=== 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: 784694e3a5332e1a8c4df163d64bc806796dde08 ################# 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 2.6.6, 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 (19.04.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/01 job id: 46200290 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/02 job id: 46200291 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/03 job id: 46200293 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/04 job id: 46200294 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/05 job id: 46200296 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/06 job id: 46200298 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/07 job id: 46200301 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/08 job id: 46200302 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/09 job id: 46200304 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/10 job id: 46200305 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/11 job id: 46200306 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/12 job id: 46200308 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/13 job id: 46200310 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/14 job id: 46200312 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/15 job id: 46200313 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/16 job id: 46200314 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/17 job id: 46200316 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/18 job id: 46200317 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/19 job id: 46200318 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/20 job id: 46200319 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/21 job id: 46200321 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/22 job id: 46200322 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/23 job id: 46200323 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/24 job id: 46200324 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/25 job id: 46200325 --- 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/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/26 job id: 46200326 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/27 job id: 46200327 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 0.039 0.043 133.461 133.462 farming_run 1 2.0 133.018 133.021 133.412 133.436 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457693E+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 230 1103589. 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 0.006 0.025 115.245 115.245 qs_energies 1 2.0 0.000 0.000 115.056 115.058 mp2_main 1 3.0 0.000 0.000 113.014 113.017 mp2_gpw_main 1 4.0 0.019 0.026 112.008 112.011 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.372 93.495 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.357 55.478 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 41.668 46.636 get_2c_integrals 1 6.0 0.008 0.009 37.138 37.842 integrate_v_rspace 273 8.0 0.434 0.448 25.049 29.689 pw_transfer 6555 10.6 0.377 0.384 27.440 27.959 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.083 26.399 grid_integrate_task_list 273 9.0 20.865 25.960 20.865 25.960 fft_wrap_pw1pw2_100 2178 12.4 1.222 1.457 23.606 23.936 compute_2c_integrals 1 7.0 0.002 0.002 19.742 19.744 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.935 19.477 mp2_eri_2c_integrate_gpw 1 9.0 2.367 2.431 18.932 19.476 rpa_ri_compute_en 1 5.0 0.019 0.022 18.527 18.639 cp_fm_cholesky_decompose 12 8.2 17.329 18.058 17.329 18.058 cholesky_decomp 1 7.0 0.000 0.000 16.243 16.971 fft3d_s 5443 13.4 16.158 16.334 16.181 16.358 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.561 10.855 15.561 calculate_wavefunction 272 8.0 5.421 5.495 12.547 13.174 rpa_num_int 1 6.0 0.002 0.025 10.626 10.626 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.587 10.591 calc_mat_Q 8 8.0 0.000 0.000 9.455 9.536 contract_S_to_Q 8 9.0 0.000 0.000 8.879 8.958 calc_potential_gpw 544 9.5 0.005 0.006 8.275 8.675 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.246 8.578 parallel_gemm_fm 14 9.1 0.000 0.000 8.469 8.564 parallel_gemm_fm_cosma 14 10.1 8.469 8.564 8.469 8.564 potential_pw2rs 545 10.0 0.107 0.109 7.689 8.343 collocate_single_gaussian 272 10.0 0.039 0.042 7.501 7.795 create_integ_mat 1 6.0 0.022 0.028 7.684 7.684 array2fm 1 7.0 0.000 0.000 6.728 7.125 pw_scatter_s 2720 13.7 4.444 4.643 4.444 4.643 pw_gather_s 2722 13.2 3.854 4.311 3.854 4.311 array2fm_buffer_send 1 8.0 2.959 3.158 2.959 3.158 pw_poisson_solve 545 10.5 1.104 1.140 2.165 2.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.008090, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 0.027 0.039 395.378 395.380 farming_run 1 2.0 394.695 394.707 395.338 395.340 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 897827128576 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 249788821 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 1.223991E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 29658893. 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 0.009 0.031 209.693 209.694 qs_energies 1 2.0 0.000 0.000 209.483 209.490 scf_env_do_scf 1 3.0 0.000 0.000 106.726 106.726 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.850 105.858 rebuild_ks_matrix 4 6.0 0.000 0.000 105.848 105.857 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.848 105.857 hfx_ks_matrix 4 8.0 0.001 0.001 105.467 105.472 integrate_four_center 4 9.0 0.143 0.463 105.466 105.471 mp2_main 1 3.0 0.000 0.000 102.469 102.476 mp2_gpw_main 1 4.0 0.033 0.045 101.620 101.626 integrate_four_center_main 4 10.0 0.120 0.563 96.458 99.713 integrate_four_center_bin 267 11.0 96.338 99.683 96.338 99.683 init_scf_loop 1 4.0 0.000 0.000 92.529 92.529 mp2_ri_gpw_compute_in 1 5.0 0.068 0.076 74.946 75.962 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.537 55.551 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 42.085 47.258 integrate_v_rspace 95 8.0 0.395 0.567 28.457 33.595 pw_transfer 2240 10.6 0.145 0.172 29.933 30.477 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.953 29.468 ao_to_mo_and_store_B_mult_1 91 7.0 10.761 29.345 10.761 29.345 grid_integrate_task_list 95 9.0 23.742 29.077 23.742 29.077 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.527 28.272 fft_wrap_pw1pw2_100 730 12.4 1.254 1.433 26.676 27.211 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.897 24.880 24.891 get_2c_integrals 1 6.0 0.000 0.000 20.316 20.345 compute_2c_integrals 1 7.0 0.003 0.004 19.296 19.306 compute_2c_integrals_loop_lm 1 8.0 0.004 0.008 18.835 19.177 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.833 18.832 19.176 fft3d_s 1823 13.4 18.456 18.807 18.470 18.820 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.196 14.196 calculate_wavefunction 91 8.0 2.008 2.041 9.741 9.993 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.593 8.769 9.284 potential_pw2rs 186 10.0 0.033 0.034 8.636 9.207 local_gemm 172 8.0 8.213 8.722 8.213 8.722 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.210 8.575 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.527 7.881 8.310 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.145 collocate_single_gaussian 91 10.0 0.017 0.024 7.879 8.099 mp_sync 37 10.5 3.669 6.516 3.669 6.516 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.433 6.355 6.433 mp_sendrecv_dm3 2068 8.0 5.905 6.323 5.905 6.323 pw_gather_s 912 13.2 4.939 5.423 4.939 5.423 pw_scatter_s 910 13.7 3.916 4.244 3.916 4.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.618282, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1506.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 449.155072E+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 811225. 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 0.022 0.049 53.503 53.505 qs_mol_dyn_low 1 2.0 0.003 0.006 53.185 53.192 qs_forces 11 3.9 0.004 0.009 53.111 53.112 qs_energies 11 4.9 0.002 0.008 51.649 51.662 scf_env_do_scf 11 5.9 0.001 0.001 45.624 45.624 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.538 43.538 qs_scf_new_mos 108 7.5 0.000 0.001 33.500 33.774 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.500 33.774 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.222 33.721 ot_scf_mini 108 9.5 0.002 0.003 31.798 31.973 multiply_cannon 2286 13.5 0.186 0.191 25.738 27.493 multiply_cannon_loop 2286 14.5 1.480 1.563 25.067 26.835 velocity_verlet 10 3.0 0.003 0.013 25.756 25.757 ot_mini 108 10.5 0.001 0.003 19.182 19.432 qs_ot_get_derivative 108 11.5 0.001 0.001 16.274 16.477 mp_waitall_1 245248 16.5 8.383 14.760 8.383 14.760 multiply_cannon_metrocomm3 54864 15.5 0.068 0.076 5.984 12.901 multiply_cannon_multrec 54864 15.5 4.244 6.648 7.449 10.851 qs_ot_get_p 119 10.4 0.001 0.002 7.964 8.310 rebuild_ks_matrix 119 8.3 0.000 0.000 7.847 7.985 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 7.846 7.985 mp_sum_l 7207 12.9 5.665 7.438 5.665 7.438 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.916 7.041 multiply_cannon_sync_h2d 54864 15.5 5.929 6.855 5.929 6.855 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.548 5.991 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.579 5.713 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.329 5.380 init_scf_run 11 5.9 0.000 0.001 4.726 4.726 scf_env_initial_rho_setup 11 6.9 0.001 0.007 4.725 4.726 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.500 4.501 sum_up_and_integrate 119 10.3 0.012 0.016 4.475 4.481 integrate_v_rspace 119 11.3 0.003 0.003 4.462 4.470 dbcsr_mm_accdrv_process 76910 16.1 1.132 1.798 3.127 4.394 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.315 4.315 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.173 4.315 calculate_rho_elec 119 8.7 0.012 0.018 4.172 4.314 cp_fm_redistribute_end 50 14.0 2.201 4.292 2.206 4.294 cp_fm_diag_elpa_base 50 14.0 2.083 4.186 2.086 4.194 calculate_dm_sparse 119 9.5 0.001 0.001 2.916 3.083 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.882 3.072 apply_single 119 13.6 0.000 0.000 2.882 3.072 rs_pw_transfer 974 11.9 0.012 0.013 2.721 2.906 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.766 2.841 ot_diis_step 108 11.5 0.006 0.009 2.666 2.667 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.491 2.590 jit_kernel_multiply 13 15.8 1.934 2.507 1.934 2.507 calculate_first_density_matrix 1 7.0 0.001 0.005 2.500 2.506 density_rs2pw 119 9.7 0.004 0.004 2.168 2.366 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.284 2.287 acc_transpose_blocks 54864 15.5 0.227 0.247 1.742 2.208 wfi_extrapolate 11 7.9 0.001 0.003 2.156 2.156 grid_integrate_task_list 119 12.3 2.029 2.134 2.029 2.134 init_scf_loop 11 6.9 0.001 0.005 2.066 2.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.990 2.060 mp_sum_d 4135 12.0 1.259 1.842 1.259 1.842 potential_pw2rs 119 12.3 0.004 0.004 1.826 1.836 pw_transfer 1439 11.6 0.052 0.056 1.717 1.788 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.642 1.712 make_m2s 4572 13.5 0.054 0.056 1.531 1.572 mp_alltoall_d11v 2130 13.8 1.271 1.504 1.271 1.504 make_images 4572 14.5 0.133 0.138 1.449 1.489 mp_waitany 12084 13.8 1.262 1.486 1.262 1.486 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.370 1.395 grid_collocate_task_list 119 9.7 1.291 1.366 1.291 1.366 fft3d_ps 1201 14.6 0.375 0.490 1.286 1.350 fft_wrap_pw1pw2_140 487 13.2 0.183 0.198 1.273 1.343 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.699 1.084 acc_transpose_blocks_kernels 54864 16.5 0.245 0.387 0.776 1.072 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.505000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 484.462592E+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 1045248. 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 0.027 0.037 38.050 38.051 qs_mol_dyn_low 1 2.0 0.003 0.004 37.826 37.834 qs_forces 11 3.9 0.002 0.003 37.765 37.766 qs_energies 11 4.9 0.002 0.007 36.081 36.085 scf_env_do_scf 11 5.9 0.001 0.001 31.066 31.066 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.650 28.651 dbcsr_multiply_generic 2286 12.5 0.099 0.101 21.026 21.377 qs_scf_new_mos 108 7.5 0.001 0.001 19.693 19.945 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.692 19.944 ot_scf_mini 108 9.5 0.002 0.003 18.818 18.996 velocity_verlet 10 3.0 0.001 0.002 17.866 17.868 multiply_cannon 2286 13.5 0.208 0.214 16.159 17.505 multiply_cannon_loop 2286 14.5 0.891 0.964 15.084 16.510 ot_mini 108 10.5 0.001 0.001 11.555 11.795 mp_waitall_1 200699 16.5 5.427 10.693 5.427 10.693 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.040 9.442 qs_ot_get_derivative 108 11.5 0.001 0.001 9.108 9.289 multiply_cannon_multrec 27432 15.5 1.993 4.236 5.863 8.600 rebuild_ks_matrix 119 8.3 0.000 0.000 7.100 7.236 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.099 7.236 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.260 6.383 dbcsr_mm_accdrv_process 47894 16.0 3.068 5.170 3.799 5.620 qs_ot_get_p 119 10.4 0.001 0.001 4.382 4.620 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.450 4.365 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.107 4.113 apply_single 119 13.6 0.000 0.000 3.107 4.113 sum_up_and_integrate 119 10.3 0.024 0.027 4.104 4.111 integrate_v_rspace 119 11.3 0.002 0.003 4.080 4.089 mp_sum_l 7207 12.9 2.063 3.972 2.063 3.972 init_scf_run 11 5.9 0.000 0.001 3.786 3.787 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.786 3.786 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.665 3.694 calculate_rho_elec 119 8.7 0.021 0.024 3.664 3.693 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.980 2.999 rs_pw_transfer 974 11.9 0.010 0.012 2.451 2.951 multiply_cannon_sync_h2d 27432 15.5 2.226 2.913 2.226 2.913 make_m2s 4572 13.5 0.052 0.054 2.361 2.572 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.553 2.553 density_rs2pw 119 9.7 0.004 0.005 2.000 2.510 make_images 4572 14.5 0.200 0.237 2.273 2.484 init_scf_loop 11 6.9 0.001 0.005 2.396 2.398 ot_diis_step 108 11.5 0.010 0.011 2.397 2.398 calculate_first_density_matrix 1 7.0 0.000 0.002 2.306 2.308 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.245 2.246 cp_fm_redistribute_end 50 14.0 1.138 2.221 1.141 2.224 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.079 2.172 cp_fm_diag_elpa_base 50 14.0 1.051 2.127 1.079 2.169 calculate_dm_sparse 119 9.5 0.000 0.001 2.020 2.096 grid_integrate_task_list 119 12.3 1.830 1.942 1.830 1.942 pw_transfer 1439 11.6 0.065 0.070 1.884 1.