannon 2485 13.5 0.576 0.616 53.956 58.038 multiply_cannon_loop 2485 14.5 0.439 0.447 49.396 50.183 ot_mini 116 10.6 0.001 0.001 39.087 39.149 init_scf_loop 11 6.9 0.000 0.000 35.660 35.661 mp_waitall_1 124680 16.7 25.541 32.542 25.541 32.542 prepare_preconditioner 11 7.9 0.000 0.000 31.718 31.748 make_preconditioner 11 8.9 0.000 0.000 31.718 31.748 rebuild_ks_matrix 127 8.3 0.001 0.001 29.983 30.084 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.019 29.983 30.083 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.651 29.903 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.184 27.278 multiply_cannon_multrec 9940 15.5 10.305 14.236 17.770 20.845 qs_ot_get_derivative 116 11.6 0.001 0.002 19.504 19.571 ot_diis_step 116 11.6 0.019 0.021 19.517 19.518 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.095 19.433 apply_single 127 13.6 0.001 0.001 19.094 19.433 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.136 19.240 make_m2s 4970 13.5 0.065 0.070 16.026 18.263 cp_fm_cholesky_invert 11 10.9 18.198 18.204 18.198 18.204 qs_ot_get_p 127 10.4 0.001 0.001 17.930 17.986 make_images 4970 14.5 2.241 2.745 15.722 17.961 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.447 15.482 calculate_rho_elec 127 8.7 0.257 0.267 15.446 15.481 sum_up_and_integrate 127 10.3 0.180 0.189 15.088 15.142 integrate_v_rspace 127 11.3 0.004 0.004 14.907 14.970 qs_ot_p2m_diag 82 11.4 0.490 0.495 14.206 14.221 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.085 13.086 multiply_cannon_sync_h2d 9940 15.5 11.537 12.277 11.537 12.277 make_images_data 4970 15.5 0.052 0.060 9.817 12.111 hybrid_alltoall_any 5155 16.4 0.834 3.765 9.624 12.020 init_scf_run 11 5.9 0.000 0.001 10.486 10.486 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.486 10.486 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.076 10.087 cp_fm_diag_elpa_base 82 14.4 9.838 9.912 10.073 10.084 pw_transfer 1535 11.6 0.085 0.095 9.959 9.993 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.737 9.777 fft_wrap_pw1pw2_140 519 13.2 0.493 0.514 8.576 8.621 grid_integrate_task_list 127 12.3 7.703 8.155 7.703 8.155 cp_fm_cholesky_decompose 22 10.9 7.935 8.088 7.935 8.088 density_rs2pw 127 9.7 0.005 0.005 7.110 7.994 fft3d_ps 1281 14.7 2.695 2.772 7.913 7.935 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.766 7.814 dbcsr_mm_accdrv_process 20590 16.1 2.738 3.507 7.093 7.762 wfi_extrapolate 11 7.9 0.001 0.001 7.386 7.386 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.325 7.193 mp_allgather_i34 2485 14.5 2.771 6.923 2.771 6.923 calculate_dm_sparse 127 9.5 0.001 0.001 6.083 6.132 mp_alltoall_d11v 2401 14.1 4.965 6.019 4.965 6.019 grid_collocate_task_list 127 9.7 5.350 6.012 5.350 6.012 dbcsr_complete_redistribute 393 12.7 2.108 2.199 5.098 5.473 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 5.308 5.309 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.236 5.241 rs_pw_transfer 1038 11.9 0.012 0.013 4.126 5.049 potential_pw2rs 127 12.3 0.026 0.026 4.588 4.603 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.846 4.246 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.179 4.206 mp_irecv_dv 28618 15.9 1.809 4.172 1.809 4.172 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.573 3.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.699 3.707 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.646 3.667 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.314 3.614 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.520 3.588 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.472 3.492 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.236000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1448.090909, yerr=54.691622 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 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 3.014648E+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 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59820. MP_Allreduce 10885 1518. MP_Sync 86 MP_Alltoall 1700 36954426. