=== 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: 14e431119277e2f0aa53cf24ac09b5860de697e5 ################# 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 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/01 job id: 51664280 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/02 job id: 51664282 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/03 job id: 51664283 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/04 job id: 51664285 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/05 job id: 51664287 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/06 job id: 51664290 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/07 job id: 51664295 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/08 job id: 51664297 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/09 job id: 51664299 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/10 job id: 51664302 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/11 job id: 51664305 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/12 job id: 51664307 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/13 job id: 51664309 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/14 job id: 51664311 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/15 job id: 51664313 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/16 job id: 51664315 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/17 job id: 51664318 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/18 job id: 51664319 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/19 job id: 51664320 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/20 job id: 51664321 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/21 job id: 51664322 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/22 job id: 51664323 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/23 job id: 51664325 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/24 job id: 51664326 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/25 job id: 51664330 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/26 job id: 51664333 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/27 job id: 51664334 --- 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/14e431119277e2f0aa53cf24ac09b5860de697e5_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 172669. 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 0.019 0.034 136.016 136.017 farming_run 1 2.0 135.465 135.466 135.987 135.990 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.477865E+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 0.009 0.029 115.797 115.797 qs_energies 1 2.0 0.001 0.001 115.587 115.590 mp2_main 1 3.0 0.000 0.000 113.388 113.392 mp2_gpw_main 1 4.0 0.027 0.033 112.339 112.342 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.361 93.716 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.601 54.957 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 41.006 46.066 get_2c_integrals 1 6.0 0.008 0.009 37.872 38.586 integrate_v_rspace 273 8.0 0.442 0.456 24.633 29.500 pw_transfer 6555 10.6 0.370 0.383 26.898 27.449 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.526 26.145 grid_integrate_task_list 273 9.0 20.535 25.872 20.535 25.872 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 23.039 23.642 compute_2c_integrals 1 7.0 0.002 0.002 19.661 19.662 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.667 19.312 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.431 18.666 19.308 rpa_ri_compute_en 1 5.0 0.020 0.038 18.874 19.077 cp_fm_cholesky_decompose 12 8.2 18.212 18.917 18.212 18.917 cholesky_decomp 1 7.0 0.000 0.000 17.055 17.784 fft3d_s 5443 13.4 16.164 16.518 16.186 16.540 ao_to_mo_and_store_B_mult_1 272 7.0 10.772 15.322 10.772 15.322 calculate_wavefunction 272 8.0 5.366 5.505 12.292 12.998 rpa_num_int 1 6.0 0.001 0.017 10.761 10.771 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.711 10.714 calc_mat_Q 8 8.0 0.000 0.000 9.498 9.600 contract_S_to_Q 8 9.0 0.000 0.000 8.924 9.023 calc_potential_gpw 544 9.5 0.004 0.005 8.250 8.796 parallel_gemm_fm 14 9.1 0.000 0.000 8.524 8.616 parallel_gemm_fm_cosma 14 10.1 8.523 8.616 8.523 8.616 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.121 8.528 potential_pw2rs 545 10.0 0.106 0.107 7.505 8.177 collocate_single_gaussian 272 10.0 0.039 0.042 7.365 7.702 create_integ_mat 1 6.0 0.020 0.029 7.651 7.661 array2fm 1 7.0 0.000 0.000 6.629 7.077 pw_scatter_s 2720 13.7 4.338 4.461 4.338 4.461 pw_gather_s 2722 13.2 3.431 3.715 3.431 3.715 array2fm_buffer_send 1 8.0 2.866 3.133 2.866 3.133 pw_poisson_solve 545 10.5 1.121 1.188 2.204 2.484 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.341950, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 200775. 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 0.036 0.047 524.479 524.480 farming_run 1 2.0 523.665 523.673 524.414 524.418 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.230733E+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 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 0.031 319.559 319.560 qs_energies 1 2.0 0.003 0.024 319.335 319.336 mp2_main 1 3.0 0.000 0.000 204.051 204.055 mp2_gpw_main 1 4.0 0.047 0.072 202.974 202.979 mp2_ri_gpw_compute_en 1 5.0 0.037 0.043 128.997 130.638 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.777 1.921 127.401 127.405 scf_env_do_scf 1 3.0 0.000 0.000 114.971 114.972 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.947 113.954 rebuild_ks_matrix 4 6.0 0.000 0.000 113.945 113.953 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.066 113.945 113.953 hfx_ks_matrix 4 8.0 0.001 0.001 113.565 113.569 integrate_four_center 4 9.0 0.153 0.466 113.564 113.568 mp2_ri_gpw_compute_en_expansio 172 7.0 0.464 0.503 111.343 111.976 local_gemm 172 8.0 110.879 111.498 110.879 111.498 integrate_four_center_main 4 10.0 0.129 0.603 101.821 105.351 integrate_four_center_bin 262 11.0 101.692 105.322 101.692 105.322 init_scf_loop 1 4.0 0.000 0.000 96.960 96.960 mp2_ri_gpw_compute_in 1 5.0 0.067 0.092 73.831 74.794 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.680 54.642 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.156 41.389 46.379 integrate_v_rspace 95 8.0 0.398 0.566 27.901 32.764 pw_transfer 2240 10.6 0.145 0.162 29.510 29.870 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.517 28.907 ao_to_mo_and_store_B_mult_1 91 7.0 10.623 28.727 10.623 28.727 grid_integrate_task_list 95 9.0 23.291 28.368 23.291 28.368 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.258 26.563 get_2c_integrals 1 6.0 0.000 0.000 20.021 20.087 compute_2c_integrals 1 7.0 0.002 0.003 19.001 19.009 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.636 18.861 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.883 18.635 18.860 fft3d_s 1823 13.4 18.483 18.797 18.497 18.811 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.010 18.010 calculate_wavefunction 91 8.0 2.006 2.035 9.611 9.817 mp2_ri_gpw_compute_en_comm 22 7.0 0.486 0.508 8.193 9.750 potential_pw2rs 186 10.0 0.033 0.034 8.426 9.026 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.155 8.503 calc_potential_gpw 182 9.5 0.002 0.002 7.921 8.123 collocate_single_gaussian 91 10.0 0.016 0.020 7.757 7.946 mp_sendrecv_dm3 2068 8.0 6.262 7.836 6.262 7.836 mp_sync 37 10.5 3.734 6.978 3.734 6.978 integrate_four_center_load 4 10.0 0.000 0.000 6.774 6.779 hfx_load_balance 1 11.0 0.000 0.000 6.774 6.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.972806, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1491.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 444.665856E+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 0.049 0.079 54.151 54.153 qs_mol_dyn_low 1 2.0 0.015 0.043 53.549 53.562 qs_forces 11 3.9 0.002 0.003 53.266 53.274 qs_energies 11 4.9 0.006 0.009 51.758 51.786 scf_env_do_scf 11 5.9 0.000 0.001 45.102 45.102 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.013 43.014 dbcsr_multiply_generic 2286 12.5 0.098 0.102 33.089 33.489 qs_scf_new_mos 108 7.5 0.000 0.001 33.051 33.362 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.050 33.362 ot_scf_mini 108 9.5 0.002 0.002 31.396 31.608 multiply_cannon 2286 13.5 0.186 0.194 26.138 27.437 multiply_cannon_loop 2286 14.5 1.811 1.902 25.483 26.826 velocity_verlet 10 3.0 0.042 0.135 25.865 25.874 ot_mini 108 10.5 0.001 0.001 18.664 18.923 qs_ot_get_derivative 108 11.5 0.001 0.001 15.674 15.857 mp_waitall_1 245248 16.5 8.152 14.038 8.152 14.038 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 5.812 12.633 multiply_cannon_multrec 54864 15.5 3.669 5.810 7.834 11.212 qs_ot_get_p 119 10.4 0.004 0.014 8.053 8.384 rebuild_ks_matrix 119 8.3 0.000 0.000 7.814 7.977 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.027 7.813 7.977 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.880 7.030 mp_sum_l 7287 12.8 5.085 6.688 5.085 6.688 multiply_cannon_sync_h2d 54864 15.5 5.207 6.452 5.207 6.452 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.442 5.846 dbcsr_mm_accdrv_process 76910 16.1 1.834 2.897 4.078 5.821 qs_ot_p2m_diag 50 11.0 0.004 0.008 5.281 5.327 init_scf_run 11 5.9 0.000 0.001 5.277 5.277 scf_env_initial_rho_setup 11 6.9 0.002 0.004 5.277 5.277 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.992 5.104 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.458 4.459 sum_up_and_integrate 119 10.3 0.001 0.003 4.411 4.417 integrate_v_rspace 119 11.3 0.002 0.003 4.400 4.407 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.258 4.258 cp_fm_redistribute_end 50 14.0 2.171 4.232 2.177 4.235 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.096 4.197 calculate_rho_elec 119 8.7 0.011 0.017 4.096 4.197 cp_fm_diag_elpa_base 50 14.0 2.052 4.139 2.056 4.146 calculate_dm_sparse 119 9.5 0.000 0.000 3.098 3.252 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.905 3.088 apply_single 119 13.6 0.000 0.000 2.905 3.088 calculate_first_density_matrix 1 7.0 0.003 0.025 3.068 3.076 jit_kernel_multiply 13 15.8 2.181 2.839 2.181 2.839 acc_transpose_blocks 54864 15.5 0.221 0.244 2.204 2.736 ot_diis_step 108 11.5 0.006 0.006 2.713 2.713 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.697 2.661 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.472 2.532 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.012 2.346 2.349 density_rs2pw 119 9.7 0.004 0.004 2.106 2.192 wfi_extrapolate 11 7.9 0.002 0.015 2.134 2.135 grid_integrate_task_list 119 12.3 2.013 2.122 2.013 2.122 init_scf_loop 11 6.9 0.000 0.000 2.072 2.072 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.008 2.062 mp_sum_d 4137 12.0 1.345 1.943 1.345 1.943 potential_pw2rs 119 12.3 0.004 0.004 1.788 1.796 pw_transfer 1439 11.6 0.051 0.056 1.627 1.692 make_m2s 4572 13.5 0.053 0.055 1.600 1.651 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.612 1.646 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.552 1.619 make_images 4572 14.5 0.132 0.138 1.518 1.569 transfer_rs2pw 487 10.6 0.005 0.006 1.378 1.505 mp_alltoall_d11v 2130 13.8 1.308 1.472 1.308 1.472 acc_transpose_blocks_sync 164592 16.5 1.189 1.427 1.189 1.427 mp_waitany 12084 13.8 1.267 1.423 1.267 1.423 grid_collocate_task_list 119 9.7 1.350 1.406 1.350 1.406 fft3d_ps 1201 14.6 0.372 0.479 1.264 1.326 transfer_pw2rs 487 13.2 0.006 0.006 1.308 1.316 fft_wrap_pw1pw2_140 487 13.2 0.108 0.116 1.206 1.274 parallel_gemm_fm 81 9.0 0.000 0.000 1.196 1.201 parallel_gemm_fm_cosma 81 10.0 1.196 1.201 1.196 1.201 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.743 1.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.153000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=423.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 480.018432E+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 1031941. 