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.862 1.865 potential_pw2rs 119 12.3 0.006 0.006 1.840 1.851 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.793 1.822 jit_kernel_multiply 9 16.2 0.679 1.736 0.679 1.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.669 1.707 make_images_data 4572 15.5 0.045 0.051 1.107 1.518 prepare_preconditioner 11 7.9 0.000 0.000 1.487 1.515 make_preconditioner 11 8.9 0.000 0.001 1.487 1.515 acc_transpose_blocks 27432 15.5 0.109 0.115 1.185 1.485 fft_wrap_pw1pw2_140 487 13.2 0.202 0.210 1.435 1.466 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.392 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.424 1.424 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.958 1.383 grid_collocate_task_list 119 9.7 1.225 1.371 1.225 1.371 fft3d_ps 1201 14.6 0.521 0.577 1.344 1.370 mp_alltoall_d11v 2130 13.8 1.120 1.299 1.120 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.259 mp_allgather_i34 2286 14.5 0.516 1.253 0.516 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.160 1.212 mp_waitany 5720 13.7 0.537 1.074 0.537 1.074 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.150 0.555 1.055 mp_sum_d 4135 12.0 0.573 1.025 0.573 1.025 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.962 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.899 0.912 acc_transpose_blocks_kernels 27432 16.5 0.180 0.268 0.652 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.051000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.545455, yerr=2.061052 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 517.922816E+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 937146. 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 0.017 0.033 31.758 31.759 qs_mol_dyn_low 1 2.0 0.003 0.004 31.491 31.499 qs_forces 11 3.9 0.003 0.009 31.435 31.436 qs_energies 11 4.9 0.001 0.006 29.767 29.770 scf_env_do_scf 11 5.9 0.001 0.001 25.077 25.078 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.580 22.582 dbcsr_multiply_generic 2286 12.5 0.095 0.098 16.057 16.135 velocity_verlet 10 3.0 0.001 0.001 14.945 14.947 qs_scf_new_mos 108 7.5 0.001 0.001 14.512 14.530 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.512 14.530 ot_scf_mini 108 9.5 0.002 0.003 13.802 13.821 multiply_cannon 2286 13.5 0.195 0.202 12.932 13.729 multiply_cannon_loop 2286 14.5 0.629 0.656 12.167 12.986 ot_mini 108 10.5 0.001 0.001 8.490 8.512 multiply_cannon_multrec 18288 15.5 1.941 2.863 6.717 7.078 qs_ot_get_derivative 108 11.5 0.001 0.002 7.004 7.022 rebuild_ks_matrix 119 8.3 0.000 0.000 6.410 6.430 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.410 6.430 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.567 5.585 dbcsr_mm_accdrv_process 38222 16.0 3.911 4.971 4.691 5.404 sum_up_and_integrate 119 10.3 0.030 0.031 3.995 4.014 integrate_v_rspace 119 11.3 0.003 0.004 3.964 3.987 init_scf_run 11 5.9 0.000 0.001 3.463 3.463 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.463 3.463 mp_waitall_1 158411 16.6 2.470 3.408 2.470 3.408 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.393 3.405 calculate_rho_elec 119 8.7 0.031 0.031 3.393 3.405 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.601 3.234 qs_ot_get_p 119 10.4 0.001 0.002 3.180 3.200 init_scf_loop 11 6.9 0.001 0.003 2.481 2.485 rs_pw_transfer 974 11.9 0.009 0.010 2.193 2.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.045 2.343 apply_single 119 13.6 0.000 0.000 2.045 2.343 calculate_first_density_matrix 1 7.0 0.000 0.002 2.266 2.267 qs_ot_p2m_diag 50 11.0 0.012 0.014 2.164 2.169 density_rs2pw 119 9.7 0.004 0.004 1.897 2.127 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.374 2.105 pw_transfer 1439 11.6 0.065 0.069 1.972 1.995 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.907 1.907 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.880 1.905 grid_integrate_task_list 119 12.3 1.803 1.901 1.803 1.901 jit_kernel_multiply 10 16.1 0.728 1.861 0.728 1.861 make_m2s 4572 13.5 0.044 0.045 1.695 1.848 calculate_dm_sparse 119 9.5 0.001 0.001 1.795 1.810 potential_pw2rs 119 12.3 0.007 0.008 1.764 1.782 make_images 4572 14.5 0.189 0.202 1.610 1.762 prepare_preconditioner 11 7.9 0.000 0.000 1.695 1.697 make_preconditioner 11 8.9 0.000 0.001 1.695 1.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.689 1.691 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.670 1.681 cp_fm_diag_elpa_base 50 14.0 1.648 1.657 1.669 1.680 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.547 1.632 multiply_cannon_sync_h2d 18288 15.5 1.419 1.593 1.419 1.593 fft_wrap_pw1pw2_140 487 13.2 0.254 0.267 1.538 1.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.551 1.559 qs_ot_get_derivative_diag 49 12.0 0.001 0.003 1.507 1.520 ot_diis_step 108 11.5 0.011 0.011 1.468 1.468 fft3d_ps 1201 14.6 0.529 0.544 1.365 1.390 mp_sum_l 7207 12.9 1.048 1.390 1.048 1.390 grid_collocate_task_list 119 9.7 1.215 1.350 1.215 1.350 acc_transpose_blocks 18288 15.5 0.076 0.078 1.259 1.283 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.179 wfi_extrapolate 11 7.9 0.001 0.001 1.142 1.142 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.988 0.989 make_images_data 4572 15.5 0.045 0.049 0.768 0.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.914 0.916 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.854 0.879 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.667 0.857 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.817 0.831 mp_alltoall_d11v 2130 13.8 0.645 0.807 0.645 0.807 mp_alltoall_z22v 1201 16.6 0.694 0.772 0.694 0.772 mp_waitany 9880 13.7 0.527 0.747 0.527 0.747 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.126 0.522 0.736 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.662 0.728 cp_fm_cholesky_invert 11 10.9 0.702 0.705 0.702 0.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.759000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.545455, yerr=1.776802 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 538.722304E+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 0.012 0.028 34.346 34.346 qs_mol_dyn_low 1 2.0 0.003 0.004 34.191 34.198 qs_forces 11 3.9 0.002 0.003 34.129 34.130 qs_energies 11 4.9 0.001 0.001 32.419 32.424 scf_env_do_scf 11 5.9 0.002 0.004 27.518 27.519 scf_env_do_scf_inner_loop 108 6.5 0.004 0.007 24.160 24.160 dbcsr_multiply_generic 2286 12.5 0.098 0.101 17.744 17.840 velocity_verlet 10 3.0 0.001 0.002 17.537 17.539 qs_scf_new_mos 108 7.5 0.001 0.001 15.816 15.867 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.815 15.867 ot_scf_mini 108 9.5 0.002 0.003 14.889 14.941 multiply_cannon 2286 13.5 0.237 0.276 14.307 14.658 multiply_cannon_loop 2286 14.5 0.932 0.965 13.391 13.730 ot_mini 108 10.5 0.001 0.001 9.024 9.088 multiply_cannon_multrec 27432 15.5 2.327 2.997 8.559 8.923 dbcsr_mm_accdrv_process 47916 15.9 5.330 7.270 6.137 7.326 qs_ot_get_derivative 108 11.5 0.001 0.002 7.221 7.273 rebuild_ks_matrix 119 8.3 0.000 0.000 6.513 6.556 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.513 6.556 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.778 5.815 sum_up_and_integrate 119 10.3 0.035 0.037 3.765 3.775 integrate_v_rspace 119 11.3 0.003 0.003 3.730 3.739 init_scf_run 11 5.9 0.000 0.001 3.587 3.588 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.587 3.587 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.470 3.505 calculate_rho_elec 119 8.7 0.040 0.046 3.470 3.505 qs_ot_get_p 119 10.4 0.001 0.001 3.287 3.365 init_scf_loop 11 6.9 0.000 0.000 3.336 3.338 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.707 3.096 prepare_preconditioner 11 7.9 0.000 0.000 2.506 2.515 make_preconditioner 11 8.9 0.000 0.000 2.506 2.515 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.108 2.437 make_m2s 4572 13.5 0.054 0.055 2.113 2.229 calculate_first_density_matrix 1 7.0 0.000 0.000 2.227 2.229 mp_waitall_1 137007 16.6 1.696 2.201 1.696 2.201 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.033 2.160 apply_single 119 13.6 0.000 0.000 2.033 2.160 rs_pw_transfer 974 11.9 0.009 0.009 1.929 2.125 make_images 4572 14.5 0.270 0.331 2.007 2.122 density_rs2pw 119 9.7 0.004 0.004 1.890 2.083 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.065 2.074 calculate_dm_sparse 119 9.5 0.000 0.000 2.014 2.066 pw_transfer 1439 11.6 0.065 0.070 1.973 2.007 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.881 1.919 grid_integrate_task_list 119 12.3 1.818 1.903 1.818 1.903 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.867 1.894 jit_kernel_multiply 10 16.0 0.747 1.817 0.747 1.817 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.801 1.802 ot_diis_step 108 11.5 0.012 0.012 1.764 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.758 1.759 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.634 1.645 fft_wrap_pw1pw2_140 487 13.2 0.287 0.301 1.555 1.593 potential_pw2rs 119 12.3 0.009 0.010 1.526 1.529 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.526 cp_fm_diag_elpa_base 50 14.0 1.484 1.500 1.515 1.524 acc_transpose_blocks 27432 15.5 0.110 0.113 1.465 1.487 fft3d_ps 1201 14.6 0.555 0.603 1.319 1.346 grid_collocate_task_list 119 9.7 1.221 1.342 1.221 1.342 wfi_extrapolate 11 7.9 0.001 0.001 1.312 1.313 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.752 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.205 1.214 mp_sum_l 7207 12.9 0.898 1.171 0.898 1.171 cp_fm_upper_to_full 72 14.2 0.813 1.157 0.813 1.157 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.075 1.092 multiply_cannon_sync_h2d 27432 15.5 1.006 1.084 1.006 1.084 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.062 1.063 dbcsr_complete_redistribute 329 12.2 0.119 0.141 0.747 1.020 make_images_data 4572 15.5 0.044 0.048 0.825 0.942 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.719 0.898 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.869 mp_alltoall_d11v 2130 13.8 0.708 0.856 0.708 0.856 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.838 0.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.828 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.556 0.825 cp_fm_cholesky_invert 11 10.9 0.726 0.729 0.726 0.729 mp_alltoall_i22 627 13.8 0.430 0.724 0.430 0.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.346000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=512.636364, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 595.877888E+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 0.013 0.031 27.717 27.717 qs_mol_dyn_low 1 2.0 0.003 0.004 27.542 27.550 qs_forces 11 3.9 0.004 0.004 27.482 27.483 qs_energies 11 4.9 0.001 0.001 25.765 25.767 scf_env_do_scf 11 5.9 0.001 0.001 20.941 20.941 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 18.511 18.512 velocity_verlet 10 3.0 0.001 0.002 14.216 14.220 dbcsr_multiply_generic 2286 12.5 0.091 0.097 12.040 12.112 qs_scf_new_mos 108 7.5 0.001 0.001 10.543 10.565 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.542 10.565 multiply_cannon 2286 13.5 0.231 0.240 9.660 10.060 ot_scf_mini 108 9.5 0.002 0.002 9.881 9.905 multiply_cannon_loop 2286 14.5 0.328 0.337 8.784 8.966 multiply_cannon_multrec 9144 15.5 1.645 1.894 5.970 6.228 rebuild_ks_matrix 119 8.3 0.000 0.000 5.909 5.932 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.909 5.932 ot_mini 108 10.5 0.001 0.001 5.528 5.557 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.250 5.272 dbcsr_mm_accdrv_process 12550 15.8 3.349 4.251 4.223 4.296 qs_ot_get_derivative 108 11.5 0.001 0.001 4.247 4.271 sum_up_and_integrate 119 10.3 0.038 0.041 3.612 3.620 integrate_v_rspace 119 11.3 0.003 0.003 3.574 3.584 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.471 3.476 calculate_rho_elec 119 8.7 0.060 0.061 3.470 3.475 init_scf_run 11 5.9 0.000 0.001 3.384 3.384 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.384 3.384 qs_ot_get_p 119 10.4 0.001 0.001 2.622 2.663 init_scf_loop 11 6.9 0.000 0.000 2.410 2.412 calculate_first_density_matrix 1 7.0 0.000 0.000 2.278 2.279 pw_transfer 1439 11.6 0.066 0.069 2.011 2.023 grid_integrate_task_list 119 12.3 1.845 1.972 1.845 1.972 jit_kernel_multiply 10 16.1 0.837 1.933 0.837 1.933 density_rs2pw 119 9.7 0.004 0.004 1.786 1.932 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.918 1.931 calculate_dm_sparse 119 9.5 0.000 0.000 1.797 1.814 mp_waitall_1 115863 16.7 1.260 1.812 1.260 1.812 make_m2s 4572 13.5 0.035 0.036 1.652 1.806 rs_pw_transfer 974 11.9 0.008 0.008 1.619 1.774 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.748 1.750 make_images 4572 14.5 0.268 0.301 1.563 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.682 1.687 make_preconditioner 11 8.9 0.000 0.000 1.682 1.687 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.615 1.616 fft_wrap_pw1pw2_140 487 13.2 0.364 0.375 1.589 1.602 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.575 1.601 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.559 1.559 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.491 1.505 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.469 1.476 grid_collocate_task_list 119 9.7 1.271 1.389 1.271 1.389 potential_pw2rs 119 12.3 0.010 0.011 1.349 1.352 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.280 1.288 cp_fm_diag_elpa_base 50 14.0 1.253 1.269 1.278 1.286 ot_diis_step 108 11.5 0.012 0.012 1.269 1.269 fft3d_ps 1201 14.6 0.559 0.569 1.251 1.262 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.232 1.233 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.226 1.231 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.148 1.166 apply_single 119 13.6 0.000 0.000 1.148 1.166 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.085 1.097 wfi_extrapolate 11 7.9 0.001 0.001 1.059 1.059 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.772 0.980 acc_transpose_blocks 9144 15.5 0.038 0.040 0.967 0.973 make_images_data 4572 15.5 0.038 0.042 0.767 0.940 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.920 mp_alltoall_d11v 2130 13.8 0.766 0.912 0.766 0.912 cp_fm_cholesky_invert 11 10.9 0.832 0.834 0.832 0.834 multiply_cannon_sync_h2d 9144 15.5 0.709 0.791 0.709 0.