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 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 0.025 0.050 299.174 299.175 qs_mol_dyn_low 1 2.0 0.003 0.005 298.649 298.651 qs_forces 11 3.9 0.004 0.005 298.528 298.532 qs_energies 11 4.9 0.002 0.002 289.176 289.195 scf_env_do_scf 11 5.9 0.002 0.002 266.975 266.986 velocity_verlet 10 3.0 0.001 0.002 215.226 215.234 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 140.570 140.571 init_scf_loop 11 6.9 0.000 0.000 126.146 126.148 prepare_preconditioner 11 7.9 0.000 0.000 121.055 121.087 make_preconditioner 11 8.9 0.000 0.000 121.055 121.087 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.633 118.212 qs_scf_new_mos 116 7.6 0.001 0.001 89.884 90.005 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.883 90.004 ot_scf_mini 116 9.6 0.004 0.004 85.134 85.195 dbcsr_multiply_generic 2485 12.5 0.212 0.221 82.471 83.037 cp_fm_upper_to_full 104 14.8 53.620 77.045 53.620 77.045 multiply_cannon 2485 13.5 0.694 0.765 58.456 58.875 multiply_cannon_loop 2485 14.5 0.473 0.493 54.836 56.150 ot_mini 116 10.6 0.001 0.001 44.023 44.086 dbcsr_complete_redistribute 393 12.7 3.974 4.013 30.008 43.587 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.633 40.247 rebuild_ks_matrix 127 8.3 0.001 0.001 38.216 38.259 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.019 38.215 38.258 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.377 37.883 mp_alltoall_i22 712 14.1 22.177 35.987 22.177 35.987 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.191 35.233 cp_fm_cholesky_invert 11 10.9 33.728 33.734 33.728 33.734 mp_waitall_1 102768 16.8 28.220 31.999 28.220 31.999 qs_ot_get_p 127 10.4 0.001 0.001 25.715 25.747 qs_ot_get_derivative 116 11.6 0.002 0.002 23.948 24.008 qs_ot_p2m_diag 82 11.4 0.868 0.873 21.743 21.773 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.658 20.672 calculate_rho_elec 127 8.7 0.479 0.480 20.657 20.672 make_m2s 4970 13.5 0.076 0.080 19.491 20.418 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 19.059 20.250 ot_diis_step 116 11.6 0.022 0.022 20.044 20.045 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.025 20.025 make_images 4970 14.5 3.694 3.816 19.012 19.943 sum_up_and_integrate 127 10.3 0.320 0.322 19.629 19.709 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.318 19.498 apply_single 127 13.6 0.001 0.001 19.318 19.498 integrate_v_rspace 127 11.3 0.004 0.004 19.309 19.389 multiply_cannon_multrec 9940 15.5 10.359 12.104 17.729 17.857 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.860 16.861 cp_fm_diag_elpa_base 82 14.4 12.414 14.067 16.856 16.856 multiply_cannon_sync_h2d 9940 15.5 15.533 15.545 15.533 15.545 pw_transfer 1535 11.6 0.092 0.092 12.311 12.316 init_scf_run 11 5.9 0.000 0.001 12.171 12.172 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.171 12.172 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.078 12.084 make_images_data 4970 15.5 0.060 0.065 10.386 12.049 hybrid_alltoall_any 5155 16.4 1.292 3.013 10.492 12.039 fft_wrap_pw1pw2_140 519 13.2 0.540 0.542 10.713 10.718 fft3d_ps 1281 14.7 2.726 2.731 10.128 10.135 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.320 9.367 cp_fm_cholesky_decompose 22 10.9 9.128 9.155 9.128 9.155 mp_alltoall_d11v 2401 14.1 8.140 9.146 8.140 9.146 wfi_extrapolate 11 7.9 0.001 0.001 9.078 9.078 dbcsr_mm_accdrv_process 20590 16.0 3.782 5.724 7.133 8.984 grid_integrate_task_list 127 12.3 8.510 8.682 8.510 8.682 density_rs2pw 127 9.7 0.005 0.005 8.322 8.451 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.080 8.080 calculate_dm_sparse 127 9.5 0.001 0.001 6.608 6.697 grid_collocate_task_list 127 9.7 6.314 6.356 6.314 6.356 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.268 6.322 rs_scatter_matrices 138 9.7 3.591 4.549 5.997 6.221 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.008 6.059 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.175000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2672.454545, yerr=173.851224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 1.261904E+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 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 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 