=== 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: 7eeddba65004ce63710e802b2136684b4a3d7a4d ################# 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/01 job id: 51743616 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/02 job id: 51743617 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/03 job id: 51743619 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/04 job id: 51743622 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/05 job id: 51743623 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/06 job id: 51743624 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/07 job id: 51743625 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/08 job id: 51743626 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/09 job id: 51743627 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/10 job id: 51743628 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/11 job id: 51743629 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/12 job id: 51743630 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/13 job id: 51743631 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/14 job id: 51743632 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/15 job id: 51743634 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/16 job id: 51743635 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/17 job id: 51743636 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/18 job id: 51743637 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/19 job id: 51743638 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/20 job id: 51743639 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/21 job id: 51743640 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/22 job id: 51743641 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/23 job id: 51743642 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/24 job id: 51743643 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/25 job id: 51743644 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/26 job id: 51743645 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/27 job id: 51743646 --- 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/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 0.023 0.055 138.708 138.709 farming_run 1 2.0 138.026 138.027 138.654 138.658 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.545105E+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 0.005 0.025 119.535 119.535 qs_energies 1 2.0 0.000 0.000 119.335 119.338 mp2_main 1 3.0 0.000 0.000 117.219 117.222 mp2_gpw_main 1 4.0 0.028 0.036 116.175 116.178 mp2_ri_gpw_compute_in 1 5.0 0.175 0.191 97.194 97.449 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 57.276 57.530 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 43.707 48.718 get_2c_integrals 1 6.0 0.008 0.009 39.315 39.743 integrate_v_rspace 273 8.0 0.438 0.454 25.541 30.101 fft_wrap_pw1pw2 5465 10.4 0.065 0.068 28.074 28.579 grid_integrate_task_list 273 9.0 20.796 26.297 20.796 26.297 fft_wrap_pw1pw2_100 2178 11.4 1.306 1.440 25.531 26.061 compute_2c_integrals 1 7.0 0.002 0.002 21.447 21.448 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.885 21.166 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.410 20.884 21.163 rpa_ri_compute_en 1 5.0 0.019 0.022 18.876 19.066 cp_fm_cholesky_decompose 12 8.2 17.802 18.204 17.802 18.204 fft3d_s 5443 12.4 16.880 17.460 16.901 17.479 cholesky_decomp 1 7.0 0.000 0.000 16.705 17.112 ao_to_mo_and_store_B_mult_1 272 7.0 10.776 15.321 10.776 15.321 calculate_wavefunction 272 8.0 5.385 5.529 12.835 13.779 calc_potential_gpw 544 9.5 0.004 0.004 10.792 11.587 rpa_num_int 1 6.0 0.000 0.005 10.709 10.719 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.621 10.647 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.817 10.543 potential_pw2rs 545 10.0 0.106 0.107 8.556 10.344 calc_mat_Q 8 8.0 0.007 0.008 9.483 9.582 contract_S_to_Q 8 9.0 0.000 0.000 8.900 8.992 collocate_single_gaussian 272 10.0 0.038 0.040 7.874 8.611 parallel_gemm_fm 14 9.1 0.000 0.000 8.500 8.604 parallel_gemm_fm_cosma 14 10.1 8.500 8.604 8.500 8.604 create_integ_mat 1 6.0 0.013 0.027 7.819 7.828 array2fm 1 7.0 0.000 0.000 6.694 7.256 pw_gather_s 2722 12.2 4.028 4.942 4.028 4.942 pw_poisson_solve 545 10.5 0.010 0.011 4.674 4.825 pw_scatter_s 2720 12.7 4.404 4.821 4.404 4.821 pw_poisson_set 548 11.5 0.018 0.019 3.348 3.532 array2fm_buffer_send 1 8.0 2.888 3.168 2.888 3.168 pw_copy 4911 11.6 2.640 2.747 2.640 2.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.177637, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2782.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 0.058 0.066 528.809 528.810 farming_run 1 2.0 527.692 527.701 528.732 528.734 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.270604E+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 0.011 0.030 324.584 324.584 qs_energies 1 2.0 0.002 0.012 324.367 324.367 mp2_main 1 3.0 0.001 0.013 209.566 209.568 mp2_gpw_main 1 4.0 0.058 0.156 208.692 208.698 mp2_ri_gpw_compute_en 1 5.0 0.038 0.041 128.446 130.661 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.773 1.924 126.752 126.758 scf_env_do_scf 1 3.0 0.000 0.000 114.514 114.514 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.646 113.653 rebuild_ks_matrix 4 6.0 0.000 0.000 113.645 113.652 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.645 113.652 hfx_ks_matrix 4 8.0 0.001 0.001 113.298 113.301 integrate_four_center 4 9.0 0.152 0.471 113.297 113.300 mp2_ri_gpw_compute_en_expansio 172 7.0 0.463 0.498 110.900 111.961 local_gemm 172 8.0 110.436 111.499 110.436 111.499 integrate_four_center_main 4 10.0 0.098 0.492 102.255 104.888 integrate_four_center_bin 264 11.0 102.158 104.821 102.158 104.821 