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.779 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.682 0.734 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.724 0.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.698 0.707 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.323 0.689 jit_kernel_transpose 5 15.6 0.608 0.611 0.608 0.611 mp_waitany 5200 13.7 0.454 0.599 0.454 0.599 mp_allgather_i34 2286 14.5 0.205 0.596 0.205 0.596 mp_alltoall_z22v 1201 16.6 0.567 0.596 0.567 0.596 rs_pw_transfer_RS2PW_140 130 11.5 0.100 0.104 0.419 0.574 qs_create_task_list 11 7.9 0.000 0.001 0.541 0.567 generate_qs_task_list 11 8.9 0.188 0.211 0.541 0.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.717000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.636364, yerr=1.610913 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 751.951872E+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 0.014 0.033 40.518 40.518 qs_mol_dyn_low 1 2.0 0.003 0.004 40.323 40.330 qs_forces 11 3.9 0.003 0.003 40.262 40.263 qs_energies 11 4.9 0.001 0.001 38.296 38.299 scf_env_do_scf 11 5.9 0.001 0.001 32.772 32.772 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.004 25.005 velocity_verlet 10 3.0 0.001 0.001 22.991 22.997 dbcsr_multiply_generic 2286 12.5 0.099 0.102 17.089 17.333 qs_scf_new_mos 108 7.5 0.001 0.001 15.439 15.538 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.438 15.538 ot_scf_mini 108 9.5 0.002 0.002 14.375 14.479 multiply_cannon 2286 13.5 0.301 0.307 13.393 14.335 multiply_cannon_loop 2286 14.5 0.342 0.348 12.167 13.147 ot_mini 108 10.5 0.001 0.001 8.651 8.771 multiply_cannon_multrec 9144 15.5 3.392 4.811 8.514 8.594 init_scf_loop 11 6.9 0.000 0.000 7.742 7.744 rebuild_ks_matrix 119 8.3 0.000 0.000 7.158 7.300 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.157 7.299 prepare_preconditioner 11 7.9 0.000 0.000 6.798 6.811 make_preconditioner 11 8.9 0.000 0.000 6.798 6.811 qs_ot_get_derivative 108 11.5 0.001 0.001 6.662 6.766 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.335 6.687 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.465 6.595 dbcsr_mm_accdrv_process 12550 15.8 4.143 5.577 4.998 6.321 cp_fm_upper_to_full 72 14.2 3.224 4.629 3.224 4.629 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.149 4.153 calculate_rho_elec 119 8.7 0.118 0.121 4.149 4.152 sum_up_and_integrate 119 10.3 0.065 0.066 3.918 3.924 integrate_v_rspace 119 11.3 0.003 0.004 3.853 3.859 init_scf_run 11 5.9 0.000 0.001 3.497 3.497 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.497 3.497 qs_ot_get_p 119 10.4 0.001 0.001 3.043 3.182 mp_waitall_1 94719 16.7 2.126 3.144 2.126 3.144 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.365 2.814 dbcsr_complete_redistribute 329 12.2 0.289 0.293 1.960 2.789 pw_transfer 1439 11.6 0.069 0.069 2.605 2.609 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.506 2.510 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.668 2.488 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.184 2.414 apply_single 119 13.6 0.000 0.000 2.184 2.414 mp_alltoall_i22 627 13.8 1.430 2.287 1.430 2.287 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.457 2.270 make_m2s 4572 13.5 0.038 0.038 2.110 2.262 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.223 2.166 calculate_dm_sparse 119 9.5 0.000 0.000 2.149 2.165 make_images 4572 14.5 0.354 0.387 1.989 2.142 fft_wrap_pw1pw2_140 487 13.2 0.617 0.621 2.123 2.128 density_rs2pw 119 9.7 0.004 0.004 2.104 2.118 calculate_first_density_matrix 1 7.0 0.000 0.000 2.075 2.077 grid_integrate_task_list 119 12.3 2.031 2.049 2.031 2.049 ot_diis_step 108 11.5 0.014 0.014 1.967 1.967 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.840 1.841 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.765 1.766 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.756 1.757 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.632 1.685 mp_sum_l 7207 12.9 1.042 1.658 1.042 1.658 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.573 1.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.516 1.558 fft3d_ps 1201 14.6 0.596 0.606 1.522 1.527 grid_collocate_task_list 119 9.7 1.468 1.486 1.468 1.486 rs_pw_transfer 974 11.9 0.009 0.009 1.410 1.430 cp_fm_cholesky_invert 11 10.9 1.383 1.386 1.383 1.386 wfi_extrapolate 11 7.9 0.001 0.001 1.359 1.359 potential_pw2rs 119 12.3 0.014 0.014 1.350 1.352 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.299 1.300 cp_fm_diag_elpa_base 50 14.0 1.154 1.207 1.298 1.298 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.230 1.235 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.037 1.207 make_images_data 4572 15.5 0.042 0.045 0.957 1.142 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.115 1.133 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.097 1.116 multiply_cannon_sync_h2d 9144 15.5 1.042 1.045 1.042 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.983 1.031 mp_alltoall_d11v 2130 13.8 1.015 1.030 1.015 1.030 jit_kernel_multiply 6 15.6 0.828 1.012 0.828 1.012 qs_create_task_list 11 7.9 0.000 0.000 0.937 0.947 generate_qs_task_list 11 8.9 0.370 0.389 0.936 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.902 0.915 acc_transpose_blocks 9144 15.5 0.039 0.039 0.894 0.903 mp_alltoall_z22v 1201 16.6 0.790 0.812 0.790 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.518000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=709.636364, yerr=13.282064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 497.004544E+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 1308054. 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 0.013 0.028 80.915 80.916 qs_mol_dyn_low 1 2.0 0.003 0.006 80.652 80.661 qs_forces 11 3.9 0.003 0.003 80.579 80.583 qs_energies 11 4.9 0.001 0.001 77.752 77.770 scf_env_do_scf 11 5.9 0.000 0.001 68.891 68.893 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.463 63.464 dbcsr_multiply_generic 2055 12.4 0.105 0.108 50.176 50.525 qs_scf_new_mos 99 7.5 0.000 0.001 46.266 46.390 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.266 46.390 ot_scf_mini 99 9.5 0.002 0.002 43.921 44.001 multiply_cannon 2055 13.4 0.183 0.188 41.895 43.065 multiply_cannon_loop 2055 14.4 1.510 1.563 40.934 42.094 velocity_verlet 10 3.0 0.001 0.001 42.041 42.042 ot_mini 99 10.5 0.001 0.001 25.887 25.968 qs_ot_get_derivative 99 11.5 0.001 0.001 19.176 19.274 multiply_cannon_multrec 49320 15.4 12.534 13.303 17.343 18.136 rebuild_ks_matrix 110 8.3 0.000 0.000 14.451 14.595 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.451 14.595 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.668 12.795 mp_waitall_1 220248 16.4 10.273 11.304 10.273 11.304 multiply_cannon_sync_h2d 49320 15.4 10.394 10.982 10.394 10.982 qs_ot_get_p 110 10.4 0.001 0.001 9.648 9.782 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.104 7.668 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.157 7.629 apply_single 110 13.6 0.000 0.001 7.156 7.628 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.272 7.569 sum_up_and_integrate 110 10.3 0.036 0.042 7.020 7.036 integrate_v_rspace 110 11.3 0.003 0.004 6.984 7.008 init_scf_run 11 5.9 0.000 0.001 6.826 6.827 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.826 6.827 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.517 6.655 calculate_rho_elec 110 8.6 0.021 0.026 6.516 6.655 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.591 6.639 ot_diis_step 99 11.5 0.005 0.006 6.511 6.511 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.690 5.691 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.529 5.572 init_scf_loop 11 6.9 0.000 0.000 5.402 5.402 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.137 5.163 cp_fm_diag_elpa_base 48 14.0 5.124 5.151 5.135 5.162 mp_sum_l 6514 12.8 4.069 5.025 4.069 5.025 dbcsr_mm_accdrv_process 87628 16.1 1.873 1.966 4.684 4.917 rs_pw_transfer 902 11.9 0.012 0.013 3.623 4.255 wfi_extrapolate 11 7.9 0.001 0.001 4.021 4.021 make_m2s 4110 13.4 0.061 0.066 3.760 3.906 density_rs2pw 110 9.6 0.004 0.005 3.327 3.893 calculate_dm_sparse 110 9.5 0.001 0.001 3.717 3.818 make_images 4110 14.4 0.177 0.192 3.664 3.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.568 3.572 multiply_cannon_metrocomm1 49320 15.4 0.061 0.063 2.206 3.551 grid_integrate_task_list 110 12.3 3.248 3.420 3.248 3.420 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.241 3.303 prepare_preconditioner 11 7.9 0.000 0.000 3.265 3.284 make_preconditioner 11 8.9 0.000 0.000 3.265 3.284 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.131 3.163 pw_transfer 1331 11.6 0.055 0.066 3.069 3.144 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.038 3.093 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.980 3.057 calculate_first_density_matrix 1 7.0 0.000 0.001 2.716 2.722 fft_wrap_pw1pw2_140 451 13.1 0.452 0.498 2.552 2.637 potential_pw2rs 110 12.3 0.006 0.007 2.572 2.592 jit_kernel_multiply 13 15.9 2.537 2.553 2.537 2.553 mp_alltoall_d11v 2046 13.8 2.037 2.427 2.037 2.427 mp_waitany 14300 13.8 1.800 2.397 1.800 2.397 grid_collocate_task_list 110 9.6 2.082 2.313 2.082 2.313 fft3d_ps 1111 14.6 0.791 0.879 2.176 2.232 acc_transpose_blocks 49320 15.4 0.208 0.215 2.099 2.168 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.912 1.930 make_images_data 4110 15.4 0.042 0.046 1.742 1.883 mp_sum_d 3889 11.9 1.308 1.875 1.308 1.875 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.524 1.821 cp_fm_cholesky_invert 11 10.9 1.776 1.780 1.776 1.780 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.646 1.670 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.916000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.272727, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.612736E+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 2973460. 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 0.015 0.032 68.201 68.202 qs_mol_dyn_low 1 2.0 0.003 0.004 67.960 67.969 qs_forces 11 3.9 0.003 0.004 67.894 67.895 qs_energies 11 4.9 0.001 0.005 64.566 64.570 scf_env_do_scf 11 5.9 0.002 0.010 56.075 56.079 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 48.590 48.592 dbcsr_multiply_generic 2055 12.4 0.113 0.119 37.227 37.365 velocity_verlet 10 3.0 0.001 0.002 35.736 35.737 qs_scf_new_mos 99 7.5 0.001 0.001 32.395 32.539 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.394 32.538 multiply_cannon 2055 13.4 0.222 0.243 30.841 31.991 ot_scf_mini 99 9.5 0.003 0.004 30.744 30.875 multiply_cannon_loop 2055 14.4 0.914 0.947 29.603 30.501 ot_mini 99 10.5 0.001 0.001 17.923 18.057 multiply_cannon_multrec 24660 15.4 7.661 9.414 13.830 15.369 rebuild_ks_matrix 110 8.3 0.000 0.000 13.769 13.834 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.768 13.833 qs_ot_get_derivative 99 11.5 0.001 0.001 12.141 12.271 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.118 12.178 mp_waitall_1 176588 16.5 7.475 9.995 7.475 9.995 multiply_cannon_sync_h2d 24660 15.4 7.004 8.047 7.004 8.047 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.079 7.734 init_scf_loop 11 6.9 0.001 0.005 7.450 7.451 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.496 7.286 apply_single 110 13.6 0.000 0.001 6.496 7.285 sum_up_and_integrate 110 10.3 0.053 0.059 6.604 6.617 integrate_v_rspace 110 11.3 0.002 0.003 6.551 6.567 qs_ot_get_p 110 10.4 0.001 0.001 6.248 6.398 dbcsr_mm_accdrv_process 52282 16.1 4.474 5.455 6.007 6.302 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.132 6.146 calculate_rho_elec 110 8.6 0.040 0.048 6.132 6.145 init_scf_run 11 5.9 0.000 0.001 6.113 6.114 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.113 6.113 ot_diis_step 99 11.5 0.010 0.010 5.735 5.735 prepare_preconditioner 11 7.9 0.000 0.000 5.378 5.398 make_preconditioner 11 8.9 0.000 0.001 5.377 5.398 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.587 5.309 make_full_inverse_cholesky 11 9.9 0.000 0.001 4.981 5.136 make_m2s 4110 13.4 0.057 0.060 4.049 4.487 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.431 4.450 make_images 4110 14.4 0.398 0.443 3.941 4.376 cp_dbcsr_syevd 48 12.0 0.003 0.006 4.007 4.007 pw_transfer 1331 11.6 0.066 0.071 3.610 3.747 density_rs2pw 110 9.6 0.005 0.006 3.253 3.693 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.503 3.640 wfi_extrapolate 11 7.9 0.001 0.001 3.478 3.478 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.458 3.470 cp_fm_diag_elpa_base 48 14.0 3.411 3.431 3.454 3.465 rs_pw_transfer 902 11.9 0.012 0.015 2.945 3.429 grid_integrate_task_list 110 12.3 3.158 3.317 3.158 3.317 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.260 3.261 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.106 3.179 fft_wrap_pw1pw2_140 451 13.1 0.521 0.535 3.008 3.147 calculate_dm_sparse 110 9.5 0.001 0.001 2.953 2.977 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.910 2.941 make_images_data 4110 15.4 0.046 0.050 2.301 2.747 hybrid_alltoall_any 4261 16.3 0.102 0.445 1.972 2.746 fft3d_ps 1111 14.6 1.110 1.334 2.475 2.625 calculate_first_density_matrix 1 7.0 0.000 0.003 2.523 2.526 cp_fm_cholesky_invert 11 10.9 2.501 2.507 2.501 2.507 grid_collocate_task_list 110 9.6 2.055 2.467 2.055 2.467 potential_pw2rs 110 12.3 0.008 0.009 2.402 2.417 mp_sum_l 6514 12.8 1.721 2.255 1.721 2.255 mp_alltoall_d11v 2046 13.8 1.647 2.036 1.647 2.036 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.926 1.947 jit_kernel_multiply 10 16.2 1.181 1.924 1.181 1.924 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.871 1.