0.016 0.037 85.166 85.168 qs_energies 1 2.0 0.000 0.000 84.696 84.703 ls_scf 1 3.0 0.000 0.000 83.788 83.795 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.627 72.795 multiply_cannon 111 7.7 0.018 0.021 55.912 57.239 multiply_cannon_loop 111 8.7 0.210 0.226 52.488 54.020 ls_scf_main 1 4.0 0.000 0.000 52.647 52.647 density_matrix_trs4 2 5.0 0.002 0.003 47.117 47.191 ls_scf_init_scf 1 4.0 0.000 0.000 28.080 28.081 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.973 27.028 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.863 24.883 mp_waitall_1 11031 10.9 22.242 24.536 22.242 24.536 multiply_cannon_multrec 2664 9.7 8.195 8.833 15.517 17.134 multiply_cannon_sync_h2d 2664 9.7 13.840 16.002 13.840 16.002 make_m2s 222 7.7 0.008 0.011 13.113 13.645 make_images 222 8.7 0.098 0.108 13.091 13.625 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.470 12.012 make_images_data 222 9.7 0.004 0.005 7.652 8.303 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.447 8.110 dbcsr_mm_accdrv_process 4760 10.4 0.520 0.650 6.939 7.876 hybrid_alltoall_any 227 10.6 0.217 1.841 6.486 7.838 dbcsr_mm_accdrv_process_sort 4760 11.4 6.217 7.083 6.217 7.083 calculate_norms 4752 9.8 5.558 6.174 5.558 6.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.130 mp_sum_l 807 5.4 3.067 4.472 3.067 4.472 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.073 3.702 mp_irecv_dv 6231 10.9 2.056 3.684 2.056 3.684 make_images_sizes 222 9.7 0.000 0.000 0.787 3.480 mp_alltoall_i44 222 10.7 0.787 3.480 0.787 3.480 arnoldi_extremal 4 6.8 0.000 0.000 3.417 3.441 arnoldi_normal_ev 4 7.8 0.001 0.003 3.417 3.441 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.252 3.406 build_subspace 16 8.4 0.009 0.012 3.308 3.312 ls_scf_post 1 4.0 0.000 0.000 3.061 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.879 2.920 dbcsr_special_finalize 555 9.7 0.005 0.006 2.317 2.745 dbcsr_merge_single_wm 555 10.7 0.464 0.593 2.310 2.737 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.446 2.674 make_images_pack 222 9.7 2.211 2.624 2.213 2.625 dbcsr_sort_data 658 11.4 2.103 2.498 2.103 2.498 dbcsr_matrix_vector_mult_local 304 10.0 2.076 2.490 2.078 2.492 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.298 2.387 buffer_matrices_ensure_size 222 8.7 1.758 2.017 1.758 2.017 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.785 1.787 rebuild_ks_matrix 3 7.3 0.000 0.000 1.776 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.776 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.168000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 2.110489E+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_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 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 0.049 0.082 91.598 91.598 qs_energies 1 2.0 0.000 0.000 90.981 90.985 ls_scf 1 3.0 0.000 0.000 89.657 89.661 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.648 75.952 multiply_cannon 111 7.7 0.028 0.043 53.360 57.402 ls_scf_main 1 4.0 0.000 0.000 55.303 55.309 multiply_cannon_loop 111 8.7 0.116 0.123 50.086 53.372 density_matrix_trs4 2 5.0 0.002 0.003 49.583 49.798 mp_waitall_1 9105 10.9 21.438 30.868 21.438 30.868 ls_scf_init_scf 1 4.0 0.000 0.002 30.843 30.845 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.555 29.650 multiply_cannon_multrec 1332 9.7 13.109 17.062 22.316 27.420 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.149 27.163 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.888 21.082 make_m2s 222 7.7 0.006 0.007 15.554 16.286 make_images 222 8.7 1.573 1.916 15.524 16.256 dbcsr_mm_accdrv_process 4041 10.4 0.277 0.476 8.805 10.361 make_images_data 222 9.7 0.004 0.004 8.978 9.963 dbcsr_mm_accdrv_process_sort 4041 11.4 8.389 9.914 8.389 9.914 hybrid_alltoall_any 227 10.6 0.521 2.454 8.381 9.609 mp_sum_l 807 5.4 5.445 9.005 5.445 9.005 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.211 7.850 mp_irecv_dv 3311 11.0 3.192 7.800 3.192 7.800 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.171 7.055 calculate_norms 2376 9.8 6.006 6.724 6.006 6.724 