init_scf_loop 1 4.0 0.000 0.000 96.800 96.800 mp2_ri_gpw_compute_in 1 5.0 0.077 0.236 80.067 81.136 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.023 56.749 57.822 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.163 44.557 49.528 integrate_v_rspace 95 8.0 0.399 0.563 28.194 32.996 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.518 31.057 fft_wrap_pw1pw2_100 730 11.4 0.647 0.743 28.224 28.823 grid_integrate_task_list 95 9.0 23.487 28.489 23.487 28.489 ao_to_mo_and_store_B_mult_1 91 7.0 10.520 28.408 10.520 28.408 get_2c_integrals 1 6.0 0.003 0.022 23.214 23.248 compute_2c_integrals 1 7.0 0.004 0.017 22.193 22.206 compute_2c_integrals_loop_lm 1 8.0 0.003 0.013 21.759 22.068 mp2_eri_2c_integrate_gpw 1 9.0 1.721 1.799 21.757 22.067 fft3d_s 1823 12.4 19.213 20.125 19.226 20.138 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.711 17.711 calc_potential_gpw 182 9.5 0.002 0.002 12.412 13.485 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.743 11.446 calculate_wavefunction 91 8.0 2.018 2.051 10.038 10.823 potential_pw2rs 186 10.0 0.033 0.035 9.032 10.392 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.525 7.976 9.099 collocate_single_gaussian 91 10.0 0.018 0.033 8.306 8.940 mp_sync 37 10.5 3.022 7.703 3.022 7.703 mp_sendrecv_dm3 2068 8.0 6.020 7.209 6.020 7.209 integrate_four_center_load 4 10.0 0.000 0.000 6.754 6.758 hfx_load_balance 1 11.0 0.000 0.000 6.754 6.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.686802, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1525.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 445.419520E+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 592243. 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 0.030 0.077 53.973 53.981 qs_mol_dyn_low 1 2.0 0.023 0.090 53.639 53.656 qs_forces 11 3.9 0.002 0.002 53.430 53.447 qs_energies 11 4.9 0.005 0.025 51.913 51.957 scf_env_do_scf 11 5.9 0.001 0.005 45.419 45.421 scf_env_do_scf_inner_loop 108 6.5 0.004 0.018 43.297 43.298 qs_scf_new_mos 108 7.5 0.000 0.001 33.316 33.590 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.316 33.589 dbcsr_multiply_generic 2286 12.5 0.092 0.098 33.158 33.578 ot_scf_mini 108 9.5 0.002 0.003 31.641 31.803 multiply_cannon 2286 13.5 0.192 0.200 26.080 27.633 multiply_cannon_loop 2286 14.5 1.811 1.919 25.404 26.974 velocity_verlet 10 3.0 0.012 0.047 26.262 26.266 ot_mini 108 10.5 0.001 0.001 18.848 19.106 qs_ot_get_derivative 108 11.5 0.001 0.001 15.935 16.125 mp_waitall_1 245248 16.5 8.260 14.256 8.260 14.256 multiply_cannon_metrocomm3 54864 15.5 0.071 0.077 5.906 12.552 multiply_cannon_multrec 54864 15.5 3.677 5.803 7.651 10.995 qs_ot_get_p 119 10.4 0.001 0.001 8.180 8.490 rebuild_ks_matrix 119 8.3 0.000 0.000 7.866 8.021 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.042 7.866 8.020 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.932 7.076 mp_sum_l 7287 12.8 5.144 6.768 5.144 6.768 multiply_cannon_sync_h2d 54864 15.5 5.179 6.198 5.179 6.198 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.500 5.995 dbcsr_mm_accdrv_process 76910 16.1 1.810 2.850 3.887 5.547 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.261 5.308 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.095 5.186 init_scf_run 11 5.9 0.000 0.001 5.003 5.004 scf_env_initial_rho_setup 11 6.9 0.001 0.006 5.003 5.003 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.497 4.498 sum_up_and_integrate 119 10.3 0.001 0.002 4.415 4.421 integrate_v_rspace 119 11.3 0.002 0.003 4.404 4.411 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.286 4.287 cp_fm_redistribute_end 50 14.0 2.188 4.259 2.195 4.263 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.091 4.223 calculate_rho_elec 119 8.7 0.013 0.033 4.091 4.222 cp_fm_diag_elpa_base 50 14.0 2.062 4.152 2.066 4.161 calculate_dm_sparse 119 9.5 0.000 0.000 2.999 3.143 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.943 3.142 apply_single 119 13.6 0.000 0.000 2.943 3.142 acc_transpose_blocks 54864 15.5 0.218 0.239 2.249 2.790 calculate_first_density_matrix 1 7.0 0.002 0.014 2.747 2.760 ot_diis_step 108 11.5 0.006 0.006 2.700 2.700 multiply_cannon_metrocomm1 54864 15.5 0.055 0.061 1.692 2.694 jit_kernel_multiply 13 15.8 2.014 2.639 2.014 2.639 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.430 2.514 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.271 2.273 density_rs2pw 119 9.7 0.003 0.004 2.091 2.187 wfi_extrapolate 11 7.9 0.001 0.001 2.160 2.160 grid_integrate_task_list 119 12.3 2.002 2.125 2.002 2.125 init_scf_loop 11 6.9 0.002 0.019 2.102 2.104 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.986 2.049 mp_sum_d 4139 12.0 1.354 1.939 1.354 1.939 potential_pw2rs 119 12.3 0.004 0.004 1.790 1.801 make_m2s 4572 13.5 0.053 0.056 1.643 1.687 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.547 1.615 make_images 4572 14.5 0.132 0.138 1.561 1.606 transfer_rs2pw 487 10.6 0.006 0.007 1.442 1.586 mp_waitany 12084 13.8 1.315 1.548 1.315 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.501 1.535 mp_alltoall_d11v 2130 13.8 1.354 1.521 1.354 1.521 acc_transpose_blocks_sync 164592 16.5 1.192 1.430 1.192 1.430 grid_collocate_task_list 119 9.7 1.350 1.428 1.350 1.428 fft3d_ps 1201 13.6 0.365 0.467 1.322 1.382 transfer_pw2rs 487 13.2 0.005 0.006 1.327 1.337 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.187 1.256 parallel_gemm_fm 81 9.0 0.000 0.000 1.122 1.143 parallel_gemm_fm_cosma 81 10.0 1.122 1.143 1.122 1.143 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.981000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 483.115008E+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 1198614. 