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 0.025 0.069 40.220 40.222 qs_mol_dyn_low 1 2.0 0.008 0.044 39.751 39.762 qs_forces 11 3.9 0.002 0.003 39.592 39.593 qs_energies 11 4.9 0.005 0.028 37.825 37.830 scf_env_do_scf 11 5.9 0.001 0.003 32.153 32.154 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.429 29.430 dbcsr_multiply_generic 2286 12.5 0.112 0.116 22.441 22.791 qs_scf_new_mos 108 7.5 0.001 0.001 20.821 21.050 qs_scf_loop_do_ot 108 8.5 0.001 0.002 20.820 21.050 ot_scf_mini 108 9.5 0.003 0.004 19.889 20.055 multiply_cannon 2286 13.5 0.208 0.217 17.321 18.793 velocity_verlet 10 3.0 0.025 0.051 18.472 18.475 multiply_cannon_loop 2286 14.5 1.185 1.246 16.168 17.649 ot_mini 108 10.5 0.001 0.001 12.262 12.495 mp_waitall_1 200699 16.5 5.691 10.962 5.691 10.962 qs_ot_get_derivative 108 11.5 0.001 0.001 9.895 10.063 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.168 9.713 multiply_cannon_multrec 27432 15.5 1.835 4.188 6.384 9.119 rebuild_ks_matrix 119 8.3 0.000 0.000 6.875 7.014 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.874 7.014 dbcsr_mm_accdrv_process 47894 16.0 3.641 6.079 4.469 6.640 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.961 6.088 qs_ot_get_p 119 10.4 0.005 0.037 4.940 5.160 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.792 4.640 init_scf_run 11 5.9 0.000 0.001 4.367 4.368 scf_env_initial_rho_setup 11 6.9 0.002 0.014 4.367 4.368 mp_sum_l 7287 12.8 2.083 4.156 2.083 4.156 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.998 4.064 apply_single 119 13.6 0.000 0.000 2.998 4.064 sum_up_and_integrate 119 10.3 0.001 0.002 3.789 3.805 integrate_v_rspace 119 11.3 0.002 0.003 3.774 3.791 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.623 3.646 calculate_rho_elec 119 8.7 0.021 0.029 3.622 3.645 qs_ot_p2m_diag 50 11.0 0.009 0.017 3.044 3.067 make_m2s 4572 13.5 0.051 0.053 2.599 2.925 make_images 4572 14.5 0.206 0.246 2.510 2.839 calculate_first_density_matrix 1 7.0 0.002 0.013 2.830 2.834 init_scf_loop 11 6.9 0.001 0.003 2.700 2.702 cp_dbcsr_syevd 50 12.0 0.003 0.006 2.643 2.649 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.314 2.400 ot_diis_step 108 11.5 0.011 0.011 2.316 2.317 calculate_dm_sparse 119 9.5 0.000 0.001 2.218 2.297 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.286 2.286 cp_fm_redistribute_end 50 14.0 1.158 2.259 1.162 2.262 multiply_cannon_sync_h2d 27432 15.5 1.705 2.226 1.705 2.226 cp_fm_diag_elpa_base 50 14.0 1.068 2.167 1.096 2.208 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.167 2.171 jit_kernel_multiply 10 16.4 0.769 2.157 0.769 2.157 acc_transpose_blocks 27432 15.5 0.113 0.119 1.671 2.038 density_rs2pw 119 9.7 0.004 0.004 1.921 2.004 pw_transfer 1439 11.6 0.064 0.068 1.926 1.972 grid_integrate_task_list 119 12.3 1.844 1.935 1.844 1.935 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.855 1.893 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.835 1.882 prepare_preconditioner 11 7.9 0.000 0.000 1.567 1.592 make_preconditioner 11 8.9 0.000 0.001 1.567 1.592 make_images_data 4572 15.5 0.047 0.054 1.229 1.563 potential_pw2rs 119 12.3 0.006 0.007 1.527 1.547 make_full_inverse_cholesky 11 9.9 0.003 0.022 1.455 1.523 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.470 1.515 fft3d_ps 1201 14.6 0.522 0.581 1.467 1.509 wfi_extrapolate 11 7.9 0.001 0.003 1.462 1.463 hybrid_alltoall_any 4725 16.4 0.053 0.114 1.096 1.426 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.396 1.405 grid_collocate_task_list 119 9.7 1.285 1.339 1.285 1.339 mp_alltoall_d11v 2130 13.8 1.215 1.336 1.215 1.336 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.206 1.255 transfer_rs2pw 487 10.6 0.005 0.006 1.102 1.186 mp_sum_d 4137 12.0 0.597 1.079 0.597 1.079 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.972 0.985 qs_energies_init_hamiltonians 11 5.9 0.004 0.018 0.964 0.966 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.714 0.952 mp_allgather_i34 2286 14.5 0.575 0.950 0.575 0.950 acc_transpose_blocks_sync 82296 16.5 0.817 0.939 0.817 0.939 transfer_pw2rs 487 13.2 0.004 0.005 0.909 0.916 mp_alltoall_z22v 1201 16.6 0.741 0.827 0.741 0.827 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.222000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=456.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 512.897024E+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 0.022 0.085 33.220 33.221 qs_mol_dyn_low 1 2.0 0.006 0.028 32.703 32.717 qs_forces 11 3.9 0.010 0.031 32.519 32.520 qs_energies 11 4.9 0.005 0.031 30.904 30.911 scf_env_do_scf 11 5.9 0.002 0.010 25.637 25.637 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 22.935 22.940 dbcsr_multiply_generic 2286 12.5 0.096 0.100 17.286 17.361 qs_scf_new_mos 108 7.5 0.001 0.001 15.329 15.346 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.329 15.345 velocity_verlet 10 3.0 0.022 0.069 15.183 15.187 multiply_cannon 2286 13.5 0.194 0.201 14.046 14.742 ot_scf_mini 108 9.5 0.002 0.003 14.575 14.589 multiply_cannon_loop 2286 14.5 0.859 0.893 13.297 14.007 ot_mini 108 10.5 0.001 0.001 9.075 9.089 qs_ot_get_derivative 108 11.5 0.001 0.001 7.592 7.606 multiply_cannon_multrec 18288 15.5 1.877 2.829 7.300 7.576 dbcsr_mm_accdrv_process 38222 16.0 5.270 6.117 5.329 6.182 rebuild_ks_matrix 119 8.3 0.000 0.000 5.974 5.990 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 5.973 5.989 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.222 5.234 init_scf_run 11 5.9 0.000 0.001 3.971 3.971 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.971 3.971 mp_waitall_1 158411 16.6 2.557 3.643 2.557 3.643 sum_up_and_integrate 119 10.3 0.001 0.003 3.549 3.560 qs_ot_get_p 119 10.4 0.001 0.002 3.530 3.554 integrate_v_rspace 119 11.3 0.003 0.003 3.536 3.549 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.938 3.529 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.239 3.243 calculate_rho_elec 119 8.7 0.031 0.033 3.239 3.242 calculate_first_density_matrix 1 7.0 0.001 0.008 2.739 2.741 init_scf_loop 11 6.9 0.002 0.013 2.683 2.703 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.464 2.442 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.987 2.338 apply_single 119 13.6 0.000 0.000 1.986 2.338 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.175 2.179 calculate_dm_sparse 119 9.5 0.000 0.000 2.027 2.033 make_m2s 4572 13.5 0.043 0.045 1.801 1.933 grid_integrate_task_list 119 12.3 1.815 1.907 1.815 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.893 1.896 pw_transfer 1439 11.6 0.065 0.068 1.877 1.892 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.881 1.883 density_rs2pw 119 9.7 0.004 0.004 1.804 1.882 make_images 4572 14.5 0.191 0.205 1.716 1.846 acc_transpose_blocks 18288 15.5 0.079 0.081 1.721 1.808 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.785 1.801 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.726 1.733 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.679 1.685 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.645 1.654 cp_fm_diag_elpa_base 50 14.0 1.622 1.634 1.644 1.652 prepare_preconditioner 11 7.9 0.000 0.000 1.648 1.650 make_preconditioner 11 8.9 0.000 0.002 1.648 1.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.507 1.590 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.458 1.475 ot_diis_step 108 11.5 0.011 0.011 1.470 1.470 mp_sum_l 7287 12.8 1.072 1.443 1.072 1.443 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.368 1.371 potential_pw2rs 119 12.3 0.007 0.008 1.351 1.359 fft3d_ps 1201 14.6 0.546 0.568 1.340 1.354 grid_collocate_task_list 119 9.7 1.235 1.313 1.235 1.313 multiply_cannon_sync_h2d 18288 15.5 1.065 1.197 1.065 1.197 wfi_extrapolate 11 7.9 0.001 0.003 1.169 1.169 transfer_rs2pw 487 10.6 0.005 0.005 0.992 1.064 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.014 1.021 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.934 0.953 make_images_data 4572 15.5 0.047 0.051 0.773 0.928 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.884 0.895 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.669 0.848 acc_transpose_blocks_sync 54864 16.5 0.738 0.828 0.738 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.824 0.826 mp_alltoall_d11v 2130 13.8 0.640 0.801 0.640 0.801 transfer_pw2rs 487 13.2 0.004 0.004 0.789 0.791 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.008 0.673 0.740 mp_alltoall_z22v 1201 16.6 0.647 0.725 0.647 0.725 cp_fm_cholesky_invert 11 10.9 0.699 0.702 0.699 0.702 arnoldi_extremal 119 11.4 0.002 0.003 0.640 0.694 arnoldi_normal_ev 119 12.4 0.015 0.103 0.638 0.690 jit_kernel_transpose 5 15.6 0.667 0.671 0.667 0.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.221000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=489.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.455616E+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 0.012 0.041 37.817 37.818 qs_mol_dyn_low 1 2.0 0.012 0.046 37.579 37.600 qs_forces 11 3.9 0.002 0.003 37.285 37.290 qs_energies 11 4.9 0.017 0.033 35.531 35.542 scf_env_do_scf 11 5.9 0.000 0.001 29.915 29.917 scf_env_do_scf_inner_loop 108 6.5 0.007 0.034 26.372 26.373 dbcsr_multiply_generic 2286 12.5 0.122 0.144 20.204 20.350 velocity_verlet 10 3.0 0.038 0.124 18.775 18.786 qs_scf_new_mos 108 7.5 0.001 0.001 18.190 18.255 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.190 18.254 ot_scf_mini 108 9.5 0.002 0.003 17.141 17.198 multiply_cannon 2286 13.5 0.218 0.226 16.394 16.737 multiply_cannon_loop 2286 14.5 1.506 1.602 15.426 15.786 ot_mini 108 10.5 0.001 0.001 10.513 10.580 multiply_cannon_multrec 27432 15.5 2.504 3.232 9.160 9.451 qs_ot_get_derivative 108 11.5 0.001 0.001 8.622 8.678 dbcsr_mm_accdrv_process 47916 15.9 5.971 7.759 6.554 7.979 rebuild_ks_matrix 119 8.3 0.000 0.000 6.438 6.484 qs_ks_build_kohn_sham_matrix 119 9.3 0.020 0.081 6.438 6.484 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.715 5.755 init_scf_run 11 5.9 0.000 0.001 4.016 4.017 scf_env_initial_rho_setup 11 6.9 0.038 0.058 4.016 4.016 qs_ot_get_p 119 10.4 0.001 0.002 3.786 3.868 sum_up_and_integrate 119 10.3 0.005 0.033 3.558 3.571 integrate_v_rspace 119 11.3 0.003 0.003 3.543 3.560 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.132 3.534 init_scf_loop 11 6.9 0.000 0.000 3.523 3.524 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.449 3.490 calculate_rho_elec 119 8.7 0.041 0.054 3.449 3.490 prepare_preconditioner 11 7.9 0.000 0.000 2.591 2.604 make_preconditioner 11 8.9 0.000 0.000 2.591 2.604 acc_transpose_blocks 27432 15.5 0.118 0.128 2.382 2.543 calculate_first_density_matrix 1 7.0 0.001 0.009 2.509 2.527 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.197 2.526 mp_waitall_1 137007 16.6 1.943 2.492 1.943 2.492 make_m2s 4572 13.5 0.053 0.057 2.352 2.459 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.358 2.386 make_images 4572 14.5 0.274 0.334 2.245 2.351 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.336 2.347 calculate_dm_sparse 119 9.5 0.000 0.000 2.281 2.336 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.121 2.244 apply_single 119 13.6 0.000 0.000 2.121 2.243 pw_transfer 1439 11.6 0.064 0.068 2.083 2.124 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.992 2.036 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.969 1.972 density_rs2pw 119 9.7 0.004 0.004 1.865 1.965 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.930 1.930 grid_integrate_task_list 119 12.3 1.815 1.891 1.815 1.891 ot_diis_step 108 11.5 0.012 0.012 1.849 1.849 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.732 1.743 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.630 1.675 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.664 1.674 cp_fm_diag_elpa_base 50 14.0 1.628 1.645 1.661 1.671 acc_transpose_blocks_sync 82296 16.5 1.389 1.541 1.389 1.541 fft3d_ps 1201 14.6 0.594 0.655 1.481 1.508 wfi_extrapolate 11 7.9 0.002 0.014 1.408 1.409 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.844 1.389 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.363 1.373 potential_pw2rs 119 12.3 0.009 0.009 1.362 1.370 grid_collocate_task_list 119 9.7 1.249 1.338 1.249 1.338 mp_sum_l 7287 12.8 1.034 1.319 1.034 1.319 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.297 1.317 jit_kernel_multiply 7 16.1 0.516 1.315 0.516 1.315 qs_energies_init_hamiltonians 11 5.9 0.005 0.010 1.255 1.259 cp_fm_upper_to_full 72 14.2 0.817 1.161 0.817 1.161 dbcsr_complete_redistribute 329 12.2 0.149 0.183 0.852 1.128 transfer_rs2pw 487 10.6 0.004 0.005 0.959 1.052 mp_alltoall_d11v 2130 13.8 0.819 1.029 0.819 1.029 make_images_data 4572 15.5 0.048 0.054 0.859 0.972 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.739 0.902 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.631 0.899 acc_transpose_blocks_kernels 27432 16.5 0.269 0.278 0.844 0.882 build_core_hamiltonian_matrix_ 11 4.9 0.005 0.020 0.807 0.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.819 mp_alltoall_z22v 1201 16.6 0.769 0.804 0.769 0.804 cp_fm_cholesky_invert 11 10.9 0.764 0.767 0.764 0.767 qs_env_update_s_mstruct 11 6.9 0.004 0.007 0.717 0.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.818000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.000000, yerr=4.786344 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 606.666752E+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 0.040 0.056 29.443 29.443 qs_mol_dyn_low 1 2.0 0.003 0.003 29.085 29.120 qs_forces 11 3.9 0.002 0.003 29.023 29.023 qs_energies 11 4.9 0.006 0.015 27.316 27.319 scf_env_do_scf 11 5.9 0.000 0.001 22.039 22.039 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.620 19.621 velocity_verlet 10 3.0 0.002 0.003 14.652 14.654 dbcsr_multiply_generic 2286 12.5 0.115 0.119 13.612 13.738 qs_scf_new_mos 108 7.5 0.001 0.001 11.943 11.970 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.942 11.969 multiply_cannon 2286 13.5 0.229 0.238 10.979 11.537 ot_scf_mini 108 9.5 0.002 0.002 11.234 11.259 multiply_cannon_loop 2286 14.5 0.640 0.666 10.056 10.303 multiply_cannon_multrec 9144 15.5 1.742 2.051 6.311 6.635 ot_mini 108 10.5 0.001 0.001 6.569 6.599 rebuild_ks_matrix 119 8.3 0.000 0.000 5.790 5.814 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.789 5.814 qs_ot_get_derivative 108 11.5 0.001 0.001 5.285 5.309 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.148 5.169 dbcsr_mm_accdrv_process 12550 15.8 3.730 4.490 4.462 4.533 init_scf_run 11 5.9 0.000 0.001 3.807 3.807 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.807 3.807 sum_up_and_integrate 119 10.3 0.001 0.001 3.432 3.437 integrate_v_rspace 119 11.3 0.003 0.003 3.422 3.427 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.377 3.382 calculate_rho_elec 119 8.7 0.060 0.061 3.376 3.381 qs_ot_get_p 119 10.4 0.001 0.002 2.785 2.821 calculate_first_density_matrix 1 7.0 0.000 0.000 2.654 2.655 init_scf_loop 11 6.9 0.000 0.000 2.399 2.400 pw_transfer 1439 11.6 0.065 0.067 2.108 2.117 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.015 2.025 jit_kernel_multiply 10 15.4 0.692 2.017 0.692 2.017 make_m2s 4572 13.5 0.033 0.035 1.825 1.986 grid_integrate_task_list 119 12.3 1.845 1.943 1.845 1.943 mp_waitall_1 115863 16.7 1.410 1.937 1.410 1.937 make_images 4572 14.5 0.269 0.301 1.736 1.894 calculate_dm_sparse 119 9.5 0.000 0.000 1.872 1.889 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.840 1.866 density_rs2pw 119 9.7 0.003 0.003 1.767 1.852 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.831 1.833 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.797 1.800 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.679 1.690 prepare_preconditioner 11 7.9 0.000 0.000 1.674 1.678 make_preconditioner 11 8.9 0.000 0.000 1.674 1.678 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.666 1.674 acc_transpose_blocks 9144 15.5 0.043 0.044 1.581 1.616 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.567 1.595 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.556 1.557 fft3d_ps 1201 14.6 0.650 0.662 1.367 1.377 grid_collocate_task_list 119 9.7 1.296 1.350 1.296 1.350 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.338 1.350 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.278 1.286 cp_fm_diag_elpa_base 50 14.0 1.251 1.267 1.276 1.