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.807 1.819 mp_waitany 10164 13.8 1.241 1.777 1.241 1.777 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.784 1.741 mp_irecv_dv 57340 16.2 0.658 1.622 0.658 1.622 acc_transpose_blocks 24660 15.4 0.112 0.114 1.561 1.586 mp_allgather_i34 2055 14.4 0.566 1.569 0.566 1.569 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.478 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.219 0.974 1.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.202000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.727273, yerr=7.817125 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 655.450112E+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 0.018 0.039 59.289 59.290 qs_mol_dyn_low 1 2.0 0.003 0.004 58.952 58.961 qs_forces 11 3.9 0.004 0.011 58.886 58.888 qs_energies 11 4.9 0.002 0.012 55.668 55.672 scf_env_do_scf 11 5.9 0.001 0.001 47.767 47.768 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 39.355 39.357 velocity_verlet 10 3.0 0.001 0.002 32.234 32.236 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.089 28.360 qs_scf_new_mos 99 7.5 0.001 0.001 24.397 24.496 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.396 24.495 multiply_cannon 2055 13.4 0.214 0.224 22.049 23.335 ot_scf_mini 99 9.5 0.002 0.003 23.186 23.306 multiply_cannon_loop 2055 14.4 0.610 0.624 20.852 21.882 ot_mini 99 10.5 0.001 0.001 13.320 13.443 rebuild_ks_matrix 110 8.3 0.000 0.000 12.355 12.499 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.355 12.498 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.874 11.006 multiply_cannon_multrec 16440 15.4 3.805 4.834 9.527 10.503 mp_waitall_1 139946 16.5 6.929 10.115 6.929 10.115 qs_ot_get_derivative 99 11.5 0.002 0.003 8.825 8.943 init_scf_loop 11 6.9 0.001 0.003 8.379 8.384 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.426 7.385 prepare_preconditioner 11 7.9 0.000 0.000 6.597 6.618 make_preconditioner 11 8.9 0.000 0.002 6.597 6.618 sum_up_and_integrate 110 10.3 0.061 0.062 6.508 6.524 integrate_v_rspace 110 11.3 0.003 0.004 6.447 6.463 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.929 6.287 qs_rho_update_rho_low 110 7.6 0.001 0.002 5.927 5.937 calculate_rho_elec 110 8.6 0.059 0.060 5.927 5.937 dbcsr_mm_accdrv_process 34862 16.1 4.621 5.179 5.575 5.759 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.009 5.500 apply_single 110 13.6 0.000 0.000 5.009 5.500 init_scf_run 11 5.9 0.000 0.001 5.494 5.495 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.494 5.494 qs_ot_get_p 110 10.4 0.001 0.001 5.118 5.283 ot_diis_step 99 11.5 0.010 0.011 4.469 4.469 make_m2s 4110 13.4 0.050 0.051 4.007 4.417 make_images 4110 14.4 0.390 0.513 3.893 4.302 multiply_cannon_sync_h2d 16440 15.4 3.678 4.298 3.678 4.298 density_rs2pw 110 9.6 0.004 0.005 3.042 4.277 rs_pw_transfer 902 11.9 0.011 0.011 2.605 3.824 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.037 3.675 pw_transfer 1331 11.6 0.066 0.072 3.616 3.624 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.562 3.565 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.509 3.519 grid_integrate_task_list 110 12.3 3.170 3.441 3.170 3.441 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.231 3.232 fft_wrap_pw1pw2_140 451 13.1 0.637 0.645 3.051 3.063 wfi_extrapolate 11 7.9 0.001 0.001 2.938 2.938 make_images_data 4110 15.4 0.043 0.047 2.334 2.843 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.086 2.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.761 2.763 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.714 2.723 cp_fm_diag_elpa_base 48 14.0 2.649 2.676 2.712 2.722 cp_fm_cholesky_invert 11 10.9 2.552 2.558 2.552 2.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.453 2.509 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.505 calculate_first_density_matrix 1 7.0 0.000 0.002 2.471 2.473 mp_waitany 17072 13.8 1.185 2.458 1.185 2.458 grid_collocate_task_list 110 9.6 2.080 2.446 2.080 2.446 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.900 2.430 fft3d_ps 1111 14.6 1.091 1.102 2.332 2.345 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.255 2.316 mp_irecv_dv 48980 15.7 0.831 2.306 0.831 2.306 potential_pw2rs 110 12.3 0.011 0.012 2.238 2.243 mp_alltoall_d11v 2046 13.8 1.697 2.150 1.697 2.150 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.178 0.903 2.116 mp_sum_l 6514 12.8 1.429 2.003 1.429 2.003 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.972 1.974 dbcsr_complete_redistribute 325 12.2 0.327 0.363 1.410 1.857 cp_fm_upper_to_full 70 14.2 1.394 1.765 1.394 1.765 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.675 1.689 cp_fm_cholesky_decompose 22 10.9 1.531 1.551 1.531 1.551 mp_allgather_i34 2055 14.4 0.518 1.520 0.518 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.376 1.492 jit_kernel_multiply 8 16.5 0.567 1.467 0.567 1.467 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.445 1.458 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.938 1.387 rs_gather_matrices 110 12.3 0.234 0.260 0.961 1.378 acc_transpose_blocks 16440 15.4 0.074 0.076 1.239 1.255 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.232 1.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.290000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.545455, yerr=9.847599 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 726.126592E+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 0.016 0.029 64.375 64.376 qs_mol_dyn_low 1 2.0 0.003 0.004 64.119 64.128 qs_forces 11 3.9 0.004 0.005 64.054 64.055 qs_energies 11 4.9 0.002 0.007 60.636 60.640 scf_env_do_scf 11 5.9 0.001 0.001 52.418 52.421 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.896 40.896 velocity_verlet 10 3.0 0.001 0.002 36.681 36.684 dbcsr_multiply_generic 2055 12.4 0.112 0.119 29.152 29.339 qs_scf_new_mos 99 7.5 0.001 0.001 25.895 26.017 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.895 26.016 ot_scf_mini 99 9.5 0.003 0.003 24.284 24.394 multiply_cannon 2055 13.4 0.245 0.259 22.298 23.399 multiply_cannon_loop 2055 14.4 0.878 0.894 20.901 21.407 multiply_cannon_multrec 24660 15.4 4.255 6.848 12.653 13.941 ot_mini 99 10.5 0.001 0.001 13.785 13.919 rebuild_ks_matrix 110 8.3 0.000 0.000 12.212 12.304 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.212 12.303 init_scf_loop 11 6.9 0.001 0.005 11.484 11.485 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.780 10.862 qs_ot_get_derivative 99 11.5 0.001 0.002 9.675 9.790 prepare_preconditioner 11 7.9 0.000 0.000 9.734 9.748 make_preconditioner 11 8.9 0.000 0.001 9.734 9.748 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.984 9.429 dbcsr_mm_accdrv_process 52304 16.0 6.854 8.195 8.249 9.209 sum_up_and_integrate 110 10.3 0.068 0.071 6.494 6.509 integrate_v_rspace 110 11.3 0.003 0.003 6.426 6.441 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.062 6.072 calculate_rho_elec 110 8.6 0.078 0.081 6.062 6.072 mp_waitall_1 121746 16.5 4.148 6.009 4.148 6.009 qs_ot_get_p 110 10.4 0.001 0.001 5.516 5.685 make_m2s 4110 13.4 0.059 0.063 5.302 5.668 make_images 4110 14.4 0.576 0.699 5.162 5.526 init_scf_run 11 5.9 0.000 0.001 5.465 5.466 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.465 5.466 cp_fm_upper_to_full 70 14.2 3.328 4.819 3.328 4.819 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.001 4.088 apply_single 110 13.6 0.000 0.000 4.001 4.088 ot_diis_step 99 11.5 0.011 0.011 4.076 4.076 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.780 3.795 pw_transfer 1331 11.6 0.066 0.076 3.731 3.762 dbcsr_complete_redistribute 325 12.2 0.419 0.464 2.620 3.747 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.624 3.660 density_rs2pw 110 9.6 0.004 0.004 3.001 3.507 grid_integrate_task_list 110 12.3 3.259 3.451 3.259 3.451 multiply_cannon_sync_h2d 24660 15.4 3.201 3.318 3.201 3.318 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.305 3.305 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.218 3.275 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.084 3.198 fft_wrap_pw1pw2_140 451 13.1 0.669 0.691 3.118 3.156 hybrid_alltoall_any 4261 16.3 0.120 0.461 2.277 3.118 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.332 3.095 make_images_data 4110 15.4 0.045 0.049 2.695 3.055 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.954 3.017 wfi_extrapolate 11 7.9 0.001 0.001 2.963 2.963 calculate_dm_sparse 110 9.5 0.001 0.001 2.893 2.920 rs_pw_transfer 902 11.9 0.010 0.011 2.346 2.861 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.741 2.841 mp_alltoall_i22 605 13.7 1.628 2.797 1.628 2.797 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.781 2.792 cp_fm_diag_elpa_base 48 14.0 2.636 2.690 2.779 2.790 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.783 2.785 cp_fm_cholesky_invert 11 10.9 2.589 2.598 2.589 2.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.438 2.466 calculate_first_density_matrix 1 7.0 0.000 0.002 2.409 2.411 grid_collocate_task_list 110 9.6 2.180 2.409 2.180 2.409 fft3d_ps 1111 14.6 1.083 1.113 2.383 2.400 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.252 2.254 potential_pw2rs 110 12.3 0.013 0.013 2.147 2.154 mp_alltoall_d11v 2046 13.8 1.711 2.033 1.711 2.033 jit_kernel_multiply 10 15.8 1.065 1.818 1.065 1.818 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.676 1.713 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.675 1.684 cp_fm_cholesky_decompose 22 10.9 1.627 1.673 1.627 1.673 mp_allgather_i34 2055 14.4 0.488 1.548 0.488 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.519 1.534 acc_transpose_blocks 24660 15.4 0.104 0.107 1.511 1.533 mp_sum_l 6514 12.8 0.877 1.530 0.877 1.530 mp_waitany 13376 13.8 1.047 1.529 1.047 1.529 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.825 1.465 mp_irecv_dv 62702 16.1 0.725 1.389 0.725 1.389 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.175 0.804 1.326 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.201 1.