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 0.039 0.073 41.586 41.587 qs_mol_dyn_low 1 2.0 0.003 0.003 40.982 40.990 qs_forces 11 3.9 0.013 0.025 40.828 40.828 qs_energies 11 4.9 0.011 0.014 39.151 39.163 scf_env_do_scf 11 5.9 0.000 0.001 33.211 33.211 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.815 30.816 dbcsr_multiply_generic 2286 12.5 0.099 0.102 24.444 24.787 qs_scf_new_mos 108 7.5 0.001 0.001 22.366 22.604 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.366 22.604 ot_scf_mini 108 9.5 0.002 0.003 21.452 21.625 multiply_cannon 2286 13.5 0.210 0.216 17.767 20.635 multiply_cannon_loop 2286 14.5 1.190 1.247 16.625 19.696 velocity_verlet 10 3.0 0.002 0.004 18.224 18.227 ot_mini 108 10.5 0.001 0.001 13.959 14.199 mp_waitall_1 200699 16.5 5.926 12.678 5.926 12.678 qs_ot_get_derivative 108 11.5 0.001 0.001 11.602 11.775 multiply_cannon_metrocomm3 27432 15.5 0.071 0.076 4.451 11.453 multiply_cannon_multrec 27432 15.5 1.825 4.149 6.185 10.019 dbcsr_mm_accdrv_process 47894 16.0 3.565 7.098 4.277 8.148 rebuild_ks_matrix 119 8.3 0.000 0.000 6.714 6.855 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.713 6.854 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.923 6.417 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.899 6.028 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 4.643 5.920 apply_single 119 13.6 0.000 0.000 4.643 5.920 mp_sum_l 7287 12.8 3.642 5.917 3.642 5.917 qs_ot_get_p 119 10.4 0.001 0.002 4.577 4.809 init_scf_run 11 5.9 0.000 0.001 4.663 4.663 scf_env_initial_rho_setup 11 6.9 0.009 0.025 4.663 4.663 sum_up_and_integrate 119 10.3 0.001 0.001 3.594 3.601 integrate_v_rspace 119 11.3 0.002 0.003 3.580 3.588 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.486 3.512 calculate_rho_elec 119 8.7 0.021 0.024 3.485 3.511 calculate_first_density_matrix 1 7.0 0.000 0.000 3.161 3.171 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.955 2.973 make_m2s 4572 13.5 0.051 0.053 2.584 2.911 make_images 4572 14.5 0.206 0.244 2.495 2.824 acc_transpose_blocks 27432 15.5 0.112 0.119 2.037 2.589 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.559 2.559 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.548 2.549 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.311 2.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.338 2.376 init_scf_loop 11 6.9 0.000 0.000 2.375 2.376 ot_diis_step 108 11.5 0.011 0.011 2.310 2.311 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.250 2.250 multiply_cannon_sync_h2d 27432 15.5 1.716 2.234 1.716 2.234 cp_fm_redistribute_end 50 14.0 1.141 2.227 1.144 2.229 cp_fm_diag_elpa_base 50 14.0 1.054 2.138 1.082 2.178 calculate_dm_sparse 119 9.5 0.000 0.001 2.051 2.132 grid_integrate_task_list 119 12.3 1.843 1.939 1.843 1.939 jit_kernel_multiply 10 16.3 0.653 1.879 0.653 1.879 density_rs2pw 119 9.7 0.003 0.004 1.786 1.855 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.627 1.656 make_images_data 4572 15.5 0.047 0.053 1.219 1.564 prepare_preconditioner 11 7.9 0.000 0.000 1.497 1.522 make_preconditioner 11 8.9 0.000 0.000 1.497 1.522 acc_transpose_blocks_kernels 27432 16.5 0.186 0.274 1.088 1.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.401 1.457 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.090 1.451 wfi_extrapolate 11 7.9 0.001 0.001 1.434 1.434 fft3d_ps 1201 13.6 0.509 0.563 1.347 1.372 potential_pw2rs 119 12.3 0.006 0.006 1.347 1.352 grid_collocate_task_list 119 9.7 1.287 1.348 1.287 1.348 fft_wrap_pw1pw2_140 487 12.2 0.047 0.048 1.292 1.323 mp_alltoall_d11v 2130 13.8 1.185 1.321 1.185 1.321 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.254 jit_kernel_transpose 5 15.5 0.902 1.243 0.902 1.243 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.241 transfer_rs2pw 487 10.6 0.005 0.005 1.066 1.141 mp_sum_d 4139 12.0 0.544 1.012 0.544 1.012 mp_allgather_i34 2286 14.5 0.565 0.951 0.565 0.951 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.935 0.945 acc_transpose_blocks_sync 82296 16.5 0.813 0.943 0.813 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.870 0.883 transfer_pw2rs 487 13.2 0.004 0.005 0.876 0.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.587000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.181818, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 512.987136E+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 0.071 0.097 34.007 34.012 qs_mol_dyn_low 1 2.0 0.003 0.004 33.408 33.426 qs_forces 11 3.9 0.016 0.020 33.325 33.330 qs_energies 11 4.9 0.002 0.002 31.757 31.759 scf_env_do_scf 11 5.9 0.000 0.001 26.846 26.846 scf_env_do_scf_inner_loop 108 6.5 0.015 0.030 24.307 24.308 dbcsr_multiply_generic 2286 12.5 0.092 0.095 18.136 18.225 qs_scf_new_mos 108 7.5 0.001 0.001 16.738 16.764 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.738 16.763 ot_scf_mini 108 9.5 0.002 0.002 15.976 16.000 multiply_cannon 2286 13.5 0.192 0.196 14.877 15.599 velocity_verlet 10 3.0 0.026 0.029 15.234 15.235 multiply_cannon_loop 2286 14.5 0.859 0.919 14.109 14.852 ot_mini 108 10.5 0.001 0.001 10.324 10.350 qs_ot_get_derivative 108 11.5 0.001 0.001 8.817 8.841 multiply_cannon_multrec 18288 15.5 1.889 2.955 7.269 8.636 dbcsr_mm_accdrv_process 38222 16.0 5.176 6.903 5.286 6.969 rebuild_ks_matrix 119 8.3 0.000 0.000 5.944 5.958 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.943 5.958 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.229 5.241 mp_waitall_1 158411 16.6 3.581 4.917 3.581 4.917 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.099 4.671 init_scf_run 11 5.9 0.000 0.001 3.711 3.711 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.711 3.711 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 2.474 3.653 sum_up_and_integrate 119 10.3 0.001 0.001 3.498 3.503 integrate_v_rspace 119 11.3 0.003 0.003 3.486 3.491 qs_ot_get_p 119 10.4 0.001 