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.281 apply_single 119 13.6 0.000 0.000 1.215 1.280 ot_diis_step 108 11.5 0.013 0.013 1.269 1.269 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.263 potential_pw2rs 119 12.3 0.010 0.010 1.243 1.245 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.244 1.245 wfi_extrapolate 11 7.9 0.001 0.001 1.101 1.101 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.782 0.998 make_images_data 4572 15.5 0.042 0.047 0.779 0.942 mp_alltoall_d11v 2130 13.8 0.800 0.941 0.800 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.926 transfer_rs2pw 487 10.6 0.004 0.005 0.791 0.877 cp_fm_cholesky_invert 11 10.9 0.835 0.838 0.835 0.838 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.816 0.822 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.807 0.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.762 0.764 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.383 0.761 acc_transpose_blocks_sync 27432 16.5 0.715 0.752 0.715 0.752 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.691 0.739 jit_kernel_transpose 5 15.6 0.689 0.703 0.689 0.703 mp_allgather_i34 2286 14.5 0.232 0.659 0.232 0.659 transfer_pw2rs 487 13.2 0.003 0.004 0.649 0.651 mp_alltoall_z22v 1201 16.6 0.592 0.624 0.592 0.624 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.443000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=576.636364, yerr=4.051324 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 758.972416E+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 0.027 0.057 44.930 44.931 qs_mol_dyn_low 1 2.0 0.003 0.003 44.513 44.527 qs_forces 11 3.9 0.002 0.002 44.421 44.421 qs_energies 11 4.9 0.002 0.003 42.380 42.383 scf_env_do_scf 11 5.9 0.001 0.001 36.099 36.099 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.903 27.904 velocity_verlet 10 3.0 0.002 0.002 24.984 24.993 dbcsr_multiply_generic 2286 12.5 0.112 0.117 19.481 19.670 qs_scf_new_mos 108 7.5 0.001 0.001 17.877 17.976 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.876 17.975 ot_scf_mini 108 9.5 0.002 0.002 16.689 16.788 multiply_cannon 2286 13.5 0.294 0.301 15.179 16.172 multiply_cannon_loop 2286 14.5 0.863 0.885 13.901 14.935 ot_mini 108 10.5 0.001 0.001 10.137 10.249 multiply_cannon_multrec 9144 15.5 3.398 4.970 8.841 8.990 qs_ot_get_derivative 108 11.5 0.001 0.001 8.099 8.197 init_scf_loop 11 6.9 0.000 0.000 8.162 8.166 rebuild_ks_matrix 119 8.3 0.000 0.000 7.508 7.646 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.507 7.645 prepare_preconditioner 11 7.9 0.000 0.000 7.101 7.114 make_preconditioner 11 8.9 0.000 0.000 7.101 7.114 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.600 6.994 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.767 6.893 dbcsr_mm_accdrv_process 12550 15.8 4.774 6.589 5.310 6.623 cp_fm_upper_to_full 72 14.2 3.254 4.707 3.254 4.707 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.418 4.424 calculate_rho_elec 119 8.7 0.117 0.120 4.417 4.423 init_scf_run 11 5.9 0.000 0.001 4.140 4.140 scf_env_initial_rho_setup 11 6.9 0.012 0.013 4.140 4.140 sum_up_and_integrate 119 10.3 0.001 0.001 4.012 4.019 integrate_v_rspace 119 11.3 0.004 0.004 4.001 4.008 qs_ot_get_p 119 10.4 0.011 0.020 3.626 3.764 mp_waitall_1 94719 16.7 2.537 3.551 2.537 3.551 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.881 3.342 pw_transfer 1439 11.6 0.068 0.068 3.213 3.220 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.115 3.122 dbcsr_complete_redistribute 329 12.2 0.285 0.294 2.084 2.944 make_m2s 4572 13.5 0.037 0.037 2.638 2.812 make_images 4572 14.5 0.354 0.389 2.517 2.693 fft_wrap_pw1pw2_140 487 13.2 0.648 0.653 2.634 2.643 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.756 2.604 calculate_first_density_matrix 1 7.0 0.000 0.000 2.563 2.564 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.245 2.526 apply_single 119 13.6 0.000 0.000 2.245 2.526 density_rs2pw 119 9.7 0.003 0.003 2.404 2.428 mp_alltoall_i22 627 13.8 1.540 2.418 1.540 2.418 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.496 2.338 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.238 2.289 calculate_dm_sparse 119 9.5 0.000 0.000 2.271 2.286 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.315 2.188 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.158 2.159 grid_integrate_task_list 119 12.3 2.085 2.117 2.085 2.117 fft3d_ps 1201 14.6 0.873 0.895 2.037 2.041 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.016 2.017 ot_diis_step 108 11.5 0.014 0.015 2.012 2.013 acc_transpose_blocks 9144 15.5 0.045 0.046 1.847 1.886 qs_energies_init_hamiltonians 11 5.9 0.006 0.006 1.798 1.799 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.796 1.797 mp_sum_l 7287 12.8 1.108 1.794 1.108 1.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.689 1.715 grid_collocate_task_list 119 9.7 1.524 1.543 1.524 1.543 cp_fm_cholesky_invert 11 10.9 1.516 1.519 1.516 1.519 wfi_extrapolate 11 7.9 0.001 0.001 1.500 1.500 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.225 1.494 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.484 1.484 cp_fm_diag_elpa_base 50 14.0 1.326 1.389 1.481 1.482 make_images_data 4572 15.5 0.046 0.050 1.193 1.447 potential_pw2rs 119 12.3 0.013 0.014 1.442 1.444 mp_alltoall_d11v 2130 13.8 1.318 1.349 1.318 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.322 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.272 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.125 1.146 acc_transpose_blocks_sync 27432 16.5 1.108 1.137 1.108 1.137 mp_alltoall_z22v 1201 16.6 1.026 1.053 1.026 1.053 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.999 1.052 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.949 0.962 qs_create_task_list 11 7.9 0.003 0.005 0.936 0.947 generate_qs_task_list 11 8.9 0.366 0.385 0.933 0.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.931000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.727273, yerr=12.613780 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 495.489024E+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 1022478. 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 0.014 0.032 84.819 84.820 qs_mol_dyn_low 1 2.0 0.009 0.053 84.316 84.374 qs_forces 11 3.9 0.003 0.005 84.086 84.088 qs_energies 11 4.9 0.002 0.009 81.224 81.238 scf_env_do_scf 11 5.9 0.000 0.001 70.502 70.504 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.936 64.936 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.593 51.996 qs_scf_new_mos 99 7.5 0.000 0.001 47.889 48.035 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.889 48.034 ot_scf_mini 99 9.5 0.002 0.002 45.486 45.645 multiply_cannon 2055 13.4 0.181 0.187 43.018 44.212 velocity_verlet 10 3.0 0.013 0.097 43.828 43.830 multiply_cannon_loop 2055 14.4 1.782 1.819 42.041 43.300 ot_mini 99 10.5 0.001 0.001 27.056 27.215 qs_ot_get_derivative 99 11.5 0.001 0.001 20.205 20.365 multiply_cannon_multrec 49320 15.4 11.374 11.949 17.514 18.210 rebuild_ks_matrix 110 8.3 0.000 0.000 14.501 14.685 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.027 14.501 14.684 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.702 12.869 mp_waitall_1 220248 16.4 11.089 12.204 11.089 12.204 multiply_cannon_sync_h2d 49320 15.4 9.592 10.264 9.592 10.264 qs_ot_get_p 110 10.4 0.001 0.001 9.865 10.031 init_scf_run 11 5.9 0.000 0.001 8.514 8.514 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.514 8.514 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.740 8.289 multiply_cannon_metrocomm3 49320 15.4 0.082 0.086 6.701 8.149 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.265 7.902 apply_single 110 13.6 0.000 0.000 7.265 7.902 sum_up_and_integrate 110 10.3 0.002 0.003 7.001 7.022 integrate_v_rspace 110 11.3 0.003 0.003 6.976 7.001 ot_diis_step 99 11.5 0.005 0.006 6.624 6.624 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.569 6.593 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.339 6.510 calculate_rho_elec 110 8.6 0.020 0.025 6.338 6.510 dbcsr_mm_accdrv_process 87628 16.1 3.069 3.137 6.008 6.294 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.761 5.761 init_scf_loop 11 6.9 0.000 0.001 5.539 5.540 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.393 5.481 mp_sum_l 6594 12.7 4.241 5.352 4.241 5.352 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.180 5.203 cp_fm_diag_elpa_base 48 14.0 5.167 5.189 5.178 5.202 calculate_first_density_matrix 1 7.0 0.000 0.003 4.367 4.377 wfi_extrapolate 11 7.9 0.003 0.018 4.043 4.043 make_m2s 4110 13.4 0.060 0.064 3.877 4.002 make_images 4110 14.4 0.177 0.189 3.782 3.911 calculate_dm_sparse 110 9.5 0.001 0.001 3.799 3.905 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 2.555 3.768 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.633 3.638 density_rs2pw 110 9.6 0.004 0.004 3.325 3.462 grid_integrate_task_list 110 12.3 3.247 3.425 3.247 3.425 prepare_preconditioner 11 7.9 0.000 0.000 3.381 3.401 make_preconditioner 11 8.9 0.000 0.000 3.381 3.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.281 3.343 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.261 3.314 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.155 3.199 pw_transfer 1331 11.6 0.054 0.064 3.093 3.175 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.005 3.090 parallel_gemm_fm 81 9.0 0.000 0.000 2.934 2.938 parallel_gemm_fm_cosma 81 10.0 2.934 2.938 2.934 2.938 jit_kernel_multiply 13 15.9 2.659 2.736 2.659 2.736 acc_transpose_blocks 49320 15.4 0.208 0.214 2.590 2.706 make_basis_sm 11 9.8 0.000 0.000 2.677 2.679 fft_wrap_pw1pw2_140 451 13.1 0.309 0.331 2.566 2.676 potential_pw2rs 110 12.3 0.005 0.006 2.609 2.633 mp_alltoall_d11v 2046 13.8 2.105 2.438 2.105 2.438 fft3d_ps 1111 14.6 0.797 0.887 2.299 2.354 grid_collocate_task_list 110 9.6 2.157 2.276 2.157 2.276 mp_waitany 14300 13.8 1.828 2.080 1.828 2.080 mp_sum_d 3891 11.9 1.388 2.033 1.388 2.033 transfer_rs2pw 451 10.6 0.005 0.006 1.875 1.992 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.929 1.952 cp_fm_cholesky_invert 11 10.9 1.873 1.877 1.873 1.877 make_images_data 4110 15.4 0.043 0.048 1.740 1.871 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.499 1.785 transfer_pw2rs 451 13.1 0.006 0.007 1.770 1.784 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.670 1.701 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.820000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=469.454545, yerr=3.627261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.431936E+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 1832379. 