multiply_cannon_sync_h2d 1332 9.7 4.878 5.864 4.878 5.864 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.033 5.271 arnoldi_extremal 4 6.8 0.000 0.000 4.613 4.637 arnoldi_normal_ev 4 7.8 0.001 0.004 4.613 4.637 build_subspace 16 8.4 0.014 0.021 4.353 4.356 ls_scf_post 1 4.0 0.000 0.000 3.511 3.515 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.121 3.359 ls_scf_store_result 1 5.0 0.000 0.000 3.223 3.333 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.222 2.738 3.224 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.204 2.792 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.615 2.699 mp_allgather_i34 111 8.7 0.969 2.533 0.969 2.533 make_images_pack 222 9.7 2.025 2.430 2.027 2.432 dbcsr_sort_data 436 11.2 1.849 2.053 1.849 2.053 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.959 1.961 rebuild_ks_matrix 3 7.3 0.000 0.000 1.946 1.949 qs_ks_build_kohn_sham_matrix 3 8.3 0.019 0.031 1.946 1.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.598000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 2.703475E+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_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 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 0.040 0.066 94.078 94.083 qs_energies 1 2.0 0.000 0.000 93.350 93.353 ls_scf 1 3.0 0.000 0.001 91.916 91.921 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.580 76.913 ls_scf_main 1 4.0 0.000 0.004 57.429 57.433 multiply_cannon 111 7.7 0.042 0.112 52.923 56.769 multiply_cannon_loop 111 8.7 0.100 0.111 49.340 53.390 density_matrix_trs4 2 5.0 0.002 0.004 51.378 51.524 mp_waitall_1 7281 11.0 24.020 34.177 24.020 34.177 ls_scf_init_scf 1 4.0 0.000 0.001 30.874 30.876 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.633 29.732 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.220 27.234 multiply_cannon_multrec 888 9.7 12.518 15.175 21.031 24.288 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.261 23.492 make_m2s 222 7.7 0.006 0.007 16.976 18.192 make_images 222 8.7 1.970 2.291 16.938 18.154 make_images_data 222 9.7 0.003 0.004 9.716 10.763 hybrid_alltoall_any 227 10.6 0.619 2.859 9.400 10.718 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.417 8.034 9.256 dbcsr_mm_accdrv_process_sort 3754 11.4 7.665 8.839 7.665 8.839 mp_sum_l 807 5.4 5.130 8.220 5.130 8.220 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.663 7.725 multiply_cannon_sync_h2d 888 9.7 6.042 7.396 6.042 7.396 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.468 7.109 mp_irecv_dv 2335 11.1 2.454 7.067 2.454 7.067 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.897 6.418 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.996 5.178 arnoldi_extremal 4 6.8 0.000 0.000 5.075 5.094 arnoldi_normal_ev 4 7.8 0.001 0.005 5.075 5.094 build_subspace 16 8.4 0.014 0.020 4.770 4.776 calculate_norms 1584 9.8 4.263 4.638 4.263 4.638 mp_allgather_i34 111 8.7 1.423 3.772 1.423 3.772 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.433 3.750 ls_scf_post 1 4.0 0.000 0.001 3.613 3.618 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.580 3.015 3.582 ls_scf_store_result 1 5.0 0.000 0.000 3.363 3.438 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.918 3.036 dbcsr_sort_data 325 11.1 1.896 2.136 1.896 2.136 make_images_pack 222 9.7 1.816 2.116 1.819 2.119 make_images_sizes 222 9.7 0.000 0.000 0.933 2.015 mp_alltoall_i44 222 10.7 0.933 2.015 0.933 2.015 dbcsr_data_release 9322 10.9 1.323 1.972 1.323 1.972 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.948 1.950 rebuild_ks_matrix 3 7.3 0.000 0.000 1.930 1.932 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.930 1.932 dbcsr_finalize 304 7.8 0.026 0.032 1.621 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.083000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2192.