=== 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: 263bc551066ec83fd9537a4c612e9ad8f6479d6e ################# 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 2023.05.001, # HDF5 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, 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. # # Last update: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/01 job id: 49179439 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/02 job id: 49179441 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/03 job id: 49179442 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/04 job id: 49179445 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/05 job id: 49179446 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/06 job id: 49179447 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/07 job id: 49179449 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/08 job id: 49179450 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/09 job id: 49179451 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/10 job id: 49179453 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/11 job id: 49179454 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/12 job id: 49179455 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/13 job id: 49179459 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/14 job id: 49179461 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/15 job id: 49179463 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/16 job id: 49179464 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/17 job id: 49179466 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/18 job id: 49179467 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/19 job id: 49179468 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/20 job id: 49179469 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/21 job id: 49179470 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/22 job id: 49179471 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/23 job id: 49179473 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/24 job id: 49179474 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/25 job id: 49179476 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/26 job id: 49179478 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/27 job id: 49179479 --- 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/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 424 8. 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 0.020 0.035 137.275 137.318 farming_run 1 2.0 136.771 136.813 137.239 137.283 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.486520E+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 1134128. MP_Allreduce 571 1938539. 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 0.024 116.510 116.547 qs_energies 1 2.0 0.000 0.000 116.313 116.349 mp2_main 1 3.0 0.000 0.000 113.818 113.854 mp2_gpw_main 1 4.0 0.027 0.033 112.467 112.502 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.127 93.509 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.769 55.153 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.166 41.167 46.270 get_2c_integrals 1 6.0 0.008 0.009 37.388 38.185 integrate_v_rspace 273 8.0 0.440 0.455 24.723 29.697 pw_transfer 6555 10.6 0.379 0.398 26.921 27.330 grid_integrate_task_list 273 9.0 20.608 26.065 20.608 26.065 fft_wrap_pw1pw2 5465 11.4 0.043 0.045 25.529 26.040 fft_wrap_pw1pw2_100 2178 12.4 1.163 1.340 23.094 23.607 compute_2c_integrals 1 7.0 0.003 0.005 19.580 19.586 rpa_ri_compute_en 1 5.0 0.020 0.021 19.221 19.470 compute_2c_integrals_loop_lm 1 8.0 0.004 0.006 18.665 19.192 mp2_eri_2c_integrate_gpw 1 9.0 2.354 2.419 18.662 19.191 cp_fm_cholesky_decompose 12 8.2 17.995 18.812 17.995 18.812 cholesky_decomp 1 7.0 0.000 0.000 16.654 17.476 fft3d_s 5443 13.4 16.176 16.569 16.199 16.592 ao_to_mo_and_store_B_mult_1 272 7.0 10.769 15.317 10.769 15.317 calculate_wavefunction 272 8.0 5.424 5.575 12.358 13.039 rpa_num_int 1 6.0 0.000 0.002 10.962 10.974 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.917 10.944 calc_mat_Q 8 8.0 0.000 0.000 9.554 9.656 contract_S_to_Q 8 9.0 0.000 0.000 8.975 9.076 calc_potential_gpw 544 9.5 0.005 0.006 8.275 8.768 parallel_gemm_fm 14 9.1 0.000 0.000 8.560 8.679 parallel_gemm_fm_cosma 14 10.1 8.560 8.679 8.560 8.679 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.131 8.468 potential_pw2rs 545 10.0 0.106 0.107 7.518 8.127 create_integ_mat 1 6.0 0.014 0.028 7.782 7.791 collocate_single_gaussian 272 10.0 0.039 0.042 7.370 7.652 array2fm 1 7.0 0.000 0.000 6.737 7.089 pw_scatter_s 2720 13.7 4.354 4.607 4.354 4.607 pw_gather_s 2722 13.2 3.434 3.687 3.434 3.687 array2fm_buffer_send 1 8.0 3.027 3.164 3.027 3.164 pw_poisson_solve 545 10.5 1.113 1.187 2.210 2.501 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.466615, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 424 9. 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 0.025 0.038 421.945 421.946 farming_run 1 2.0 421.065 421.075 421.902 421.906 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.234596E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 168297. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 217.585 217.585 qs_energies 1 2.0 0.003 0.021 217.360 217.372 scf_env_do_scf 1 3.0 0.000 0.000 115.147 115.147 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.201 114.208 rebuild_ks_matrix 4 6.0 0.000 0.000 114.200 114.207 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 114.200 114.207 hfx_ks_matrix 4 8.0 0.001 0.001 113.854 113.858 integrate_four_center 4 9.0 0.152 0.479 113.853 113.857 integrate_four_center_main 4 10.0 0.124 0.606 101.922 105.785 integrate_four_center_bin 267 11.0 101.798 105.756 101.798 105.756 mp2_main 1 3.0 0.001 0.016 101.912 101.927 mp2_gpw_main 1 4.0 0.053 0.087 100.962 100.978 init_scf_loop 1 4.0 0.000 0.000 96.909 96.909 mp2_ri_gpw_compute_in 1 5.0 0.069 0.108 73.999 75.048 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.013 53.803 54.852 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 41.557 46.594 integrate_v_rspace 95 8.0 0.397 0.565 28.008 32.902 pw_transfer 2240 10.6 0.147 0.173 29.536 29.969 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.542 28.983 ao_to_mo_and_store_B_mult_1 91 7.0 10.562 28.946 10.562 28.946 mp2_ri_gpw_compute_en 1 5.0 0.057 0.069 26.757 28.814 grid_integrate_task_list 95 9.0 23.428 28.534 23.428 28.534 fft_wrap_pw1pw2_100 730 12.4 1.275 1.471 26.301 26.760 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.857 1.906 25.038 25.046 get_2c_integrals 1 6.0 0.003 0.030 20.102 20.130 compute_2c_integrals 1 7.0 0.003 0.004 19.080 19.087 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.629 18.958 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.851 18.628 18.957 fft3d_s 1823 13.4 18.489 18.905 18.504 18.920 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.236 18.236 calculate_wavefunction 91 8.0 2.015 2.044 9.633 9.869 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.583 8.783 9.352 potential_pw2rs 186 10.0 0.033 0.034 8.412 8.982 local_gemm 172 8.0 8.227 8.793 8.227 8.793 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.150 8.581 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.516 8.008 8.423 calc_potential_gpw 182 9.5 0.002 0.003 7.914 8.162 collocate_single_gaussian 91 10.0 0.017 0.031 7.759 8.039 mp_sync 37 10.5 3.901 7.447 3.901 7.447 integrate_four_center_load 4 10.0 0.000 0.000 6.755 6.758 hfx_load_balance 1 11.0 0.000 0.000 6.755 6.758 mp_sendrecv_dm3 2068 8.0 6.046 6.466 6.046 6.466 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.437 6.344 6.437 pw_gather_s 912 13.2 4.501 4.933 4.501 4.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.959407, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 452.124672E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 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 0.018 0.041 54.908 54.913 qs_mol_dyn_low 1 2.0 0.005 0.014 54.494 54.504 qs_forces 11 3.9 0.002 0.003 54.398 54.401 qs_energies 11 4.9 0.001 0.002 52.896 52.908 scf_env_do_scf 11 5.9 0.000 0.001 46.836 46.837 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.693 44.693 qs_scf_new_mos 108 7.5 0.000 0.001 34.420 34.725 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.419 34.725 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.979 34.458 ot_scf_mini 108 9.5 0.002 0.002 32.690 32.860 multiply_cannon 2286 13.5 0.190 0.196 26.282 27.859 multiply_cannon_loop 2286 14.5 1.827 1.928 25.572 27.209 velocity_verlet 10 3.0 0.001 0.001 26.833 26.833 ot_mini 108 10.5 0.001 0.001 19.755 20.015 qs_ot_get_derivative 108 11.5 0.001 0.001 16.667 16.854 mp_waitall_1 245248 16.5 8.524 14.565 8.524 14.565 multiply_cannon_metrocomm3 54864 15.5 0.072 0.079 6.053 12.809 multiply_cannon_multrec 54864 15.5 3.625 5.707 7.683 11.079 qs_ot_get_p 119 10.4 0.001 0.002 8.196 8.501 rebuild_ks_matrix 119 8.3 0.000 0.000 7.961 8.100 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 7.961 8.100 mp_sum_l 7287 12.8 5.626 7.462 5.626 7.462 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.008 7.129 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.830 6.309 multiply_cannon_sync_h2d 54864 15.5 5.127 6.273 5.127 6.273 dbcsr_mm_accdrv_process 76910 16.1 1.851 2.885 3.973 5.647 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.319 5.439 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.348 5.402 init_scf_run 11 5.9 0.000 0.001 4.728 4.728 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.728 4.728 sum_up_and_integrate 119 10.3 0.001 0.002 4.546 4.552 integrate_v_rspace 119 11.3 0.002 0.002 4.535 4.542 cp_dbcsr_syevd 50 12.0 0.003 0.005 4.489 4.489 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.291 4.445 calculate_rho_elec 119 8.7 0.011 0.016 4.291 4.444 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.276 4.277 cp_fm_redistribute_end 50 14.0 2.181 4.247 2.188 4.250 cp_fm_diag_elpa_base 50 14.0 2.056 4.133 2.060 4.142 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.970 3.171 apply_single 119 13.6 0.000 0.000 2.969 3.171 calculate_dm_sparse 119 9.5 0.000 0.001 3.001 3.167 multiply_cannon_metrocomm1 54864 15.5 0.054 0.059 1.738 2.913 ot_diis_step 108 11.5 0.006 0.006 2.767 2.767 acc_transpose_blocks 54864 15.5 0.229 0.254 2.225 2.754 jit_kernel_multiply 13 15.8 2.058 2.673 2.058 2.673 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.548 2.606 calculate_first_density_matrix 1 7.0 0.000 0.001 2.477 2.480 density_rs2pw 119 9.7 0.004 0.004 2.269 2.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.246 2.249 wfi_extrapolate 11 7.9 0.001 0.001 2.179 2.179 init_scf_loop 11 6.9 0.000 0.000 2.123 2.124 grid_integrate_task_list 119 12.3 2.026 2.109 2.026 2.109 mp_sum_d 4135 12.0 1.399 2.041 1.399 2.041 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.922 1.969 potential_pw2rs 119 12.3 0.004 0.004 1.901 1.917 pw_transfer 1439 11.6 0.052 0.057 1.779 1.850 make_m2s 4572 13.5 0.054 0.057 1.768 1.812 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.704 1.776 make_images 4572 14.5 0.134 0.139 1.684 1.730 mp_alltoall_d11v 2130 13.8 1.406 1.622 1.406 1.622 transfer_rs2pw 487 10.6 0.005 0.006 1.427 1.519 fft3d_ps 1201 14.6 0.378 0.487 1.419 1.487 mp_waitany 12084 13.8 1.274 1.476 1.274 1.476 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.435 1.453 acc_transpose_blocks_sync 164592 16.5 1.201 1.436 1.201 1.436 grid_collocate_task_list 119 9.7 1.352 1.429 1.352 1.429 transfer_pw2rs 487 13.2 0.006 0.006 1.379 1.394 fft_wrap_pw1pw2_140 487 13.2 0.141 0.153 1.318 1.389 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.804 1.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.913000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.501248E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1420507. 