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.376000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.818182, yerr=9.073620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 828.960768E+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 0.015 0.033 54.359 54.361 qs_mol_dyn_low 1 2.0 0.003 0.004 54.108 54.117 qs_forces 11 3.9 0.003 0.003 54.002 54.003 qs_energies 11 4.9 0.001 0.001 50.322 50.326 scf_env_do_scf 11 5.9 0.001 0.001 42.071 42.071 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 34.503 34.504 velocity_verlet 10 3.0 0.001 0.002 30.798 30.804 dbcsr_multiply_generic 2055 12.4 0.104 0.109 22.389 22.573 qs_scf_new_mos 99 7.5 0.001 0.001 19.572 19.621 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.571 19.621 multiply_cannon 2055 13.4 0.245 0.258 17.100 18.538 ot_scf_mini 99 9.5 0.002 0.002 18.349 18.381 multiply_cannon_loop 2055 14.4 0.318 0.331 15.748 16.117 rebuild_ks_matrix 110 8.3 0.000 0.000 11.851 11.871 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.850 11.871 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.519 10.537 ot_mini 99 10.5 0.001 0.001 10.089 10.120 multiply_cannon_multrec 8220 15.4 3.196 4.526 7.411 8.324 init_scf_loop 11 6.9 0.000 0.000 7.523 7.524 mp_waitall_1 103326 16.6 5.785 7.314 5.785 7.314 sum_up_and_integrate 110 10.3 0.080 0.081 6.460 6.472 integrate_v_rspace 110 11.3 0.003 0.003 6.379 6.391 qs_ot_get_derivative 99 11.5 0.001 0.002 6.309 6.340 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.214 6.228 calculate_rho_elec 110 8.6 0.114 0.115 6.213 6.228 prepare_preconditioner 11 7.9 0.000 0.000 5.876 5.880 make_preconditioner 11 8.9 0.000 0.000 5.876 5.880 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.453 5.526 init_scf_run 11 5.9 0.000 0.001 5.056 5.056 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.056 5.056 dbcsr_mm_accdrv_process 17442 15.9 2.810 3.627 4.086 5.012 qs_ot_get_p 110 10.4 0.001 0.001 4.376 4.411 make_m2s 4110 13.4 0.039 0.040 4.104 4.388 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.966 4.378 make_images 4110 14.4 0.638 0.694 3.975 4.261 pw_transfer 1331 11.6 0.066 0.072 3.883 3.895 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.775 3.790 ot_diis_step 99 11.5 0.012 0.012 3.761 3.761 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.688 3.711 apply_single 110 13.6 0.000 0.000 3.688 3.711 grid_integrate_task_list 110 12.3 3.352 3.541 3.352 3.541 density_rs2pw 110 9.6 0.004 0.005 2.967 3.343 fft_wrap_pw1pw2_140 451 13.1 0.835 0.847 3.277 3.299 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.108 3.111 multiply_cannon_sync_h2d 8220 15.4 2.926 3.104 2.926 3.104 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.816 2.816 cp_fm_cholesky_invert 11 10.9 2.803 2.807 2.803 2.807 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.294 2.744 make_images_data 4110 15.4 0.038 0.043 2.351 2.739 wfi_extrapolate 11 7.9 0.001 0.001 2.717 2.718 qs_energies_init_hamiltonians 11 5.9 0.027 0.029 2.713 2.714 grid_collocate_task_list 110 9.6 2.278 2.537 2.278 2.537 rs_pw_transfer 902 11.9 0.010 0.011 2.110 2.504 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.496 2.497 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.496 fft3d_ps 1111 14.6 1.134 1.154 2.319 2.332 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.320 2.328 cp_fm_diag_elpa_base 48 14.0 2.265 2.286 2.318 2.326 calculate_first_density_matrix 1 7.0 0.000 0.000 2.244 2.245 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.159 2.174 potential_pw2rs 110 12.3 0.015 0.016 2.017 2.021 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.766 1.991 mp_alltoall_d11v 2046 13.8 1.587 1.876 1.587 1.876 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.825 1.838 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.728 1.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.660 1.666 cp_fm_cholesky_decompose 22 10.9 1.642 1.661 1.642 1.661 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.520 1.642 mp_allgather_i34 2055 14.4 0.537 1.635 0.537 1.635 dbcsr_complete_redistribute 325 12.2 0.563 0.597 1.392 1.481 mp_waitany 9240 13.8 1.045 1.454 1.045 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.442 1.452 qs_create_task_list 11 7.9 0.003 0.004 1.233 1.339 generate_qs_task_list 11 8.9 0.376 0.444 1.230 1.335 rs_gather_matrices 110 12.3 0.325 0.369 0.922 1.203 jit_kernel_multiply 7 15.9 0.969 1.200 0.969 1.200 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.791 1.188 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.086 1.105 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.361000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=777.818182, yerr=13.388042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 1.353290E+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 0.017 0.038 86.317 86.317 qs_mol_dyn_low 1 2.0 0.003 0.004 86.016 86.025 qs_forces 11 3.9 0.003 0.003 85.946 85.947 qs_energies 11 4.9 0.001 0.001 81.793 81.794 scf_env_do_scf 11 5.9 0.001 0.001 71.835 71.835 velocity_verlet 10 3.0 0.001 0.001 55.489 55.496 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.336 43.336 init_scf_loop 11 6.9 0.000 0.000 28.427 28.428 dbcsr_multiply_generic 2055 12.4 0.119 0.124 27.946 28.017 prepare_preconditioner 11 7.9 0.000 0.000 26.428 26.436 make_preconditioner 11 8.9 0.000 0.000 26.428 26.436 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.628 25.897 qs_scf_new_mos 99 7.5 0.001 0.001 25.366 25.400 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.365 25.399 ot_scf_mini 99 9.5 0.002 0.002 23.643 23.664 multiply_cannon 2055 13.4 0.343 0.363 21.189 22.039 multiply_cannon_loop 2055 14.4 0.339 0.343 19.441 19.802 cp_fm_upper_to_full 70 14.2 13.047 18.562 13.047 18.562 rebuild_ks_matrix 110 8.3 0.000 0.000 13.913 13.938 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.912 13.938 ot_mini 99 10.5 0.001 0.001 13.227 13.249 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.592 12.615 dbcsr_complete_redistribute 325 12.2 1.019 1.049 7.340 10.662 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.335 9.655 multiply_cannon_multrec 8220 15.4 4.377 4.573 9.522 9.621 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.784 9.069 qs_ot_get_derivative 99 11.5 0.001 0.001 8.776 8.797 mp_alltoall_i22 605 13.7 5.401 8.732 5.401 8.732 mp_waitall_1 84994 16.7 7.191 7.955 7.191 7.955 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.607 7.644 calculate_rho_elec 110 8.6 0.225 0.225 7.606 7.644 sum_up_and_integrate 110 10.3 0.150 0.152 7.202 7.215 integrate_v_rspace 110 11.3 0.004 0.004 7.051 7.064 init_scf_run 11 5.9 0.000 0.001 5.686 5.686 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.685 5.685 make_m2s 4110 13.4 0.043 0.044 5.090 5.582 make_images 4110 14.4 0.881 0.937 4.903 5.395 qs_ot_get_p 110 10.4 0.001 0.001 5.325 5.348 dbcsr_mm_accdrv_process 11614 15.7 3.276 3.620 5.004 5.239 cp_fm_cholesky_invert 11 10.9 5.141 5.146 5.141 5.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.572 5.048 apply_single 110 13.6 0.000 0.000 4.572 5.047 pw_transfer 1331 11.6 0.075 0.076 4.967 4.975 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.570 4.972 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.850 4.857 ot_diis_step 99 11.5 0.015 0.016 4.437 4.437 fft_wrap_pw1pw2_140 451 13.1 1.339 1.345 4.244 4.253 multiply_cannon_sync_h2d 8220 15.4 3.946 3.951 3.946 3.951 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.854 3.862 grid_integrate_task_list 110 12.3 3.652 3.710 3.652 3.710 density_rs2pw 110 9.6 0.004 0.004 3.657 3.692 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.672 3.673 hybrid_alltoall_any 4261 16.3 0.257 0.558 2.779 3.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.423 3.423 make_images_data 4110 15.4 0.041 0.044 2.758 3.345 wfi_extrapolate 11 7.9 0.001 0.001 3.293 3.293 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.768 3.209 calculate_dm_sparse 110 9.5 0.001 0.001 3.086 3.109 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.845 2.846 cp_fm_diag_elpa_base 48 14.0 2.303 2.506 2.843 2.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.822 2.824 fft3d_ps 1111 14.6 1.304 1.320 2.775 2.791 grid_collocate_task_list 110 9.6 2.626 2.649 2.626 2.649 potential_pw2rs 110 12.3 0.022 0.022 2.354 2.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.343 2.352 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.308 2.320 calculate_first_density_matrix 1 7.0 0.000 0.000 2.273 2.274 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.192 2.254 rs_pw_transfer 902 11.9 0.011 0.011 2.211 2.233 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.171 cp_fm_cholesky_decompose 22 10.9 1.944 1.967 1.944 1.967 mp_alltoall_d11v 2046 13.8 1.859 1.957 1.859 1.957 qs_create_task_list 11 7.9 0.000 0.000 1.896 1.945 generate_qs_task_list 11 8.9 0.735 0.789 1.896 1.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.922 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.715 1.747 jit_kernel_multiply 10 15.3 1.529 1.736 1.529 1.736 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.317000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1211.181818, yerr=55.740989 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 627.462144E+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 2939583. 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 0.016 0.038 205.429 205.429 qs_mol_dyn_low 1 2.0 0.003 0.004 205.007 205.020 qs_forces 11 3.9 0.006 0.012 204.694 204.695 qs_energies 11 4.9 0.002 0.008 199.090 199.104 scf_env_do_scf 11 5.9 0.001 0.002 182.582 182.585 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 162.386 162.388 dbcsr_multiply_generic 2507 12.6 0.179 0.183 123.833 124.478 qs_scf_new_mos 117 7.6 0.001 0.001 122.689 122.860 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.688 122.859 velocity_verlet 10 3.0 0.001 0.001 122.638 122.641 ot_scf_mini 117 9.6 0.003 0.004 116.053 116.183 multiply_cannon 2507 13.6 0.239 0.248 100.845 102.287 multiply_cannon_loop 2507 14.6 2.077 2.141 98.692 99.944 ot_mini 117 10.6 0.001 0.001 65.204 65.353 multiply_cannon_multrec 60168 15.6 33.541 36.162 41.641 43.515 qs_ot_get_derivative 117 11.6 0.001 0.001 40.400 40.536 rebuild_ks_matrix 128 8.3 0.001 0.001 33.543 33.748 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.542 33.747 mp_waitall_1 267128 16.5 27.863 30.406 27.863 30.406 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.122 30.276 multiply_cannon_sync_h2d 60168 15.6 27.588 30.103 27.588 30.103 qs_ot_get_p 128 10.4 0.001 0.001 28.950 29.215 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.337 25.199 apply_single 128 13.6 0.001 0.001 24.336 25.199 ot_diis_step 117 11.6 0.007 0.008 24.547 24.548 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.545 22.621 init_scf_loop 11 6.9 0.001 0.004 20.124 20.126 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.999 20.000 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.357 19.498 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.765 17.671 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.076 17.109 cp_fm_diag_elpa_base 83 14.4 17.011 17.042 17.072 17.106 prepare_preconditioner 11 7.9 0.000 0.000 15.538 15.601 make_preconditioner 11 8.9 0.000 0.001 15.538 15.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.766 14.929 sum_up_and_integrate 128 10.3 0.090 0.108 14.075 14.095 integrate_v_rspace 128 11.3 0.004 0.005 13.985 14.006 make_m2s 5014 13.6 0.104 0.113 13.573 13.879 make_images 5014 14.6 0.399 0.417 13.392 13.708 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.492 13.631 calculate_rho_elec 128 8.7 0.046 0.065 13.492 13.630 init_scf_run 11 5.9 0.000 0.001 12.371 12.372 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.371 12.372 density_rs2pw 128 9.7 0.006 0.008 6.925 10.070 mp_sum_l 7870 13.0 8.153 9.136 8.153 9.136 wfi_extrapolate 11 7.9 0.001 0.001 9.092 9.092 cp_fm_cholesky_invert 11 10.9 8.810 8.818 8.810 8.818 rs_pw_transfer 1046 11.9 0.017 0.019 5.619 8.766 calculate_dm_sparse 128 9.5 0.001 0.001 8.437 8.505 dbcsr_mm_accdrv_process 124484 16.2 3.091 3.241 7.661 8.142 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.852 7.931 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.745 7.872 pw_transfer 1547 11.6 0.074 0.092 7.643 7.817 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.439 7.613 grid_integrate_task_list 128 12.3 7.043 7.556 7.043 7.556 make_images_data 5014 15.6 0.067 0.072 6.723 7.555 hybrid_alltoall_any 5200 16.5 0.290 2.266 5.870 7.171 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 5.693 7.124 fft_wrap_pw1pw2_140 523 13.2 1.272 1.319 6.557 6.731 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.711 6.722 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.835 5.952 mp_waitany 16020 13.9 2.691 5.847 2.691 5.847 grid_collocate_task_list 128 9.7 4.550 5.766 4.550 5.766 fft3d_ps 1291 14.7 2.139 2.779 5.209 5.535 mp_alltoall_d11v 2415 14.1 4.108 5.379 4.108 5.379 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.292 2.127 5.289 cp_fm_cholesky_decompose 22 10.9 4.548 4.561 4.548 4.561 potential_pw2rs 128 12.3 0.009 0.011 4.507 4.528 mp_sum_d 4465 12.1 3.409 4.251 3.409 4.251 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.429000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.181818, yerr=6.561848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 828.944384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826487232 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4040 57684. MP_Allreduce 11098 963. MP_Sync 87 MP_Alltoall 1983 6482264. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 189.661 189.662 qs_mol_dyn_low 1 2.0 0.003 0.004 189.217 189.231 qs_forces 11 3.9 0.004 0.005 189.115 189.116 qs_energies 11 4.9 0.001 0.005 182.320 182.330 scf_env_do_scf 11 5.9 0.001 0.001 165.984 165.994 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.556 133.558 velocity_verlet 10 3.0 0.002 0.005 119.879 119.881 dbcsr_multiply_generic 2527 12.6 0.196 0.203 97.705 98.852 qs_scf_new_mos 118 7.6 0.001 0.001 93.369 93.971 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.368 93.971 ot_scf_mini 118 9.6 0.004 0.004 88.519 89.048 multiply_cannon 2527 13.6 0.483 0.541 77.729 82.491 multiply_cannon_loop 2527 14.6 1.243 1.278 74.503 77.070 ot_mini 118 10.6 0.001 0.001 49.920 50.533 mp_waitall_1 216438 16.6 24.809 39.460 24.809 39.460 multiply_cannon_multrec 30324 15.6 21.970 25.826 31.596 35.652 rebuild_ks_matrix 129 8.3 0.001 0.001 33.421 33.934 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.421 33.933 init_scf_loop 11 6.9 0.001 0.005 32.342 32.345 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.072 30.557 multiply_cannon_metrocomm3 30324 15.6 0.091 0.097 16.071 29.246 qs_ot_get_derivative 118 11.6 0.001 0.002 27.663 28.213 prepare_preconditioner 11 7.9 0.000 0.000 27.951 28.034 make_preconditioner 11 8.9 0.000 0.001 27.951 28.034 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.648 27.200 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.300 23.415 apply_single 129 13.6 0.001 0.001 22.300 23.414 ot_diis_step 118 11.6 0.014 0.015 22.080 22.082 multiply_cannon_sync_h2d 30324 15.6 19.241 21.867 19.241 21.867 qs_ot_get_p 129 10.4 0.001 0.002 20.295 20.915 cp_fm_cholesky_invert 11 10.9 16.354 16.366 16.354 16.366 make_m2s 5054 13.6 0.089 0.096 14.296 15.862 make_images 5054 14.6 1.162 1.353 14.088 15.656 qs_ot_p2m_diag 83 11.4 0.187 0.215 15.544 15.577 sum_up_and_integrate 129 10.3 0.117 0.135 14.717 14.753 integrate_v_rspace 129 11.3 0.004 0.004 14.600 14.642 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.416 14.416 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.073 14.107 calculate_rho_elec 129 8.7 0.088 0.105 14.073 14.106 init_scf_run 11 5.9 0.000 0.001 11.532 11.533 scf_env_initial_rho_setup 11 6.9 0.002 0.002 11.531 11.533 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.328 11.355 cp_fm_diag_elpa_base 83 14.4 11.077 11.146 11.322 11.353 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.800 11.163 multiply_cannon_metrocomm4 27797 15.6 0.103 0.120 3.852 10.706 make_images_data 5054 15.6 0.065 0.074 8.733 10.508 mp_irecv_dv 70031 16.3 3.649 10.310 3.649 10.310 density_rs2pw 129 9.7 0.006 0.007 7.397 9.934 hybrid_alltoall_any 5240 16.5 0.344 1.514 7.473 9.885 dbcsr_mm_accdrv_process 62734 16.2 4.470 5.356 9.079 9.696 pw_transfer 1559 11.6 0.087 0.098 8.803 8.862 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.576 8.635 wfi_extrapolate 11 7.9 0.001 0.001 8.357 8.357 rs_pw_transfer 1054 12.0 0.015 0.016 5.651 8.192 fft_wrap_pw1pw2_140 527 13.2 1.347 1.374 7.611 7.675 grid_integrate_task_list 129 12.3 7.226 7.645 7.226 7.645 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.274 7.028 cp_fm_cholesky_decompose 22 10.9 6.743 6.819 6.743 6.819 calculate_dm_sparse 129 9.5 0.001 0.001 6.511 6.654 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.174 6.184 mp_sum_l 7930 13.1 3.882 5.909 3.882 5.909 grid_collocate_task_list 129 9.7 4.736 5.907 4.736 5.907 fft3d_ps 1301 14.7 2.835 3.018 5.860 5.907 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.297 5.450 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.366 5.428 mp_waitany 11836 13.9 2.505 5.058 2.505 5.058 mp_allgather_i34 2527 14.6 1.787 4.958 1.787 4.958 potential_pw2rs 129 12.3 0.016 0.019 4.931 4.946 mp_alltoall_d11v 2423 14.1 4.075 4.830 4.075 4.830 rs_pw_transfer_RS2PW_140 140 11.5 0.354 0.377 2.098 4.619 mp_sum_d 4507 12.1 2.633 3.975 2.633 3.975 dbcsr_complete_redistribute 395 12.7 0.792 0.894 3.082 3.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.662000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.272727, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/17/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 2.950886E+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 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 940.937216E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11057 1004. MP_Sync 87 MP_Alltoall 1724 9394173. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.049 175.039 175.040 qs_mol_dyn_low 1 2.0 0.003 0.004 174.463 174.476 qs_forces 11 3.9 0.004 0.005 174.364 174.373 qs_energies 11 4.9 0.002 0.006 167.724 167.736 scf_env_do_scf 11 5.9 0.001 0.002 152.199 152.200 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.651 117.652 velocity_verlet 10 3.0 0.001 0.002 112.964 112.966 dbcsr_multiply_generic 2527 12.6 0.182 0.186 81.844 83.137 qs_scf_new_mos 118 7.6 0.001 0.001 79.157 79.499 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.156 79.498 ot_scf_mini 118 9.6 0.004 0.005 75.018 75.465 multiply_cannon 2527 13.6 0.507 0.526 62.046 67.343 multiply_cannon_loop 2527 14.6 0.858 0.883 58.987 61.973 ot_mini 118 10.6 0.001 0.001 42.730 43.174 mp_waitall_1 171878 16.6 24.963 35.038 24.963 35.038 init_scf_loop 11 6.9 0.001 0.004 34.452 34.453 rebuild_ks_matrix 129 8.3 0.001 0.001 30.971 31.434 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.970 31.433 prepare_preconditioner 11 7.9 0.000 0.000 30.391 30.443 make_preconditioner 11 8.9 0.000 0.001 30.391 30.443 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.082 29.488 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.897 28.336 multiply_cannon_multrec 20216 15.6 13.667 17.022 22.332 25.778 multiply_cannon_metrocomm3 20216 15.6 0.058 0.061 15.177 24.997 qs_ot_get_derivative 118 11.6 0.002 0.002 22.853 23.296 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.988 21.111 apply_single 129 13.6 0.001 0.001 19.987 21.110 ot_diis_step 118 11.6 0.018 0.018 19.774 19.774 qs_ot_get_p 129 10.4 0.001 0.005 17.652 18.167 multiply_cannon_sync_h2d 20216 15.6 14.381 16.207 14.381 16.207 make_m2s 5054 13.6 0.080 0.086 14.708 15.722 make_images 5054 14.6 1.187 1.281 14.476 15.487 sum_up_and_integrate 129 10.3 0.134 0.146 14.705 14.730 integrate_v_rspace 129 11.3 0.004 0.004 14.570 14.597 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.388 14.431 calculate_rho_elec 129 8.7 0.132 0.147 14.387 14.430 cp_fm_cholesky_invert 11 10.9 14.098 14.107 14.098 14.107 qs_ot_p2m_diag 83 11.4 0.265 0.274 13.352 13.359 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.362 12.364 make_images_data 5054 15.6 0.060 0.068 9.104 10.509 init_scf_run 11 5.9 0.000 0.001 10.453 10.453 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.453 10.453 hybrid_alltoall_any 5240 16.5 0.435 1.984 7.836 10.071 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.378 9.399 cp_fm_diag_elpa_base 83 14.4 8.978 9.124 9.375 9.396 multiply_cannon_metrocomm4 17689 15.6 0.062 0.072 3.474 9.329 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.913 9.218 density_rs2pw 129 9.7 0.006 0.007 7.152 9.112 mp_irecv_dv 50625 16.2 3.352 9.080 3.352 9.080 pw_transfer 1559 11.6 0.087 0.104 8.936 9.035 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.710 8.814 dbcsr_mm_accdrv_process 41830 16.2 4.343 5.047 8.108 8.232 fft_wrap_pw1pw2_140 527 13.2 1.426 1.456 7.747 7.868 grid_integrate_task_list 129 12.3 7.341 7.804 7.341 7.804 wfi_extrapolate 11 7.9 0.001 0.001 7.411 7.411 cp_fm_cholesky_decompose 22 10.9 7.207 7.230 7.207 7.230 cp_fm_upper_to_full 105 14.8 5.721 7.177 5.721 7.177 rs_pw_transfer 1054 12.0 0.014 0.015 5.085 7.048 dbcsr_complete_redistribute 395 12.7 1.168 1.196 4.377 6.123 fft3d_ps 1301 14.7 2.763 2.989 5.807 5.887 grid_collocate_task_list 129 9.7 4.941 5.866 4.941 5.866 calculate_dm_sparse 129 9.5 0.001 0.001 5.726 5.826 mp_alltoall_d11v 2423 14.1 4.268 5.454 4.268 5.454 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.410 5.417 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.658 5.357 mp_sum_l 7930 13.1 3.255 5.008 3.255 5.008 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.232 4.977 potential_pw2rs 129 12.3 0.021 0.024 4.681 4.691 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.536 4.676 mp_allgather_i34 2527 14.6 1.567 4.484 1.567 4.484 mp_waitany 11836 13.9 2.368 4.385 2.368 4.385 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.001 4.020 transfer_fm_to_dbcsr 11 9.9 0.019 0.027 2.290 4.015 rs_pw_transfer_RS2PW_140 140 11.5 0.332 0.353 1.943 3.883 mp_alltoall_i22 718 14.1 1.940 3.828 1.940 3.828 qs_energies_init_hamiltonians 11 5.9 0.008 0.014 3.804 3.805 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.579 3.625 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.040000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.636364, yerr=10.271520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 1.135239E+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 0.019 0.037 186.979 186.983 qs_mol_dyn_low 1 2.0 0.003 0.006 186.568 186.580 qs_forces 11 3.9 0.005 0.013 186.462 186.468 qs_energies 11 4.9 0.002 0.007 179.300 179.310 scf_env_do_scf 11 5.9 0.001 0.002 162.665 162.674 velocity_verlet 10 3.0 0.001 0.002 124.626 124.629 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 116.712 116.713 dbcsr_multiply_generic 2527 12.6 0.189 0.198 79.415 80.129 qs_scf_new_mos 118 7.6 0.001 0.001 79.007 79.284 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.007 79.283 ot_scf_mini 118 9.6 0.003 0.004 74.484 74.765 multiply_cannon 2527 13.6 0.566 0.596 55.095 57.617 multiply_cannon_loop 2527 14.6 1.181 1.212 51.290 52.944 init_scf_loop 11 6.9 0.002 0.009 45.834 45.835 ot_mini 118 10.6 0.001 0.001 42.343 42.621 prepare_preconditioner 11 7.9 0.000 0.000 41.723 41.740 make_preconditioner 11 8.9 0.000 0.002 41.723 41.740 make_full_inverse_cholesky 11 9.9 0.012 0.021 35.372 40.386 multiply_cannon_multrec 30324 15.6 14.369 19.577 26.263 30.868 rebuild_ks_matrix 129 8.3 0.001 0.001 30.033 30.219 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.024 30.033 30.218 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.088 27.255 mp_waitall_1 149060 16.7 17.040 25.871 17.040 25.871 qs_ot_get_derivative 118 11.6 0.002 0.002 22.679 22.962 make_m2s 5054 13.6 0.096 0.099 19.990 21.056 make_images 5054 14.6 1.956 2.280 19.683 20.751 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.058 19.593 apply_single 129 13.6 0.001 0.001 19.058 19.592 ot_diis_step 118 11.6 0.018 0.018 19.536 19.538 qs_ot_get_p 129 10.4 0.001 0.002 18.341 18.618 cp_fm_upper_to_full 105 14.8 11.467 16.880 11.467 16.880 cp_fm_cholesky_invert 11 10.9 15.783 15.792 15.783 15.792 multiply_cannon_metrocomm3 30324 15.6 0.048 0.051 6.380 15.182 sum_up_and_integrate 129 10.3 0.141 0.153 14.767 14.795 integrate_v_rspace 129 11.3 0.004 0.005 14.626 14.656 qs_rho_update_rho_low 129 7.7 0.001 0.002 14.612 14.645 calculate_rho_elec 129 8.7 0.175 0.191 14.611 14.645 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.258 14.311 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.004 13.006 dbcsr_complete_redistribute 395 12.7 1.505 1.626 9.033 12.845 multiply_cannon_sync_h2d 30324 15.6 11.881 12.697 11.881 12.697 make_images_data 5054 15.6 0.062 0.067 10.730 12.592 dbcsr_mm_accdrv_process 62756 16.2 7.396 8.365 11.463 11.977 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.684 11.472 hybrid_alltoall_any 5240 16.5 0.529 2.219 9.685 11.468 init_scf_run 11 5.9 0.000 0.001 10.682 10.683 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.681 10.683 transfer_fm_to_dbcsr 11 9.9 0.002 0.009 6.328 10.059 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.829 9.848 cp_fm_diag_elpa_base 83 14.4 8.867 9.170 9.823 9.841 mp_alltoall_i22 718 14.1 5.582 9.522 5.582 9.522 qs_ot_get_derivative_diag 77 12.4 0.004 0.014 9.320 9.515 pw_transfer 1559 11.6 0.087 0.101 9.189 9.265 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.960 9.043 density_rs2pw 129 9.7 0.006 0.007 6.959 8.228 fft_wrap_pw1pw2_140 527 13.2 1.572 1.605 7.994 8.085 grid_integrate_task_list 129 12.3 7.585 7.967 7.585 7.967 wfi_extrapolate 11 7.9 0.001 0.001 7.597 7.598 cp_fm_cholesky_decompose 22 10.9 7.369 7.470 7.369 7.470 multiply_cannon_metrocomm4 25270 15.6 0.079 0.089 2.752 7.122 mp_irecv_dv 76685 16.2 2.604 6.852 2.604 6.852 calculate_dm_sparse 129 9.5 0.001 0.001 6.197 6.310 fft3d_ps 1301 14.7 2.859 2.942 5.860 5.920 rs_pw_transfer 1054 12.0 0.013 0.015 4.580 5.918 grid_collocate_task_list 129 9.7 5.120 5.907 5.120 5.907 mp_alltoall_d11v 2423 14.1 4.990 5.528 4.990 5.528 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.378 5.427 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.445 4.537 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.498 4.499 potential_pw2rs 129 12.3 0.023 0.023 4.480 4.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.345 4.395 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.170 4.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.983000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1075.181818, yerr=17.124242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.529393E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 172.647 172.648 qs_mol_dyn_low 1 2.0 0.003 0.004 172.273 172.285 qs_forces 11 3.9 0.004 0.005 172.164 172.172 qs_energies 11 4.9 0.001 0.001 164.650 164.661 scf_env_do_scf 11 5.9 0.001 0.002 147.431 147.435 velocity_verlet 10 3.0 0.001 0.002 113.620 113.625 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 112.010 112.012 dbcsr_multiply_generic 2507 12.6 0.182 0.194 74.384 75.029 qs_scf_new_mos 117 7.6 0.001 0.001 73.556 73.698 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.555 73.697 ot_scf_mini 117 9.6 0.003 0.004 69.128 69.305 multiply_cannon 2507 13.6 0.587 0.625 54.456 59.225 multiply_cannon_loop 2507 14.6 0.441 0.457 49.860 51.188 ot_mini 117 10.6 0.001 0.001 39.763 39.942 init_scf_loop 11 6.9 0.000 0.000 35.275 35.278 mp_waitall_1 125778 16.7 26.261 33.813 26.261 33.813 prepare_preconditioner 11 7.9 0.000 0.000 31.342 31.367 make_preconditioner 11 8.9 0.000 0.000 31.342 31.367 rebuild_ks_matrix 128 8.3 0.001 0.001 29.814 29.970 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.019 29.813 29.969 make_full_inverse_cholesky 11 9.9 0.017 0.028 29.253 29.500 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.091 27.237 multiply_cannon_multrec 10028 15.6 10.368 15.174 17.618 22.036 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 12.660 20.607 ot_diis_step 117 11.6 0.020 0.023 20.108 20.108 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.584 20.075 apply_single 128 13.6 0.001 0.001 19.584 20.075 qs_ot_get_derivative 117 11.6 0.002 0.006 19.589 19.766 make_m2s 5014 13.6 0.067 0.071 16.291 18.699 make_images 5014 14.6 2.298 2.657 15.987 18.384 cp_fm_cholesky_invert 11 10.9 17.944 17.950 17.944 17.950 qs_ot_get_p 128 10.4 0.001 0.001 16.405 16.564 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.226 15.280 calculate_rho_elec 128 8.7 0.256 0.267 15.225 15.279 sum_up_and_integrate 128 10.3 0.181 0.191 14.889 14.939 integrate_v_rspace 128 11.3 0.004 0.004 14.708 14.758 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.680 12.696 make_images_data 5014 15.6 0.052 0.060 9.975 12.557 multiply_cannon_sync_h2d 10028 15.6 11.505 12.177 11.505 12.177 hybrid_alltoall_any 5200 16.5 0.819 3.627 9.803 12.139 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.565 11.567 init_scf_run 11 5.9 0.000 0.001 10.416 10.416 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.415 10.416 pw_transfer 1547 11.6 0.086 0.095 9.369 9.393 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.144 9.177 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.614 8.626 cp_fm_diag_elpa_base 83 14.4 8.372 8.448 8.603 8.616 fft_wrap_pw1pw2_140 523 13.2 1.912 1.950 8.119 8.152 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.399 8.118 grid_integrate_task_list 128 12.3 7.755 8.112 7.755 8.112 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.870 8.002 cp_fm_cholesky_decompose 22 10.9 7.756 7.877 7.756 7.877 mp_allgather_i34 2507 14.6 2.783 7.842 2.783 7.842 density_rs2pw 128 9.7 0.006 0.006 6.777 7.667 dbcsr_mm_accdrv_process 20762 16.1 2.620 3.414 6.878 7.513 wfi_extrapolate 11 7.9 0.001 0.001 7.462 7.462 calculate_dm_sparse 128 9.5 0.001 0.001 6.070 6.195 grid_collocate_task_list 128 9.7 5.394 6.056 5.394 6.056 fft3d_ps 1291 14.7 2.762 2.837 5.620 5.654 mp_alltoall_d11v 2415 14.1 4.943 5.587 4.943 5.587 dbcsr_complete_redistribute 395 12.7 2.131 2.202 5.047 5.453 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.281 5.282 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.884 5.