0.001 3.406 3.433 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.148 3.151 calculate_rho_elec 119 8.7 0.031 0.032 3.147 3.150 init_scf_loop 11 6.9 0.000 0.000 2.522 2.522 calculate_first_density_matrix 1 7.0 0.000 0.000 2.496 2.496 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.988 2.361 apply_single 119 13.6 0.000 0.000 1.988 2.360 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.280 2.286 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.980 1.981 make_m2s 4572 13.5 0.043 0.045 1.828 1.954 grid_integrate_task_list 119 12.3 1.804 1.927 1.804 1.927 calculate_dm_sparse 119 9.5 0.000 0.001 1.862 1.874 make_images 4572 14.5 0.193 0.206 1.743 1.868 density_rs2pw 119 9.7 0.003 0.003 1.722 1.830 prepare_preconditioner 11 7.9 0.000 0.000 1.772 1.775 make_preconditioner 11 8.9 0.000 0.000 1.772 1.775 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.725 1.734 cp_fm_diag_elpa_base 50 14.0 1.701 1.715 1.723 1.732 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.721 1.730 acc_transpose_blocks 18288 15.5 0.077 0.080 1.620 1.726 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.629 1.709 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.661 1.662 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.610 1.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.502 1.509 ot_diis_step 108 11.5 0.011 0.011 1.490 1.490 mp_sum_l 7287 12.8 1.072 1.467 1.072 1.467 grid_collocate_task_list 119 9.7 1.232 1.325 1.232 1.325 fft3d_ps 1201 13.6 0.534 0.555 1.295 1.306 potential_pw2rs 119 12.3 0.007 0.008 1.296 1.299 fft_wrap_pw1pw2_140 487 12.2 0.061 0.062 1.288 1.297 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.196 1.200 wfi_extrapolate 11 7.9 0.001 0.001 1.164 1.165 multiply_cannon_sync_h2d 18288 15.5 1.035 1.152 1.035 1.152 transfer_rs2pw 487 10.6 0.005 0.005 0.970 1.072 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.956 0.975 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 make_images_data 4572 15.5 0.047 0.051 0.784 0.927 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.682 0.854 mp_alltoall_d11v 2130 13.8 0.692 0.845 0.692 0.845 acc_transpose_blocks_sync 54864 16.5 0.740 0.840 0.740 0.840 transfer_pw2rs 487 13.2 0.004 0.004 0.836 0.840 acc_transpose_blocks_kernels 18288 16.5 0.215 0.222 0.783 0.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.786 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.673 0.742 cp_fm_cholesky_invert 11 10.9 0.725 0.728 0.725 0.728 parallel_gemm_fm 81 9.0 0.000 0.000 0.702 0.705 parallel_gemm_fm_cosma 81 10.0 0.702 0.705 0.702 0.705 mp_alltoall_z22v 1201 15.6 0.614 0.684 0.614 0.684 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.012000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=489.272727, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 555.532288E+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 0.024 0.096 37.326 37.343 qs_mol_dyn_low 1 2.0 0.011 0.068 36.947 36.956 qs_forces 11 3.9 0.002 0.002 36.646 36.647 qs_energies 11 4.9 0.002 0.004 34.940 34.946 scf_env_do_scf 11 5.9 0.002 0.034 29.792 29.794 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.394 26.394 dbcsr_multiply_generic 2286 12.5 0.099 0.101 20.568 20.669 qs_scf_new_mos 108 7.5 0.001 0.001 18.514 18.578 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.513 18.577 velocity_verlet 10 3.0 0.006 0.039 18.348 18.358 ot_scf_mini 108 9.5 0.002 0.004 17.483 17.536 multiply_cannon 2286 13.5 0.217 0.223 16.965 17.322 multiply_cannon_loop 2286 14.5 1.539 1.595 16.018 16.328 ot_mini 108 10.5 0.001 0.001 11.102 11.163 multiply_cannon_multrec 27432 15.5 2.454 3.148 9.476 10.166 qs_ot_get_derivative 108 11.5 0.001 0.001 9.243 9.298 dbcsr_mm_accdrv_process 47916 15.9 6.272 8.586 6.919 8.650 rebuild_ks_matrix 119 8.3 0.000 0.000 6.272 6.321 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.272 6.320 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.560 5.602 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.985 4.392 init_scf_run 11 5.9 0.000 0.001 3.650 3.650 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.650 3.650 qs_ot_get_p 119 10.4 0.001 0.001 3.542 3.617 sum_up_and_integrate 119 10.3 0.001 0.001 3.427 3.436 integrate_v_rspace 119 11.3 0.002 0.003 3.416 3.425 init_scf_loop 11 6.9 0.000 0.001 3.348 3.348 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.231 3.268 calculate_rho_elec 119 8.7 0.040 0.046 3.231 3.267 mp_waitall_1 137007 16.6 2.100 3.127 2.100 3.127 prepare_preconditioner 11 7.9 0.000 0.000 2.483 2.491 make_preconditioner 11 8.9 0.000 0.001 2.483 2.491 acc_transpose_blocks 27432 15.5 0.115 0.117 2.414 2.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.100 2.418 make_m2s 4572 13.5 0.053 0.055 2.219 2.344 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.304 2.332 calculate_first_density_matrix 1 7.0 0.000 0.002 2.273 2.275 make_images 4572 14.5 0.272 0.332 2.113 2.236 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.110 2.232 apply_single 119 13.6 0.000 0.000 2.110 2.232 calculate_dm_sparse 119 9.5 0.000 0.000 2.134 2.182 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 1.154 2.174 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.149 2.159 jit_kernel_multiply 6 15.9 0.581 1.948 0.581 1.948 grid_integrate_task_list 119 12.3 1.825 1.903 1.825 1.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.852 1.854 ot_diis_step 108 11.5 0.012 0.012 1.817 1.817 density_rs2pw 119 9.7 0.004 0.005 1.701 1.787 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.746 1.747 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.708 1.