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 3.343647E+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_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 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 0.077 0.144 96.553 96.576 qs_energies 1 2.0 0.000 0.000 95.823 95.829 ls_scf 1 3.0 0.000 0.000 94.177 94.181 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.005 78.289 ls_scf_main 1 4.0 0.000 0.000 58.532 58.537 multiply_cannon 111 7.7 0.056 0.110 51.325 55.794 density_matrix_trs4 2 5.0 0.002 0.003 52.373 52.525 multiply_cannon_loop 111 8.7 0.114 0.125 46.221 49.298 ls_scf_init_scf 1 4.0 0.000 0.000 32.418 32.419 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.208 31.271 mp_waitall_1 6369 11.0 22.322 28.880 22.322 28.880 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.763 28.777 multiply_cannon_multrec 1332 9.7 14.137 16.978 21.978 25.625 make_m2s 222 7.7 0.007 0.007 21.003 22.461 make_images 222 8.7 3.135 3.594 20.953 22.413 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.009 16.583 make_images_data 222 9.7 0.004 0.004 11.665 13.279 hybrid_alltoall_any 227 10.6 0.798 3.808 10.967 12.603 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.411 7.481 8.985 dbcsr_mm_accdrv_process_sort 3641 11.4 7.105 8.569 7.105 8.569 mp_sum_l 807 5.4 4.126 7.684 4.126 7.684 multiply_cannon_sync_h2d 1332 9.7 5.542 6.132 5.542 6.132 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.129 5.974 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.090 5.962 mp_irecv_dv 3229 10.9 2.067 5.916 2.067 5.916 arnoldi_extremal 4 6.8 0.000 0.000 5.193 5.207 arnoldi_normal_ev 4 7.8 0.001 0.005 5.193 5.206 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.477 4.877 build_subspace 16 8.4 0.014 0.021 4.851 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.772 mp_allgather_i34 111 8.7 2.224 4.754 2.224 4.754 calculate_norms 2376 9.8 4.191 4.527 4.191 4.527 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.570 3.859 dbcsr_matrix_vector_mult_local 304 10.0 3.173 3.670 3.175 3.672 dbcsr_sort_data 658 11.4 3.094 3.532 3.094 3.532 dbcsr_special_finalize 555 9.7 0.006 0.007 2.845 3.266 dbcsr_merge_single_wm 555 10.7 0.540 0.668 2.836 3.258 ls_scf_post 1 4.0 0.000 0.000 3.227 3.232 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.044 3.118 ls_scf_store_result 1 5.0 0.000 0.000 2.970 3.032 dbcsr_data_release 10477 10.7 1.570 2.405 1.570 2.405 dbcsr_finalize 304 7.8 0.049 0.061 1.794 1.988 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.967 1.969 rebuild_ks_matrix 3 7.3 0.000 0.000 1.944 1.946 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.944 1.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.576000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 4.675265E+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_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 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 0.049 0.058 98.935 98.936 qs_energies 1 2.0 0.000 0.000 98.131 98.135 ls_scf 1 3.0 0.000 0.000 96.167 96.177 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.492 77.675 ls_scf_main 1 4.0 0.000 0.000 61.707 61.708 multiply_cannon 111 7.7 0.094 0.184 55.070 60.160 density_matrix_trs4 2 5.0 0.002 0.003 54.498 54.585 multiply_cannon_loop 111 8.7 0.069 0.076 50.478 52.050 mp_waitall_1 5436 11.0 26.228 32.327 26.228 32.327 ls_scf_init_scf 1 4.0 0.000 0.000 30.856 30.861 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.465 29.506 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.298 27.310 multiply_cannon_multrec 444 9.7 13.900 16.088 20.927 23.215 make_m2s 222 7.7 0.004 0.005 17.627 20.177 make_images 222 8.7 3.717 4.436 17.565 20.117 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.880 16.094 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.274 15.544 make_images_data 222 9.7 0.003 0.004 9.875 12.458 hybrid_alltoall_any 227 10.6 0.789 3.763 9.695 12.353 multiply_cannon_sync_h2d 444 9.7 6.498 7.857 6.498 7.857 dbcsr_mm_accdrv_process 3003 10.4 0.165 0.345 6.722 7.846 dbcsr_mm_accdrv_process_sort 3003 11.4 6.407 7.498 6.407 7.498 mp_allgather_i34 111 8.7 2.822 6.986 2.822 6.986 arnoldi_extremal 4 6.8 0.000 0.000 5.820 5.832 arnoldi_normal_ev 4 7.8 0.003 0.005 5.819 5.832 build_subspace 16 8.4 0.015 0.020 5.424 5.434 mp_sum_l 807 5.4 2.841 4.981 2.841 4.981 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.590 4.780 