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.178 5.191 mp_irecv_dv 28860 15.9 1.849 5.174 1.849 5.174 rs_pw_transfer 1046 11.9 0.013 0.013 4.029 4.954 potential_pw2rs 128 12.3 0.026 0.027 4.332 4.344 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.110 4.147 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.831 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.709 3.737 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.627 3.640 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.316 3.628 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.481 3.572 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.419 3.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.648000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1432.636364, yerr=46.885301 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.119280E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 300.042 300.042 qs_mol_dyn_low 1 2.0 0.003 0.004 299.507 299.518 qs_forces 11 3.9 0.004 0.005 299.411 299.417 qs_energies 11 4.9 0.001 0.001 290.072 290.081 scf_env_do_scf 11 5.9 0.001 0.001 267.789 267.811 velocity_verlet 10 3.0 0.001 0.001 217.007 217.015 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 141.826 141.829 init_scf_loop 11 6.9 0.000 0.000 125.715 125.722 prepare_preconditioner 11 7.9 0.000 0.000 120.636 120.676 make_preconditioner 11 8.9 0.000 0.000 120.636 120.676 make_full_inverse_cholesky 11 9.9 0.036 0.039 96.032 117.781 qs_scf_new_mos 118 7.6 0.001 0.001 90.905 91.057 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.905 91.056 ot_scf_mini 118 9.6 0.004 0.004 86.030 86.135 dbcsr_multiply_generic 2535 12.6 0.214 0.226 84.336 85.226 cp_fm_upper_to_full 106 14.8 54.085 77.731 54.085 77.731 multiply_cannon 2535 13.6 0.716 0.779 59.434 60.421 multiply_cannon_loop 2535 14.6 0.473 0.481 55.677 57.683 ot_mini 118 10.6 0.001 0.001 45.144 45.266 dbcsr_complete_redistribute 397 12.7 4.008 4.048 30.197 44.119 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.860 40.792 rebuild_ks_matrix 129 8.3 0.001 0.001 38.541 38.707 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.540 38.706 transfer_fm_to_dbcsr 11 9.9 0.027 0.033 24.558 38.317 mp_alltoall_i22 720 14.1 22.426 36.573 22.426 36.573 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.565 35.719 mp_waitall_1 104820 16.8 28.965 33.135 28.965 33.135 cp_fm_cholesky_invert 11 10.9 33.086 33.092 33.086 33.092 qs_ot_get_p 129 10.4 0.001 0.001 25.331 25.444 qs_ot_get_derivative 118 11.6 0.002 0.002 24.232 24.342 qs_ot_p2m_diag 84 11.4 0.890 0.894 21.259 21.288 multiply_cannon_metrocomm3 10140 15.6 0.024 0.025 19.357 21.278 make_m2s 5070 13.6 0.077 0.081 20.110 21.099 ot_diis_step 118 11.6 0.022 0.023 20.869 20.870 make_images 5070 14.6 3.783 3.893 19.626 20.617 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.460 20.477 calculate_rho_elec 129 8.7 0.485 0.486 20.459 20.477 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.937 20.354 apply_single 129 13.6 0.001 0.001 19.936 20.354 sum_up_and_integrate 129 10.3 0.323 0.324 19.627 19.716 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.530 19.532 integrate_v_rspace 129 11.3 0.004 0.004 19.304 19.396 multiply_cannon_multrec 10140 15.6 10.554 12.404 17.962 18.066 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.350 16.351 cp_fm_diag_elpa_base 84 14.4 11.838 13.536 16.346 16.346 multiply_cannon_sync_h2d 10140 15.6 15.780 15.793 15.780 15.793 hybrid_alltoall_any 5257 16.5 1.310 3.058 11.080 12.862 make_images_data 5070 15.6 0.060 0.064 10.852 12.527 init_scf_run 11 5.9 0.000 0.001 12.202 12.203 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.202 12.203 pw_transfer 1559 11.6 0.094 0.094 11.340 11.344 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.102 11.107 mp_alltoall_d11v 2429 14.1 8.452 10.081 8.452 10.081 fft_wrap_pw1pw2_140 527 13.2 3.098 3.126 9.888 9.900 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.480 9.556 wfi_extrapolate 11 7.9 0.001 0.001 9.092 9.093 dbcsr_mm_accdrv_process 20958 16.1 3.794 5.758 7.169 9.015 grid_integrate_task_list 129 12.3 8.632 8.814 8.632 8.814 cp_fm_cholesky_decompose 22 10.9 8.734 8.784 8.734 8.784 density_rs2pw 129 9.7 0.006 0.006 7.927 8.059 qs_energies_init_hamiltonians 11 5.9 0.007 0.012 8.036 8.037 calculate_dm_sparse 129 9.5 0.001 0.001 6.704 6.777 grid_collocate_task_list 129 9.7 6.410 6.440 6.410 6.440 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.313 6.357 rs_scatter_matrices 140 9.7 3.643 4.622 6.084 6.317 fft3d_ps 1301 14.7 2.865 2.878 6.136 6.161 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.031 6.160 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.042000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2801.454545, yerr=169.770092 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 1.255707E+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 30200558. 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 0.028 0.053 84.959 84.962 qs_energies 1 2.0 0.000 0.000 84.276 84.289 ls_scf 1 3.0 0.000 0.001 83.349 83.362 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.516 72.712 multiply_cannon 111 7.7 0.017 0.020 55.837 57.083 multiply_cannon_loop 111 8.7 0.207 0.223 52.467 53.952 ls_scf_main 1 4.0 0.000 0.001 52.200 52.212 density_matrix_trs4 2 5.0 0.002 0.003 46.755 46.835 ls_scf_init_scf 1 4.0 0.000 0.002 28.109 28.110 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.027 27.079 mp_waitall_1 11031 10.9 22.480 25.571 22.480 25.571 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.945 24.966 multiply_cannon_multrec 2664 9.7 8.166 8.946 15.443 17.165 multiply_cannon_sync_h2d 2664 9.7 13.632 15.038 13.632 15.038 make_m2s 222 7.7 0.008 0.011 13.067 13.581 make_images 222 8.7 0.098 0.108 13.045 13.562 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.769 12.511 make_images_data 222 9.7 0.004 0.005 7.660 8.273 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.387 8.065 hybrid_alltoall_any 227 10.6 0.215 1.838 6.589 8.052 dbcsr_mm_accdrv_process 4760 10.4 0.502 0.602 6.898 7.846 dbcsr_mm_accdrv_process_sort 4760 11.4 6.196 7.074 6.196 7.074 calculate_norms 4752 9.8 5.504 6.115 5.504 6.115 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.106 mp_sum_l 807 5.4 3.083 4.461 3.083 4.461 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.075 3.834 mp_irecv_dv 6231 10.9 2.059 3.800 2.059 3.800 make_images_sizes 222 9.7 0.000 0.000 0.735 3.564 mp_alltoall_i44 222 10.7 0.735 3.564 0.735 3.564 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.322 3.374 arnoldi_extremal 4 6.8 0.000 0.000 3.180 3.204 arnoldi_normal_ev 4 7.8 0.001 0.003 3.180 3.204 build_subspace 16 8.4 0.009 0.012 3.089 3.091 ls_scf_post 1 4.0 0.000 0.001 3.040 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.319 2.750 dbcsr_merge_single_wm 555 10.7 0.450 0.590 2.310 2.741 make_images_pack 222 9.7 2.207 2.623 2.209 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.306 2.553 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.461 2.066 2.462 dbcsr_sort_data 658 11.4 2.118 2.461 2.118 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.247 2.323 buffer_matrices_ensure_size 222 8.7 1.748 2.016 1.748 2.016 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.765 1.767 rebuild_ks_matrix 3 7.3 0.000 0.000 1.756 1.757 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.756 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.962000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 2.087416E+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 0.033 0.047 91.879 91.881 qs_energies 1 2.0 0.000 0.000 91.430 91.435 ls_scf 1 3.0 0.000 0.000 90.119 90.124 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.918 76.288 multiply_cannon 111 7.7 0.029 0.040 53.490 57.693 ls_scf_main 1 4.0 0.000 0.000 55.501 55.507 multiply_cannon_loop 111 8.7 0.116 0.122 50.217 53.702 density_matrix_trs4 2 5.0 0.002 0.003 49.710 49.955 ls_scf_init_scf 1 4.0 0.000 0.000 31.062 31.063 mp_waitall_1 9105 10.9 21.539 30.694 21.539 30.694 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.591 29.707 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.204 27.213 multiply_cannon_multrec 1332 9.7 13.186 16.560 22.352 26.810 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.954 21.018 make_m2s 222 7.7 0.006 0.007 15.591 16.335 make_images 222 8.7 1.579 1.950 15.561 16.306 dbcsr_mm_accdrv_process 4041 10.4 0.300 0.452 8.763 10.315 make_images_data 222 9.7 0.004 0.004 8.974 10.000 dbcsr_mm_accdrv_process_sort 4041 11.4 8.357 9.863 8.357 9.863 hybrid_alltoall_any 227 10.6 0.521 2.464 8.342 9.556 mp_sum_l 807 5.4 5.517 8.849 5.517 8.849 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.193 7.708 mp_irecv_dv 3311 11.0 3.174 7.658 3.174 7.658 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.262 7.081 calculate_norms 2376 9.8 5.999 6.707 5.999 6.707 multiply_cannon_sync_h2d 1332 9.7 4.849 6.221 4.849 6.221 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.140 5.350 arnoldi_extremal 4 6.8 0.000 0.000 4.611 4.645 arnoldi_normal_ev 4 7.8 0.001 0.005 4.611 4.645 build_subspace 16 8.4 0.014 0.021 4.347 4.351 ls_scf_post 1 4.0 0.000 0.000 3.556 3.561 ls_scf_store_result 1 5.0 0.000 0.000 3.250 3.379 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.114 3.345 dbcsr_matrix_vector_mult_local 304 10.0 2.726 3.205 2.728 3.207 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.239 2.792 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.591 2.687 mp_allgather_i34 111 8.7 0.974 2.514 0.974 2.514 make_images_pack 222 9.7 2.030 2.461 2.032 2.463 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.145 2.147 rebuild_ks_matrix 3 7.3 0.000 0.000 2.132 2.134 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.132 2.134 dbcsr_sort_data 436 11.2 1.856 2.101 1.856 2.101 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.881000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1740.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 2.761576E+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 0.033 0.050 94.822 94.823 qs_energies 1 2.0 0.000 0.000 94.231 94.238 ls_scf 1 3.0 0.000 0.000 92.835 92.840 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.383 77.688 ls_scf_main 1 4.0 0.000 0.000 58.114 58.118 multiply_cannon 111 7.7 0.044 0.098 53.159 57.196 multiply_cannon_loop 111 8.7 0.100 0.111 49.639 53.742 density_matrix_trs4 2 5.0 0.002 0.003 52.132 52.345 mp_waitall_1 7281 11.0 24.431 35.119 24.431 35.119 ls_scf_init_scf 1 4.0 0.000 0.000 31.115 31.118 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.687 29.746 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.299 27.312 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.455 24.170 multiply_cannon_multrec 888 9.7 12.553 15.296 21.021 24.040 make_m2s 222 7.7 0.006 0.007 17.316 18.538 make_images 222 8.7 1.970 2.304 17.278 18.497 hybrid_alltoall_any 227 10.6 0.621 2.866 9.545 11.083 make_images_data 222 9.7 0.003 0.004 9.937 10.924 mp_sum_l 807 5.4 5.345 9.536 5.345 9.536 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.415 8.005 9.253 dbcsr_mm_accdrv_process_sort 3754 11.4 7.636 8.838 7.636 8.838 multiply_cannon_sync_h2d 888 9.7 6.118 7.615 6.118 7.615 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.143 7.423 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.657 7.043 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.460 6.916 mp_irecv_dv 2335 11.1 2.444 6.871 2.444 6.871 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.955 5.183 arnoldi_extremal 4 6.8 0.000 0.000 5.030 5.047 arnoldi_normal_ev 4 7.8 0.001 0.005 5.030 5.047 build_subspace 16 8.4 0.014 0.020 4.730 4.737 calculate_norms 1584 9.8 4.261 4.647 4.261 4.647 mp_allgather_i34 111 8.7 1.360 3.873 1.360 3.873 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.406 3.733 ls_scf_post 1 4.0 0.000 0.000 3.606 3.613 dbcsr_matrix_vector_mult_local 304 10.0 2.995 3.561 2.997 3.563 ls_scf_store_result 1 5.0 0.000 0.000 3.323 3.429 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.814 2.900 make_images_sizes 222 9.7 0.000 0.000 1.048 2.187 mp_alltoall_i44 222 10.7 1.047 2.187 1.047 2.187 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.167 2.169 rebuild_ks_matrix 3 7.3 0.000 0.000 2.149 2.150 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.149 2.150 dbcsr_sort_data 325 11.1 1.900 2.145 1.900 2.145 make_images_pack 222 9.7 1.819 2.099 1.821 2.102 dbcsr_data_release 9322 10.9 1.305 2.006 1.305 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.823000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2155.