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 0.010 0.027 41.007 41.008 qs_mol_dyn_low 1 2.0 0.003 0.003 40.768 40.778 qs_forces 11 3.9 0.004 0.007 40.674 40.677 qs_energies 11 4.9 0.002 0.003 38.991 38.998 scf_env_do_scf 11 5.9 0.000 0.001 33.317 33.317 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.791 30.791 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.343 23.752 qs_scf_new_mos 108 7.5 0.001 0.001 21.849 22.081 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.848 22.081 ot_scf_mini 108 9.5 0.002 0.003 20.884 21.052 multiply_cannon 2286 13.5 0.209 0.216 17.806 19.418 velocity_verlet 10 3.0 0.001 0.003 19.078 19.079 multiply_cannon_loop 2286 14.5 1.197 1.255 16.587 18.410 ot_mini 108 10.5 0.001 0.001 12.877 13.115 mp_waitall_1 200699 16.5 6.056 11.202 6.056 11.202 qs_ot_get_derivative 108 11.5 0.001 0.001 10.452 10.624 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.404 9.853 multiply_cannon_multrec 27432 15.5 1.858 4.293 6.484 9.495 rebuild_ks_matrix 119 8.3 0.000 0.000 6.941 7.096 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.941 7.095 dbcsr_mm_accdrv_process 47894 16.0 3.714 6.077 4.546 6.891 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.115 6.259 qs_ot_get_p 119 10.4 0.001 0.002 4.984 5.205 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.894 4.741 init_scf_run 11 5.9 0.000 0.001 4.435 4.435 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.435 4.435 mp_sum_l 7287 12.8 2.292 4.397 2.292 4.397 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.033 4.081 apply_single 119 13.6 0.000 0.000 3.032 4.081 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.791 3.827 calculate_rho_elec 119 8.7 0.021 0.024 3.790 3.826 sum_up_and_integrate 119 10.3 0.001 0.001 3.816 3.823 integrate_v_rspace 119 11.3 0.002 0.003 3.802 3.808 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.192 3.216 make_m2s 4572 13.5 0.053 0.054 2.798 3.131 make_images 4572 14.5 0.207 0.248 2.709 3.045 calculate_first_density_matrix 1 7.0 0.001 0.002 2.859 2.861 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.772 2.772 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.511 2.599 init_scf_loop 11 6.9 0.000 0.000 2.501 2.501 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.412 2.413 calculate_dm_sparse 119 9.5 0.000 0.001 2.320 2.402 cp_fm_redistribute_end 50 14.0 1.223 2.379 1.228 2.383 ot_diis_step 108 11.5 0.011 0.011 2.359 2.360 cp_fm_diag_elpa_base 50 14.0 1.121 2.273 1.150 2.314 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.247 2.249 multiply_cannon_sync_h2d 27432 15.5 1.676 2.191 1.676 2.191 acc_transpose_blocks 27432 15.5 0.113 0.120 1.763 2.150 density_rs2pw 119 9.7 0.004 0.004 2.055 2.117 jit_kernel_multiply 10 16.4 0.774 2.096 0.774 2.096 pw_transfer 1439 11.6 0.065 0.068 1.951 1.983 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.940 1.980 grid_integrate_task_list 119 12.3 1.837 1.944 1.837 1.944 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.859 1.894 make_images_data 4572 15.5 0.048 0.054 1.300 1.685 prepare_preconditioner 11 7.9 0.000 0.000 1.611 1.637 make_preconditioner 11 8.9 0.000 0.000 1.611 1.637 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.509 1.568 potential_pw2rs 119 12.3 0.006 0.006 1.544 1.554 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.155 1.539 fft3d_ps 1201 14.6 0.523 0.578 1.495 1.525 wfi_extrapolate 11 7.9 0.001 0.001 1.518 1.518 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.476 1.486 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.448 1.483 mp_alltoall_d11v 2130 13.8 1.309 1.439 1.309 1.439 grid_collocate_task_list 119 9.7 1.291 1.350 1.291 1.350 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.265 1.310 transfer_rs2pw 487 10.6 0.005 0.006 1.181 1.246 mp_sum_d 4135 12.0 0.608 1.095 0.608 1.095 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.807 1.060 mp_allgather_i34 2286 14.5 0.641 1.054 0.641 1.054 transfer_pw2rs 487 13.2 0.005 0.005 0.968 0.976 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.945 0.949 acc_transpose_blocks_sync 82296 16.5 0.818 0.949 0.818 0.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.899 make_images_sizes 4572 15.5 0.005 0.005 0.629 0.898 mp_alltoall_i44 4572 16.5 0.624 0.894 0.624 0.894 mp_alltoall_z22v 1201 16.6 0.748 0.828 0.748 0.828 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.008000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 519.061504E+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 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.035 34.343 34.348 qs_mol_dyn_low 1 2.0 0.003 0.003 34.016 34.023 qs_forces 11 3.9 0.002 0.002 33.943 33.945 qs_energies 11 4.9 0.003 0.009 32.365 32.367 scf_env_do_scf 11 5.9 0.001 0.002 27.566 27.566 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.951 24.952 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.937 18.058 qs_scf_new_mos 108 7.5 0.001 0.001 16.857 16.875 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.857 16.875 ot_scf_mini 108 9.5 0.002 0.004 16.053 16.075 velocity_verlet 10 3.0 0.001 0.003 15.907 15.909 multiply_cannon 2286 13.5 0.194 0.198 13.920 14.688 multiply_cannon_loop 2286 14.5 0.855 0.891 13.095 13.775 ot_mini 108 10.5 0.001 0.001 9.921 9.951 qs_ot_get_derivative 108 11.5 0.001 0.001 8.357 8.380 multiply_cannon_multrec 18288 15.5 1.881 2.876 6.956 7.278 rebuild_ks_matrix 119 8.3 0.000 0.000 6.200 6.225 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.199 6.224 dbcsr_mm_accdrv_process 38222 16.0 4.871 5.664 4.981 5.727 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.475 5.499 mp_waitall_1 158411 16.6 2.998 3.932 2.998 3.932 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.976 3.887 qs_ot_get_p 119 10.4 0.001 0.003 3.792 3.833 sum_up_and_integrate 119 10.3 0.001 0.001 3.725 3.731 integrate_v_rspace 119 11.3 0.002 0.003 3.712 3.719 init_scf_run 11 5.9 0.000 0.002 3.580 3.580 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.580 3.580 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.480 3.489 calculate_rho_elec 119 8.7 0.031 0.031 3.480 3.489 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.345 2.716 apply_single 119 13.6 0.000 0.000 2.345 2.715 init_scf_loop 11 6.9 0.001 0.004 2.597 2.598 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.536 2.543 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.588 2.436 calculate_first_density_matrix 1 7.0 0.001 0.006 2.282 2.283 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.205 2.206 make_m2s 4572 13.5 0.045 0.046 1.990 2.155 density_rs2pw 119 9.7 0.004 0.004 2.014 2.121 make_images 4572 14.5 0.194 0.207 1.904 2.068 mp_sum_l 7287 12.8 1.640 2.021 1.640 2.021 pw_transfer 1439 11.6 0.065 0.069 1.979 1.993 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.918 1.932 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.907 1.919 cp_fm_diag_elpa_base 50 14.0 1.881 1.895 1.904 1.917 calculate_dm_sparse 119 9.5 0.000 0.001 1.897 1.909 grid_integrate_task_list 119 12.3 1.797 1.905 1.797 1.905 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.887 1.902 prepare_preconditioner 11 7.9 0.000 0.000 1.821 1.825 make_preconditioner 11 8.9 0.000 0.001 1.821 1.825 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.663 1.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.732 1.734 acc_transpose_blocks 18288 15.5 0.080 0.082 1.644 1.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.532 1.539 ot_diis_step 108 11.5 0.012 0.012 1.531 1.532 potential_pw2rs 119 12.3 0.007 0.009 1.500 1.508 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.484 1.500 fft3d_ps 1201 14.6 0.532 0.549 1.456 1.469 grid_collocate_task_list 119 9.7 1.234 1.315 1.234 1.315 wfi_extrapolate 11 7.9 0.001 0.001 1.240 1.240 transfer_rs2pw 487 10.6 0.005 0.005 1.119 1.234 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.191 1.197 multiply_cannon_sync_h2d 18288 15.5 1.020 1.180 1.020 1.180 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.024 1.037 make_images_data 4572 15.5 0.048 0.052 0.841 1.014 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 0.984 0.988 mp_alltoall_d11v 2130 13.8 0.800 0.971 0.800 0.971 transfer_pw2rs 487 13.2 0.004 0.004 0.918 0.922 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.726 0.912 mp_alltoall_z22v 1201 16.6 0.762 0.833 0.762 0.833 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.811 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.806 acc_transpose_blocks_sync 54864 16.5 0.734 0.794 0.734 0.794 cp_fm_cholesky_invert 11 10.9 0.748 0.753 0.748 0.753 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 0.674 0.748 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.503 0.735 mp_waitany 9880 13.7 0.578 0.692 0.578 0.692 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.348000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.181818, yerr=0.935966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 556.691456E+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 63489. MP_Allreduce 10154 346. 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 0.010 0.028 38.570 38.570 qs_mol_dyn_low 1 2.0 0.003 0.003 38.377 38.385 qs_forces 11 3.9 0.002 0.003 38.132 38.136 qs_energies 11 4.9 0.001 0.002 36.401 36.405 scf_env_do_scf 11 5.9 0.000 0.001 31.259 31.261 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.762 27.763 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.882 21.055 velocity_verlet 10 3.0 0.002 0.003 19.385 19.394 qs_scf_new_mos 108 7.5 0.001 0.001 19.314 19.375 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.313 19.374 ot_scf_mini 108 9.5 0.002 0.003 18.232 18.287 multiply_cannon 2286 13.5 0.219 0.227 16.660 17.064 multiply_cannon_loop 2286 14.5 1.514 1.628 15.614 16.008 ot_mini 108 10.5 0.001 0.001 11.148 11.213 multiply_cannon_multrec 27432 15.5 2.595 3.381 9.210 9.547 qs_ot_get_derivative 108 11.5 0.001 0.001 9.183 9.239 dbcsr_mm_accdrv_process 47916 15.9 6.096 7.907 6.512 8.073 rebuild_ks_matrix 119 8.3 0.000 0.000 6.547 6.602 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.547 6.601 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.813 5.865 qs_ot_get_p 119 10.4 0.001 0.002 4.138 4.220 init_scf_run 11 5.9 0.000 0.002 3.762 3.763 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.762 3.763 sum_up_and_integrate 119 10.3 0.001 0.001 3.663 3.673 integrate_v_rspace 119 11.3 0.003 0.003 3.652 3.662 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.240 3.635 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.534 3.571 calculate_rho_elec 119 8.7 0.040 0.046 3.533 3.571 init_scf_loop 11 6.9 0.000 0.000 3.476 3.476 mp_waitall_1 137007 16.6 2.223 3.029 2.223 3.029 make_m2s 4572 13.5 0.055 0.057 2.513 2.657 prepare_preconditioner 11 7.9 0.000 0.000 2.624 2.632 make_preconditioner 11 8.9 0.000 0.000 2.624 2.632 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.539 2.567 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.235 2.554 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.536 2.546 make_images 4572 14.5 0.275 0.335 2.403 2.544 acc_transpose_blocks 27432 15.5 0.119 0.124 2.393 2.516 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.150 2.301 apply_single 119 13.6 0.000 0.000 2.150 2.301 calculate_first_density_matrix 1 7.0 0.000 0.002 2.296 2.298 calculate_dm_sparse 119 9.5 0.000 0.000 2.208 2.263 pw_transfer 1439 11.6 0.065 0.069 2.089 2.126 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.097 2.099 density_rs2pw 119 9.7 0.004 0.004 1.954 2.089 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.997 2.037 grid_integrate_task_list 119 12.3 1.827 1.926 1.827 1.926 ot_diis_step 108 11.5 0.012 0.013 1.908 1.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.887 1.889 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.797 1.806 cp_fm_diag_elpa_base 50 14.0 1.761 1.778 1.795 1.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.668 1.686 mp_sum_l 7287 12.8 1.282 1.671 1.282 1.671 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.581 1.622 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.918 1.583 fft3d_ps 1201 14.6 0.557 0.607 1.516 1.543 acc_transpose_blocks_sync 82296 16.5 1.396 1.523 1.396 1.523 potential_pw2rs 119 12.3 0.009 0.009 1.422 1.427 wfi_extrapolate 11 7.9 0.001 0.001 1.410 1.410 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.352 1.366 grid_collocate_task_list 119 9.7 1.247 1.326 1.247 1.326 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.267 cp_fm_upper_to_full 72 14.2 0.811 1.149 0.811 1.149 transfer_rs2pw 487 10.6 0.004 0.005 0.996 1.138 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.112 1.115 make_images_data 4572 15.5 0.048 0.051 0.957 1.105 dbcsr_complete_redistribute 329 12.2 0.129 0.160 0.823 1.092 jit_kernel_multiply 6 16.1 0.350 1.069 0.350 1.069 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.807 1.004 mp_alltoall_d11v 2130 13.8 0.859 0.957 0.859 0.957 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.610 0.876 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.872 mp_alltoall_z22v 1201 16.6 0.840 0.869 0.840 0.869 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.849 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.831 cp_fm_cholesky_invert 11 10.9 0.806 0.810 0.806 0.810 transfer_pw2rs 487 13.2 0.004 0.004 0.803 0.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.570000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.636364, yerr=3.141300 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 622.882816E+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 63488. MP_Allreduce 10154 346. 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 0.014 0.033 55.402 55.404 qs_mol_dyn_low 1 2.0 0.003 0.003 55.142 55.153 qs_forces 11 3.9 0.002 0.002 55.068 55.069 qs_energies 11 4.9 0.002 0.003 52.777 52.782 scf_env_do_scf 11 5.9 0.000 0.001 45.839 45.840 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 39.525 39.527 velocity_verlet 10 3.0 0.002 0.003 29.153 29.157 qs_scf_new_mos 108 7.5 0.001 0.001 24.481 24.547 qs_scf_loop_do_ot 108 8.5 0.001 0.001 24.480 24.546 ot_scf_mini 108 9.5 0.002 0.002 23.195 23.255 dbcsr_multiply_generic 2286 12.5 0.098 0.103 22.648 23.118 multiply_cannon 2286 13.5 0.228 0.236 15.237 16.429 multiply_cannon_loop 2286 14.5 0.643 0.662 13.691 14.854 ot_mini 108 10.5 0.001 0.001 12.558 12.676 mp_waitall_1 115863 16.7 9.322 11.493 9.322 11.493 qs_ot_get_derivative 108 11.5 0.001 0.001 10.584 10.654 rebuild_ks_matrix 119 8.3 0.000 0.000 10.395 10.526 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 10.395 10.526 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.289 9.406 qs_ot_get_p 119 10.4 0.002 0.002 7.471 7.567 qs_rho_update_rho_low 119 7.7 0.001 0.001 7.331 7.382 calculate_rho_elec 119 8.7 0.060 0.061 7.330 7.381 sum_up_and_integrate 119 10.3 0.001 0.001 7.177 7.188 integrate_v_rspace 119 11.3 0.003 0.003 7.167 7.178 multiply_cannon_multrec 9144 15.5 1.697 1.853 6.176 6.473 init_scf_loop 11 6.9 0.000 0.000 6.254 6.259 make_m2s 4572 13.5 0.035 0.036 5.184 5.667 make_images 4572 14.5 0.272 0.303 5.093 5.575 density_rs2pw 119 9.7 0.003 0.004 5.257 5.323 init_scf_run 11 5.9 0.000 0.001 5.052 5.052 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.051 5.052 prepare_preconditioner 11 7.9 0.000 0.000 4.967 4.977 make_preconditioner 11 8.9 0.000 0.000 4.967 4.977 multiply_cannon_metrocomm1 9144 15.5 0.024 0.025 4.141 4.976 qs_ot_p2m_diag 50 11.0 0.022 0.023 4.792 4.