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.647 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.513 1.523 acc_transpose_blocks_sync 82296 16.5 1.446 1.521 1.446 1.521 cp_fm_diag_elpa_base 50 14.0 1.477 1.494 1.509 1.518 fft_wrap_pw1pw2_140 487 12.2 0.071 0.075 1.411 1.447 fft3d_ps 1201 13.6 0.576 0.640 1.369 1.392 grid_collocate_task_list 119 9.7 1.249 1.331 1.249 1.331 wfi_extrapolate 11 7.9 0.001 0.001 1.322 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.281 1.300 mp_sum_l 7287 12.8 0.973 1.299 0.973 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.230 1.240 potential_pw2rs 119 12.3 0.008 0.009 1.236 1.239 qs_energies_init_hamiltonians 11 5.9 0.009 0.016 1.178 1.179 cp_fm_upper_to_full 72 14.2 0.809 1.145 0.809 1.145 dbcsr_complete_redistribute 329 12.2 0.130 0.190 0.792 1.056 transfer_rs2pw 487 10.6 0.004 0.005 0.927 1.017 make_images_data 4572 15.5 0.047 0.051 0.811 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.008 0.801 0.888 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.690 0.873 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.598 0.861 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.825 0.843 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.800 0.805 mp_alltoall_d11v 2130 13.8 0.697 0.804 0.697 0.804 mp_alltoall_i22 627 13.8 0.458 0.764 0.458 0.764 transfer_pw2rs 487 13.2 0.004 0.004 0.750 0.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.343000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.000000, yerr=3.765875 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 622.317568E+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 0.061 0.104 29.625 29.629 qs_mol_dyn_low 1 2.0 0.004 0.022 29.174 29.200 qs_forces 11 3.9 0.003 0.004 29.109 29.111 qs_energies 11 4.9 0.002 0.002 27.416 27.420 scf_env_do_scf 11 5.9 0.000 0.001 22.001 22.001 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.477 19.478 velocity_verlet 10 3.0 0.013 0.014 14.605 14.607 dbcsr_multiply_generic 2286 12.5 0.108 0.129 13.574 13.662 qs_scf_new_mos 108 7.5 0.001 0.001 11.991 12.022 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.990 12.021 multiply_cannon 2286 13.5 0.224 0.233 10.863 11.360 ot_scf_mini 108 9.5 0.002 0.002 11.282 11.310 multiply_cannon_loop 2286 14.5 0.642 0.661 9.925 10.093 ot_mini 108 10.5 0.001 0.001 6.441 6.474 multiply_cannon_multrec 9144 15.5 1.730 1.947 6.233 6.442 rebuild_ks_matrix 119 8.3 0.000 0.000 5.704 5.723 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.704 5.723 qs_ot_get_derivative 108 11.5 0.001 0.001 5.134 5.163 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.078 5.096 dbcsr_mm_accdrv_process 12550 15.8 3.689 4.406 4.391 4.448 init_scf_run 11 5.9 0.000 0.001 3.970 3.970 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.969 3.970 sum_up_and_integrate 119 10.3 0.001 0.001 3.352 3.356 integrate_v_rspace 119 11.3 0.003 0.003 3.342 3.346 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.231 3.236 calculate_rho_elec 119 8.7 0.060 0.061 3.231 3.236 qs_ot_get_p 119 10.4 0.001 0.001 2.992 3.019 calculate_first_density_matrix 1 7.0 0.000 0.000 2.826 2.827 init_scf_loop 11 6.9 0.000 0.000 2.505 2.506 make_m2s 4572 13.5 0.033 0.035 1.897 2.041 make_images 4572 14.5 0.270 0.300 1.807 1.949 grid_integrate_task_list 119 12.3 1.864 1.945 1.864 1.945 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.923 1.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.902 1.903 mp_waitall_1 115863 16.7 1.400 1.867 1.400 1.867 calculate_dm_sparse 119 9.5 0.000 0.000 1.834 1.854 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.756 1.775 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.753 1.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.724 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.714 1.718 make_preconditioner 11 8.9 0.000 0.000 1.714 1.718 density_rs2pw 119 9.7 0.003 0.003 1.619 1.679 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.674 1.675 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.605 1.629 acc_transpose_blocks 9144 15.5 0.041 0.042 1.592 1.622 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.439 1.450 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.374 1.383 cp_fm_diag_elpa_base 50 14.0 1.346 1.363 1.373 1.381 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.365 1.381 fft3d_ps 1201 13.6 0.636 0.652 1.366 1.375 grid_collocate_task_list 119 9.7 1.296 1.350 1.296 1.350 ot_diis_step 108 11.5 0.013 0.013 1.294 1.294 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.234 1.258 apply_single 119 13.6 0.000 0.000 1.234 1.257 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.229 1.232 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.229 jit_kernel_multiply 6 15.3 0.663 1.199 0.663 1.199 potential_pw2rs 119 12.3 0.010 0.010 1.164 1.166 wfi_extrapolate 11 7.9 0.001 0.001 1.093 1.093 hybrid_alltoall_any 4725 16.4 0.066 0.176 0.793 1.023 make_images_data 4572 15.5 0.042 0.047 0.812 0.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.926 mp_alltoall_d11v 2130 13.8 0.795 0.890 0.795 0.890 transfer_rs2pw 487 10.6 0.004 0.005 0.802 0.866 cp_fm_cholesky_invert 11 10.9 0.859 0.861 0.859 0.861 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.823 0.832 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.823 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.750 acc_transpose_blocks_sync 27432 16.5 0.713 0.746 0.713 0.746 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.678 0.728 jit_kernel_transpose 5 15.6 0.705 0.716 0.705 0.716 mp_allgather_i34 2286 14.5 0.247 0.688 0.247 0.688 transfer_pw2rs 487 13.2 0.003 0.004 0.663 0.666 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.351 0.652 mp_alltoall_z22v 1201 15.6 0.605 0.643 0.605 0.643 parallel_gemm_fm 81 9.0 0.000 0.000 0.609 0.611 parallel_gemm_fm_cosma 81 10.0 0.609 0.611 0.609 0.611 