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 3.312943E+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 0.040 0.053 97.054 97.055 qs_energies 1 2.0 0.000 0.000 96.405 96.411 ls_scf 1 3.0 0.000 0.001 94.763 94.771 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.549 78.866 ls_scf_main 1 4.0 0.000 0.002 58.959 58.960 multiply_cannon 111 7.7 0.052 0.148 51.812 55.771 density_matrix_trs4 2 5.0 0.002 0.003 52.900 53.007 multiply_cannon_loop 111 8.7 0.114 0.124 46.783 49.953 ls_scf_init_scf 1 4.0 0.000 0.002 32.541 32.543 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.329 31.395 mp_waitall_1 6369 11.0 23.035 30.029 23.035 30.029 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.867 28.880 multiply_cannon_multrec 1332 9.7 14.183 17.014 21.996 25.295 make_m2s 222 7.7 0.007 0.007 21.143 22.522 make_images 222 8.7 3.125 3.601 21.092 22.474 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.446 16.878 make_images_data 222 9.7 0.004 0.004 11.848 13.485 hybrid_alltoall_any 227 10.6 0.796 3.784 11.203 12.853 dbcsr_mm_accdrv_process 3641 10.4 0.236 0.410 7.453 8.968 dbcsr_mm_accdrv_process_sort 3641 11.4 7.079 8.552 7.079 8.552 mp_sum_l 807 5.4 4.017 7.110 4.017 7.110 multiply_cannon_sync_h2d 1332 9.7 5.527 6.406 5.527 6.406 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.072 6.060 mp_irecv_dv 3229 10.9 2.048 5.987 2.048 5.987 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.099 5.546 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.572 5.347 arnoldi_extremal 4 6.8 0.000 0.000 5.191 5.204 arnoldi_normal_ev 4 7.8 0.001 0.005 5.191 5.204 build_subspace 16 8.4 0.014 0.021 4.840 4.847 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.505 4.663 calculate_norms 2376 9.8 4.212 4.515 4.212 4.515 mp_allgather_i34 111 8.7 2.150 4.379 2.150 4.379 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.532 3.832 dbcsr_matrix_vector_mult_local 304 10.0 3.153 3.655 3.156 3.657 dbcsr_sort_data 658 11.4 3.062 3.496 3.062 3.496 ls_scf_post 1 4.0 0.000 0.001 3.263 3.270 dbcsr_special_finalize 555 9.7 0.006 0.007 2.799 3.221 dbcsr_merge_single_wm 555 10.7 0.531 0.669 2.791 3.213 ls_scf_store_result 1 5.0 0.000 0.000 2.960 3.069 ls_scf_dm_to_ks 2 5.0 0.000 0.002 2.985 3.044 dbcsr_data_release 10477 10.7 1.583 2.435 1.583 2.435 dbcsr_finalize 304 7.8 0.049 0.061 1.800 2.007 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.951 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.055000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2663.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 4.680278E+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 0.048 0.066 98.970 98.971 qs_energies 1 2.0 0.000 0.000 98.201 98.205 ls_scf 1 3.0 0.000 0.000 96.277 96.286 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.963 78.145 ls_scf_main 1 4.0 0.000 0.000 62.023 62.025 multiply_cannon 111 7.7 0.102 0.200 55.393 60.904 density_matrix_trs4 2 5.0 0.002 0.003 54.904 55.026 multiply_cannon_loop 111 8.7 0.069 0.075 51.011 52.525 mp_waitall_1 5436 11.0 26.773 32.631 26.773 32.631 ls_scf_init_scf 1 4.0 0.000 0.000 30.682 30.686 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.424 29.463 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.259 multiply_cannon_multrec 444 9.7 13.966 16.745 20.959 22.537 make_m2s 222 7.7 0.004 0.005 17.814 20.185 make_images 222 8.7 3.716 4.414 17.752 20.125 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.224 16.466 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.337 14.281 make_images_data 222 9.7 0.003 0.004 10.015 12.377 hybrid_alltoall_any 227 10.6 0.790 3.782 9.858 12.251 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.347 6.708 7.838 multiply_cannon_sync_h2d 444 9.7 6.601 7.647 6.601 7.647 dbcsr_mm_accdrv_process_sort 3003 11.4 6.394 7.488 6.394 7.488 mp_allgather_i34 111 8.7 2.620 7.002 2.620 7.002 arnoldi_extremal 4 6.8 0.000 0.000 5.701 5.711 arnoldi_normal_ev 4 7.8 0.002 0.005 5.701 5.711 build_subspace 16 8.4 0.015 0.020 5.315 5.324 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.553 4.737 mp_sum_l 807 5.4 2.809 4.494 2.809 4.494 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.695 4.389 mp_irecv_dv 1241 11.2 1.676 4.354 1.676 4.354 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.089 4.280 dbcsr_matrix_vector_mult_local 304 10.0 3.629 4.094 3.631 4.096 calculate_norms 792 9.8 3.544 3.638 3.544 3.638 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.530 3.625 ls_scf_post 1 4.0 0.000 0.000 3.572 3.576 make_images_sizes 222 9.7 0.000 0.000 1.160 3.501 mp_alltoall_i44 222 10.7 1.160 3.500 1.160 3.500 ls_scf_store_result 1 5.0 0.000 0.000 3.350 3.400 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.931 3.288 dbcsr_finalize 304 7.8 0.062 0.078 2.194 2.282 dbcsr_merge_all 275 8.9 0.473 0.519 2.046 2.120 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.079 2.081 rebuild_ks_matrix 3 7.3 0.000 0.000 2.046 2.049 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.046 2.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.971000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3628.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_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 8.813265E+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 0.092 0.113 107.689 107.690 qs_energies 1 2.0 0.000 0.000 106.267 106.280 ls_scf 1 3.0 0.000 0.000 103.372 103.384 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.358 77.481 ls_scf_main 1 4.0 0.000 0.000 65.691 65.692 density_matrix_trs4 2 5.0 0.002 0.003 56.556 56.619 multiply_cannon 111 7.7 0.154 0.248 49.951 51.871 multiply_cannon_loop 111 8.7 0.067 0.069 46.463 47.481 ls_scf_init_scf 1 4.0 0.000 0.000 34.009 34.009 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.467 32.486 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.646 29.658 mp_waitall_1 4527 11.1 22.158 26.130 22.158 26.130 make_m2s 222 7.7 0.005 0.005 23.794 24.789 make_images 222 8.7 4.589 4.962 23.688 24.681 multiply_cannon_multrec 444 9.7 17.887 18.469 22.472 22.978 make_images_data 222 9.7 0.003 0.004 13.047 15.483 hybrid_alltoall_any 227 10.6 1.660 3.623 12.851 15.471 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.531 11.061 multiply_cannon_sync_h2d 444 9.7 8.850 8.911 8.850 8.911 arnoldi_extremal 4 6.8 0.000 0.000 7.138 7.153 arnoldi_normal_ev 4 7.8 0.003 0.009 7.137 7.153 build_subspace 16 8.4 0.026 0.036 6.592 6.602 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.287 5.414 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.005 5.268 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.111 5.201 dbcsr_matrix_vector_mult_local 304 10.0 4.881 5.156 4.883 5.159 dbcsr_mm_accdrv_process 1814 10.4 0.279 0.327 4.417 4.552 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.246 4.114 4.246 ls_scf_post 1 4.0 0.000 0.000 3.673 3.684 mp_allgather_i34 111 8.7 1.123 3.526 1.123 3.526 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.435 make_images_sizes 222 9.7 0.000 0.000 1.414 3.433 mp_alltoall_i44 222 10.7 1.413 3.433 1.413 3.433 calculate_norms 792 9.8 3.238 3.270 3.238 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.116 dbcsr_merge_all 275 8.9 0.892 0.922 2.864 2.903 dbcsr_data_release 12724 10.6 2.326 2.888 2.326 2.888 dbcsr_complete_redistribute 5 7.6 1.431 1.461 2.754 2.874 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 2.865 2.865 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.520 dbcsr_sort_data 325 11.1 2.438 2.490 2.438 2.490 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.457 2.458 rebuild_ks_matrix 3 7.3 0.000 0.000 2.391 2.393 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.391 2.393 dbcsr_new_transposed 4 7.5 0.242 0.251 2.269 2.286 dbcsr_frobenius_norm 74 6.6 2.058 2.133 2.190 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.209 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.690000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6861.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/784694e3a5332e1a8c4df163d64bc806796dde08_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 582.193152E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 639962389. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.050 229.943 229.944 qs_mol_dyn_low 1 2.0 0.003 0.007 229.034 229.056 qs_forces 5 3.8 0.005 0.006 228.967 228.969 qs_energies 5 4.8 0.000 0.001 226.031 226.052 scf_env_do_scf 5 5.8 0.000 0.001 207.738 207.741 scf_env_do_scf_inner_loop 105 6.6 0.002 0.007 181.632 181.633 qs_scf_new_mos 105 7.6 0.000 0.001 140.993 141.173 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.993 141.172 dbcsr_multiply_generic 1445 12.2 0.127 0.139 135.105 135.499 ot_scf_mini 105 9.6 0.003 0.003 131.322 131.454 multiply_cannon 1445 13.2 0.275 0.288 116.488 118.698 multiply_cannon_loop 1445 14.2 2.374 2.510 114.772 116.183 velocity_verlet 4 3.0 0.001 0.001 105.999 106.000 ot_mini 105 10.6 0.001 0.001 59.260 59.401 multiply_cannon_multrec 69360 15.2 31.853 37.132 39.442 44.759 qs_ot_get_p 112 10.4 0.001 0.001 42.456 42.754 mp_waitall_1 488190 16.1 33.557 41.666 33.557 41.666 qs_ot_get_derivative 55 11.6 0.001 0.001 37.656 37.808 multiply_cannon_sync_h2d 69360 15.2 30.861 35.419 30.861 35.419 multiply_cannon_metrocomm3 69360 15.2 0.194 0.204 24.668 33.510 qs_ot_p2m_diag 40 11.0 0.020 0.030 31.603 31.674 rebuild_ks_matrix 110 8.4 0.000 0.000 29.138 29.309 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.013 29.138 29.308 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.554 28.554 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.761 26.917 init_scf_loop 7 6.6 0.000 0.000 26.076 26.077 cp_fm_syevd 40 13.0 0.000 0.000 23.453 23.609 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.963 23.228 apply_single 62 13.6 0.000 0.000 22.963 23.228 prepare_preconditioner 7 7.6 0.000 0.000 21.236 21.273 make_preconditioner 7 8.6 0.000 0.000 21.236 21.273 ot_new_cg_direction 55 11.6 0.001 0.001 20.930 20.930 cp_fm_redistribute_end 40 14.0 9.395 18.745 9.400 18.746 cp_fm_syevd_base 40 14.0 9.339 18.693 9.339 18.693 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.402 17.736 calculate_rho_elec 110 8.6 0.030 0.033 17.402 17.736 init_scf_run 5 5.8 0.000 0.000 15.687 15.687 scf_env_initial_rho_setup 5 6.8 0.002 0.005 15.686 15.687 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.486 14.661 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.259 14.322 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.652 13.739 rs_pw_transfer 690 11.5 0.011 0.012 11.599 12.933 density_rs2pw 110 9.6 0.006 0.007 11.499 12.799 mp_sum_l 4684 12.4 11.827 12.710 11.827 12.710 pw_transfer 1645 12.4 0.084 0.112 12.088 12.333 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.945 12.195 calculate_dm_sparse 110 9.5 0.000 0.001 11.289 11.437 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 11.117 11.134 acc_transpose_blocks 69360 15.2 0.360 0.378 10.212 10.829 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 10.682 10.755 calculate_first_density_matrix 1 7.0 0.000 0.001 10.723 10.747 qs_vxc_create 110 10.4 0.002 0.004 10.468 10.519 fft_wrap_pw1pw2_240 915 15.0 1.190 1.288 10.155 10.377 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 9.925 9.994 cp_fm_cholesky_invert 7 10.6 9.986 9.994 9.986 9.994 check_diag 80 13.5 8.589 8.860 9.421 9.562 acc_transpose_blocks_kernels 69360 16.2 0.848 0.891 8.316 8.844 multiply_cannon_metrocomm1 69360 15.2 0.090 0.096 4.421 8.835 sum_up_and_integrate 60 10.3 0.028 0.031 8.711 8.722 integrate_v_rspace 60 11.3 0.002 0.002 8.683 8.695 fft3d_pb 915 16.0 2.388 2.667 8.220 8.485 dbcsr_mm_accdrv_process 154766 15.8 3.983 4.140 7.457 8.249 jit_kernel_transpose 5 15.0 7.468 7.974 7.468 7.974 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.300 7.553 xc_vxc_pw_create 60 11.3 0.039 0.050 7.005 7.055 make_m2s 2890 13.2 0.078 0.087 6.298 6.854 make_full_single_inverse 7 9.6 0.001 0.001 6.717 6.754 make_images 2890 14.2 0.241 0.259 6.193 6.748 xc_pw_derive 510 13.4 0.005 0.007 6.206 6.290 mp_waitany 7680 13.5 4.404 5.813 4.404 5.813 mp_alltoall_z22v 2340 17.7 5.296 5.574 5.296 5.574 multiply_cannon_metrocomm4 67915 15.2 0.181 0.196 2.058 5.190 potential_pw2rs 60 12.3 0.003 0.003 4.881 4.914 wfi_extrapolate 5 7.8 0.000 0.001 4.643 4.643 mp_irecv_dv 143422 16.1 1.756 4.611 1.756 4.611 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.944000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=555.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 784694e3a5332e1a8c4df163d64bc806796dde08 Summary: empty Status: OK