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.675 4.789 dbcsr_mm_accdrv_process 12550 15.8 3.811 4.463 4.367 4.559 potential_pw2rs 119 12.3 0.010 0.011 4.435 4.498 pw_transfer 1439 11.6 0.066 0.068 4.256 4.338 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.262 4.269 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 4.162 4.245 make_images_data 4572 15.5 0.042 0.046 2.469 3.802 fft3d_ps 1201 14.6 0.584 0.599 3.578 3.653 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.542 3.582 hybrid_alltoall_any 4725 16.4 0.065 0.175 2.125 3.532 cp_fm_diag_elpa 50 13.0 0.000 0.000 3.447 3.454 cp_fm_diag_elpa_base 50 14.0 3.405 3.432 3.443 3.449 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 3.368 3.411 transfer_rs2pw 487 10.6 0.004 0.004 3.229 3.350 make_images_sizes 4572 15.5 0.005 0.005 1.925 3.139 mp_alltoall_i44 4572 16.5 1.920 3.135 1.920 3.135 transfer_pw2rs 487 13.2 0.004 0.004 3.062 3.133 mp_alltoall_z22v 1201 16.6 2.857 2.922 2.857 2.922 fft_wrap_pw1pw2_140 487 13.2 0.321 0.330 2.792 2.877 calculate_dm_sparse 119 9.5 0.000 0.000 2.698 2.742 mp_sum_l 7287 12.8 1.794 2.720 1.794 2.720 cp_fm_cholesky_invert 11 10.9 2.558 2.563 2.558 2.563 calculate_first_density_matrix 1 7.0 0.000 0.001 2.475 2.479 wfi_extrapolate 11 7.9 0.001 0.001 2.464 2.465 mp_sendrecv_dv 5355 12.7 2.364 2.420 2.364 2.420 mp_alltoall_d11v 2130 13.8 2.212 2.351 2.212 2.351 grid_integrate_task_list 119 12.3 2.019 2.095 2.019 2.095 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.003 2.011 ot_diis_step 108 11.5 0.013 0.013 1.886 1.889 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.731 1.840 apply_single 119 13.6 0.000 0.000 1.731 1.840 mp_allgather_i34 2286 14.5 0.854 1.827 0.854 1.827 cp_fm_cholesky_decompose 22 10.9 1.690 1.696 1.690 1.696 yz_to_x 606 15.1 0.053 0.056 1.558 1.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.585 1.626 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.602 1.618 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.567 1.596 dbcsr_complete_redistribute 329 12.2 0.156 0.166 1.396 1.509 grid_collocate_task_list 119 9.7 1.427 1.501 1.427 1.501 acc_transpose_blocks 9144 15.5 0.041 0.043 1.431 1.470 x_to_yz 595 16.2 0.071 0.075 1.423 1.432 transfer_rs2pw_50 119 11.7 0.123 0.128 1.206 1.367 transfer_pw2rs_50 119 14.3 0.109 0.120 1.185 1.339 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.336 1.337 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.242 1.255 arnoldi_extremal 119 11.4 0.002 0.002 1.174 1.239 arnoldi_normal_ev 119 12.4 0.002 0.002 1.173 1.237 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=55.404000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.272727, yerr=7.568967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 768.352256E+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 63723. MP_Allreduce 10154 429. 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 0.012 0.034 47.524 47.525 qs_mol_dyn_low 1 2.0 0.003 0.003 47.276 47.284 qs_forces 11 3.9 0.003 0.006 47.195 47.199 qs_energies 11 4.9 0.002 0.003 45.091 45.095 scf_env_do_scf 11 5.9 0.001 0.001 38.718 38.718 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.192 30.195 velocity_verlet 10 3.0 0.002 0.002 26.737 26.743 dbcsr_multiply_generic 2286 12.5 0.104 0.107 21.240 21.371 qs_scf_new_mos 108 7.5 0.001 0.001 19.780 19.874 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.779 19.873 ot_scf_mini 108 9.5 0.002 0.002 18.507 18.616 multiply_cannon 2286 13.5 0.303 0.309 15.987 16.728 multiply_cannon_loop 2286 14.5 0.864 0.879 14.508 15.185 ot_mini 108 10.5 0.001 0.001 11.047 11.156 multiply_cannon_multrec 9144 15.5 3.385 4.663 8.825 8.922 qs_ot_get_derivative 108 11.5 0.001 0.001 8.818 8.915 init_scf_loop 11 6.9 0.000 0.000 8.491 8.498 rebuild_ks_matrix 119 8.3 0.000 0.000 7.804 7.916 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.015 7.804 7.915 prepare_preconditioner 11 7.9 0.000 0.000 7.422 7.433 make_preconditioner 11 8.9 0.000 0.000 7.422 7.433 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.918 7.288 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.003 7.105 dbcsr_mm_accdrv_process 12550 15.8 4.443 5.854 5.302 6.585 cp_fm_upper_to_full 72 14.2 3.195 4.639 3.195 4.639 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.578 4.590 calculate_rho_elec 119 8.7 0.118 0.121 4.578 4.589 mp_waitall_1 94719 16.7 3.511 4.440 3.511 4.440 qs_ot_get_p 119 10.4 0.003 0.003 4.205 4.343 sum_up_and_integrate 119 10.3 0.001 0.001 4.262 4.271 integrate_v_rspace 119 11.3 0.003 0.003 4.252 4.261 init_scf_run 11 5.9 0.000 0.002 4.166 4.166 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.165 4.166 make_m2s 4572 13.5 0.038 0.039 3.296 3.640 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.185 3.580 make_images 4572 14.5 0.356 0.387 3.175 3.520 pw_transfer 1439 11.6 0.069 0.069 3.096 3.102 dbcsr_complete_redistribute 329 12.2 0.283 0.287 2.179 3.028 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.997 3.004 calculate_dm_sparse 119 9.5 0.000 0.000 2.686 2.708 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.341 2.668 apply_single 119 13.6 0.000 0.000 2.340 2.667 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.808 2.657 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.455 2.501 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.465 2.476 fft_wrap_pw1pw2_140 487 13.2 0.576 0.582 2.468 2.476 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.706 2.465 calculate_first_density_matrix 1 7.0 0.001 0.003 2.453 2.456 density_rs2pw 119 9.7 0.004 0.004 2.429 2.443 mp_alltoall_i22 627 13.8 1.510 2.362 1.510 2.362 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.499 2.340 ot_diis_step 108 11.5 0.015 0.015 2.191 2.194 grid_integrate_task_list 119 12.3 2.167 2.183 2.167 2.183 acc_transpose_blocks 9144 15.5 0.045 0.045 2.044 2.178 mp_sum_l 7287 12.8 1.405 2.128 1.405 2.128 make_images_data 4572 15.5 0.046 0.049 1.631 2.119 fft3d_ps 1201 14.6 0.601 0.612 2.093 2.099 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.604 2.075 cp_dbcsr_syevd 50 12.0 0.004 0.005 2.038 2.040 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.947 1.948 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.854 1.858 cp_fm_cholesky_invert 11 10.9 1.813 1.818 1.813 1.818 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.681 1.682 cp_fm_diag_elpa_base 50 14.0 1.511 1.586 1.678 1.678 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.622 1.655 wfi_extrapolate 11 7.9 0.001 0.001 1.635 1.636 potential_pw2rs 119 12.3 0.014 0.014 1.620 1.625 grid_collocate_task_list 119 9.7 1.584 1.598 1.584 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.590 1.595 mp_alltoall_d11v 2130 13.8 1.513 1.553 1.513 1.553 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.507 1.515 mp_alltoall_z22v 1201 16.6 1.354 1.374 1.354 1.374 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.183 1.199 acc_transpose_blocks_sync 27432 16.5 1.105 1.122 1.105 1.122 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.988 1.037 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.024 1.033 jit_kernel_multiply 6 15.6 0.829 1.029 0.829 1.029 acc_transpose_blocks_kernels 9144 16.5 0.121 0.124 0.879 0.994 mp_allgather_i34 2286 14.5 0.429 0.990 0.429 0.990 qs_create_task_list 11 7.9 0.001 0.001 0.977 0.989 generate_qs_task_list 11 8.9 0.390 0.409 0.977 0.989 make_images_sizes 4572 15.5 0.005 0.005 0.560 0.973 mp_alltoall_i44 4572 16.5 0.556 0.968 0.556 0.968 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.525000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.000000, yerr=13.544539 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 501.256192E+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 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. 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 0.008 0.026 80.063 80.064 qs_mol_dyn_low 1 2.0 0.003 0.003 79.800 79.814 qs_forces 11 3.9 0.003 0.005 79.701 79.702 qs_energies 11 4.9 0.001 0.002 76.855 76.875 scf_env_do_scf 11 5.9 0.000 0.001 68.078 68.080 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 62.568 62.569 dbcsr_multiply_generic 2055 12.4 0.106 0.108 49.808 50.181 qs_scf_new_mos 99 7.5 0.000 0.001 45.751 45.881 qs_scf_loop_do_ot 99 8.5 0.000 0.001 45.751 45.881 ot_scf_mini 99 9.5 0.002 0.002 43.381 43.493 multiply_cannon 2055 13.4 0.187 0.192 42.012 42.767 velocity_verlet 10 3.0 0.001 0.001 41.802 41.803 multiply_cannon_loop 2055 14.4 1.777 1.830 41.041 41.777 ot_mini 99 10.5 0.001 0.001 25.441 25.554 qs_ot_get_derivative 99 11.5 0.001 0.001 18.646 18.790 multiply_cannon_multrec 49320 15.4 11.546 12.302 17.516 18.213 rebuild_ks_matrix 110 8.3 0.000 0.000 14.374 14.546 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.373 14.546 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.601 12.737 mp_waitall_1 220248 16.4 9.880 10.674 9.880 10.674 multiply_cannon_sync_h2d 49320 15.4 9.766 10.443 9.766 10.443 qs_ot_get_p 110 10.4 0.001 0.002 9.553 9.710 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.234 7.787 apply_single 110 13.6 0.000 0.001 7.234 7.787 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.909 7.428 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.024 7.048 sum_up_and_integrate 110 10.3 0.002 0.003 6.888 6.900 integrate_v_rspace 110 11.3 0.003 0.003 6.863 6.881 init_scf_run 11 5.9 0.000 0.001 6.740 6.740 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.740 6.740 ot_diis_step 99 11.5 0.006 0.006 6.617 6.617 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.490 6.511 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.179 6.280 calculate_rho_elec 110 8.6 0.020 0.025 6.179 6.279 dbcsr_mm_accdrv_process 87628 16.1 3.006 3.134 5.838 6.107 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.633 5.633 init_scf_loop 11 6.9 0.000 0.000 5.485 5.485 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.284 5.365 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.071 5.093 cp_fm_diag_elpa_base 48 14.0 5.057 5.079 5.069 5.091 mp_sum_l 6594 12.7 3.517 4.299 3.517 4.299 wfi_extrapolate 11 7.9 0.001 0.001 4.017 4.017 make_m2s 4110 13.4 0.062 0.067 3.822 3.938 calculate_dm_sparse 110 9.5 0.001 0.001 3.747 3.845 make_images 4110 14.4 0.177 0.189 3.725 3.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.557 3.561 density_rs2pw 110 9.6 0.004 0.005 3.268 3.504 grid_integrate_task_list 110 12.3 3.246 3.408 3.246 3.408 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 2.095 3.384 prepare_preconditioner 11 7.9 0.000 0.000 3.346 3.365 make_preconditioner 11 8.9 0.000 0.000 3.346 3.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.216 3.257 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.108 3.163 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.088 3.147 pw_transfer 1331 11.6 0.055 0.068 2.984 3.062 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.895 2.975 acc_transpose_blocks 49320 15.4 0.210 0.221 2.540 2.631 calculate_first_density_matrix 1 7.0 0.000 0.001 2.624 2.629 jit_kernel_multiply 13 15.9 2.554 2.569 2.554 2.569 fft_wrap_pw1pw2_140 451 13.1 0.392 0.432 2.467 2.555 potential_pw2rs 110 12.3 0.005 0.006 2.520 2.534 mp_alltoall_d11v 2046 13.8 1.972 2.423 1.972 2.423 grid_collocate_task_list 110 9.6 2.162 2.291 2.162 2.291 fft3d_ps 1111 14.6 0.795 0.882 2.188 2.245 transfer_rs2pw 451 10.6 0.005 0.006 1.858 2.084 mp_waitany 14300 13.8 1.803 2.081 1.803 2.081 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.903 1.937 mp_sum_d 3889 11.9 1.274 1.876 1.274 1.876 make_images_data 4110 15.4 0.043 0.047 1.726 1.859 cp_fm_cholesky_invert 11 10.9 1.838 1.842 1.838 1.842 hybrid_alltoall_any 4261 16.3 0.085 0.482 1.504 1.776 transfer_pw2rs 451 13.1 0.006 0.007 1.716 1.723 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.651 1.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.064000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.984810 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 591.917056E+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 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3217469. 