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 0.041 0.056 70.792 70.792 qs_mol_dyn_low 1 2.0 0.003 0.003 70.234 70.243 qs_forces 11 3.9 0.003 0.005 70.160 70.164 qs_energies 11 4.9 0.011 0.028 66.845 66.854 scf_env_do_scf 11 5.9 0.000 0.001 56.556 56.560 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.832 48.833 dbcsr_multiply_generic 2055 12.4 0.127 0.146 37.980 38.170 velocity_verlet 10 3.0 0.004 0.013 35.913 35.914 qs_scf_new_mos 99 7.5 0.001 0.001 33.070 33.182 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.069 33.181 multiply_cannon 2055 13.4 0.224 0.241 31.370 32.412 ot_scf_mini 99 9.5 0.003 0.003 31.401 31.523 multiply_cannon_loop 2055 14.4 1.158 1.180 30.236 30.934 ot_mini 99 10.5 0.001 0.001 18.310 18.435 multiply_cannon_multrec 24660 15.4 7.014 8.580 14.193 15.806 rebuild_ks_matrix 110 8.3 0.000 0.000 13.467 13.572 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.466 13.572 qs_ot_get_derivative 99 11.5 0.001 0.001 12.560 12.683 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.841 11.935 mp_waitall_1 176588 16.5 7.442 9.926 7.442 9.926 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.088 8.107 init_scf_run 11 5.9 0.000 0.001 7.853 7.853 scf_env_initial_rho_setup 11 6.9 0.005 0.008 7.852 7.853 init_scf_loop 11 6.9 0.000 0.000 7.690 7.690 multiply_cannon_sync_h2d 24660 15.4 6.387 7.501 6.387 7.501 dbcsr_mm_accdrv_process 52282 16.1 5.584 6.365 7.007 7.353 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.419 7.006 apply_single 110 13.6 0.000 0.001 6.418 7.006 qs_ot_get_p 110 10.4 0.009 0.034 6.595 6.742 sum_up_and_integrate 110 10.3 0.002 0.003 6.332 6.342 integrate_v_rspace 110 11.3 0.003 0.003 6.305 6.317 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.854 5.863 calculate_rho_elec 110 8.6 0.039 0.047 5.853 5.863 ot_diis_step 99 11.5 0.010 0.010 5.702 5.703 prepare_preconditioner 11 7.9 0.000 0.000 5.560 5.586 make_preconditioner 11 8.9 0.000 0.000 5.560 5.586 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.719 5.461 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.109 5.256 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.566 4.587 make_m2s 4110 13.4 0.056 0.059 4.207 4.534 make_images 4110 14.4 0.407 0.467 4.095 4.418 calculate_first_density_matrix 1 7.0 0.000 0.000 4.254 4.260 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.101 4.101 pw_transfer 1331 11.6 0.065 0.074 3.544 3.692 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.438 3.590 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.531 3.541 cp_fm_diag_elpa_base 48 14.0 3.484 3.508 3.529 3.539 wfi_extrapolate 11 7.9 0.001 0.001 3.497 3.497 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.452 3.454 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.293 3.365 density_rs2pw 110 9.6 0.004 0.004 3.110 3.345 grid_integrate_task_list 110 12.3 3.152 3.327 3.152 3.327 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.105 3.142 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.940 3.094 calculate_dm_sparse 110 9.5 0.001 0.001 2.992 3.018 fft3d_ps 1111 14.6 1.117 1.356 2.531 2.693 make_images_data 4110 15.4 0.049 0.054 2.284 2.625 cp_fm_cholesky_invert 11 10.9 2.573 2.580 2.573 2.580 hybrid_alltoall_any 4261 16.3 0.105 0.455 1.957 2.564 mp_sum_l 6594 12.7 1.756 2.527 1.756 2.527 make_basis_sm 11 9.8 0.000 0.000 2.463 2.466 parallel_gemm_fm 81 9.0 0.000 0.000 2.432 2.440 parallel_gemm_fm_cosma 81 10.0 2.432 2.439 2.432 2.439 grid_collocate_task_list 110 9.6 2.168 2.334 2.168 2.334 potential_pw2rs 110 12.3 0.008 0.009 2.219 2.228 acc_transpose_blocks 24660 15.4 0.116 0.120 2.064 2.106 mp_alltoall_d11v 2046 13.8 1.762 2.089 1.762 2.089 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.981 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.886 1.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.782 1.791 jit_kernel_multiply 9 16.4 1.053 1.768 1.053 1.768 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.767 1.677 transfer_rs2pw 451 10.6 0.006 0.007 1.438 1.625 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.558 1.569 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.489 mp_irecv_dv 57340 16.2 0.639 1.458 0.639 1.458 mp_allgather_i34 2055 14.4 0.441 1.448 0.441 1.448 dbcsr_complete_redistribute 325 12.2 0.235 0.287 1.181 1.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.792000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.909091, yerr=8.338896 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 658.808832E+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 0.032 0.067 60.102 60.108 qs_mol_dyn_low 1 2.0 0.008 0.022 59.650 59.671 qs_forces 11 3.9 0.003 0.003 59.478 59.480 qs_energies 11 4.9 0.005 0.032 56.308 56.321 scf_env_do_scf 11 5.9 0.001 0.001 47.903 47.903 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.530 39.530 velocity_verlet 10 3.0 0.024 0.091 32.262 32.270 dbcsr_multiply_generic 2055 12.4 0.126 0.131 28.675 28.895 qs_scf_new_mos 99 7.5 0.001 0.001 24.980 25.078 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.980 25.077 ot_scf_mini 99 9.5 0.002 0.003 23.737 23.845 multiply_cannon 2055 13.4 0.211 0.221 22.517 23.657 multiply_cannon_loop 2055 14.4 0.813 0.838 21.373 22.436 ot_mini 99 10.5 0.001 0.001 13.594 13.706 rebuild_ks_matrix 110 8.3 0.000 0.000 12.089 12.206 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.018 12.088 12.205 multiply_cannon_multrec 16440 15.4 3.738 5.039 9.986 11.247 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.634 10.740 mp_waitall_1 139946 16.5 6.761 9.819 6.761 9.819 qs_ot_get_derivative 99 11.5 0.001 0.001 9.207 9.316 init_scf_loop 11 6.9 0.001 0.003 8.338 8.339 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.207 7.132 prepare_preconditioner 11 7.9 0.000 0.000 6.607 6.622 make_preconditioner 11 8.9 0.000 0.001 6.607 6.622 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.946 6.283 sum_up_and_integrate 110 10.3 0.002 0.003 6.236 6.252 integrate_v_rspace 110 11.3 0.003 0.003 6.210 6.226 dbcsr_mm_accdrv_process 34862 16.1 5.377 5.786 6.093 6.206 init_scf_run 11 5.9 0.000 0.001 5.801 5.801 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.801 5.801 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.690 5.699 calculate_rho_elec 110 8.6 0.059 0.060 5.690 5.699 qs_ot_get_p 110 10.4 0.001 0.002 5.336 5.483 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.877 5.302 apply_single 110 13.6 0.000 0.000 4.877 5.301 make_m2s 4110 13.4 0.049 0.051 4.131 4.505 make_images 4110 14.4 0.399 0.525 4.015 4.389 ot_diis_step 99 11.5 0.011 0.011 4.357 4.357 multiply_cannon_sync_h2d 16440 15.4 3.268 3.928 3.268 3.928 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.188 3.788 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.739 3.743 pw_transfer 1331 11.6 0.065 0.073 3.574 3.585 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.467 3.481 grid_integrate_task_list 110 12.3 3.178 3.395 3.178 3.395 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.372 3.373 density_rs2pw 110 9.6 0.004 0.004 2.939 3.141 fft_wrap_pw1pw2_140 451 13.1 0.460 0.473 3.011 3.026 wfi_extrapolate 11 7.9 0.002 0.010 2.965 2.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.870 2.871 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.840 2.851 cp_fm_diag_elpa_base 48 14.0 2.773 2.806 2.838 2.850 make_images_data 4110 15.4 0.047 0.051 2.331 2.815 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.126 2.774 calculate_first_density_matrix 1 7.0 0.001 0.004 2.745 2.748 calculate_dm_sparse 110 9.5 0.001 0.001 2.567 2.598 cp_fm_cholesky_invert 11 10.9 2.585 2.590 2.585 2.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.524 2.586 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.405 2.466 fft3d_ps 1111 14.6 1.123 1.148 2.384 2.392 grid_collocate_task_list 110 9.6 2.216 2.385 2.216 2.385 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.868 2.381 mp_irecv_dv 48980 15.7 0.792 2.243 0.792 2.243 qs_energies_init_hamiltonians 11 5.9 0.010 0.025 2.132 2.135 mp_alltoall_d11v 2046 13.8 1.758 2.098 1.758 2.098 potential_pw2rs 110 12.3 0.011 0.011 2.092 2.097 mp_sum_l 6594 12.7 1.425 2.076 1.425 2.076 dbcsr_complete_redistribute 325 12.2 0.330 0.372 1.425 1.903 cp_fm_upper_to_full 70 14.2 1.393 1.761 1.393 1.761 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.647 1.660 acc_transpose_blocks 16440 15.4 0.078 0.081 1.605 1.648 cp_fm_cholesky_decompose 22 10.9 1.523 1.546 1.523 1.546 mp_allgather_i34 2055 14.4 0.457 1.525 0.457 1.525 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.506 1.518 transfer_rs2pw 451 10.6 0.005 0.006 1.299 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.012 1.368 1.480 mp_waitany 17072 13.8 1.180 1.457 1.180 1.457 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.969 1.420 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.186 1.276 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.261 1.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.108000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.090909, yerr=7.585327 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 734.105600E+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 0.078 0.158 66.328 66.333 qs_mol_dyn_low 1 2.0 0.015 0.057 65.677 65.722 qs_forces 11 3.9 0.004 0.006 65.545 65.564 qs_energies 11 4.9 0.004 0.010 62.111 62.130 scf_env_do_scf 11 5.9 0.001 0.002 53.517 53.520 scf_env_do_scf_inner_loop 99 6.5 0.010 0.029 41.806 41.807 velocity_verlet 10 3.0 0.003 0.005 36.986 37.006 dbcsr_multiply_generic 2055 12.4 0.118 0.125 30.643 30.860 qs_scf_new_mos 99 7.5 0.001 0.001 27.246 27.343 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.246 27.342 ot_scf_mini 99 9.5 0.003 0.004 25.567 25.672 multiply_cannon 2055 13.4 0.238 0.257 23.588 24.940 multiply_cannon_loop 2055 14.4 1.395 1.449 22.169 22.746 ot_mini 99 10.5 0.001 0.001 14.722 14.845 multiply_cannon_multrec 24660 15.4 4.089 6.631 13.111 14.363 rebuild_ks_matrix 110 8.3 0.000 0.000 11.994 12.110 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 11.994 12.110 init_scf_loop 11 6.9 0.004 0.019 11.673 11.675 qs_ot_get_derivative 99 11.5 0.001 0.001 10.559 10.661 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.552 10.654 dbcsr_mm_accdrv_process 52304 16.0 7.703 8.915 8.864 9.799 prepare_preconditioner 11 7.9 0.000 0.000 9.724 9.742 make_preconditioner 11 8.9 0.000 0.001 9.724 9.742 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.943 9.419 sum_up_and_integrate 110 10.3 0.002 0.003 6.216 6.233 integrate_v_rspace 110 11.3 0.003 0.003 6.190 6.206 qs_ot_get_p 110 10.4 0.001 0.002 5.888 6.038 mp_waitall_1 121746 16.5 4.160 5.986 4.160 5.986 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.792 5.803 calculate_rho_elec 110 8.6 0.078 0.081 5.791 5.803 make_m2s 4110 13.4 0.059 0.062 5.422 5.793 init_scf_run 11 5.9 0.000 0.001 5.697 5.699 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.697 5.699 make_images 4110 14.4 0.577 0.701 5.283 5.650 cp_fm_upper_to_full 70 14.2 3.366 4.885 3.366 4.885 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.049 4.145 apply_single 110 13.6 0.000 0.000 4.049 4.145 ot_diis_step 99 11.5 0.011 0.011 4.124 4.124 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.874 3.890 pw_transfer 1331 11.6 0.065 0.071 3.741 3.783 dbcsr_complete_redistribute 325 12.2 0.422 0.482 2.640 3.773 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.635 3.680 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.507 3.571 grid_integrate_task_list 110 12.3 3.301 3.427 3.301 3.427 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.297 3.354 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.318 3.318 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.121 3.242 fft_wrap_pw1pw2_140 451 13.1 0.530 0.541 3.157 3.205 hybrid_alltoall_any 4261 16.3 0.123 0.459 2.248 3.119 make_images_data 4110 15.4 0.049 0.053 2.724 3.077 density_rs2pw 110 9.6 0.004 0.004 2.901 3.051 multiply_cannon_metrocomm3 24660 15.4 0.037 0.040 1.403 3.039 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.010 wfi_extrapolate 11 7.9 0.002 0.013 2.978 2.978 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.884 mp_alltoall_i22 605 13.7 1.680 2.870 1.680 2.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.860 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.800 2.811 cp_fm_diag_elpa_base 48 14.0 2.654 2.714 2.798 2.809 calculate_first_density_matrix 1 7.0 0.001 0.003 2.624 2.628 cp_fm_cholesky_invert 11 10.9 2.557 2.565 2.557 2.565 multiply_cannon_sync_h2d 24660 15.4 2.393 2.544 2.393 2.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.503 2.541 fft3d_ps 1111 14.6 1.151 1.189 2.462 2.483 acc_transpose_blocks 24660 15.4 0.111 0.116 2.388 2.480 grid_collocate_task_list 110 9.6 2.262 2.461 2.262 2.461 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.400 2.419 potential_pw2rs 110 12.3 0.012 0.013 2.027 2.037 mp_alltoall_d11v 2046 13.8 1.724 1.860 1.724 1.860 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.781 1.812 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.710 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.685 1.698 mp_allgather_i34 2055 14.4 0.489 1.671 0.489 1.671 cp_fm_cholesky_decompose 22 10.9 1.606 1.644 1.606 1.644 mp_sum_l 6594 12.7 1.001 1.595 1.001 1.595 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.552 jit_kernel_multiply 8 16.0 0.811 1.550 0.811 1.550 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.858 1.513 acc_transpose_blocks_sync 73980 16.4 1.356 1.472 1.356 1.472 qs_env_update_s_mstruct 11 6.9 0.006 0.014 1.339 1.445 transfer_rs2pw 451 10.6 0.005 0.006 1.230 1.435 mp_irecv_dv 62702 16.1 0.751 1.426 0.751 1.426 mp_waitany 13376 13.8 1.106 1.361 1.106 1.