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.629000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.636364, yerr=4.227028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 771.076096E+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 0.078 0.128 43.110 43.125 qs_mol_dyn_low 1 2.0 0.003 0.004 42.489 42.510 qs_forces 11 3.9 0.015 0.015 42.398 42.399 qs_energies 11 4.9 0.018 0.019 40.400 40.403 scf_env_do_scf 11 5.9 0.001 0.001 34.132 34.132 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.307 26.308 velocity_verlet 10 3.0 0.002 0.002 23.799 23.808 dbcsr_multiply_generic 2286 12.5 0.100 0.101 19.456 19.683 qs_scf_new_mos 108 7.5 0.001 0.001 17.153 17.249 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.152 17.249 multiply_cannon 2286 13.5 0.296 0.305 15.503 16.481 ot_scf_mini 108 9.5 0.002 0.002 16.002 16.099 multiply_cannon_loop 2286 14.5 0.859 0.889 14.255 15.234 ot_mini 108 10.5 0.001 0.001 9.743 9.858 multiply_cannon_multrec 9144 15.5 3.367 4.615 8.750 8.887 init_scf_loop 11 6.9 0.000 0.000 7.798 7.800 qs_ot_get_derivative 108 11.5 0.001 0.001 7.680 7.777 rebuild_ks_matrix 119 8.3 0.000 0.000 6.945 7.087 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.945 7.086 prepare_preconditioner 11 7.9 0.000 0.000 6.845 6.859 make_preconditioner 11 8.9 0.000 0.000 6.845 6.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.412 6.738 dbcsr_mm_accdrv_process 12550 15.8 4.521 6.491 5.245 6.536 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.264 6.392 cp_fm_upper_to_full 72 14.2 3.237 4.613 3.237 4.613 init_scf_run 11 5.9 0.000 0.001 4.199 4.199 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.199 4.199 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.884 3.889 calculate_rho_elec 119 8.7 0.118 0.122 3.883 3.888 sum_up_and_integrate 119 10.3 0.001 0.001 3.659 3.666 integrate_v_rspace 119 11.3 0.003 0.003 3.649 3.656 qs_ot_get_p 119 10.4 0.016 0.016 3.350 3.486 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.798 3.230 mp_waitall_1 94719 16.7 2.221 3.167 2.221 3.167 dbcsr_complete_redistribute 329 12.2 0.283 0.288 1.956 2.770 calculate_first_density_matrix 1 7.0 0.000 0.000 2.757 2.759 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.243 2.554 apply_single 119 13.6 0.000 0.000 2.243 2.554 make_m2s 4572 13.5 0.037 0.037 2.377 2.525 acc_transpose_blocks 9144 15.5 0.043 0.044 2.428 2.481 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.658 2.475 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.437 2.438 make_images 4572 14.5 0.352 0.384 2.256 2.403 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.351 2.357 mp_alltoall_i22 627 13.8 1.407 2.260 1.407 2.260 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.428 2.238 calculate_dm_sparse 119 9.5 0.000 0.000 2.217 2.232 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.180 2.218 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.247 2.148 grid_integrate_task_list 119 12.3 2.087 2.107 2.087 2.107 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.037 2.088 ot_diis_step 108 11.5 0.014 0.014 2.037 2.037 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.004 2.009 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.958 1.961 density_rs2pw 119 9.7 0.003 0.003 1.906 1.936 mp_sum_l 7287 12.8 1.044 1.811 1.044 1.811 fft3d_ps 1201 13.6 0.858 0.881 1.803 1.810 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.777 1.777 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.605 1.606 grid_collocate_task_list 119 9.7 1.527 1.542 1.527 1.542 cp_fm_cholesky_invert 11 10.9 1.421 1.424 1.421 1.424 wfi_extrapolate 11 7.9 0.001 0.001 1.382 1.382 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.127 1.358 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.325 1.325 cp_fm_diag_elpa_base 50 14.0 1.181 1.233 1.324 1.324 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.302 1.322 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 1.278 1.283 make_images_data 4572 15.5 0.045 0.049 1.054 1.239 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.229 mp_alltoall_d11v 2130 13.8 1.157 1.175 1.157 1.175 potential_pw2rs 119 12.3 0.013 0.014 1.167 1.169 jit_kernel_transpose 5 15.6 1.160 1.163 1.160 1.163 acc_transpose_blocks_sync 27432 16.5 1.090 1.143 1.090 1.143 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.102 1.119 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.997 1.061 jit_kernel_multiply 6 15.5 0.694 1.014 0.694 1.014 qs_create_task_list 11 7.9 0.006 0.011 0.938 0.948 generate_qs_task_list 11 8.9 0.367 0.389 0.932 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.125000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.636364, yerr=12.227188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 497.614848E+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 848091. 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 0.035 0.120 81.623 81.626 qs_mol_dyn_low 1 2.0 0.023 0.086 81.128 81.151 qs_forces 11 3.9 0.004 0.017 80.905 80.925 qs_energies 11 4.9 0.005 0.029 78.031 78.073 scf_env_do_scf 11 5.9 0.001 0.001 68.878 68.881 scf_env_do_scf_inner_loop 99 6.5 0.004 0.069 63.345 63.345 dbcsr_multiply_generic 2055 12.4 0.104 0.106 50.712 51.026 qs_scf_new_mos 99 7.5 0.000 0.001 46.623 46.757 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.622 46.757 ot_scf_mini 99 9.5 0.002 0.003 44.256 44.365 multiply_cannon 2055 13.4 0.185 0.188 42.455 43.465 velocity_verlet 10 3.0 0.031 0.115 42.589 42.595 multiply_cannon_loop 2055 14.4 1.764 1.804 41.536 42.517 ot_mini 99 10.5 0.001 0.001 26.315 26.417 qs_ot_get_derivative 99 11.5 0.001 0.001 19.509 19.594 multiply_cannon_multrec 49320 15.4 11.494 12.096 17.458 18.192 rebuild_ks_matrix 110 8.3 0.000 0.000 14.501 14.617 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.027 