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 0.033 120.436 120.437 qs_mol_dyn_low 1 2.0 0.003 0.003 120.124 120.136 qs_forces 11 3.9 0.003 0.003 119.599 119.603 qs_energies 11 4.9 0.002 0.003 115.452 115.502 scf_env_do_scf 11 5.9 0.000 0.001 104.078 104.081 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 85.906 85.909 velocity_verlet 10 3.0 0.001 0.002 63.570 63.582 qs_scf_new_mos 99 7.5 0.001 0.001 60.793 61.221 qs_scf_loop_do_ot 99 8.5 0.001 0.001 60.792 61.221 ot_scf_mini 99 9.5 0.003 0.003 57.966 58.418 dbcsr_multiply_generic 2055 12.4 0.122 0.126 54.868 58.300 multiply_cannon 2055 13.4 0.227 0.249 37.854 46.410 multiply_cannon_loop 2055 14.4 1.181 1.217 34.783 43.818 mp_waitall_1 176588 16.5 18.199 32.192 18.199 32.192 ot_mini 99 10.5 0.001 0.001 32.039 32.185 qs_ot_get_derivative 99 11.5 0.001 0.001 24.384 24.839 multiply_cannon_metrocomm3 24660 15.4 0.072 0.078 9.778 23.783 rebuild_ks_matrix 110 8.3 0.000 0.000 18.867 19.273 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 18.867 19.272 init_scf_loop 11 6.9 0.000 0.000 18.052 18.055 qs_ot_get_p 110 10.4 0.002 0.005 16.819 17.167 qs_ks_update_qs_env 110 7.6 0.001 0.001 16.674 17.038 prepare_preconditioner 11 7.9 0.000 0.000 15.373 15.395 make_preconditioner 11 8.9 0.000 0.000 15.372 15.395 multiply_cannon_multrec 24660 15.4 6.265 8.128 13.379 15.349 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.699 15.029 qs_ot_p2m_diag 48 11.0 0.029 0.044 12.130 12.281 qs_rho_update_rho_low 110 7.6 0.001 0.001 11.383 11.581 calculate_rho_elec 110 8.6 0.039 0.047 11.383 11.581 cp_dbcsr_syevd 48 12.0 0.003 0.004 11.035 11.039 sum_up_and_integrate 110 10.3 0.002 0.003 10.878 10.893 integrate_v_rspace 110 11.3 0.002 0.003 10.851 10.864 mp_sum_l 6594 12.7 7.718 10.077 7.718 10.077 make_m2s 4110 13.4 0.058 0.060 8.679 9.680 make_images 4110 14.4 0.407 0.464 8.566 9.564 cp_fm_diag_elpa 48 13.0 0.000 0.000 9.418 9.431 cp_fm_diag_elpa_base 48 14.0 9.316 9.369 9.415 9.426 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.192 9.018 cp_fm_cholesky_invert 11 10.9 8.395 8.439 8.395 8.439 init_scf_run 11 5.9 0.000 0.001 8.434 8.435 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.434 8.435 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.065 8.211 apply_single 110 13.6 0.000 0.001 7.065 8.211 density_rs2pw 110 9.6 0.004 0.005 7.787 7.997 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.737 7.949 dbcsr_mm_accdrv_process 52282 16.1 5.526 6.202 6.961 7.364 pw_transfer 1331 11.6 0.067 0.074 6.799 7.180 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 6.691 7.073 ot_diis_step 99 11.5 0.010 0.011 7.039 7.043 multiply_cannon_sync_h2d 24660 15.4 5.644 6.966 5.644 6.966 potential_pw2rs 110 12.3 0.008 0.009 6.209 6.446 fft3d_ps 1111 14.6 1.114 1.363 5.806 6.218 make_images_data 4110 15.4 0.049 0.053 4.631 5.929 wfi_extrapolate 11 7.9 0.001 0.001 5.631 5.632 mp_alltoall_d11v 2046 13.8 4.288 5.102 4.288 5.102 hybrid_alltoall_any 4261 16.3 0.106 0.447 3.916 5.059 cp_fm_cholesky_decompose 22 10.9 4.927 4.939 4.927 4.939 multiply_cannon_metrocomm1 24660 15.4 0.036 0.038 1.463 4.616 transfer_rs2pw 451 10.6 0.006 0.007 4.386 4.584 fft_wrap_pw1pw2_140 451 13.1 0.442 0.457 4.196 4.574 mp_alltoall_z22v 1111 16.6 4.175 4.465 4.175 4.465 calculate_dm_sparse 110 9.5 0.001 0.001 4.260 4.340 qs_ot_get_orbitals 99 10.5 0.001 0.001 4.006 4.295 transfer_pw2rs 451 13.1 0.005 0.006 4.030 4.259 mp_allgather_i34 2055 14.4 2.377 4.177 2.377 4.177 mp_sum_d 3889 11.9 2.919 3.952 2.919 3.952 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.889 3.898 make_images_sizes 4110 15.4 0.005 0.005 2.588 3.798 mp_alltoall_i44 4110 16.4 2.584 3.793 2.584 3.793 grid_integrate_task_list 110 12.3 3.044 3.358 3.044 3.358 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.122 3.269 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.900 2.880 dbcsr_complete_redistribute 325 12.2 0.230 0.282 2.432 2.873 yz_to_x 451 15.2 0.088 0.102 2.501 2.787 mp_sendrecv_dv 10340 12.6 2.551 2.759 2.551 2.759 dbcsr_dot_sd 1091 11.9 0.216 0.260 2.012 2.682 mp_irecv_dv 57340 16.2 0.777 2.662 0.777 2.662 calculate_first_density_matrix 1 7.0 0.000 0.001 2.627 2.636 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.383 2.418 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=120.437000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.545455, yerr=7.278407 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 666.378240E+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 66421. MP_Allreduce 9774 562. 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 0.025 0.055 103.682 103.683 qs_mol_dyn_low 1 2.0 0.003 0.004 103.243 103.255 qs_forces 11 3.9 0.004 0.007 103.129 103.131 qs_energies 11 4.9 0.003 0.017 99.141 99.176 scf_env_do_scf 11 5.9 0.001 0.001 88.361 88.362 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 69.770 69.773 velocity_verlet 10 3.0 0.009 0.016 56.460 56.466 qs_scf_new_mos 99 7.5 0.001 0.001 47.852 48.084 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.851 48.083 dbcsr_multiply_generic 2055 12.4 0.115 0.119 43.353 46.011 ot_scf_mini 99 9.5 0.002 0.003 45.722 45.959 multiply_cannon 2055 13.4 0.220 0.230 28.605 35.151 multiply_cannon_loop 2055 14.4 0.826 0.848 25.800 33.077 mp_waitall_1 139946 16.5 15.805 26.350 15.805 26.350 ot_mini 99 10.5 0.001 0.001 24.516 24.820 qs_ot_get_derivative 99 11.5 0.001 0.001 18.606 18.834 init_scf_loop 11 6.9 0.001 0.004 18.478 18.482 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 7.498 18.050 rebuild_ks_matrix 110 8.3 0.000 0.000 16.764 17.698 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 16.763 17.698 prepare_preconditioner 11 7.9 0.000 0.000 16.041 16.084 make_preconditioner 11 8.9 0.000 0.002 16.041 16.084 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.793 15.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.110 15.626 qs_ot_get_p 110 10.4 0.002 0.003 14.335 14.673 multiply_cannon_multrec 16440 15.4 3.322 4.146 9.638 10.463 qs_ot_p2m_diag 48 11.0 0.042 0.044 9.951 10.024 sum_up_and_integrate 110 10.3 0.001 0.002 9.902 9.916 integrate_v_rspace 110 11.3 0.003 0.003 9.876 9.890 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.774 9.876 calculate_rho_elec 110 8.6 0.059 0.059 9.774 9.876 cp_dbcsr_syevd 48 12.0 0.003 0.003 9.133 9.136 make_m2s 4110 13.4 0.050 0.053 7.946 9.097 make_images 4110 14.4 0.400 0.521 7.827 8.977 cp_fm_cholesky_invert 11 10.9 8.502 8.535 8.502 8.535 mp_sum_l 6594 12.7 6.187 8.402 6.187 8.402 init_scf_run 11 5.9 0.000 0.002 7.832 7.832 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.831 7.832 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.767 7.778 cp_fm_diag_elpa_base 48 14.0 7.631 7.716 7.762 7.772 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.955 6.570 density_rs2pw 110 9.6 0.004 0.005 6.339 6.506 dbcsr_mm_accdrv_process 34862 16.1 5.438 5.916 6.175 6.332 make_images_data 4110 15.4 0.046 0.050 4.582 6.252 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.473 6.192 apply_single 110 13.6 0.000 0.000 5.473 6.191 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.050 6.179 pw_transfer 1331 11.6 0.067 0.076 5.893 5.964 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.784 5.858 ot_diis_step 99 11.5 0.011 0.012 5.513 5.515 potential_pw2rs 110 12.3 0.011 0.012 5.228 5.305 hybrid_alltoall_any 4261 16.3 0.109 0.379 3.915 5.192 wfi_extrapolate 11 7.9 0.001 0.001 5.036 5.037 fft3d_ps 1111 14.6 1.103 1.128 4.734 4.809 cp_fm_cholesky_decompose 22 10.9 4.488 4.506 4.488 4.506 mp_alltoall_d11v 2046 13.8 3.923 4.462 3.923 4.462 fft_wrap_pw1pw2_140 451 13.1 0.571 0.579 4.127 4.203 mp_allgather_i34 2055 14.4 2.104 3.747 2.104 3.747 transfer_rs2pw 451 10.6 0.005 0.006 3.510 3.735 multiply_cannon_sync_h2d 16440 15.4 3.093 3.625 3.093 3.625 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 1.860 3.619 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.585 3.588 transfer_pw2rs 451 13.1 0.005 0.005 3.487 3.576 calculate_dm_sparse 110 9.5 0.001 0.001 3.510 3.558 mp_alltoall_z22v 1111 16.6 3.412 3.494 3.412 3.494 grid_integrate_task_list 110 12.3 3.153 3.357 3.153 3.357 make_images_sizes 4110 15.4 0.005 0.005 1.932 3.348 mp_alltoall_i44 4110 16.4 1.928 3.343 1.928 3.343 dbcsr_complete_redistribute 325 12.2 0.318 0.361 2.645 3.280 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.737 3.030 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.781 2.969 multiply_cannon_metrocomm4 14385 15.4 0.049 0.056 0.946 2.770 mp_sum_d 3887 11.9 2.017 2.690 2.017 2.690 mp_irecv_dv 48980 15.7 0.874 2.637 0.874 2.637 calculate_first_density_matrix 1 7.0 0.001 0.002 2.621 2.625 grid_collocate_task_list 110 9.6 2.222 2.415 2.222 2.415 mp_waitany 17072 13.8 2.103 2.321 2.103 2.321 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.675 2.317 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.179 2.271 mp_alltoall_i22 605 13.7 1.364 2.194 1.364 2.194 cp_fm_upper_to_full 70 14.2 1.593 2.191 1.593 2.191 dbcsr_dot_sd 1091 11.9 0.295 0.302 1.635 2.112 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.108 2.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=103.683000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.000000, yerr=8.323024 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 739.135488E+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 66419. MP_Allreduce 9774 603. 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 0.013 0.032 109.262 109.264 qs_mol_dyn_low 1 2.0 0.003 0.003 108.966 108.979 qs_forces 11 3.9 0.003 0.005 108.703 108.706 qs_energies 11 4.9 0.001 0.002 104.595 104.607 scf_env_do_scf 11 5.9 0.000 0.001 93.719 93.722 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 70.303 70.305 velocity_verlet 10 3.0 0.002 0.004 61.172 61.183 qs_scf_new_mos 99 7.5 0.001 0.001 48.224 48.365 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.223 48.365 ot_scf_mini 99 9.5 0.002 0.003 45.845 45.963 dbcsr_multiply_generic 2055 12.4 0.119 0.125 42.131 43.599 multiply_cannon 2055 13.4 0.239 0.264 27.789 31.206 multiply_cannon_loop 2055 14.4 1.420 1.493 24.908 29.101 ot_mini 99 10.5 0.001 0.001 23.914 24.068 init_scf_loop 11 6.9 0.000 0.000 23.306 23.309 prepare_preconditioner 11 7.9 0.000 0.000 21.023 21.059 make_preconditioner 11 8.9 0.000 0.000 21.023 21.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 19.000 20.587 mp_waitall_1 121746 16.5 11.721 19.007 11.721 19.007 qs_ot_get_derivative 99 11.5 0.001 0.001 18.638 18.757 rebuild_ks_matrix 110 8.3 0.000 0.000 16.397 16.641 qs_ks_build_kohn_sham_matrix 110 9.3 0.025 0.038 16.397 16.641 qs_ot_get_p 110 10.4 0.002 0.002 15.185 15.369 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.517 14.728 multiply_cannon_multrec 24660 15.4 3.947 6.940 12.907 14.265 qs_ot_p2m_diag 48 11.0 0.055 0.064 11.272 11.338 make_m2s 4110 13.4 0.061 0.064 9.900 10.977 make_images 4110 14.4 0.577 0.700 9.755 10.833 cp_dbcsr_syevd 48 12.0 0.003 0.004 10.253 10.258 qs_rho_update_rho_low 110 7.6 0.001 0.001 10.103 10.165 calculate_rho_elec 110 8.6 0.078 0.081 10.103 10.164 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 3.018 9.933 sum_up_and_integrate 110 10.3 0.001 0.002 9.806 9.819 dbcsr_mm_accdrv_process 52304 16.0 7.648 8.806 8.809 9.792 integrate_v_rspace 110 11.3 0.003 0.003 9.781 9.792 cp_fm_cholesky_invert 11 10.9 9.709 9.740 9.709 9.740 cp_fm_diag_elpa 48 13.0 0.000 0.000 8.608 8.617 cp_fm_diag_elpa_base 48 14.0 8.333 8.409 8.602 8.611 init_scf_run 11 5.9 0.000 0.001 7.517 7.517 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.516 7.517 make_images_data 4110 15.4 0.049 0.053 5.524 7.179 density_rs2pw 110 9.6 0.004 0.004 6.568 6.702 pw_transfer 1331 11.6 0.067 0.076 6.372 6.494 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 6.264 6.389 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.222 6.285 hybrid_alltoall_any 4261 16.3 0.123 0.461 4.559 6.018 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.773 5.861 cp_fm_upper_to_full 70 14.2 3.878 5.545 3.878 5.545 fft3d_ps 1111 14.6 1.100 1.134 5.155 5.257 potential_pw2rs 110 12.3 0.012 0.013 5.122 5.196 dbcsr_complete_redistribute 325 12.2 0.404 0.450 3.883 5.142 cp_fm_cholesky_decompose 22 10.9 5.078 5.122 5.078 5.122 mp_sum_l 6594 12.7 3.762 5.084 3.762 5.084 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.560 5.072 apply_single 110 13.6 0.000 0.000 4.559 5.072 ot_diis_step 99 11.5 0.011 0.012 4.977 4.978 wfi_extrapolate 11 7.9 0.001 0.001 4.834 4.834 fft_wrap_pw1pw2_140 451 13.1 0.602 0.623 4.587 4.714 mp_alltoall_d11v 2046 13.8 4.008 4.291 4.008 4.291 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.805 4.052 mp_alltoall_z22v 1111 16.6 3.832 3.961 3.832 3.961 calculate_dm_sparse 110 9.5 0.001 0.001 3.795 3.852 transfer_rs2pw 451 10.6 0.005 0.006 3.484 3.590 make_images_sizes 4110 15.4 0.005 0.005 2.062 3.529 mp_alltoall_i44 4110 16.4 2.058 3.524 2.058 3.524 mp_alltoall_i22 605 13.7 2.164 3.459 2.164 3.459 grid_integrate_task_list 110 12.3 3.263 3.428 3.263 3.428 transfer_pw2rs 451 13.1 0.005 0.005 3.200 3.278 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.013 3.272 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.264 3.268 mp_allgather_i34 2055 14.4 1.964 3.162 1.964 3.162 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.019 3.136 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 1.542 3.103 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.613 2.699 acc_transpose_blocks 24660 15.4 0.114 0.117 2.415 2.630 calculate_first_density_matrix 1 7.0 0.000 0.001 2.529 2.532 multiply_cannon_sync_h2d 24660 15.4 2.299 2.441 2.299 2.441 grid_collocate_task_list 110 9.6 2.269 2.412 2.269 2.412 mp_waitany 13376 13.8 2.116 2.403 2.116 2.403 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.386 2.390 yz_to_x 561 15.1 0.092 0.099 2.112 2.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=109.264000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.727273, yerr=10.410579 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 853.848064E+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 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 93.753 93.755 qs_mol_dyn_low 1 2.0 0.003 0.003 93.432 93.443 qs_forces 11 3.9 0.002 0.003 93.021 93.022 qs_energies 11 4.9 0.002 0.004 88.705 88.717 scf_env_do_scf 11 5.9 0.000 0.001 77.661 77.661 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 59.999 60.002 velocity_verlet 10 3.0 0.039 0.040 52.445 52.468 qs_scf_new_mos 99 7.5 0.001 0.001 38.651 38.802 qs_scf_loop_do_ot 99 8.5 0.001 0.001 38.650 38.802 ot_scf_mini 99 9.5 0.002 0.002 36.759 36.859 dbcsr_multiply_generic 2055 12.4 0.111 0.117 34.544 35.300 multiply_cannon 2055 13.4 0.248 0.261 23.497 25.599 multiply_cannon_loop 2055 14.4 0.608 0.634 20.626 22.439 ot_mini 99 10.5 0.001 0.001 18.471 18.579 mp_waitall_1 103326 16.6 13.255 18.218 13.255 18.218 init_scf_loop 11 6.9 0.000 0.000 17.560 17.563 rebuild_ks_matrix 110 8.3 0.000 0.000 15.981 16.296 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 15.981 16.296 prepare_preconditioner 11 7.9 0.000 0.000 15.327 15.343 make_preconditioner 11 8.9 0.000 0.000 15.327 15.