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.333000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.636364, yerr=10.789496 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 856.760320E+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 0.019 0.032 58.615 58.615 qs_mol_dyn_low 1 2.0 0.003 0.003 58.244 58.269 qs_forces 11 3.9 0.019 0.026 58.042 58.043 qs_energies 11 4.9 0.001 0.002 54.370 54.378 scf_env_do_scf 11 5.9 0.000 0.001 44.347 44.347 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.260 36.261 velocity_verlet 10 3.0 0.006 0.007 31.823 31.835 dbcsr_multiply_generic 2055 12.4 0.117 0.128 23.559 23.812 qs_scf_new_mos 99 7.5 0.001 0.001 21.487 21.569 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.486 21.568 ot_scf_mini 99 9.5 0.002 0.002 20.222 20.291 multiply_cannon 2055 13.4 0.241 0.253 17.876 18.934 multiply_cannon_loop 2055 14.4 0.604 0.625 16.598 17.012 rebuild_ks_matrix 110 8.3 0.000 0.000 11.708 11.787 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.707 11.786 ot_mini 99 10.5 0.001 0.001 11.072 11.141 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.424 10.495 multiply_cannon_multrec 8220 15.4 3.220 4.859 7.699 8.908 init_scf_loop 11 6.9 0.000 0.000 8.036 8.038 mp_waitall_1 103326 16.6 5.934 7.621 5.934 7.621 qs_ot_get_derivative 99 11.5 0.001 0.001 7.266 7.333 init_scf_run 11 5.9 0.000 0.001 6.688 6.688 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.688 6.688 prepare_preconditioner 11 7.9 0.000 0.000 6.285 6.295 make_preconditioner 11 8.9 0.000 0.000 6.285 6.295 sum_up_and_integrate 110 10.3 0.001 0.002 6.278 6.292 integrate_v_rspace 110 11.3 0.003 0.003 6.251 6.265 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.126 6.145 calculate_rho_elec 110 8.6 0.113 0.114 6.125 6.144 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.867 5.942 qs_ot_get_p 110 10.4 0.023 0.065 5.272 5.353 dbcsr_mm_accdrv_process 17442 15.9 3.136 4.048 4.337 5.310 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.114 4.677 make_m2s 4110 13.4 0.037 0.038 4.241 4.433 make_images 4110 14.4 0.645 0.706 4.110 4.299 pw_transfer 1331 11.6 0.065 0.070 4.149 4.159 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.041 4.055 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.814 3.884 apply_single 110 13.6 0.000 0.000 3.813 3.884 calculate_first_density_matrix 1 7.0 0.000 0.000 3.846 3.847 ot_diis_step 99 11.5 0.012 0.012 3.777 3.778 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.725 3.729 grid_integrate_task_list 110 12.3 3.371 3.549 3.371 3.549 fft_wrap_pw1pw2_140 451 13.1 0.718 0.731 3.479 3.493 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.372 3.373 density_rs2pw 110 9.6 0.004 0.004 3.085 3.151 cp_fm_cholesky_invert 11 10.9 3.094 3.098 3.094 3.098 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.781 2.789 cp_fm_diag_elpa_base 48 14.0 2.725 2.750 2.780 2.787 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.749 2.753 wfi_extrapolate 11 7.9 0.001 0.001 2.743 2.743 hybrid_alltoall_any 4261 16.3 0.199 0.847 2.257 2.657 fft3d_ps 1111 14.6 1.293 1.298 2.586 2.607 calculate_dm_sparse 110 9.5 0.001 0.001 2.553 2.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.540 2.542 make_images_data 4110 15.4 0.041 0.047 2.225 2.526 multiply_cannon_sync_h2d 8220 15.4 2.344 2.512 2.344 2.512 grid_collocate_task_list 110 9.6 2.367 2.435 2.367 2.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.157 2.183 make_basis_sm 11 9.8 0.000 0.000 2.136 2.138 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.084 2.134 parallel_gemm_fm 81 9.0 0.000 0.000 2.113 2.115 parallel_gemm_fm_cosma 81 10.0 2.112 2.115 2.112 2.115 potential_pw2rs 110 12.3 0.015 0.015 2.063 2.067 mp_alltoall_d11v 2046 13.8 1.814 2.061 1.814 2.061 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.003 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.963 1.998 cp_fm_cholesky_decompose 22 10.9 1.726 1.738 1.726 1.738 qs_env_update_s_mstruct 11 6.9 0.005 0.013 1.573 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.615 1.623 dbcsr_complete_redistribute 325 12.2 0.551 0.577 1.481 1.567 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.752 1.512 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.495 1.509 mp_allgather_i34 2055 14.4 0.446 1.496 0.446 1.496 acc_transpose_blocks 8220 15.4 0.041 0.042 1.393 1.445 jit_kernel_multiply 8 15.6 0.880 1.409 0.880 1.409 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.480 1.390 mp_irecv_dv 24056 15.7 0.453 1.342 0.453 1.342 qs_create_task_list 11 7.9 0.005 0.012 1.226 1.328 generate_qs_task_list 11 8.9 0.374 0.442 1.221 1.318 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.257 1.295 transfer_rs2pw 451 10.6 0.005 0.005 1.175 1.256 mp_waitany 9240 13.8 1.088 1.182 1.088 1.182 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.615000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.545455, yerr=12.942977 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 1.373176E+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 0.073 0.128 88.283 88.284 qs_mol_dyn_low 1 2.0 0.003 0.003 87.573 87.595 qs_forces 11 3.9 0.003 0.003 87.506 87.506 qs_energies 11 4.9 0.001 0.002 83.363 83.365 scf_env_do_scf 11 5.9 0.000 0.001 72.864 72.864 velocity_verlet 10 3.0 0.007 0.009 55.886 55.892 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.548 44.549 dbcsr_multiply_generic 2055 12.4 0.123 0.128 29.443 29.546 init_scf_loop 11 6.9 0.000 0.000 28.246 28.247 qs_scf_new_mos 99 7.5 0.001 0.001 26.666 26.703 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.665 26.702 prepare_preconditioner 11 7.9 0.000 0.000 26.228 26.236 make_preconditioner 11 8.9 0.000 0.000 26.227 26.236 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.376 25.710 ot_scf_mini 99 9.5 0.002 0.002 24.868 24.894 multiply_cannon 2055 13.4 0.329 0.354 22.447 23.212 multiply_cannon_loop 2055 14.4 0.823 0.849 20.675 20.972 cp_fm_upper_to_full 70 14.2 12.802 18.379 12.802 18.379 ot_mini 99 10.5 0.001 0.001 14.060 14.090 rebuild_ks_matrix 110 8.3 0.000 0.000 13.936 13.973 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.936 13.972 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.639 12.672 dbcsr_complete_redistribute 325 12.2 1.008 1.043 7.411 10.760 multiply_cannon_multrec 8220 15.4 4.082 4.268 9.818 9.953 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.410 9.756 qs_ot_get_derivative 99 11.5 0.001 0.001 9.496 9.524 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.836 9.151 mp_alltoall_i22 605 13.7 5.463 8.838 5.463 8.838 mp_waitall_1 84994 16.7 7.372 8.159 7.372 8.159 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.457 7.495 calculate_rho_elec 110 8.6 0.223 0.224 7.456 7.495 sum_up_and_integrate 110 10.3 0.002 0.002 6.911 6.927 integrate_v_rspace 110 11.3 0.004 0.004 6.883 6.899 init_scf_run 11 5.9 0.000 0.001 6.206 6.206 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.206 6.206 dbcsr_mm_accdrv_process 11614 15.7 3.881 4.181 5.587 5.834 pw_transfer 1331 11.6 0.074 0.075 5.807 5.814 make_m2s 4110 13.4 0.043 0.043 5.257 5.747 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.690 5.697 qs_ot_get_p 110 10.4 0.002 0.002 5.533 5.564 make_images 4110 14.4 0.885 0.940 5.069 5.557 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.693 5.164 apply_single 110 13.6 0.000 0.000 4.693 5.164 cp_fm_cholesky_invert 11 10.9 5.117 5.122 5.117 5.122 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.794 5.037 fft_wrap_pw1pw2_140 451 13.1 1.367 1.372 5.000 5.009 ot_diis_step 99 11.5 0.015 0.016 4.547 4.547 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.976 3.986 density_rs2pw 110 9.6 0.004 0.004 3.926 3.947 grid_integrate_task_list 110 12.3 3.675 3.728 3.675 3.728 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 3.686 3.687 hybrid_alltoall_any 4261 16.3 0.264 0.569 2.768 3.488 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.459 3.460 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.992 3.433 fft3d_ps 1111 14.6 1.883 1.896 3.394 3.399 make_images_data 4110 15.4 0.045 0.048 2.740 3.343 wfi_extrapolate 11 7.9 0.001 0.001 3.318 3.318 calculate_dm_sparse 110 9.5 0.001 0.001 3.205 3.226 multiply_cannon_sync_h2d 8220 15.4 3.125 3.157 3.125 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.919 2.921 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.880 2.881 cp_fm_diag_elpa_base 48 14.0 2.347 2.543 2.879 2.879 calculate_first_density_matrix 1 7.0 0.000 0.000 2.765 2.766 grid_collocate_task_list 110 9.6 2.667 2.697 2.667 2.697 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.636 2.652 potential_pw2rs 110 12.3 0.021 0.021 2.442 2.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.405 2.426 qs_env_update_s_mstruct 11 6.9 0.004 0.009 2.197 2.256 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.181 mp_alltoall_d11v 2046 13.8 2.011 2.075 2.011 2.075 cp_fm_cholesky_decompose 22 10.9 1.942 1.973 1.942 1.973 qs_create_task_list 11 7.9 0.001 0.002 1.872 1.916 generate_qs_task_list 11 8.9 0.728 0.780 1.871 1.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.899 acc_transpose_blocks 8220 15.4 0.042 0.042 1.751 1.785 jit_kernel_multiply 10 15.2 1.508 1.774 1.508 1.774 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.732 1.772 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.284000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1239.090909, yerr=47.511634 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 622.768128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 3201307. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.055 208.257 208.258 qs_mol_dyn_low 1 2.0 0.003 0.003 207.620 207.634 qs_forces 11 3.9 0.005 0.006 207.528 207.529 qs_energies 11 4.9 0.003 0.006 201.929 201.951 scf_env_do_scf 11 5.9 0.001 0.001 184.934 184.937 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 164.353 164.355 dbcsr_multiply_generic 2485 12.5 0.177 0.181 126.529 127.219 qs_scf_new_mos 116 7.6 0.001 0.001 125.193 125.514 qs_scf_loop_do_ot 116 8.6 0.001 0.001 125.192 125.513 velocity_verlet 10 3.0 0.002 0.011 122.731 122.732 ot_scf_mini 116 9.6 0.003 0.003 118.517 118.851 multiply_cannon 2485 13.5 0.231 0.242 101.707 103.789 multiply_cannon_loop 2485 14.5 2.389 2.451 99.502 101.417 ot_mini 116 10.6 0.001 0.001 67.200 67.542 multiply_cannon_multrec 59640 15.5 31.667 33.467 41.478 43.611 qs_ot_get_derivative 116 11.6 0.001 0.001 42.376 42.714 rebuild_ks_matrix 127 8.3 0.001 0.001 33.416 33.842 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.415 33.841 mp_waitall_1 264810 16.5 29.594 32.906 29.594 32.906 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.983 30.376 qs_ot_get_p 127 10.4 0.001 0.002 29.027 29.420 multiply_cannon_sync_h2d 59640 15.5 25.971 28.524 25.971 28.524 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.349 25.149 apply_single 127 13.6 0.001 0.001 24.349 25.149 ot_diis_step 116 11.6 0.007 0.008 24.536 24.537 qs_ot_p2m_diag 82 11.4 0.078 0.090 22.380 22.459 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 20.667 20.911 init_scf_loop 11 6.9 0.000 0.000 20.505 20.507 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.566 19.567 multiply_cannon_metrocomm3 59640 15.5 0.115 0.121 16.440 19.404 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.565 16.602 cp_fm_diag_elpa_base 82 14.4 16.491 16.531 16.561 16.599 prepare_preconditioner 11 7.9 0.000 0.000 15.906 15.967 make_preconditioner 11 8.9 0.000 0.000 15.906 15.967 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.136 15.292 make_m2s 4970 13.5 0.103 0.112 14.106 14.551 make_images 4970 14.5 0.393 0.412 13.923 14.378 sum_up_and_integrate 127 10.3 0.002 0.005 13.905 13.921 integrate_v_rspace 127 11.3 0.003 0.004 13.848 13.866 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.094 13.245 calculate_rho_elec 127 8.7 0.045 0.063 13.094 13.244 init_scf_run 11 5.9 0.000 0.001 12.764 12.765 scf_env_initial_rho_setup 11 6.9 0.003 0.009 12.764 12.765 mp_sum_l 7884 12.9 9.486 11.284 9.486 11.284 dbcsr_mm_accdrv_process 123452 16.2 4.746 4.862 9.384 9.856 wfi_extrapolate 11 7.9 0.001 0.001 9.136 9.137 cp_fm_cholesky_invert 11 10.9 9.049 9.057 9.049 9.057 multiply_cannon_metrocomm1 59640 15.5 0.094 0.099 6.589 8.607 calculate_dm_sparse 127 9.5 0.001 0.001 8.481 8.572 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.170 8.242 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.037 8.234 pw_transfer 1535 11.6 0.072 0.084 7.642 7.829 make_images_data 4970 15.5 0.066 0.073 6.878 7.797 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.442 7.626 grid_integrate_task_list 127 12.3 6.973 7.548 6.973 7.548 density_rs2pw 127 9.7 0.006 0.006 6.734 7.369 hybrid_alltoall_any 5155 16.4 0.289 2.231 5.989 7.131 fft_wrap_pw1pw2_140 519 13.2 0.850 0.889 6.531 6.701 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.640 6.650 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.765 5.860 fft3d_ps 1281 14.7 2.179 2.829 5.466 5.794 mp_alltoall_d11v 2401 14.1 4.430 5.342 4.430 5.342 grid_collocate_task_list 127 9.7 4.803 5.241 4.803 5.241 cp_fm_cholesky_decompose 22 10.9 4.680 4.694 4.680 4.694 potential_pw2rs 127 12.3 0.009 0.010 4.448 4.463 mp_sum_d 4456 12.1 3.451 4.234 3.451 4.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.258000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=588.636364, yerr=5.431771 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 832.823296E+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 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5191244. 