14.501 14.617 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.744 12.842 mp_waitall_1 220248 16.4 10.512 11.317 10.512 11.317 multiply_cannon_sync_h2d 49320 15.4 9.725 10.306 9.725 10.306 qs_ot_get_p 110 10.4 0.001 0.001 9.476 9.590 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.429 7.954 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.299 7.912 apply_single 110 13.6 0.000 0.000 7.299 7.912 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.482 7.549 init_scf_run 11 5.9 0.000 0.001 6.867 6.867 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.866 6.867 sum_up_and_integrate 110 10.3 0.002 0.009 6.809 6.833 integrate_v_rspace 110 11.3 0.003 0.004 6.783 6.809 ot_diis_step 99 11.5 0.005 0.006 6.612 6.613 qs_ot_p2m_diag 48 11.0 0.013 0.027 6.357 6.391 qs_rho_update_rho_low 110 7.6 0.000 0.001 5.997 6.134 calculate_rho_elec 110 8.6 0.020 0.025 5.997 6.134 dbcsr_mm_accdrv_process 87628 16.1 3.005 3.080 5.833 6.103 init_scf_loop 11 6.9 0.001 0.004 5.506 5.506 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.502 5.503 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.356 5.428 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.943 4.973 cp_fm_diag_elpa_base 48 14.0 4.931 4.960 4.941 4.972 mp_sum_l 6594 12.7 4.037 4.751 4.037 4.751 wfi_extrapolate 11 7.9 0.002 0.007 3.982 3.982 make_m2s 4110 13.4 0.060 0.065 3.841 3.950 make_images 4110 14.4 0.177 0.190 3.746 3.860 calculate_dm_sparse 110 9.5 0.001 0.001 3.751 3.846 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.284 3.497 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.491 3.494 grid_integrate_task_list 110 12.3 3.240 3.381 3.240 3.381 prepare_preconditioner 11 7.9 0.000 0.000 3.238 3.256 make_preconditioner 11 8.9 0.000 0.001 3.238 3.256 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.177 3.227 density_rs2pw 110 9.6 0.003 0.004 3.074 3.214 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.146 3.206 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.020 3.072 calculate_first_density_matrix 1 7.0 0.001 0.004 2.770 2.777 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 2.599 2.664 acc_transpose_blocks 49320 15.4 0.202 0.211 2.523 2.580 jit_kernel_multiply 13 15.9 2.549 2.564 2.549 2.564 mp_alltoall_d11v 2046 13.8 2.016 2.501 2.016 2.501 potential_pw2rs 110 12.3 0.006 0.006 2.415 2.439 grid_collocate_task_list 110 9.6 2.160 2.287 2.160 2.287 fft_wrap_pw1pw2_140 451 12.1 0.093 0.099 2.214 2.280 fft3d_ps 1111 13.6 0.732 0.818 2.139 2.195 mp_waitany 14300 13.8 1.931 2.134 1.931 2.134 transfer_rs2pw 451 10.6 0.006 0.006 1.938 2.118 mp_sum_d 3893 11.9 1.421 1.987 1.421 1.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.909 1.929 make_images_data 4110 15.4 0.043 0.047 1.732 1.867 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.498 1.735 cp_fm_cholesky_invert 11 10.9 1.729 1.733 1.729 1.733 transfer_pw2rs 451 13.1 0.006 0.007 1.697 1.710 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.680 1.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.626000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=470.909091, yerr=3.175318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 589.946880E+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 3014449. 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 0.043 0.109 69.923 69.934 qs_mol_dyn_low 1 2.0 0.003 0.004 69.237 69.262 qs_forces 11 3.9 0.015 0.019 69.171 69.172 qs_energies 11 4.9 0.002 0.003 65.829 65.844 scf_env_do_scf 11 5.9 0.000 0.001 56.586 56.589 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.049 49.049 dbcsr_multiply_generic 2055 12.4 0.112 0.116 38.520 38.695 velocity_verlet 10 3.0 0.003 0.005 35.923 35.924 qs_scf_new_mos 99 7.5 0.001 0.001 33.324 33.472 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.323 33.471 multiply_cannon 2055 13.4 0.224 0.242 31.892 33.075 ot_scf_mini 99 9.5 0.003 0.003 31.653 31.793 multiply_cannon_loop 2055 14.4 1.158 1.184 30.618 31.422 ot_mini 99 10.5 0.001 0.001 18.448 18.592 multiply_cannon_multrec 24660 15.4 6.973 8.636 14.091 15.571 rebuild_ks_matrix 110 8.3 0.000 0.000 13.567 13.679 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.566 13.679 qs_ot_get_derivative 99 11.5 0.001 0.001 12.683 12.823 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.940 12.040 mp_waitall_1 176588 16.5 7.491 9.869 7.491 9.869 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.087 7.977 init_scf_loop 11 6.9 0.000 0.000 7.503 7.503 dbcsr_mm_accdrv_process 52282 16.1 5.532 6.376 6.945 7.274 multiply_cannon_sync_h2d 24660 15.4 6.379 7.204 6.379 7.204 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.450 7.153 apply_single 110 13.6 0.000 0.001 6.449 7.153 init_scf_run 11 5.9 0.000 0.001 6.853 6.854 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.853 6.853 qs_ot_get_p 110 10.4 0.001 0.002 6.567 6.726 sum_up_and_integrate 110 10.3 0.001 0.002 6.304 6.314 integrate_v_rspace 110 11.3 0.002 0.003 6.277 6.287 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.717 5.727 calculate_rho_elec 110 8.6 0.039 0.047 5.716 5.726 ot_diis_step 99 11.5 0.010 0.010 5.717 5.718 prepare_preconditioner 11 7.9 0.000 0.000 5.492 5.515 make_preconditioner 11 8.9 0.000 0.000 5.492 5.515 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.745 5.458 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.063 5.220 make_m2s 4110 13.4 0.056 0.061 4.206 4.672 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.648 4.670 make_images 4110 14.4 0.407 0.465 4.094 4.555 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.185 4.185 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.790 3.792 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.603 3.616 cp_fm_diag_elpa_base 48 14.0 3.556 3.577 3.600 3.