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.549 14.861 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.211 14.496 qs_ot_get_derivative 99 11.5 0.001 0.001 13.802 13.899 qs_ot_get_p 110 10.4 0.002 0.004 12.759 12.905 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.633 9.667 calculate_rho_elec 110 8.6 0.114 0.114 9.632 9.667 qs_ot_p2m_diag 48 11.0 0.082 0.085 9.570 9.587 sum_up_and_integrate 110 10.3 0.001 0.002 9.530 9.541 integrate_v_rspace 110 11.3 0.003 0.003 9.503 9.514 make_m2s 4110 13.4 0.039 0.040 7.979 9.282 make_images 4110 14.4 0.653 0.710 7.846 9.150 cp_fm_cholesky_invert 11 10.9 8.897 8.913 8.897 8.913 cp_dbcsr_syevd 48 12.0 0.003 0.004 8.792 8.795 multiply_cannon_multrec 8220 15.4 3.199 4.420 7.818 8.777 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.513 7.534 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.465 7.473 cp_fm_diag_elpa_base 48 14.0 7.356 7.432 7.460 7.469 init_scf_run 11 5.9 0.000 0.001 7.315 7.315 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.315 7.315 pw_transfer 1331 11.6 0.066 0.071 6.550 6.610 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 6.442 6.506 make_images_data 4110 15.4 0.041 0.047 4.530 6.456 density_rs2pw 110 9.6 0.004 0.004 5.889 6.051 hybrid_alltoall_any 4261 16.3 0.200 0.839 4.026 5.805 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 4.165 5.609 dbcsr_mm_accdrv_process 17442 15.9 3.195 4.287 4.474 5.461 fft3d_ps 1111 14.6 1.152 1.174 5.110 5.168 fft_wrap_pw1pw2_140 451 13.1 0.773 0.786 5.075 5.142 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.360 4.678 apply_single 110 13.6 0.000 0.000 4.360 4.678 wfi_extrapolate 11 7.9 0.001 0.001 4.627 4.627 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.542 4.598 ot_diis_step 99 11.5 0.012 0.012 4.534 4.535 potential_pw2rs 110 12.3 0.015 0.015 4.420 4.453 cp_fm_cholesky_decompose 22 10.9 4.428 4.442 4.428 4.442 mp_alltoall_d11v 2046 13.8 3.944 4.162 3.944 4.162 mp_allgather_i34 2055 14.4 2.024 4.114 2.024 4.114 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.971 4.016 mp_alltoall_z22v 1111 16.6 3.713 3.761 3.713 3.761 grid_integrate_task_list 110 12.3 3.558 3.669 3.558 3.669 make_images_sizes 4110 15.4 0.005 0.005 1.838 3.396 mp_alltoall_i44 4110 16.4 1.834 3.391 1.834 3.391 mp_sum_l 6594 12.7 2.324 3.382 2.324 3.382 calculate_dm_sparse 110 9.5 0.001 0.001 3.299 3.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.136 3.138 transfer_rs2pw 451 10.6 0.005 0.005 2.780 2.928 dbcsr_complete_redistribute 325 12.2 0.542 0.563 2.721 2.897 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.823 2.825 grid_collocate_task_list 110 9.6 2.497 2.588 2.497 2.588 transfer_pw2rs 451 13.1 0.004 0.005 2.539 2.567 calculate_first_density_matrix 1 7.0 0.001 0.001 2.529 2.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.419 2.532 multiply_cannon_sync_h2d 8220 15.4 2.344 2.444 2.344 2.444 mp_waitany 9240 13.8 2.029 2.218 2.029 2.218 qs_ot_get_orbitals 99 10.5 0.000 0.001 2.105 2.216 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.115 2.147 yz_to_x 561 15.1 0.102 0.104 2.057 2.103 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.005 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.943 1.994 x_to_yz 550 16.1 0.130 0.132 1.888 1.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=93.755000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.545455, yerr=11.211712 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 1.390703E+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 67098. MP_Allreduce 9752 812. 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 0.043 0.070 134.594 134.595 qs_mol_dyn_low 1 2.0 0.003 0.003 134.083 134.095 qs_forces 11 3.9 0.003 0.004 133.973 133.975 qs_energies 11 4.9 0.001 0.002 128.980 128.999 scf_env_do_scf 11 5.9 0.001 0.001 116.175 116.176 velocity_verlet 10 3.0 0.002 0.002 83.200 83.209 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 72.652 72.659 qs_scf_new_mos 99 7.5 0.001 0.001 46.946 47.136 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.945 47.135 ot_scf_mini 99 9.5 0.002 0.002 44.559 44.745 init_scf_loop 11 6.9 0.000 0.000 43.380 43.386 dbcsr_multiply_generic 2055 12.4 0.124 0.129 41.157 42.670 prepare_preconditioner 11 7.9 0.000 0.000 40.614 40.648 make_preconditioner 11 8.9 0.000 0.000 40.613 40.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 34.475 39.887 multiply_cannon 2055 13.4 0.345 0.374 26.856 31.362 multiply_cannon_loop 2055 14.4 0.835 0.852 24.575 28.744 ot_mini 99 10.5 0.001 0.001 22.511 22.832 mp_waitall_1 84994 16.7 15.210 21.365 15.210 21.365 rebuild_ks_matrix 110 8.3 0.000 0.001 18.937 19.459 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 18.936 19.458 cp_fm_upper_to_full 70 14.2 12.867 18.533 12.867 18.533 qs_ks_update_qs_env 110 7.6 0.001 0.001 17.054 17.527 qs_ot_get_derivative 99 11.5 0.001 0.001 16.512 16.688 cp_fm_cholesky_invert 11 10.9 15.660 15.679 15.660 15.679 qs_ot_get_p 110 10.4 0.002 0.002 15.011 15.323 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 8.270 12.855 dbcsr_complete_redistribute 325 12.2 1.005 1.036 8.746 12.141 qs_rho_update_rho_low 110 7.6 0.001 0.001 11.639 11.687 calculate_rho_elec 110 8.6 0.224 0.224 11.638 11.687 qs_ot_p2m_diag 48 11.0 0.151 0.156 11.331 11.356 sum_up_and_integrate 110 10.3 0.002 0.002 10.538 10.554 copy_fm_to_dbcsr 174 11.2 0.001 0.001 7.125 10.540 integrate_v_rspace 110 11.3 0.003 0.003 10.510 10.526 cp_dbcsr_syevd 48 12.0 0.003 0.003 10.489 10.494 make_m2s 4110 13.4 0.044 0.044 9.680 10.342 make_images 4110 14.4 0.894 0.949 9.486 10.147 multiply_cannon_multrec 8220 15.4 4.121 4.280 9.865 9.976 mp_alltoall_i22 605 13.7 5.921 9.573 5.921 9.573 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.123 9.485 cp_fm_diag_elpa 48 13.0 0.000 0.000 9.067 9.068 cp_fm_diag_elpa_base 48 14.0 8.417 8.714 9.059 9.059 pw_transfer 1331 11.6 0.075 0.075 8.853 8.902 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 8.735 8.786 init_scf_run 11 5.9 0.000 0.001 7.982 7.983 scf_env_initial_rho_setup 11 6.9 0.002 0.003 7.982 7.982 make_images_data 4110 15.4 0.045 0.048 5.323 7.464 density_rs2pw 110 9.6 0.004 0.004 7.240 7.270 fft_wrap_pw1pw2_140 451 13.1 1.277 1.284 7.079 7.122 fft3d_ps 1111 14.6 1.317 1.323 6.791 6.839 hybrid_alltoall_any 4261 16.3 0.264 0.577 4.781 6.838 mp_sum_l 6594 12.7 3.725 6.039 3.725 6.039 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.572 6.025 apply_single 110 13.6 0.000 0.000 5.572 6.025 dbcsr_mm_accdrv_process 11614 15.7 3.933 4.911 5.589 5.824 ot_diis_step 99 11.5 0.016 0.016 5.778 5.783 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.052 5.418 wfi_extrapolate 11 7.9 0.001 0.001 5.339 5.340 cp_fm_cholesky_decompose 22 10.9 5.297 5.315 5.297 5.315 mp_alltoall_z22v 1111 16.6 5.215 5.259 5.215 5.259 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.119 5.222 potential_pw2rs 110 12.3 0.021 0.021 5.025 5.057 mp_alltoall_d11v 2046 13.8 4.827 4.938 4.827 4.938 grid_integrate_task_list 110 12.3 3.872 3.990 3.872 3.990 calculate_dm_sparse 110 9.5 0.001 0.001 3.855 3.912 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.835 3.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.698 3.701 make_images_sizes 4110 15.4 0.005 0.005 2.186 3.674 mp_alltoall_i44 4110 16.4 2.182 3.669 2.182 3.669 multiply_cannon_sync_h2d 8220 15.4 3.130 3.164 3.130 3.164 yz_to_x 561 15.1 0.111 0.111 2.886 2.928 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.754 2.890 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.718 2.878 grid_collocate_task_list 110 9.6 2.818 2.854 2.818 2.854 transfer_rs2pw 451 10.6 0.006 0.006 2.770 2.819 qs_ot_get_orbitals 99 10.5 0.000 0.000 2.593 2.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=134.595000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1248.818182, yerr=55.970476 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 630.484992E+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 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2008934. 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 0.034 210.241 210.243 qs_mol_dyn_low 1 2.0 0.003 0.004 209.677 209.690 qs_forces 11 3.9 0.005 0.006 209.577 209.578 qs_energies 11 4.9 0.003 0.006 203.948 203.966 scf_env_do_scf 11 5.9 0.001 0.001 187.246 187.250 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 166.330 166.332 dbcsr_multiply_generic 2507 12.6 0.180 0.183 126.678 127.203 qs_scf_new_mos 117 7.6 0.001 0.001 126.450 126.782 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.449 126.781 velocity_verlet 10 3.0 0.001 0.002 125.533 125.534 ot_scf_mini 117 9.6 0.003 0.003 119.706 120.057 multiply_cannon 2507 13.6 0.240 0.247 101.881 103.007 multiply_cannon_loop 2507 14.6 2.417 2.469 99.676 100.879 ot_mini 117 10.6 0.001 0.001 67.209 67.549 multiply_cannon_multrec 60168 15.6 31.698 33.692 41.557 43.420 qs_ot_get_derivative 117 11.6 0.001 0.001 42.411 42.752 rebuild_ks_matrix 128 8.3 0.001 0.001 33.841 34.055 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.022 33.840 34.054 mp_waitall_1 267128 16.5 29.364 32.739 29.364 32.739 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.396 30.595 qs_ot_get_p 128 10.4 0.010 0.042 30.116 30.554 multiply_cannon_sync_h2d 60168 15.6 26.240 28.027 26.240 28.027 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.312 25.033 apply_single 128 13.6 0.001 0.001 24.312 25.032 ot_diis_step 117 11.6 0.008 0.008 24.516 24.518 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.461 23.550 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.751 21.053 init_scf_loop 11 6.9 0.000 0.001 20.836 20.838 cp_dbcsr_syevd 83 12.4 0.005 0.007 20.697 20.708 multiply_cannon_metrocomm3 60168 15.6 0.118 0.122 16.257 18.045 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.599 17.633 cp_fm_diag_elpa_base 83 14.4 17.522 17.555 17.593 17.627 prepare_preconditioner 11 7.9 0.000 0.000 16.233 16.273 make_preconditioner 11 8.9 0.000 0.000 16.233 16.272 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.457 15.618 make_m2s 5014 13.6 0.106 0.113 14.034 14.531 make_images 5014 14.6 0.398 0.418 13.850 14.357 sum_up_and_integrate 128 10.3 0.002 0.005 14.159 14.182 integrate_v_rspace 128 11.3 0.004 0.004 14.100 14.127 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.391 13.539 calculate_rho_elec 128 8.7 0.045 0.064 13.390 13.538 init_scf_run 11 5.9 0.000 0.004 12.484 12.484 scf_env_initial_rho_setup 11 6.9 0.003 0.005 12.483 12.484 mp_sum_l 7950 12.9 9.510 10.686 9.510 10.686 dbcsr_mm_accdrv_process 124484 16.2 4.803 4.943 9.426 9.952 cp_fm_cholesky_invert 11 10.9 9.314 9.322 9.314 9.322 wfi_extrapolate 11 7.9 0.001 0.001 9.244 9.244 calculate_dm_sparse 128 9.5 0.001 0.001 8.556 8.697 multiply_cannon_metrocomm1 60168 15.6 0.091 0.095 6.435 8.448 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.216 8.332 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.955 8.205 pw_transfer 1547 11.6 0.075 0.090 7.789 7.973 make_images_data 5014 15.6 0.067 0.073 6.886 7.767 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.586 7.764 density_rs2pw 128 9.7 0.006 0.007 6.959 7.583 grid_integrate_task_list 128 12.3 7.071 7.546 7.071 7.546 hybrid_alltoall_any 5200 16.5 0.296 2.267 6.023 7.195 fft_wrap_pw1pw2_140 523 13.2 1.128 1.172 6.582 6.750 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.730 6.741 fft3d_ps 1291 14.7 2.198 2.845 5.590 5.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.805 5.923 mp_alltoall_d11v 2415 14.1 4.478 5.440 4.478 5.440 grid_collocate_task_list 128 9.7 4.846 5.232 4.846 5.232 cp_fm_cholesky_decompose 22 10.9 4.718 4.735 4.718 4.735 potential_pw2rs 128 12.3 0.009 0.010 4.624 4.651 mp_sum_d 4465 12.1 3.525 4.240 3.525 4.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.243000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.545455, yerr=6.457746 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 840.110080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5229771. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 250.753 250.756 qs_mol_dyn_low 1 2.0 0.003 0.003 250.340 250.354 qs_forces 11 3.9 0.004 0.005 249.673 249.677 qs_energies 11 4.9 0.002 0.004 242.144 242.158 scf_env_do_scf 11 5.9 0.001 0.001 222.961 222.971 scf_env_do_scf_inner_loop 117 6.6 0.013 0.031 171.664 171.671 velocity_verlet 10 3.0 0.001 0.002 158.519 158.533 qs_scf_new_mos 117 7.6 0.001 0.001 121.870 122.290 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.869 122.289 ot_scf_mini 117 9.6 0.004 0.004 116.805 117.299 dbcsr_multiply_generic 2507 12.6 0.194 0.199 105.444 106.745 multiply_cannon 2507 13.6 0.511 0.574 79.269 84.291 multiply_cannon_loop 2507 14.6 1.585 1.648 74.850 77.675 ot_mini 117 10.6 0.001 0.001 56.854 57.266 init_scf_loop 11 6.9 0.000 0.000 51.142 51.147 prepare_preconditioner 11 7.9 0.000 0.000 45.989 46.047 make_preconditioner 11 8.9 0.000 0.000 45.989 46.046 mp_waitall_1 214728 16.6 32.182 45.746 32.182 45.746 make_full_inverse_cholesky 11 9.9 0.000 0.000 44.234 45.080 qs_ot_get_p 128 10.4 0.007 0.029 41.071 41.583 rebuild_ks_matrix 128 8.3 0.001 0.001 39.007 39.613 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 39.007 39.613 multiply_cannon_multrec 30084 15.6 20.847 25.991 31.438 36.885 qs_ks_update_qs_env 128 7.6 0.001 0.001 35.102 35.666 qs_ot_get_derivative 117 11.6 0.001 0.002 34.488 34.963 qs_ot_p2m_diag 83 11.4 0.189 0.218 34.240 34.281 cp_dbcsr_syevd 83 12.4 0.006 0.006 32.491 32.498 multiply_cannon_metrocomm3 30084 15.6 0.096 0.104 15.949 28.929 cp_fm_diag_elpa 83 13.4 0.000 0.000 27.682 27.713 cp_fm_diag_elpa_base 83 14.4 27.201 27.456 27.672 27.699 cp_fm_cholesky_invert 11 10.9 25.763 25.781 25.763 25.781 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.444 23.580 apply_single 128 13.6 0.001 0.001 22.443 23.579 ot_diis_step 117 11.6 0.015 0.015 22.182 22.185 make_m2s 5014 13.6 0.089 0.094 19.546 21.464 make_images 5014 14.6 1.170 1.371 19.335 21.254 sum_up_and_integrate 128 10.3 0.002 0.004 19.867 19.902 integrate_v_rspace 128 11.3 0.003 0.004 19.807 19.845 qs_rho_update_rho_low 128 7.7 0.001 0.001 19.629 19.672 calculate_rho_elec 128 8.7 0.087 0.104 19.628 19.672 multiply_cannon_sync_h2d 30084 15.6 17.814 19.564 17.814 19.564 cp_fm_cholesky_decompose 22 10.9 14.668 14.751 14.668 14.751 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 14.258 14.608 make_images_data 5014 15.6 0.066 0.076 11.470 14.045 init_scf_run 11 5.9 0.000 0.001 13.827 13.828 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.827 13.828 density_rs2pw 128 9.7 0.006 0.007 13.006 13.439 hybrid_alltoall_any 5200 16.5 0.351 1.542 9.645 12.680 pw_transfer 1547 11.6 0.086 0.100 11.804 12.000 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 11.580 11.777 multiply_cannon_metrocomm4 27577 15.6 0.106 0.118 3.926 10.795 dbcsr_mm_accdrv_process 62242 16.2 5.575 6.383 10.047 10.606 wfi_extrapolate 11 7.9 0.001 0.001 10.484 10.485 mp_irecv_dv 69486 16.3 3.726 10.395 3.726 10.395 potential_pw2rs 128 12.3 0.016 0.018 9.892 10.131 fft3d_ps 1291 14.7 2.835 3.022 9.153 9.339 fft_wrap_pw1pw2_140 523 13.2 1.199 1.228 8.875 9.066 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.346 8.117 transfer_rs2pw 523 10.6 0.007 0.008 7.141 7.758 mp_alltoall_d11v 2415 14.1 6.354 7.703 6.354 7.703 grid_integrate_task_list 128 12.3 6.974 7.547 6.974 7.547 mp_sum_l 7950 12.9 4.839 7.016 4.839 7.016 calculate_dm_sparse 128 9.5 0.001 0.001 6.800 6.923 transfer_pw2rs 523 13.3 0.007 0.008 6.421 6.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.514 6.527 mp_allgather_i34 2507 14.6 2.937 6.370 2.937 6.370 make_images_sizes 5014 15.6 0.007 0.007 3.443 6.276 mp_alltoall_i44 5014 16.6 3.436 6.270 3.436 6.270 dbcsr_complete_redistribute 395 12.7 0.767 0.861 4.869 6.043 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.911 5.975 mp_alltoall_z22v 1291 16.7 5.344 5.608 5.344 5.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.337 5.472 grid_collocate_task_list 128 9.7 5.010 5.355 5.010 5.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=250.756000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.727273, yerr=3.106910 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 950.034432E+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 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. 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 0.012 0.032 176.085 176.089 qs_mol_dyn_low 1 2.0 0.003 0.003 175.665 175.682 qs_forces 11 3.9 0.004 0.004 175.314 175.317 qs_energies 11 4.9 0.001 0.002 168.748 168.758 scf_env_do_scf 11 5.9 0.001 0.001 153.265 153.266 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.067 118.068 velocity_verlet 10 3.0 0.001 0.002 113.654 113.663 dbcsr_multiply_generic 2527 12.6 0.182 0.190 82.114 83.179 qs_scf_new_mos 118 7.6 0.001 0.001 80.761 81.049 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.760 81.048 ot_scf_mini 118 9.6 0.003 0.004 76.474 76.828 multiply_cannon 2527 13.6 0.512 0.531 62.379 67.247 multiply_cannon_loop 2527 14.6 1.139 1.197 59.430 61.993 ot_mini 118 10.6 0.001 0.001 42.777 43.106 init_scf_loop 11 6.9 0.000 0.000 35.098 35.100 mp_waitall_1 171878 16.6 24.418 34.936 24.418 34.936 prepare_preconditioner 11 7.9 0.000 0.000 31.111 31.156 make_preconditioner 11 8.9 0.000 0.000 31.111 31.156 rebuild_ks_matrix 129 8.3 0.001 0.001 30.333 30.744 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 30.333 30.743 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.703 30.179 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.319 27.689 multiply_cannon_multrec 20216 15.6 13.353 16.284 22.751 25.747 multiply_cannon_metrocomm3 20216 15.6 0.063 0.066 15.010 24.777 qs_ot_get_derivative 118 11.6 0.002 0.002 23.106 23.460 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.700 20.771 apply_single 129 13.6 0.001 0.001 19.699 20.770 ot_diis_step 118 11.6 0.018 0.019 19.562 19.562 qs_ot_get_p 129 10.4 0.002 0.003 19.091 19.520 make_m2s 5054 13.6 0.081 0.084 14.723 15.921 make_images 5054 14.6 1.161 1.254 14.487 15.683 multiply_cannon_sync_h2d 20216 15.6 13.703 15.573 13.703 15.573 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.762 14.771 cp_fm_cholesky_invert 11 10.9 14.425 14.435 14.425 14.435 sum_up_and_integrate 129 10.3 0.002 0.002 14.037 14.065 integrate_v_rspace 129 11.3 0.003 0.004 13.976 14.006 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.694 13.738 calculate_rho_elec 129 8.7 0.131 0.146 13.694 13.738 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.702 13.705 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.503 10.529 cp_fm_diag_elpa_base 83 14.4 10.088 10.253 10.499 10.524 make_images_data 5054 15.6 0.063 0.073 8.744 10.448 init_scf_run 11 5.9 0.000 0.001 10.443 10.443 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.442 10.443 hybrid_alltoall_any 5240 16.5 0.452 2.053 7.479 9.816 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.032 9.288 multiply_cannon_metrocomm4 17689 15.6 0.068 0.077 3.444 9.247 mp_irecv_dv 50625 16.2 3.314 8.993 3.314 8.993 dbcsr_mm_accdrv_process 41830 16.2 5.641 5.971 8.846 8.979 pw_transfer 1559 11.6 0.086 0.104 8.765 8.877 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.539 8.657 grid_integrate_task_list 129 12.3 7.342 7.788 7.342 7.788 fft_wrap_pw1pw2_140 527 13.2 1.311 1.350 7.564 7.696 cp_fm_cholesky_decompose 22 10.9 7.393 7.416 7.393 7.416 wfi_extrapolate 11 7.9 0.001 0.001 7.415 7.415 cp_fm_upper_to_full 105 14.8 5.823 7.351 5.823 7.351 density_rs2pw 129 9.7 0.006 0.006 6.965 7.347 dbcsr_complete_redistribute 395 12.7 1.171 1.198 4.675 6.483 calculate_dm_sparse 129 9.5 0.001 0.001 5.892 6.000 fft3d_ps 1301 14.7 2.786 3.014 5.897 5.977 grid_collocate_task_list 129 9.7 5.199 5.603 5.199 5.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.436 5.441 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.793 5.426 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.490 5.292 mp_alltoall_d11v 2423 14.1 4.414 4.877 4.414 4.877 mp_allgather_i34 2527 14.6 1.442 4.710 1.442 4.710 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.517 4.647 mp_sum_l 8010 12.9 3.157 4.577 3.157 4.577 potential_pw2rs 129 12.3 0.020 0.022 4.383 4.396 transfer_fm_to_dbcsr 11 9.9 0.020 0.025 2.388 4.132 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.938 3.969 mp_alltoall_i22 718 14.1 2.015 3.943 2.015 3.943 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.748 3.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.517 3.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.089000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.272727, yerr=12.997139 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.154712E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58275. MP_Allreduce 11070 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.047 266.141 266.146 qs_mol_dyn_low 1 2.0 0.003 0.003 265.241 265.653 qs_forces 11 3.9 0.005 0.007 265.050 265.059 qs_energies 11 4.9 0.001 0.002 257.061 257.077 scf_env_do_scf 11 5.9 0.001 0.001 236.746 236.757 velocity_verlet 10 3.0 0.002 0.002 173.118 173.123 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 166.363 166.366 qs_scf_new_mos 116 7.6 0.001 0.001 119.652 120.014 qs_scf_loop_do_ot 116 8.6 0.001 0.001 119.651 120.013 ot_scf_mini 116 9.6 0.003 0.003 114.617 114.950 dbcsr_multiply_generic 2485 12.5 0.189 0.196 92.113 93.125 init_scf_loop 11 6.9 0.000 0.000 70.180 70.183 prepare_preconditioner 11 7.9 0.000 0.000 65.352 65.376 make_preconditioner 11 8.9 0.000 0.000 65.352 65.376 make_full_inverse_cholesky 11 9.9 0.011 0.023 58.191 63.716 multiply_cannon 2485 13.5 0.512 0.543 58.816 61.920 multiply_cannon_loop 2485 14.5 1.810 1.909 54.197 56.673 ot_mini 116 10.6 0.001 0.001 52.241 52.552 qs_ot_get_p 127 10.4 0.006 0.013 47.061 47.390 qs_ot_p2m_diag 82 11.4 0.339 0.386 40.378 40.459 cp_dbcsr_syevd 82 12.4 0.006 0.006 38.291 38.301 mp_waitall_1 146592 16.7 26.770 36.519 26.770 36.519 rebuild_ks_matrix 127 8.3 0.001 0.001 35.631 35.934 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 35.631 35.934 cp_fm_diag_elpa 82 13.4 0.000 0.000 33.083 33.095 cp_fm_diag_elpa_base 82 14.4 31.627 32.069 33.067 33.077 qs_ot_get_derivative 116 11.6 0.002 0.002 32.147 32.485 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.139 32.414 multiply_cannon_multrec 29820 15.5 13.279 18.127 26.122 30.659 cp_fm_cholesky_invert 11 10.9 28.707 28.725 28.707 28.725 make_m2s 4970 13.5 0.096 0.101 26.980 27.839 make_images 4970 14.5 1.931 2.251 26.671 27.533 ot_diis_step 116 11.6 0.018 0.018 19.912 19.916 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.293 19.897 apply_single 127 13.6 0.001 0.001 19.293 19.896 qs_rho_update_rho_low 127 7.7 0.001 0.001 19.417 19.509 calculate_rho_elec 127 8.7 0.171 0.188 19.417 19.509 sum_up_and_integrate 127 10.3 0.002 0.003 19.331 19.354 integrate_v_rspace 127 11.3 0.003 0.004 19.272 19.297 cp_fm_upper_to_full 104 14.8 13.053 19.067 13.053 19.067 make_images_data 4970 15.5 0.066 0.073 15.043 17.399 cp_fm_cholesky_decompose 22 10.9 16.271 16.359 16.271 16.359 dbcsr_complete_redistribute 393 12.7 1.490 1.626 11.406 15.608 hybrid_alltoall_any 5155 16.4 0.525 2.210 12.956 15.326 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.529 14.805 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 14.158 14.397 init_scf_run 11 5.9 0.000 0.001 13.626 13.627 scf_env_initial_rho_setup 11 6.9 0.002 0.004 13.626 13.627 copy_fm_to_dbcsr 208 11.6 0.001 0.002 9.239 13.448 pw_transfer 1535 11.6 0.086 0.102 12.746 12.983 dbcsr_mm_accdrv_process 61748 16.2 8.459 9.377 12.429 12.925 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 12.521 12.763 density_rs2pw 127 9.7 0.006 0.006 11.861 12.101 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 7.139 11.185 multiply_cannon_sync_h2d 29820 15.5 10.278 10.774 10.278 10.774 mp_alltoall_i22 712 14.1 6.413 10.592 6.413 10.592 fft_wrap_pw1pw2_140 519 13.2 1.429 1.460 10.135 10.397 wfi_extrapolate 11 7.9 0.001 0.001 10.220 10.220 fft3d_ps 1281 14.7 2.826 2.913 9.726 9.924 mp_alltoall_d11v 2401 14.1 8.370 8.982 8.370 8.982 potential_pw2rs 127 12.3 0.023 0.023 8.643 8.906 grid_integrate_task_list 127 12.3 7.389 7.855 7.389 7.855 make_images_sizes 4970 15.5 0.006 0.007 3.759 7.335 mp_alltoall_i44 4970 16.5 3.752 7.328 3.752 7.328 calculate_dm_sparse 127 9.5 0.001 0.001 6.826 6.919 mp_alltoall_z22v 1281 16.7 6.323 6.557 6.323 6.557 multiply_cannon_metrocomm4 24850 15.5 0.085 0.094 2.858 6.551 mp_irecv_dv 75445 16.2 2.717 6.288 2.717 6.288 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.152 6.200 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.997 6.087 transfer_rs2pw 519 10.6 0.006 0.007 5.577 5.958 mp_sum_l 7884 12.9 4.440 5.794 4.440 5.794 grid_collocate_task_list 127 9.7 5.249 5.617 5.249 5.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=266.146000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.727273, yerr=19.287730 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 5.910121E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.529151E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592338480 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848050. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.042 281.860 281.861 qs_mol_dyn_low 1 2.0 0.003 0.003 281.272 281.287 qs_forces 11 3.9 0.004 0.005 281.125 281.129 qs_energies 11 4.9 0.003 0.006 272.520 272.536 scf_env_do_scf 11 5.9 0.001 0.001 250.676 250.684 velocity_verlet 10 3.0 0.002 0.002 182.282 182.290 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 182.286 182.289 qs_scf_new_mos 118 7.6 0.001 0.001 132.518 132.757 qs_scf_loop_do_ot 118 8.6 0.001 0.001 132.517 132.757 ot_scf_mini 118 9.6 0.003 0.004 126.957 127.091 dbcsr_multiply_generic 2529 12.6 0.186 0.201 92.083 93.458 multiply_cannon 2529 13.6 0.571 0.604 63.402 70.746 init_scf_loop 11 6.9 0.000 0.000 68.134 68.141 prepare_preconditioner 11 7.9 0.000 0.000 63.474 63.511 make_preconditioner 11 8.9 0.000 0.000 63.474 63.511 make_full_inverse_cholesky 11 9.9 0.018 0.026 60.673 61.334 multiply_cannon_loop 2529 14.6 0.819 0.860 57.844 61.178 qs_ot_get_p 129 10.4 0.002 0.003 58.768 59.110 ot_mini 118 10.6 0.001 0.001 52.677 52.948 qs_ot_p2m_diag 84 11.4 0.502 0.510 51.791 51.824 mp_waitall_1 126876 16.7 39.166 51.119 39.166 51.119 cp_dbcsr_syevd 84 12.4 0.006 0.006 49.577 49.586 cp_fm_diag_elpa 84 13.4 0.000 0.000 42.947 42.956 cp_fm_diag_elpa_base 84 14.4 42.396 42.753 42.936 42.945 cp_fm_cholesky_invert 11 10.9 39.172 39.193 39.172 39.193 rebuild_ks_matrix 129 8.3 0.001 0.001 36.764 37.120 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.020 36.763 37.120 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.441 33.763 qs_ot_get_derivative 118 11.6 0.001 0.002 33.031 33.147 multiply_cannon_metrocomm3 10116 15.6 0.026 0.028 14.462 25.068 make_m2s 5058 13.6 0.067 0.071 22.078 25.041 make_images 5058 14.6 2.183 2.632 21.771 24.739 multiply_cannon_multrec 10116 15.6 10.328 14.568 17.966 21.715 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.047 21.086 calculate_rho_elec 129 8.7 0.258 0.269 21.047 21.085 sum_up_and_integrate 129 10.3 0.002 0.002 19.918 19.961 integrate_v_rspace 129 11.3 0.004 0.004 19.857 19.902 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.017 19.786 apply_single 129 13.6 0.001 0.001 19.017 19.786 ot_diis_step 118 11.6 0.020 0.021 19.403 19.404 make_images_data 5058 15.6 0.056 0.067 13.289 17.826 cp_fm_cholesky_decompose 22 10.9 17.618 17.755 17.618 17.755 hybrid_alltoall_any 5245 16.5 0.847 3.790 11.931 15.833 pw_transfer 1559 11.6 0.087 0.095 14.625 14.738 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 14.400 14.518 init_scf_run 11 5.9 0.000 0.001 14.189 14.190 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.189 14.190 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 14.061 14.166 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 9.990 13.728 density_rs2pw 129 9.7 0.005 0.006 12.345 12.601 fft_wrap_pw1pw2_140 527 13.2 1.815 1.844 11.581 11.702 fft3d_ps 1301 14.7 2.803 2.872 11.105 11.198 multiply_cannon_sync_h2d 10116 15.6 10.690 11.180 10.690 11.180 wfi_extrapolate 11 7.9 0.001 0.001 10.760 10.762 mp_alltoall_d11v 2429 14.1 9.264 10.126 9.264 10.126 mp_allgather_i34 2529 14.6 3.726 8.959 3.726 8.959 potential_pw2rs 129 12.3 0.026 0.027 8.636 8.707 dbcsr_complete_redistribute 397 12.7 2.090 2.145 7.693 8.388 grid_integrate_task_list 129 12.3 7.771 8.218 7.771 8.218 dbcsr_mm_accdrv_process 20934 16.1 3.330 4.549 7.299 8.008 mp_alltoall_z22v 1301 16.7 7.674 7.860 7.674 7.860 calculate_dm_sparse 129 9.5 0.001 0.001 7.382 7.566 make_images_sizes 5058 15.6 0.007 0.008 3.428 7.426 mp_alltoall_i44 5058 16.6 3.421 7.419 3.421 7.419 mp_sum_l 8016 12.9 4.377 6.327 4.377 6.327 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.086 6.107 grid_collocate_task_list 129 9.7 5.627 5.914 5.627 5.914 check_diag 84 13.4 2.158 2.203 5.781 5.885 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.558 5.644 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=281.861000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1440.909091, yerr=14.749142 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.215954E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3989 59322. MP_Allreduce 11050 1505. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.065 288.204 288.206 qs_mol_dyn_low 1 2.0 0.003 0.003 287.569 287.611 qs_forces 11 3.9 0.005 0.006 287.305 287.307 qs_energies 11 4.9 0.002 0.002 278.446 278.455 scf_env_do_scf 11 5.9 0.001 0.001 255.758 255.764 velocity_verlet 10 3.0 0.004 0.008 207.644 207.671 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.059 133.061 init_scf_loop 11 6.9 0.000 0.000 122.420 122.424 prepare_preconditioner 11 7.9 0.000 0.000 117.630 117.652 make_preconditioner 11 8.9 0.000 0.000 117.630 117.652 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.564 114.779 qs_scf_new_mos 116 7.6 0.001 0.001 89.961 90.023 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.961 90.022 ot_scf_mini 116 9.6 0.004 0.004 85.092 85.119 dbcsr_multiply_generic 2485 12.5 0.211 0.224 82.019 82.540 cp_fm_upper_to_full 104 14.8 51.693 73.699 51.693 73.699 multiply_cannon 2485 13.5 0.666 0.695 59.388 60.523 multiply_cannon_loop 2485 14.5 1.044 1.079 55.423 56.826 ot_mini 116 10.6 0.001 0.001 44.326 44.351 dbcsr_complete_redistribute 393 12.7 3.996 4.027 29.032 41.398 copy_fm_to_dbcsr 208 11.6 0.001 0.001 25.516 37.836 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.023 35.228 rebuild_ks_matrix 127 8.3 0.001 0.001 33.692 33.718 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.019 33.691 33.717 cp_fm_cholesky_invert 11 10.9 33.650 33.657 33.650 33.657 mp_alltoall_i22 712 14.1 20.789 33.008 20.789 33.008 mp_waitall_1 102768 16.8 28.145 32.373 28.145 32.373 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.244 31.281 qs_ot_get_p 127 10.4 0.002 0.002 25.748 25.807 qs_ot_get_derivative 116 11.6 0.002 0.002 24.778 24.802 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.703 21.736 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 19.376 20.693 cp_dbcsr_syevd 82 12.4 0.006 0.007 19.901 19.905 make_m2s 4970 13.5 0.077 0.081 18.077 19.534 ot_diis_step 116 11.6 0.021 0.022 19.518 19.519 make_images 4970 14.5 3.051 3.249 17.599 19.059 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.879 18.987 apply_single 127 13.6 0.001 0.001 18.879 18.987 multiply_cannon_multrec 9940 15.5 10.242 12.103 18.176 18.237 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.704 16.705 cp_fm_diag_elpa_base 82 14.4 12.374 13.946 16.700 16.700 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.420 16.437 calculate_rho_elec 127 8.7 0.476 0.477 16.420 16.437 sum_up_and_integrate 127 10.3 0.002 0.002 15.976 16.066 integrate_v_rspace 127 11.3 0.004 0.004 15.914 16.003 multiply_cannon_sync_h2d 9940 15.5 14.187 14.220 14.187 14.220 hybrid_alltoall_any 5155 16.4 1.300 3.013 10.206 12.337 init_scf_run 11 5.9 0.000 0.001 12.284 12.286 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.284 12.285 make_images_data 4970 15.5 0.065 0.069 10.044 12.185 pw_transfer 1535 11.6 0.094 0.095 11.364 11.367 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.128 11.133 fft_wrap_pw1pw2_140 519 13.2 3.022 3.127 9.916 9.920 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.696 9.719 dbcsr_mm_accdrv_process 20590 16.0 4.354 6.127 7.687 9.544 cp_fm_cholesky_decompose 22 10.9 8.992 9.005 8.992 9.005 wfi_extrapolate 11 7.9 0.001 0.001 8.965 8.965 grid_integrate_task_list 127 12.3 8.496 8.692 8.496 8.692 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.213 8.216 density_rs2pw 127 9.7 0.005 0.005 7.495 7.540 mp_alltoall_d11v 2401 14.1 6.905 7.024 6.905 7.024 calculate_dm_sparse 127 9.5 0.001 0.001 6.807 6.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.474 6.506 fft3d_ps 1281 14.7 2.832 2.842 6.413 6.461 grid_collocate_task_list 127 9.7 6.367 6.438 6.367 6.438 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.158 6.221 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.206000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2904.000000, yerr=160.639631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 1.259950E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 17919672. 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 0.021 0.049 86.068 86.069 qs_energies 1 2.0 0.000 0.001 85.532 85.544 ls_scf 1 3.0 0.000 0.002 84.552 84.565 dbcsr_multiply_generic 111 6.7 0.017 0.031 73.040 73.203 multiply_cannon 111 7.7 0.018 0.025 55.916 57.149 multiply_cannon_loop 111 8.7 0.228 0.243 52.506 53.943 ls_scf_main 1 4.0 0.000 0.014 52.767 52.768 density_matrix_trs4 2 5.0 0.002 0.005 47.163 47.238 ls_scf_init_scf 1 4.0 0.000 0.003 28.714 28.716 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.408 27.461 mp_waitall_1 11031 10.9 22.672 25.544 22.672 25.544 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.131 25.155 multiply_cannon_multrec 2664 9.7 8.172 8.914 15.583 17.377 multiply_cannon_sync_h2d 2664 9.7 13.600 15.302 13.600 15.302 make_m2s 222 7.7 0.009 0.011 13.405 13.886 make_images 222 8.7 0.099 0.108 13.383 13.867 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.690 12.366 make_images_data 222 9.7 0.004 0.006 7.934 8.482 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.376 8.408 hybrid_alltoall_any 227 10.6 0.216 1.843 6.805 8.169 dbcsr_mm_accdrv_process 4760 10.4 0.599 0.702 7.028 8.061 dbcsr_mm_accdrv_process_sort 4760 11.4 6.222 7.130 6.222 7.130 calculate_norms 4752 9.8 5.509 6.179 5.509 6.179 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.054 5.196 mp_sum_l 887 5.1 3.189 4.694 3.189 4.694 make_images_sizes 222 9.7 0.000 0.000 0.802 3.740 mp_alltoall_i44 222 10.7 0.801 3.740 0.801 3.740 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.067 3.530 mp_irecv_dv 6231 10.9 2.051 3.502 2.051 3.502 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.398 3.434 arnoldi_extremal 4 6.8 0.000 0.000 3.368 3.401 arnoldi_normal_ev 4 7.8 0.001 0.004 3.368 3.401 build_subspace 16 8.4 0.009 0.012 3.234 3.242 ls_scf_post 1 4.0 0.002 0.013 3.070 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.866 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.319 2.865 dbcsr_merge_single_wm 555 10.7 0.455 0.572 2.310 2.856 make_images_pack 222 9.7 2.204 2.626 2.206 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.352 2.600 dbcsr_sort_data 658 11.4 2.114 2.584 2.114 2.584 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.469 2.072 2.471 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.345 2.436 buffer_matrices_ensure_size 222 8.7 1.745 2.115 1.745 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.856 1.857 rebuild_ks_matrix 3 7.3 0.000 0.000 1.847 1.848 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.847 1.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.069000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 2.172113E+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 266673. MP_Allreduce 3138 10075. 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 0.020 0.037 91.169 91.171 qs_energies 1 2.0 0.000 0.000 90.698 90.702 ls_scf 1 3.0 0.000 0.000 89.284 89.288 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.144 75.452 multiply_cannon 111 7.7 0.026 0.039 53.081 57.680 ls_scf_main 1 4.0 0.000 0.000 55.306 55.311 multiply_cannon_loop 111 8.7 0.136 0.148 50.355 54.116 density_matrix_trs4 2 5.0 0.002 0.003 49.464 49.697 mp_waitall_1 9105 10.9 21.622 30.568 21.622 30.568 ls_scf_init_scf 1 4.0 0.000 0.000 30.210 30.212 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.998 29.084 multiply_cannon_multrec 1332 9.7 13.228 17.666 22.526 28.099 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.643 26.652 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.991 20.995 make_m2s 222 7.7 0.006 0.007 15.374 15.894 make_images 222 8.7 1.369 1.703 15.344 15.864 dbcsr_mm_accdrv_process 4041 10.4 0.347 0.563 8.902 10.454 dbcsr_mm_accdrv_process_sort 4041 11.4 8.405 9.925 8.405 9.925 make_images_data 222 9.7 0.004 0.004 9.031 9.843 hybrid_alltoall_any 227 10.6 0.542 2.549 8.402 9.634 mp_sum_l 887 5.1 5.218 8.893 5.218 8.893 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.191 7.958 mp_irecv_dv 3311 11.0 3.171 7.909 3.171 7.909 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.942 7.019 calculate_norms 2376 9.8 6.025 6.693 6.025 6.693 multiply_cannon_sync_h2d 1332 9.7 4.836 6.183 4.836 6.183 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.028 5.243 arnoldi_extremal 4 6.8 0.000 0.000 4.802 4.822 arnoldi_normal_ev 4 7.8 0.001 0.005 4.802 4.821 build_subspace 16 8.4 0.014 0.021 4.537 4.542 ls_scf_post 1 4.0 0.000 0.000 3.768 3.771 ls_scf_store_result 1 5.0 0.000 0.000 3.451 3.591 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.180 3.445 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.219 2.743 3.220 mp_allgather_i34 111 8.7 0.722 2.978 0.722 2.978 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.677 2.793 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.240 2.670 dbcsr_data_new 4174 10.1 2.114 2.401 2.114 2.401 make_images_pack 222 9.7 1.820 2.133 1.822 2.135 dbcsr_sort_data 436 11.2 1.853 2.112 1.853 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.892 1.894 rebuild_ks_matrix 3 7.3 0.000 0.000 1.878 1.881 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.878 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.171000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 2.918126E+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 265448. MP_Allreduce 3138 10896. 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 0.024 0.043 93.235 93.238 qs_energies 1 2.0 0.000 0.000 92.659 92.662 ls_scf 1 3.0 0.000 0.000 91.233 91.237 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.691 76.045 multiply_cannon 111 7.7 0.031 0.059 52.317 56.864 ls_scf_main 1 4.0 0.000 0.000 56.816 56.820 multiply_cannon_loop 111 8.7 0.117 0.130 49.546 53.402 density_matrix_trs4 2 5.0 0.002 0.003 50.800 50.963 mp_waitall_1 7281 11.0 24.250 34.479 24.250 34.479 ls_scf_init_scf 1 4.0 0.000 0.000 30.675 30.678 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.473 29.549 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.107 27.120 multiply_cannon_multrec 888 9.7 12.621 15.068 21.224 24.246 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.467 23.358 make_m2s 222 7.7 0.006 0.007 16.656 17.382 make_images 222 8.7 1.582 1.837 16.618 17.342 make_images_data 222 9.7 0.004 0.004 9.877 10.983 hybrid_alltoall_any 227 10.6 0.640 2.958 9.447 10.581 dbcsr_mm_accdrv_process 3754 10.4 0.314 0.492 8.117 9.368 dbcsr_mm_accdrv_process_sort 3754 11.4 7.678 8.876 7.678 8.876 mp_sum_l 887 5.1 5.036 8.594 5.036 8.594 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.545 7.550 multiply_cannon_sync_h2d 888 9.7 6.024 7.232 6.024 7.232 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.212 mp_irecv_dv 2335 11.1 2.457 7.160 2.457 7.160 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.812 6.612 arnoldi_extremal 4 6.8 0.000 0.000 5.245 5.263 arnoldi_normal_ev 4 7.8 0.001 0.005 5.245 5.263 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.844 5.058 build_subspace 16 8.4 0.014 0.020 4.925 4.937 calculate_norms 1584 9.8 4.327 4.908 4.327 4.908 mp_allgather_i34 111 8.7 0.886 3.855 0.886 3.855 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.490 3.805 ls_scf_post 1 4.0 0.000 0.000 3.743 3.747 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.619 3.029 3.621 ls_scf_store_result 1 5.0 0.000 0.000 3.486 3.568 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.900 3.001 make_images_sizes 222 9.7 0.000 0.000 0.999 2.508 mp_alltoall_i44 222 10.7 0.999 2.508 0.999 2.508 dbcsr_data_new 4116 9.9 2.112 2.440 2.112 2.440 dbcsr_sort_data 325 11.1 1.834 2.051 1.834 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.915 rebuild_ks_matrix 3 7.3 0.000 0.000 1.893 1.895 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.893 1.895 make_images_pack 222 9.7 1.622 1.879 1.625 1.