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 0.120 0.244 193.497 193.519 qs_mol_dyn_low 1 2.0 0.021 0.069 191.942 191.954 qs_forces 11 3.9 0.004 0.004 191.038 191.058 qs_energies 11 4.9 0.006 0.040 184.283 184.313 scf_env_do_scf 11 5.9 0.001 0.001 166.938 166.947 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 134.045 134.047 velocity_verlet 10 3.0 0.037 0.086 120.410 120.429 dbcsr_multiply_generic 2507 12.6 0.194 0.228 97.282 98.549 qs_scf_new_mos 117 7.6 0.001 0.001 95.537 96.102 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.536 96.102 ot_scf_mini 117 9.6 0.004 0.004 90.734 91.359 multiply_cannon 2507 13.6 0.504 0.572 77.073 81.801 multiply_cannon_loop 2507 14.6 1.567 1.638 74.050 77.213 ot_mini 117 10.6 0.001 0.001 50.322 50.873 multiply_cannon_multrec 30084 15.6 21.025 26.553 31.790 37.606 mp_waitall_1 214728 16.6 23.818 37.351 23.818 37.351 rebuild_ks_matrix 128 8.3 0.001 0.001 32.175 32.865 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 32.174 32.864 init_scf_loop 11 6.9 0.000 0.000 32.802 32.803 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.934 29.562 multiply_cannon_metrocomm3 30084 15.6 0.098 0.104 15.547 29.310 qs_ot_get_derivative 117 11.6 0.001 0.002 28.306 28.930 prepare_preconditioner 11 7.9 0.000 0.000 28.510 28.577 make_preconditioner 11 8.9 0.000 0.000 28.510 28.577 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.204 27.748 qs_ot_get_p 128 10.4 0.001 0.002 22.527 23.246 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.103 23.240 apply_single 128 13.6 0.001 0.001 22.103 23.240 ot_diis_step 117 11.6 0.014 0.015 21.842 21.844 multiply_cannon_sync_h2d 30084 15.6 18.068 20.443 18.068 20.443 qs_ot_p2m_diag 83 11.4 0.190 0.222 17.573 17.610 cp_fm_cholesky_invert 11 10.9 16.644 16.656 16.644 16.656 cp_dbcsr_syevd 83 12.4 0.006 0.006 16.342 16.344 make_m2s 5014 13.6 0.086 0.091 14.208 15.413 make_images 5014 14.6 1.172 1.370 13.993 15.199 sum_up_and_integrate 128 10.3 0.002 0.005 14.037 14.067 integrate_v_rspace 128 11.3 0.003 0.004 13.977 14.009 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.404 13.440 calculate_rho_elec 128 8.7 0.087 0.104 13.404 13.439 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.083 13.117 cp_fm_diag_elpa_base 83 14.4 12.826 12.931 13.077 13.106 init_scf_run 11 5.9 0.000 0.001 12.300 12.301 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.300 12.301 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.243 11.721 multiply_cannon_metrocomm4 27577 15.6 0.100 0.113 3.828 10.916 dbcsr_mm_accdrv_process 62242 16.2 5.458 6.143 10.208 10.768 mp_irecv_dv 69486 16.3 3.624 10.510 3.624 10.510 make_images_data 5014 15.6 0.067 0.077 8.222 9.786 hybrid_alltoall_any 5200 16.5 0.350 1.549 6.942 9.203 pw_transfer 1547 11.6 0.084 0.101 8.637 8.700 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.415 8.481 wfi_extrapolate 11 7.9 0.006 0.043 8.408 8.409 density_rs2pw 128 9.7 0.006 0.006 7.112 7.703 grid_integrate_task_list 128 12.3 7.153 7.494 7.153 7.494 fft_wrap_pw1pw2_140 523 13.2 0.934 0.957 7.416 7.494 cp_fm_cholesky_decompose 22 10.9 7.061 7.250 7.061 7.250 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.263 7.133 calculate_dm_sparse 128 9.5 0.001 0.001 6.469 6.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.306 6.316 fft3d_ps 1291 14.7 2.827 3.008 5.971 6.027 mp_sum_l 7950 12.9 4.293 5.986 4.293 5.986 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.349 5.534 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.418 5.493 grid_collocate_task_list 128 9.7 5.000 5.404 5.000 5.404 mp_alltoall_d11v 2415 14.1 4.215 4.990 4.215 4.990 potential_pw2rs 128 12.3 0.016 0.019 4.519 4.539 mp_allgather_i34 2507 14.6 1.527 4.499 1.527 4.499 mp_sum_d 4472 12.1 2.723 4.087 2.723 4.087 dbcsr_complete_redistribute 395 12.7 0.769 0.854 3.196 4.051 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.519000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.454545, yerr=6.020626 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 935.399424E+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 0.077 0.107 175.941 175.941 qs_mol_dyn_low 1 2.0 0.003 0.003 175.154 175.177 qs_forces 11 3.9 0.004 0.005 175.063 175.065 qs_energies 11 4.9 0.001 0.002 168.541 168.551 scf_env_do_scf 11 5.9 0.001 0.001 152.617 152.618 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 117.929 117.929 velocity_verlet 10 3.0 0.001 0.002 112.640 112.642 dbcsr_multiply_generic 2527 12.6 0.188 0.195 82.331 83.413 qs_scf_new_mos 118 7.6 0.001 0.001 80.601 80.954 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.600 80.954 ot_scf_mini 118 9.6 0.003 0.004 76.307 76.696 multiply_cannon 2527 13.6 0.508 0.526 62.558 67.466 multiply_cannon_loop 2527 14.6 1.136 1.214 59.680 62.330 ot_mini 118 10.6 0.001 0.001 42.770 43.171 init_scf_loop 11 6.9 0.000 0.000 34.591 34.592 mp_waitall_1 171878 16.6 24.778 34.400 24.778 34.400 rebuild_ks_matrix 129 8.3 0.001 0.001 30.362 30.845 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.029 30.362 30.844 prepare_preconditioner 11 7.9 0.000 0.000 30.616 30.664 make_preconditioner 11 8.9 0.000 0.000 30.616 30.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.275 29.670 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.395 27.830 multiply_cannon_multrec 20216 15.6 12.996 16.803 22.472 26.327 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.572 25.154 qs_ot_get_derivative 118 11.6 0.001 0.002 22.895 23.299 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.980 21.004 apply_single 129 13.6 0.001 0.001 19.980 21.003 ot_diis_step 118 11.6 0.018 0.019 19.764 19.765 qs_ot_get_p 129 10.4 0.001 0.002 18.780 19.266 make_m2s 5054 13.6 0.078 0.084 14.714 15.838 multiply_cannon_sync_h2d 20216 15.6 13.758 15.778 13.758 15.778 make_images 5054 14.6 1.167 1.248 14.477 15.597 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.524 14.531 cp_fm_cholesky_invert 11 10.9 14.268 14.277 14.268 14.277 sum_up_and_integrate 129 10.3 0.002 0.003 13.987 14.011 integrate_v_rspace 129 11.3 0.003 0.004 13.926 13.951 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.619 13.649 calculate_rho_elec 129 8.7 0.131 0.146 13.618 13.648 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.461 13.462 init_scf_run 11 5.9 0.000 0.001 10.853 10.853 scf_env_initial_rho_setup 11 6.9 0.011 0.029 10.853 10.853 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.278 10.299 cp_fm_diag_elpa_base 83 14.4 9.868 10.018 10.275 10.296 make_images_data 5054 15.6 0.065 0.074 8.752 10.240 hybrid_alltoall_any 5240 16.5 0.451 2.062 7.537 9.654 multiply_cannon_metrocomm4 17689 15.6 0.066 0.075 3.558 9.593 mp_irecv_dv 50625 16.2 3.426 9.332 3.426 9.332 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.949 9.229 dbcsr_mm_accdrv_process 41830 16.2 5.673 6.041 8.920 9.102 pw_transfer 1559 11.6 0.084 0.102 8.833 8.953 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.607 8.734 grid_integrate_task_list 129 12.3 7.406 7.832 7.406 7.832 fft_wrap_pw1pw2_140 527 13.2 1.075 1.116 7.659 7.776 wfi_extrapolate 11 7.9 0.001 0.001 7.403 7.403 density_rs2pw 129 9.7 0.006 0.006 6.947 7.379 cp_fm_upper_to_full 105 14.8 5.740 7.236 5.740 7.236 cp_fm_cholesky_decompose 22 10.9 7.199 7.226 7.199 7.226 dbcsr_complete_redistribute 395 12.7 1.158 1.187 4.530 6.314 calculate_dm_sparse 129 9.5 0.001 0.001 5.898 6.011 fft3d_ps 1301 14.7 2.857 3.113 5.891 6.001 grid_collocate_task_list 129 9.7 5.192 5.479 5.192 5.479 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.437 5.443 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.649 5.361 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.371 5.151 mp_alltoall_d11v 2423 14.1 4.238 5.065 4.238 5.065 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.564 4.687 mp_sum_l 8010 12.9 3.216 4.625 3.216 4.625 mp_allgather_i34 2527 14.6 1.363 4.515 1.363 4.515 potential_pw2rs 129 12.3 0.020 0.022 4.372 4.383 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.010 4.049 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.322 4.044 mp_alltoall_i22 718 14.1 1.898 3.836 1.898 3.836 qs_energies_init_hamiltonians 11 5.9 0.004 0.014 3.780 3.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.941000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.909091, yerr=8.897952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 1.140421E+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 3991 58361. MP_Allreduce 11055 1084. MP_Sync 86 MP_Alltoall 1700 12496361. 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 0.017 0.037 186.870 186.870 qs_mol_dyn_low 1 2.0 0.003 0.003 186.022 186.051 qs_forces 11 3.9 0.004 0.004 185.903 185.907 qs_energies 11 4.9 0.002 0.003 178.815 178.825 scf_env_do_scf 11 5.9 0.001 0.001 161.635 161.643 velocity_verlet 10 3.0 0.017 0.018 123.017 123.019 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.619 115.620 dbcsr_multiply_generic 2485 12.5 0.192 0.197 79.929 80.644 qs_scf_new_mos 116 7.6 0.001 0.001 79.669 79.972 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.668 79.971 ot_scf_mini 116 9.6 0.003 0.003 75.157 75.474 multiply_cannon 2485 13.5 0.546 0.576 55.528 58.837 multiply_cannon_loop 2485 14.5 1.807 1.910 51.926 53.567 init_scf_loop 11 6.9 0.000 0.000 45.894 45.895 ot_mini 116 10.6 0.001 0.001 42.524 42.840 prepare_preconditioner 11 7.9 0.000 0.000 41.798 41.831 make_preconditioner 11 8.9 0.000 0.000 41.798 41.831 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.536 40.529 multiply_cannon_multrec 29820 15.5 13.378 18.675 26.382 31.332 rebuild_ks_matrix 127 8.3 0.001 0.001 29.113 29.441 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.112 29.441 mp_waitall_1 146592 16.7 17.049 26.907 17.049 26.907 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.238 26.531 qs_ot_get_derivative 116 11.6 0.001 0.002 23.014 23.338 make_m2s 4970 13.5 0.090 0.093 20.111 21.107 make_images 4970 14.5 1.945 2.240 19.816 20.809 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.934 19.539 apply_single 127 13.6 0.001 0.001 18.934 19.538 ot_diis_step 116 11.6 0.017 0.018 19.396 19.397 qs_ot_get_p 127 10.4 0.002 0.002 18.851 19.216 cp_fm_upper_to_full 104 14.8 11.515 16.928 11.515 16.928 cp_fm_cholesky_invert 11 10.9 15.863 15.872 15.863 15.872 multiply_cannon_metrocomm3 29820 15.5 0.048 0.051 6.483 14.868 qs_ot_p2m_diag 82 11.4 0.339 0.386 14.749 14.797 sum_up_and_integrate 127 10.3 0.002 0.003 13.878 13.902 integrate_v_rspace 127 11.3 0.003 0.004 13.818 13.844 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.623 13.655 calculate_rho_elec 127 8.7 0.171 0.186 13.623 13.655 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.361 13.363 dbcsr_mm_accdrv_process 61748 16.2 8.391 9.243 12.568 13.053 dbcsr_complete_redistribute 393 12.7 1.516 1.640 9.024 12.750 make_images_data 4970 15.5 0.065 0.072 10.770 12.502 hybrid_alltoall_any 5155 16.4 0.526 2.180 9.601 11.514 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.680 11.413 multiply_cannon_sync_h2d 29820 15.5 10.413 11.355 10.413 11.355 init_scf_run 11 5.9 0.000 0.001 11.135 11.136 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.134 11.136 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.201 10.216 cp_fm_diag_elpa_base 82 14.4 9.248 9.568 10.195 10.209 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.241 9.894 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.575 9.807 mp_alltoall_i22 712 14.1 5.470 9.167 5.470 9.167 pw_transfer 1535 11.6 0.085 0.101 9.097 9.163 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.873 8.943 fft_wrap_pw1pw2_140 519 13.2 1.237 1.272 7.912 7.998 grid_integrate_task_list 127 12.3 7.468 7.786 7.468 7.786 wfi_extrapolate 11 7.9 0.001 0.001 7.658 7.658 multiply_cannon_metrocomm4 24850 15.5 0.080 0.091 2.844 7.570 cp_fm_cholesky_decompose 22 10.9 7.459 7.566 7.459 7.566 mp_irecv_dv 75445 16.2 2.688 7.292 2.688 7.292 density_rs2pw 127 9.7 0.005 0.006 6.731 7.148 calculate_dm_sparse 127 9.5 0.001 0.001 6.205 6.294 fft3d_ps 1281 14.7 2.972 3.015 5.947 6.006 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.599 5.651 grid_collocate_task_list 127 9.7 5.240 5.587 5.240 5.587 mp_alltoall_d11v 2401 14.1 4.896 5.318 4.896 5.318 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.493 4.592 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.469 4.567 qs_energies_init_hamiltonians 11 5.9 0.032 0.038 4.486 4.486 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.212 4.273 potential_pw2rs 127 12.3 0.022 0.023 4.235 4.245 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.870000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.636364, yerr=16.204785 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.524625E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 58881. MP_Allreduce 10965 1170. MP_Sync 86 MP_Alltoall 1700 18828171. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.039 170.859 170.859 qs_mol_dyn_low 1 2.0 0.003 0.003 169.959 169.978 qs_forces 11 3.9 0.004 0.004 169.853 169.855 qs_energies 11 4.9 0.001 0.002 162.370 162.378 scf_env_do_scf 11 5.9 0.001 0.001 144.674 144.684 velocity_verlet 10 3.0 0.010 0.012 111.242 111.246 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.558 108.559 dbcsr_multiply_generic 2485 12.5 0.183 0.189 71.748 72.053 qs_scf_new_mos 116 7.6 0.001 0.001 71.247 71.341 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.246 71.340 ot_scf_mini 116 9.6 0.003 0.004 66.842 66.911 multiply_cannon 2485 13.5 0.558 0.591 53.050 56.435 multiply_cannon_loop 2485 14.5 0.799 0.831 50.110 50.772 ot_mini 116 10.6 0.001 0.001 37.327 37.400 init_scf_loop 11 6.9 0.000 0.000 35.966 35.968 prepare_preconditioner 11 7.9 0.000 0.000 32.047 32.076 make_preconditioner 11 8.9 0.000 0.000 32.047 32.076 mp_waitall_1 124680 16.7 24.637 30.541 24.637 30.541 make_full_inverse_cholesky 11 9.9 0.013 0.026 29.945 30.241 rebuild_ks_matrix 127 8.3 0.001 0.001 29.329 29.398 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.329 29.398 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.683 26.744 multiply_cannon_multrec 9940 15.5 10.351 14.433 18.005 21.430 qs_ot_get_derivative 116 11.6 0.001 0.002 20.510 20.589 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.193 19.455 cp_fm_cholesky_invert 11 10.9 18.217 18.223 18.217 18.223 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.834 16.966 apply_single 127 13.6 0.001 0.001 16.834 16.965 qs_ot_get_p 127 10.4 0.002 0.003 16.862 16.957 ot_diis_step 116 11.6 0.020 0.020 16.746 16.746 make_m2s 4970 13.5 0.062 0.066 14.862 15.694 make_images 4970 14.5 2.148 2.567 14.558 15.388 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.462 14.499 calculate_rho_elec 127 8.7 0.253 0.264 14.461 14.498 sum_up_and_integrate 127 10.3 0.002 0.002 14.310 14.357 integrate_v_rspace 127 11.3 0.004 0.004 14.249 14.299 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.147 13.165 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.993 11.993 multiply_cannon_sync_h2d 9940 15.5 10.779 11.246 10.779 11.246 init_scf_run 11 5.9 0.000 0.001 10.813 10.813 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.813 10.813 pw_transfer 1535 11.6 0.084 0.092 10.154 10.197 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.932 9.980 make_images_data 4970 15.5 0.054 0.065 8.544 9.889 hybrid_alltoall_any 5155 16.4 0.836 3.753 8.362 9.750 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.165 9.177 cp_fm_diag_elpa_base 82 14.4 8.919 9.007 9.161 9.173 fft_wrap_pw1pw2_140 519 13.2 1.669 1.701 8.731 8.783 cp_fm_cholesky_decompose 22 10.9 8.181 8.290 8.181 8.290 dbcsr_mm_accdrv_process 20590 16.1 3.344 4.476 7.307 8.047 grid_integrate_task_list 127 12.3 7.703 8.035 7.703 8.035 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.969 8.024 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.630 7.518 density_rs2pw 127 9.7 0.005 0.005 7.176 7.517 wfi_extrapolate 11 7.9 0.001 0.001 7.398 7.398 fft3d_ps 1281 14.7 3.121 3.235 6.393 6.437 calculate_dm_sparse 127 9.5 0.001 0.001 6.101 6.166 grid_collocate_task_list 127 9.7 5.518 5.751 5.518 5.751 dbcsr_complete_redistribute 393 12.7 2.101 2.178 5.280 5.704 mp_alltoall_d11v 2401 14.1 4.761 5.409 4.761 5.409 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.314 5.324 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 5.304 5.305 mp_allgather_i34 2485 14.5 1.138 4.570 1.138 4.570 potential_pw2rs 127 12.3 0.026 0.026 4.468 4.475 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.172 4.221 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.725 4.058 mp_irecv_dv 28618 15.9 1.686 3.984 1.686 3.984 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.589 3.898 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.443 3.782 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.668 3.745 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.562 3.595 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.433 3.552 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.539 3.548 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.859000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1436.272727, yerr=11.592246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 11.786057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.170791E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144969E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318395E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592214928 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.082 0.090 295.208 295.209 qs_mol_dyn_low 1 2.0 0.003 0.003 294.121 294.138 qs_forces 11 3.9 0.009 0.013 293.754 293.756 qs_energies 11 4.9 0.002 0.002 284.839 284.848 scf_env_do_scf 11 5.9 0.001 0.001 261.587 261.598 velocity_verlet 10 3.0 0.014 0.015 213.032 213.072 scf_env_do_scf_inner_loop 118 6.6 0.011 0.016 136.018 136.020 init_scf_loop 11 6.9 0.000 0.000 125.298 125.301 prepare_preconditioner 11 7.9 0.000 0.000 120.423 120.443 make_preconditioner 11 8.9 0.000 0.000 120.423 120.443 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.237 117.545 qs_scf_new_mos 118 7.6 0.001 0.001 90.832 90.926 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.831 90.925 ot_scf_mini 118 9.6 0.004 0.004 85.959 85.998 dbcsr_multiply_generic 2529 12.6 0.231 0.241 83.043 83.655 cp_fm_upper_to_full 106 14.8 54.276 77.413 54.276 77.413 multiply_cannon 2529 13.6 0.687 0.720 59.643 60.901 multiply_cannon_loop 2529 14.6 1.060 1.088 55.508 57.012 ot_mini 118 10.6 0.001 0.001 45.203 45.263 dbcsr_complete_redistribute 397 12.7 3.986 4.064 30.137 43.478 copy_fm_to_dbcsr 210 11.7 0.001 0.001 26.622 39.857 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.143 37.227 mp_alltoall_i22 720 14.1 21.892 35.165 21.892 35.165 rebuild_ks_matrix 129 8.3 0.001 0.001 34.958 35.015 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.958 35.015 cp_fm_cholesky_invert 11 10.9 33.068 33.074 33.068 33.074 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.458 32.529 mp_waitall_1 104580 16.8 28.155 32.439 28.155 32.439 qs_ot_get_p 129 10.4 0.009 0.017 25.801 25.915 qs_ot_get_derivative 118 11.6 0.002 0.002 24.989 25.026 qs_ot_p2m_diag 84 11.4 0.890 0.896 21.655 21.686 multiply_cannon_metrocomm3 10116 15.6 0.025 0.026 18.859 20.363 make_m2s 5058 13.6 0.074 0.077 18.723 20.278 ot_diis_step 118 11.6 0.022 0.023 20.180 20.181 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.829 19.831 make_images 5058 14.6 3.092 3.280 18.238 19.793 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.417 19.569 apply_single 129 13.6 0.001 0.001 19.417 19.569 multiply_cannon_multrec 10116 15.6 10.476 12.344 18.515 18.646 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.467 17.485 calculate_rho_elec 129 8.7 0.483 0.485 17.467 17.484 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.641 16.641 cp_fm_diag_elpa_base 84 14.4 12.097 13.705 16.636 16.637 sum_up_and_integrate 129 10.3 0.002 0.002 16.523 16.610 integrate_v_rspace 129 11.3 0.005 0.005 16.459 16.547 multiply_cannon_sync_h2d 10116 15.6 14.404 14.432 14.404 14.432 pw_transfer 1559 11.6 0.096 0.097 13.323 13.336 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 13.078 13.096 hybrid_alltoall_any 5245 16.5 1.319 3.068 10.828 12.993 make_images_data 5058 15.6 0.064 0.069 10.613 12.823 init_scf_run 11 5.9 0.000 0.001 12.677 12.678 scf_env_initial_rho_setup 11 6.9 0.013 0.021 12.676 12.678 fft_wrap_pw1pw2_140 527 13.2 3.270 3.336 11.655 11.685 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.881 9.908 dbcsr_mm_accdrv_process 20934 16.1 4.303 6.275 7.786 9.704 wfi_extrapolate 11 7.9 0.001 0.001 8.963 8.963 cp_fm_cholesky_decompose 22 10.9 8.834 8.862 8.834 8.862 grid_integrate_task_list 129 12.3 8.620 8.816 8.620 8.816 density_rs2pw 129 9.7 0.005 0.005 8.434 8.483 qs_energies_init_hamiltonians 11 5.9 0.045 0.072 8.387 8.389 fft3d_ps 1301 14.7 4.006 4.046 7.454 7.517 mp_alltoall_d11v 2429 14.1 6.861 7.035 6.861 7.035 calculate_dm_sparse 129 9.5 0.001 0.001 6.808 6.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.492 6.557 grid_collocate_task_list 129 9.7 6.469 6.544 6.469 6.544 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.173 6.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.209000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2880.000000, yerr=154.036005 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 1.251906E+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 26727079. 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 0.047 0.072 85.411 85.411 qs_energies 1 2.0 0.000 0.000 84.462 84.470 ls_scf 1 3.0 0.000 0.000 83.504 83.512 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.435 72.604 multiply_cannon 111 7.7 0.017 0.020 55.897 56.948 multiply_cannon_loop 111 8.7 0.228 0.243 52.485 53.655 ls_scf_main 1 4.0 0.000 0.000 52.240 52.251 density_matrix_trs4 2 5.0 0.002 0.003 46.717 46.784 ls_scf_init_scf 1 4.0 0.000 0.000 28.234 28.236 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.050 27.113 mp_waitall_1 11031 10.9 22.167 25.082 22.167 25.082 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.854 24.876 multiply_cannon_multrec 2664 9.7 8.143 8.857 15.608 17.269 multiply_cannon_sync_h2d 2664 9.7 13.688 15.416 13.688 15.416 make_m2s 222 7.7 0.009 0.011 12.961 13.444 make_images 222 8.7 0.098 0.108 12.939 13.424 multiply_cannon_metrocomm1 2664 9.7 0.010 0.012 9.546 12.653 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.409 8.623 make_images_data 222 9.7 0.004 0.005 7.545 8.117 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.701 7.078 8.048 hybrid_alltoall_any 227 10.6 0.216 1.855 6.528 7.946 dbcsr_mm_accdrv_process_sort 4760 11.4 6.285 7.184 6.285 7.184 calculate_norms 4752 9.8 5.515 6.235 5.515 6.235 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.009 5.204 mp_sum_l 887 5.1 3.045 4.708 3.045 4.708 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.065 3.816 mp_irecv_dv 6231 10.9 2.048 3.783 2.048 3.783 make_images_sizes 222 9.7 0.000 0.000 0.708 3.612 mp_alltoall_i44 222 10.7 0.707 3.612 0.707 3.612 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.236 3.548 arnoldi_extremal 4 6.8 0.000 0.000 3.197 3.220 arnoldi_normal_ev 4 7.8 0.001 0.003 3.197 3.220 build_subspace 16 8.4 0.009 0.012 3.102 3.104 ls_scf_post 1 4.0 0.000 0.000 3.030 3.038 ls_scf_store_result 1 5.0 0.000 0.000 2.850 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.352 2.854 dbcsr_merge_single_wm 555 10.7 0.460 0.566 2.344 2.845 make_images_pack 222 9.7 2.209 2.622 2.211 2.623 dbcsr_sort_data 658 11.4 2.143 2.591 2.143 2.591 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.306 2.552 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.459 2.066 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.309 2.398 buffer_matrices_ensure_size 222 8.7 1.765 2.129 1.765 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.806 1.807 rebuild_ks_matrix 3 7.3 0.000 0.000 1.797 1.798 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.008 1.796 1.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.411000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1123.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 2.134876E+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 0.199 0.244 90.649 90.654 qs_energies 1 2.0 0.000 0.000 89.232 89.258 ls_scf 1 3.0 0.000 0.000 87.743 87.770 dbcsr_multiply_generic 111 6.7 0.015 0.015 73.800 74.184 multiply_cannon 111 7.7 0.028 0.044 52.651 56.287 ls_scf_main 1 4.0 0.000 0.000 54.165 54.169 multiply_cannon_loop 111 8.7 0.136 0.147 49.825 52.589 density_matrix_trs4 2 5.0 0.002 0.003 48.527 48.766 ls_scf_init_scf 1 4.0 0.000 0.000 29.836 29.838 mp_waitall_1 9105 10.9 20.480 29.133 20.480 29.133 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.637 28.714 multiply_cannon_multrec 1332 9.7 13.314 17.025 22.665 27.488 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.158 26.168 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.317 20.365 make_m2s 222 7.7 0.006 0.008 14.933 15.521 make_images 222 8.7 1.373 1.698 14.902 15.491 dbcsr_mm_accdrv_process 4041 10.4 0.336 0.520 8.949 10.490 dbcsr_mm_accdrv_process_sort 4041 11.4 8.465 9.965 8.465 9.965 make_images_data 222 9.7 0.004 0.005 8.652 9.588 hybrid_alltoall_any 227 10.6 0.541 2.553 8.106 9.517 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.181 7.619 mp_irecv_dv 3311 11.0 3.160 7.564 3.160 7.564 mp_sum_l 887 5.1 4.793 7.305 4.793 7.305 calculate_norms 2376 9.8 6.061 6.736 6.061 6.736 multiply_cannon_sync_h2d 1332 9.7 4.914 6.304 4.914 6.304 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.584 5.715 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.869 5.055 arnoldi_extremal 4 6.8 0.000 0.000 4.561 4.574 arnoldi_normal_ev 4 7.8 0.001 0.005 4.561 4.574 build_subspace 16 8.4 0.014 0.021 4.313 4.316 ls_scf_post 1 4.0 0.000 0.000 3.742 3.768 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.525 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.105 3.336 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.209 2.735 3.211 mp_allgather_i34 111 8.7 0.809 3.034 0.809 3.034 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.158 2.931 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.608 2.684 dbcsr_data_new 4174 10.1 2.120 2.432 2.120 2.432 make_images_pack 222 9.7 1.821 2.110 1.824 2.113 dbcsr_sort_data 436 11.2 1.838 2.096 1.838 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.848 1.850 rebuild_ks_matrix 3 7.3 0.000 0.000 1.835 1.838 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.835 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.654000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1766.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 2.911560E+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 0.193 0.468 94.738 94.739 qs_energies 1 2.0 0.000 0.000 92.378 92.394 ls_scf 1 3.0 0.001 0.006 90.708 90.728 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.238 75.617 ls_scf_main 1 4.0 0.000 0.000 56.668 56.673 multiply_cannon 111 7.7 0.032 0.110 52.077 55.814 multiply_cannon_loop 111 8.7 0.116 0.130 49.296 53.627 density_matrix_trs4 2 5.0 0.002 0.003 50.788 50.941 mp_waitall_1 7281 11.0 23.446 32.770 23.446 32.770 ls_scf_init_scf 1 4.0 0.000 0.002 30.210 30.213 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.978 29.069 matrix_sqrt_Newton_Schulz 2 6.5 0.015 0.042 26.635 26.652 multiply_cannon_multrec 888 9.7 12.691 15.376 21.348 24.631 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.029 22.504 make_m2s 222 7.7 0.006 0.007 16.263 17.053 make_images 222 8.7 1.585 1.862 16.225 17.014 make_images_data 222 9.7 0.004 0.004 9.458 10.562 hybrid_alltoall_any 227 10.6 0.638 2.947 8.989 10.131 dbcsr_mm_accdrv_process 3754 10.4 0.313 0.490 8.175 9.413 dbcsr_mm_accdrv_process_sort 3754 11.4 7.738 8.923 7.738 8.923 mp_sum_l 887 5.1 5.248 8.695 5.248 8.695 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.193 multiply_cannon_sync_h2d 888 9.7 5.981 7.140 5.981 7.140 mp_irecv_dv 2335 11.1 2.457 7.131 2.457 7.131 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.000 6.891 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.604 6.675 arnoldi_extremal 4 6.8 0.000 0.003 5.069 5.107 arnoldi_normal_ev 4 7.8 0.011 0.032 5.069 5.107 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.768 4.980 calculate_norms 1584 9.8 4.393 4.869 4.393 4.869 build_subspace 16 8.4 0.014 0.021 4.726 4.734 ls_scf_post 1 4.0 0.002 0.015 3.830 3.848 mp_allgather_i34 111 8.7 0.900 3.819 0.900 3.819 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.403 3.737 dbcsr_matrix_vector_mult_local 304 10.0 3.004 3.578 3.007 3.580 ls_scf_store_result 1 5.0 0.000 0.000 3.430 3.506 ls_scf_dm_to_ks 2 5.0 0.002 0.029 2.833 2.940 dbcsr_data_new 4116 9.9 2.111 2.460 2.111 2.460 dbcsr_sort_data 325 11.1 1.884 2.187 1.884 2.187 make_images_sizes 222 9.7 0.000 0.000 0.968 2.179 mp_alltoall_i44 222 10.7 0.968 2.179 0.968 2.