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.173 4.357 dbcsr_matrix_vector_mult_local 304 10.0 3.671 4.150 3.673 4.152 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.666 3.832 mp_irecv_dv 1241 11.2 1.647 3.801 1.647 3.801 calculate_norms 792 9.8 3.586 3.741 3.586 3.741 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.962 3.733 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.602 3.696 ls_scf_post 1 4.0 0.000 0.000 3.605 3.609 make_images_sizes 222 9.7 0.000 0.000 1.117 3.521 mp_alltoall_i44 222 10.7 1.117 3.521 1.117 3.521 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.429 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.293 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.195 2.197 rebuild_ks_matrix 3 7.3 0.000 0.000 2.163 2.164 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.162 2.164 dbcsr_merge_all 275 8.9 0.475 0.531 2.052 2.128 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.936000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3633.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ed9166ef7988979478d40eb36b21c1c136d249d6_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 8.759226E+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_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.100 0.151 107.953 107.965 qs_energies 1 2.0 0.000 0.000 106.477 106.490 ls_scf 1 3.0 0.000 0.000 103.524 103.536 dbcsr_multiply_generic 111 6.7 0.023 0.026 76.712 76.835 ls_scf_main 1 4.0 0.000 0.000 65.844 65.845 density_matrix_trs4 2 5.0 0.002 0.003 56.436 56.507 multiply_cannon 111 7.7 0.153 0.267 49.301 51.341 multiply_cannon_loop 111 8.7 0.067 0.069 45.913 46.298 ls_scf_init_scf 1 4.0 0.000 0.000 33.952 33.952 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.302 32.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.505 29.515 mp_waitall_1 4527 11.1 21.539 25.293 21.539 25.293 make_m2s 222 7.7 0.005 0.005 23.715 24.639 make_images 222 8.7 4.583 5.005 23.608 24.531 multiply_cannon_multrec 444 9.7 17.851 18.521 22.445 22.898 hybrid_alltoall_any 227 10.6 1.660 3.622 12.649 15.613 make_images_data 222 9.7 0.003 0.003 12.939 15.357 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.021 10.734 multiply_cannon_sync_h2d 444 9.7 8.889 9.039 8.889 9.039 arnoldi_extremal 4 6.8 0.000 0.000 7.417 7.428 arnoldi_normal_ev 4 7.8 0.003 0.009 7.417 7.428 build_subspace 16 8.4 0.026 0.036 6.834 6.845 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.438 5.617 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.360 5.451 dbcsr_matrix_vector_mult_local 304 10.0 4.978 5.292 4.981 5.295 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.275 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.321 4.422 4.550 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.253 4.120 4.253 ls_scf_post 1 4.0 0.000 0.000 3.728 3.740 make_images_sizes 222 9.7 0.000 0.000 1.453 3.670 mp_alltoall_i44 222 10.7 1.452 3.670 1.452 3.670 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.423 calculate_norms 792 9.8 3.200 3.270 3.200 3.270 mp_allgather_i34 111 8.7 1.035 3.236 1.035 3.236 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.180 dbcsr_merge_all 275 8.9 0.891 0.916 2.871 2.961 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.922 2.922 dbcsr_complete_redistribute 5 7.6 1.429 1.465 2.772 2.889 dbcsr_data_release 12724 10.6 2.329 2.845 2.329 2.845 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.625 2.626 matrix_ls_to_qs 2 6.0 0.000 0.000 2.444 2.572 rebuild_ks_matrix 3 7.3 0.000 0.000 2.559 2.560 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.559 2.560 dbcsr_sort_data 325 11.1 2.440 2.517 2.440 2.517 dbcsr_new_transposed 4 7.5 0.246 0.258 2.325 2.356 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.187 2.225 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.204 dbcsr_add_anytype 103 7.2 0.860 0.894 2.132 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.965000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6922.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ed9166ef7988979478d40eb36b21c1c136d249d6 Summary: empty Status: OK