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.238000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2285.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 3.328717E+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 266673. MP_Allreduce 3138 13030. 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 0.030 0.046 100.214 100.215 qs_energies 1 2.0 0.000 0.000 99.548 99.552 ls_scf 1 3.0 0.000 0.000 97.861 97.866 dbcsr_multiply_generic 111 6.7 0.017 0.019 79.474 79.741 ls_scf_main 1 4.0 0.000 0.000 60.501 60.502 multiply_cannon 111 7.7 0.036 0.072 52.299 57.294 density_matrix_trs4 2 5.0 0.002 0.003 53.953 54.111 multiply_cannon_loop 111 8.7 0.154 0.174 47.177 49.985 ls_scf_init_scf 1 4.0 0.000 0.000 34.084 34.087 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.789 32.848 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.301 30.312 mp_waitall_1 6369 11.0 23.565 30.117 23.565 30.117 multiply_cannon_multrec 1332 9.7 14.121 17.632 22.043 25.247 make_m2s 222 7.7 0.007 0.007 21.494 22.894 make_images 222 8.7 3.143 3.612 21.443 22.844 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.578 18.034 make_images_data 222 9.7 0.004 0.004 12.051 13.881 hybrid_alltoall_any 227 10.6 0.799 3.770 11.384 13.144 dbcsr_mm_accdrv_process 3641 10.4 0.296 0.488 7.562 9.068 dbcsr_mm_accdrv_process_sort 3641 11.4 7.118 8.573 7.118 8.573 mp_sum_l 887 5.1 4.100 7.432 4.100 7.432 arnoldi_extremal 4 6.8 0.000 0.000 6.812 6.826 arnoldi_normal_ev 4 7.8 0.001 0.005 6.812 6.826 build_subspace 16 8.4 0.014 0.020 6.451 6.461 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.101 6.098 multiply_cannon_sync_h2d 1332 9.7 5.421 6.039 5.421 6.039 mp_irecv_dv 3229 10.9 2.075 6.012 2.075 6.012 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.191 5.938 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.740 5.276 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.646 4.830 mp_allgather_i34 111 8.7 2.235 4.614 2.235 4.614 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.189 4.588 calculate_norms 2376 9.8 4.177 4.544 4.177 4.544 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.699 3.201 3.701 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.305 3.355 ls_scf_post 1 4.0 0.000 0.000 3.276 3.281 dbcsr_sort_data 658 11.4 3.019 3.280 3.019 3.280 dbcsr_special_finalize 555 9.7 0.006 0.007 2.763 3.157 dbcsr_merge_single_wm 555 10.7 0.534 0.656 2.754 3.148 ls_scf_store_result 1 5.0 0.000 0.000 3.024 3.078 mp_sum_dv 2129 10.8 2.229 2.624 2.229 2.624 dbcsr_data_release 10477 10.7 1.593 2.437 1.593 2.437 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.094 2.096 rebuild_ks_matrix 3 7.3 0.000 0.000 2.070 2.072 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.070 2.072 make_images_sizes 222 9.7 0.000 0.000 0.829 2.059 mp_alltoall_i44 222 10.7 0.829 2.059 0.829 2.059 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.215000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2676.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 4.752093E+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 265536. MP_Allreduce 3129 15263. 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 0.037 0.052 96.082 96.084 qs_energies 1 2.0 0.000 0.000 95.255 95.261 ls_scf 1 3.0 0.000 0.000 93.243 93.248 dbcsr_multiply_generic 111 6.7 0.017 0.018 72.515 72.674 ls_scf_main 1 4.0 0.000 0.000 58.761 58.761 multiply_cannon 111 7.7 0.073 0.151 53.417 57.229 multiply_cannon_loop 111 8.7 0.089 0.098 50.784 52.066 density_matrix_trs4 2 5.0 0.002 0.003 51.284 51.352 ls_scf_init_scf 1 4.0 0.000 0.000 31.002 31.003 mp_waitall_1 5436 11.0 25.720 29.834 25.720 29.834 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.638 29.667 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.007 27.485 27.505 multiply_cannon_multrec 444 9.7 13.723 16.296 20.811 22.789 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 12.190 16.712 make_m2s 222 7.7 0.005 0.005 14.437 15.302 make_images 222 8.7 2.039 2.477 14.369 15.233 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.605 12.461 hybrid_alltoall_any 227 10.6 0.798 3.805 8.654 10.214 make_images_data 222 9.7 0.003 0.004 8.722 10.114 multiply_cannon_sync_h2d 444 9.7 6.665 7.945 6.665 7.945 dbcsr_mm_accdrv_process 3003 10.4 0.354 0.409 6.782 7.914 dbcsr_mm_accdrv_process_sort 3003 11.4 6.416 7.513 6.416 7.513 arnoldi_extremal 4 6.8 0.000 0.000 7.307 7.328 arnoldi_normal_ev 4 7.8 0.002 0.005 7.307 7.328 build_subspace 16 8.4 0.014 0.019 6.878 6.889 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.769 4.916 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.619 4.787 mp_sum_l 887 5.1 2.683 4.344 2.683 4.344 dbcsr_matrix_vector_mult_local 304 10.0 3.720 4.208 3.723 4.210 make_images_sizes 222 9.7 0.000 0.000 1.334 4.127 mp_alltoall_i44 222 10.7 1.334 4.127 1.334 4.127 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.563 4.035 mp_irecv_dv 1241 11.2 1.547 4.012 1.547 4.012 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.830 3.895 mp_allgather_i34 111 8.7 1.209 3.790 1.209 3.790 calculate_norms 792 9.8 3.588 3.687 3.588 3.687 ls_scf_post 1 4.0 0.000 0.000 3.480 3.485 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.871 3.367 ls_scf_store_result 1 5.0 0.000 0.000 3.266 3.303 dbcsr_finalize 304 7.8 0.062 0.077 2.204 2.364 mp_sum_dv 2129 10.8 2.146 2.296 2.146 2.296 dbcsr_data_new 4608 9.7 1.793 2.270 1.793 2.270 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.215 2.216 dbcsr_merge_all 275 8.9 0.481 0.542 2.061 2.202 rebuild_ks_matrix 3 7.3 0.000 0.000 2.183 2.184 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.183 2.184 qs_energies_init_hamiltonians 1 3.0 0.000 0.002 1.996 1.996 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=96.084000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 8.883044E+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 284089. MP_Allreduce 3123 21388. 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 0.073 0.092 114.291 114.293 qs_energies 1 2.0 0.000 0.000 112.858 112.863 ls_scf 1 3.0 0.000 0.000 109.872 109.876 dbcsr_multiply_generic 111 6.7 0.023 0.028 79.972 80.278 ls_scf_main 1 4.0 0.000 0.000 69.171 69.172 density_matrix_trs4 2 5.0 0.002 0.003 59.243 59.340 multiply_cannon 111 7.7 0.128 0.192 50.352 54.664 multiply_cannon_loop 111 8.7 0.099 0.101 47.290 49.540 ls_scf_init_scf 1 4.0 0.000 0.000 36.629 36.632 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.902 34.927 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 32.089 32.097 mp_waitall_1 4527 11.1 24.869 30.639 24.869 30.639 make_m2s 222 7.7 0.005 0.005 24.414 25.437 make_images 222 8.7 3.570 3.873 24.306 25.326 multiply_cannon_multrec 444 9.7 17.839 18.639 22.483 23.298 hybrid_alltoall_any 227 10.6 1.655 3.611 13.771 17.556 make_images_data 222 9.7 0.003 0.004 14.178 16.908 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 12.040 14.415 arnoldi_extremal 4 6.8 0.000 0.000 9.229 9.240 arnoldi_normal_ev 4 7.8 0.003 0.009 9.229 9.240 multiply_cannon_sync_h2d 444 9.7 8.794 8.845 8.794 8.845 build_subspace 16 8.4 0.024 0.035 8.649 8.662 dbcsr_matrix_vector_mult 304 9.0 0.016 0.031 6.397 6.668 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.711 5.796 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.236 5.505 dbcsr_matrix_vector_mult_local 304 10.0 5.115 5.453 5.118 5.456 make_images_sizes 222 9.7 0.000 0.000 2.107 4.955 mp_alltoall_i44 222 10.7 2.106 4.955 2.106 4.955 dbcsr_mm_accdrv_process 1814 10.4 0.293 0.351 4.464 4.589 mp_sum_l 887 5.1 2.843 4.470 2.843 4.470 dbcsr_mm_accdrv_process_sort 1814 11.4 4.125 4.254 4.125 4.254 ls_scf_post 1 4.0 0.000 0.000 4.071 4.077 mp_allgather_i34 111 8.7 1.166 3.967 1.166 3.967 ls_scf_store_result 1 5.0 0.000 0.000 3.752 3.826 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.006 3.331 calculate_norms 792 9.8 3.229 3.277 3.229 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.074 3.166 dbcsr_complete_redistribute 5 7.6 1.434 1.472 2.994 3.097 mp_alltoall_d11v 48 9.2 3.036 3.074 3.036 3.074 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 2.956 2.956 dbcsr_merge_all 275 8.9 0.883 0.909 2.859 2.944 dbcsr_new_transposed 4 7.5 0.241 0.251 2.721 2.730 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.721 2.723 matrix_ls_to_qs 2 6.0 0.000 0.000 2.576 2.681 rebuild_ks_matrix 3 7.3 0.000 0.000 2.654 2.656 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 2.654 2.656 dbcsr_sort_data 325 11.1 2.437 2.502 2.437 2.502 dbcsr_redistribute 4 8.5 1.349 1.393 2.453 2.460 mp_sum_dv 2129 10.8 2.231 2.452 2.231 2.452 dbcsr_data_new 6591 9.6 1.871 2.297 1.871 2.297 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=114.293000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6944.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/263bc551066ec83fd9537a4c612e9ad8f6479d6e_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 593.313792E+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 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 971801278. 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 0.045 0.070 498.061 498.064 qs_mol_dyn_low 1 2.0 0.003 0.013 496.513 496.594 qs_forces 5 3.8 0.005 0.007 496.368 496.373 qs_energies 5 4.8 0.002 0.050 491.046 491.150 scf_env_do_scf 5 5.8 0.000 0.001 470.741 470.746 scf_env_do_scf_inner_loop 105 6.6 0.003 0.031 395.003 395.017 qs_scf_new_mos 105 7.6 0.000 0.001 307.281 308.355 qs_scf_loop_do_ot 105 8.6 0.001 0.001 307.281 308.355 ot_scf_mini 105 9.6 0.003 0.004 293.872 294.734 velocity_verlet 4 3.0 0.005 0.025 186.463 186.470 qs_ot_get_p 112 10.4 0.001 0.001 176.368 177.022 dbcsr_multiply_generic 1445 12.2 0.125 0.132 169.453 176.534 qs_ot_p2m_diag 40 11.0 0.020 0.036 156.457 156.700 multiply_cannon 1445 13.2 0.282 0.300 134.389 154.386 cp_dbcsr_syevd 40 12.0 0.002 0.002 151.787 151.794 multiply_cannon_loop 1445 14.2 2.864 3.021 128.766 149.287 cp_fm_syevd 40 13.0 0.000 0.000 141.973 142.155 cp_fm_redistribute_end 40 14.0 66.863 133.309 66.893 133.330 cp_fm_syevd_base 40 14.0 66.374 132.838 66.374 132.838 mp_waitall_1 488190 16.1 72.823 104.419 72.823 104.419 multiply_cannon_metrocomm3 69360 15.2 0.199 0.213 40.459 82.884 ot_mini 105 10.6 0.001 0.001 80.309 80.661 init_scf_loop 7 6.6 0.000 0.007 75.596 75.600 prepare_preconditioner 7 7.6 0.000 0.000 67.425 67.483 make_preconditioner 7 8.6 0.000 0.001 67.425 67.483 qs_ot_get_derivative 55 11.6 0.001 0.001 57.375 57.659 make_full_inverse_cholesky 7 9.6 0.000 0.000 53.466 53.576 multiply_cannon_metrocomm1 69360 15.2 0.091 0.099 12.593 53.031 rebuild_ks_matrix 110 8.4 0.000 0.000 49.184 50.452 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.041 49.183 50.451 qs_rho_update_rho_low 110 7.6 0.001 0.001 49.033 50.393 calculate_rho_elec 110 8.6 0.029 0.032 49.032 50.392 qs_ks_update_qs_env 112 7.6 0.001 0.001 45.105 46.257 multiply_cannon_multrec 69360 15.2 26.570 32.432 36.667 42.303 density_rs2pw 110 9.6 0.005 0.007 39.889 40.550 cp_fm_cholesky_invert 7 10.6 35.387 35.460 35.387 35.460 transfer_rs2pw 445 10.6 0.007 0.008 34.051 34.655 multiply_cannon_sync_h2d 69360 15.2 24.437 30.078 24.437 30.078 mp_sendrecv_dv 168740 12.6 27.425 27.779 27.425 27.779 transfer_rs2pw_30 110 11.6 1.236 1.309 24.259 26.312 mp_sum_l 4764 12.2 19.695 24.951 19.695 24.951 sum_up_and_integrate 60 10.3 0.001 0.002 24.759 24.786 integrate_v_rspace 60 11.3 0.001 0.002 24.741 24.770 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.698 24.127 apply_single 62 13.6 0.000 0.000 23.698 24.127 qs_ot_get_orbitals 105 10.6 0.001 0.001 20.958 22.164 ot_new_cg_direction 55 11.6 0.001 0.001 21.822 21.828 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 21.074 21.315 potential_pw2rs 60 12.3 0.003 0.003 19.257 19.399 pw_transfer 1645 12.4 0.080 0.100 18.141 18.805 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 18.002 18.671 transfer_pw2rs 245 13.2 0.003 0.004 17.269 17.445 cp_fm_cholesky_decompose 14 10.2 17.263 17.276 17.263 17.276 check_diag 80 13.5 8.593 8.888 16.416 16.587 init_scf_run 5 5.8 0.000 0.004 16.574 16.576 scf_env_initial_rho_setup 5 6.8 0.003 0.005 16.574 16.576 make_m2s 2890 13.2 0.077 0.083 15.066 16.200 make_images 2890 14.2 0.242 0.261 14.961 16.092 calculate_dm_sparse 110 9.5 0.000 0.001 15.240 15.983 fft_wrap_pw1pw2_240 915 15.0 1.106 1.218 14.502 15.127 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 14.887 15.002 fft3d_pb 915 16.0 2.353 2.598 12.750 13.607 qs_vxc_create 110 10.4 0.002 0.003 13.302 13.395 transfer_pw2rs_30 60 14.3 0.273 0.315 12.126 13.249 mp_alltoall_d11v 1300 13.8 11.254 13.040 11.254 13.040 make_full_single_inverse 7 9.6 0.001 0.001 12.870 12.928 mp_alltoall_z22v 2340 17.7 10.282 11.876 10.282 11.876 dbcsr_mm_accdrv_process 154766 15.8 6.347 6.649 9.982 10.898 make_images_sizes 2890 15.2 0.004 0.004 8.784 10.287 mp_alltoall_i44 2890 16.2 8.781 10.283 8.781 10.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=498.064000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.370460 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 263bc551066ec83fd9537a4c612e9ad8f6479d6e Summary: empty Status: OK