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.739000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2224.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 3.318886E+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 0.058 0.086 99.592 99.594 qs_energies 1 2.0 0.000 0.000 98.687 98.702 ls_scf 1 3.0 0.001 0.008 96.849 96.865 dbcsr_multiply_generic 111 6.7 0.016 0.017 80.365 80.620 ls_scf_main 1 4.0 0.000 0.000 59.026 59.027 multiply_cannon 111 7.7 0.049 0.126 53.516 57.235 density_matrix_trs4 2 5.0 0.002 0.003 52.815 52.944 multiply_cannon_loop 111 8.7 0.154 0.168 48.503 51.237 ls_scf_init_scf 1 4.0 0.000 0.002 34.551 34.554 ls_scf_init_matrix_S 1 5.0 0.000 0.002 33.250 33.307 mp_waitall_1 6369 11.0 23.304 31.733 23.304 31.733 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 30.769 30.783 multiply_cannon_multrec 1332 9.7 14.062 16.981 22.044 25.111 make_m2s 222 7.7 0.007 0.008 21.262 22.649 make_images 222 8.7 3.146 3.610 21.212 22.601 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.730 17.967 make_images_data 222 9.7 0.004 0.004 11.933 13.512 hybrid_alltoall_any 227 10.6 0.796 3.754 11.239 13.006 dbcsr_mm_accdrv_process 3641 10.4 0.297 0.488 7.622 9.167 dbcsr_mm_accdrv_process_sort 3641 11.4 7.178 8.671 7.178 8.671 mp_sum_l 887 5.1 4.013 7.150 4.013 7.150 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.093 6.050 multiply_cannon_sync_h2d 1332 9.7 5.403 6.013 5.403 6.013 mp_irecv_dv 3229 10.9 2.067 5.972 2.067 5.972 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.001 5.405 arnoldi_extremal 4 6.8 0.000 0.000 5.320 5.335 arnoldi_normal_ev 4 7.8 0.002 0.006 5.320 5.335 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.467 5.084 build_subspace 16 8.4 0.014 0.021 4.968 4.976 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.601 4.780 calculate_norms 2376 9.8 4.223 4.624 4.223 4.624 mp_allgather_i34 111 8.7 2.113 4.382 2.113 4.382 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.588 3.900 dbcsr_matrix_vector_mult_local 304 10.0 3.169 3.674 3.171 3.676 dbcsr_sort_data 658 11.4 3.064 3.515 3.064 3.515 ls_scf_post 1 4.0 0.002 0.015 3.272 3.287 dbcsr_special_finalize 555 9.7 0.006 0.007 2.803 3.239 dbcsr_merge_single_wm 555 10.7 0.533 0.664 2.794 3.231 ls_scf_dm_to_ks 2 5.0 0.001 0.008 3.041 3.103 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.056 compute_matrix_preconditioner 1 6.0 0.001 0.001 2.487 2.495 dbcsr_data_release 10477 10.7 1.594 2.414 1.594 2.414 dbcsr_finalize 304 7.8 0.049 0.062 1.807 2.016 acc_transpose_blocks 1332 9.7 0.008 0.009 1.702 2.011 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.992 1.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.594000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 4.716519E+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 0.096 0.114 92.936 92.936 qs_energies 1 2.0 0.000 0.000 91.785 91.786 ls_scf 1 3.0 0.000 0.000 89.820 89.824 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.905 71.063 ls_scf_main 1 4.0 0.000 0.000 56.507 56.507 multiply_cannon 111 7.7 0.080 0.122 52.825 56.076 multiply_cannon_loop 111 8.7 0.088 0.094 50.224 51.528 density_matrix_trs4 2 5.0 0.002 0.003 49.639 49.705 ls_scf_init_scf 1 4.0 0.000 0.000 29.927 29.930 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.565 28.593 mp_waitall_1 5436 11.0 24.473 28.136 24.473 28.136 matrix_sqrt_Newton_Schulz 2 6.5 0.081 0.090 26.403 26.417 multiply_cannon_multrec 444 9.7 13.644 16.412 20.810 22.042 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.948 15.801 make_m2s 222 7.7 0.005 0.005 13.448 14.400 make_images 222 8.7 2.042 2.480 13.381 14.330 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.149 12.960 make_images_data 222 9.7 0.003 0.004 8.194 9.703 hybrid_alltoall_any 227 10.6 0.802 3.812 8.136 9.691 multiply_cannon_sync_h2d 444 9.7 6.784 8.032 6.784 8.032 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.408 6.849 7.972 dbcsr_mm_accdrv_process_sort 3003 11.4 6.486 7.574 6.486 7.574 arnoldi_extremal 4 6.8 0.000 0.000 6.067 6.078 arnoldi_normal_ev 4 7.8 0.191 0.195 6.067 6.078 build_subspace 16 8.4 0.015 0.019 5.466 5.477 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.315 4.466 mp_sum_l 887 5.1 2.630 4.396 2.630 4.396 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.148 4.359 dbcsr_matrix_vector_mult_local 304 10.0 3.666 4.142 3.668 4.144 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.538 3.706 mp_allgather_i34 111 8.7 1.157 3.690 1.157 3.690 calculate_norms 792 9.8 3.607 3.679 3.607 3.679 mp_irecv_dv 1241 11.2 1.527 3.670 1.527 3.670 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.497 3.541 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.871 3.406 ls_scf_post 1 4.0 0.000 0.000 3.386 3.388 ls_scf_store_result 1 5.0 0.000 0.000 3.157 3.189 make_images_sizes 222 9.7 0.000 0.000 0.866 3.183 mp_alltoall_i44 222 10.7 0.866 3.182 0.866 3.182 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.304 dbcsr_data_new 4608 9.7 1.799 2.275 1.799 2.275 dbcsr_merge_all 275 8.9 0.480 0.526 2.056 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.070 2.071 rebuild_ks_matrix 3 7.3 0.000 0.000 2.038 2.039 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.015 2.038 2.039 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.949 1.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.936000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 8.818209E+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 0.143 0.160 106.347 106.348 qs_energies 1 2.0 0.000 0.000 104.380 104.382 ls_scf 1 3.0 0.000 0.000 101.334 101.337 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.769 74.923 ls_scf_main 1 4.0 0.000 0.000 63.542 63.542 density_matrix_trs4 2 5.0 0.002 0.003 54.515 54.596 multiply_cannon 111 7.7 0.184 0.270 48.508 50.509 multiply_cannon_loop 111 8.7 0.098 0.098 45.505 46.209 ls_scf_init_scf 1 4.0 0.000 0.000 33.980 33.981 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.370 32.385 matrix_sqrt_Newton_Schulz 2 6.5 0.045 0.053 29.546 29.553 mp_waitall_1 4527 11.1 21.699 25.708 21.699 25.708 make_m2s 222 7.7 0.005 0.005 22.574 23.664 make_images 222 8.7 3.579 3.876 22.466 23.554 multiply_cannon_multrec 444 9.7 17.881 18.460 22.572 23.183 hybrid_alltoall_any 227 10.6 1.656 3.649 12.817 15.730 make_images_data 222 9.7 0.003 0.004 12.948 15.099 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.163 10.966 multiply_cannon_sync_h2d 444 9.7 8.791 8.840 8.791 8.840 arnoldi_extremal 4 6.8 0.000 0.000 7.368 7.373 arnoldi_normal_ev 4 7.8 0.135 0.141 7.368 7.373 build_subspace 16 8.4 0.026 0.036 6.691 6.698 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.367 5.521 dbcsr_matrix_vector_mult_local 304 10.0 4.971 5.265 4.974 5.268 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.160 5.250 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.884 5.153 dbcsr_mm_accdrv_process 1814 10.4 0.222 0.356 4.507 4.636 dbcsr_mm_accdrv_process_sort 1814 11.4 4.170 4.301 4.170 4.301 ls_scf_post 1 4.0 0.000 0.000 3.813 3.814 mp_allgather_i34 111 8.7 1.123 3.631 1.123 3.631 make_images_sizes 222 9.7 0.000 0.000 1.464 3.598 mp_alltoall_i44 222 10.7 1.464 3.598 1.464 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.517 3.552 calculate_norms 792 9.8 3.229 3.274 3.229 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.165 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 3.015 3.015 dbcsr_merge_all 275 8.9 0.884 0.912 2.866 2.943 dbcsr_complete_redistribute 5 7.6 1.435 1.476 2.754 2.865 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.508 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.506 2.508 dbcsr_sort_data 325 11.1 2.442 2.505 2.442 2.505 rebuild_ks_matrix 3 7.3 0.000 0.000 2.439 2.441 qs_ks_build_kohn_sham_matrix 3 8.3 0.022 0.025 2.439 2.441 dbcsr_new_transposed 4 7.5 0.242 0.253 2.285 2.306 dbcsr_data_new 6591 9.6 1.818 2.269 1.818 2.269 dbcsr_frobenius_norm 74 6.6 2.057 2.134 2.190 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.126 2.206 dbcsr_data_release 12724 10.6 1.985 2.181 1.985 2.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.348000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6994.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14e431119277e2f0aa53cf24ac09b5860de697e5_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 584.830976E+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 703486506. 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 0.074 0.198 230.707 230.713 qs_mol_dyn_low 1 2.0 0.029 0.292 228.997 229.026 qs_forces 5 3.8 0.015 0.028 228.720 228.746 qs_energies 5 4.8 0.003 0.035 225.705 225.778 scf_env_do_scf 5 5.8 0.001 0.015 211.527 211.529 scf_env_do_scf_inner_loop 105 6.6 0.007 0.055 183.427 183.428 qs_scf_new_mos 105 7.6 0.000 0.001 143.142 143.369 qs_scf_loop_do_ot 105 8.6 0.002 0.025 143.141 143.369 ot_scf_mini 105 9.6 0.004 0.019 133.280 133.435 dbcsr_multiply_generic 1445 12.2 0.149 0.187 132.330 132.797 multiply_cannon 1445 13.2 0.276 0.324 112.869 115.243 multiply_cannon_loop 1445 14.2 2.838 2.980 111.106 112.446 velocity_verlet 4 3.0 0.013 0.093 107.647 107.648 ot_mini 105 10.6 0.001 0.001 60.648 60.806 multiply_cannon_multrec 69360 15.2 29.793 35.668 39.876 45.495 qs_ot_get_p 112 10.4 0.001 0.001 42.561 42.881 mp_waitall_1 488190 16.1 34.982 41.927 34.982 41.927 qs_ot_get_derivative 55 11.6 0.001 0.001 38.926 39.066 multiply_cannon_metrocomm3 69360 15.2 0.201 0.213 25.715 33.357 multiply_cannon_sync_h2d 69360 15.2 29.043 33.155 29.043 33.155 qs_ot_p2m_diag 40 11.0 0.020 0.032 31.669 31.739 rebuild_ks_matrix 110 8.4 0.000 0.000 29.534 29.680 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.094 29.534 29.679 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.378 28.379 init_scf_loop 7 6.6 0.002 0.051 28.069 28.070 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.166 27.302 cp_fm_syevd 40 13.0 0.000 0.001 23.243 23.374 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.985 23.268 apply_single 62 13.6 0.000 0.000 22.984 23.268 prepare_preconditioner 7 7.6 0.000 0.000 22.915 22.942 make_preconditioner 7 8.6 0.002 0.052 22.915 22.942 ot_new_cg_direction 55 11.6 0.001 0.001 21.000 21.005 cp_fm_redistribute_end 40 14.0 9.268 18.487 9.273 18.488 cp_fm_syevd_base 40 14.0 9.208 18.430 9.208 18.430 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.937 17.264 calculate_rho_elec 110 8.6 0.030 0.046 16.937 17.264 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.408 15.480 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.829 14.978 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.174 14.289 mp_sum_l 4764 12.2 12.348 13.129 12.348 13.129 pw_transfer 1645 12.4 0.079 0.094 12.163 12.344 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 12.025 12.208 calculate_dm_sparse 110 9.5 0.000 0.001 11.466 11.674 density_rs2pw 110 9.6 0.005 0.006 11.089 11.526 init_scf_run 5 5.8 0.000 0.001 10.965 10.966 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.965 10.966 qs_vxc_create 110 10.4 0.005 0.022 10.847 10.879 cp_fm_cholesky_invert 7 10.6 10.808 10.816 10.808 10.816 fft_wrap_pw1pw2_240 915 15.0 0.834 0.907 10.606 10.787 dbcsr_mm_accdrv_process 154766 15.8 6.263 6.407 9.951 10.756 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.430 10.508 check_diag 80 13.5 8.683 8.935 9.514 9.645 fft3d_pb 915 16.0 2.395 2.575 8.789 8.968 sum_up_and_integrate 60 10.3 0.001 0.009 8.580 8.591 integrate_v_rspace 60 11.3 0.002 0.002 8.562 8.575 transfer_rs2pw 445 10.6 0.007 0.009 8.012 8.477 xc_rho_set_and_dset_create 110 12.4 0.079 0.180 7.605 7.855 multiply_cannon_metrocomm1 69360 15.2 0.098 0.103 4.769 7.508 xc_vxc_pw_create 60 11.3 0.039 0.071 7.299 7.334 make_m2s 2890 13.2 0.078 0.087 6.703 7.296 make_full_single_inverse 7 9.6 0.003 0.043 7.182 7.210 make_images 2890 14.2 0.236 0.256 6.596 7.190 xc_pw_derive 510 13.4 0.005 0.006 6.606 6.681 acc_transpose_blocks 69360 15.2 0.357 0.390 5.768 6.211 mp_alltoall_z22v 2340 17.7 5.791 6.091 5.791 6.091 calculate_first_density_matrix 1 7.0 0.001 0.017 5.875 5.891 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.017 5.755 5.771 mp_waitany 7680 13.5 4.628 5.344 4.628 5.344 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.249 5.314 potential_pw2rs 60 12.3 0.002 0.003 4.737 4.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.713000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=553.800000, yerr=3.815757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 14e431119277e2f0aa53cf24ac09b5860de697e5 Summary: empty Status: OK