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.457 3.499 wfi_extrapolate 11 7.9 0.001 0.001 3.487 3.487 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.323 3.416 grid_integrate_task_list 110 12.3 3.164 3.337 3.164 3.337 calculate_first_density_matrix 1 7.0 0.000 0.000 3.266 3.268 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.115 3.254 density_rs2pw 110 9.6 0.004 0.004 2.956 3.168 calculate_dm_sparse 110 9.5 0.001 0.001 3.006 3.033 fft_wrap_pw1pw2_140 451 12.1 0.110 0.116 2.635 2.771 make_images_data 4110 15.4 0.048 0.053 2.265 2.744 hybrid_alltoall_any 4261 16.3 0.105 0.462 1.998 2.625 fft3d_ps 1111 13.6 1.069 1.270 2.494 2.621 cp_fm_cholesky_invert 11 10.9 2.566 2.573 2.566 2.573 acc_transpose_blocks 24660 15.4 0.113 0.117 2.516 2.556 mp_sum_l 6594 12.7 1.838 2.423 1.838 2.423 grid_collocate_task_list 110 9.6 2.168 2.331 2.168 2.331 potential_pw2rs 110 12.3 0.008 0.008 2.171 2.181 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.962 1.983 jit_kernel_multiply 10 16.3 1.043 1.965 1.043 1.965 mp_alltoall_d11v 2046 13.8 1.754 1.956 1.754 1.956 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.888 1.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.790 1.803 transfer_rs2pw 451 10.6 0.007 0.007 1.467 1.739 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.568 1.577 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.789 1.574 mp_allgather_i34 2055 14.4 0.581 1.548 0.581 1.548 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.492 acc_transpose_blocks_kernels 24660 16.4 0.317 0.326 1.458 1.475 mp_irecv_dv 57340 16.2 0.658 1.451 0.658 1.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.934000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.909091, yerr=7.204223 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 665.120768E+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 0.073 0.119 61.556 61.576 qs_mol_dyn_low 1 2.0 0.005 0.034 60.993 61.007 qs_forces 11 3.9 0.005 0.007 60.549 60.551 qs_energies 11 4.9 0.007 0.050 57.352 57.359 scf_env_do_scf 11 5.9 0.001 0.002 49.198 49.198 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.644 40.644 velocity_verlet 10 3.0 0.003 0.019 32.686 32.699 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.876 30.190 qs_scf_new_mos 99 7.5 0.001 0.001 26.326 26.419 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.325 26.419 ot_scf_mini 99 9.5 0.004 0.014 25.073 25.181 multiply_cannon 2055 13.4 0.211 0.219 22.869 24.433 multiply_cannon_loop 2055 14.4 0.813 0.848 21.662 23.203 ot_mini 99 10.5 0.001 0.001 14.874 14.987 rebuild_ks_matrix 110 8.3 0.000 0.000 12.083 12.261 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.082 12.261 multiply_cannon_multrec 16440 15.4 3.619 4.769 9.953 11.570 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.630 10.793 qs_ot_get_derivative 99 11.5 0.001 0.001 10.367 10.479 mp_waitall_1 139946 16.5 7.097 10.432 7.097 10.432 init_scf_loop 11 6.9 0.001 0.004 8.520 8.521 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.560 8.035 dbcsr_mm_accdrv_process 34862 16.1 5.515 6.710 6.179 7.160 prepare_preconditioner 11 7.9 0.000 0.000 6.726 6.743 make_preconditioner 11 8.9 0.000 0.001 6.726 6.743 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.787 6.647 apply_single 110 13.6 0.000 0.000 5.787 6.647 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.034 6.380 sum_up_and_integrate 110 10.3 0.001 0.002 6.140 6.155 integrate_v_rspace 110 11.3 0.003 0.004 6.114 6.130 qs_ot_get_p 110 10.4 0.005 0.033 5.398 5.532 init_scf_run 11 5.9 0.000 0.001 5.467 5.467 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.467 5.467 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.449 5.462 calculate_rho_elec 110 8.6 0.058 0.058 5.448 5.461 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.503 4.916 make_m2s 4110 13.4 0.049 0.051 4.116 4.534 ot_diis_step 99 11.5 0.011 0.011 4.476 4.476 make_images 4110 14.4 0.399 0.527 4.001 4.418 multiply_cannon_sync_h2d 16440 15.4 3.232 3.892 3.232 3.892 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.724 3.727 grid_integrate_task_list 110 12.3 3.204 3.426 3.204 3.426 cp_dbcsr_syevd 48 12.0 0.006 0.031 3.354 3.354 mp_sum_l 6594 12.7 2.285 3.271 2.285 3.271 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.052 3.059 wfi_extrapolate 11 7.9 0.001 0.001 2.941 2.941 density_rs2pw 110 9.6 0.004 0.006 2.700 2.884 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.817 2.831 make_images_data 4110 15.4 0.045 0.049 2.311 2.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.810 2.814 cp_fm_diag_elpa_base 48 14.0 2.729 2.756 2.792 2.803 hybrid_alltoall_any 4261 16.3 0.109 0.380 2.062 2.791 cp_fm_cholesky_invert 11 10.9 2.633 2.640 2.633 2.640 fft_wrap_pw1pw2_140 451 12.1 0.127 0.131 2.615 2.625 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.560 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.904 2.547 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.493 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.413 2.471 calculate_first_density_matrix 1 7.0 0.000 0.003 2.434 2.437 mp_irecv_dv 48980 15.7 0.827 2.413 0.827 2.413 fft3d_ps 1111 13.6 1.087 1.101 2.364 2.378 grid_collocate_task_list 110 9.6 2.218 2.358 2.218 2.358 qs_energies_init_hamiltonians 11 5.9 0.010 0.022 2.131 2.135 mp_alltoall_d11v 2046 13.8 1.771 2.070 1.771 2.070 potential_pw2rs 110 12.3 0.010 0.011 1.978 1.984 dbcsr_complete_redistribute 325 12.2 0.352 0.399 1.511 1.945 cp_fm_upper_to_full 70 14.2 1.389 1.761 1.389 1.761 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.666 acc_transpose_blocks 16440 15.4 0.075 0.077 1.572 1.638 mp_allgather_i34 2055 14.4 0.516 1.565 0.516 1.565 cp_fm_cholesky_decompose 22 10.9 1.539 1.563 1.539 1.563 transfer_rs2pw 451 10.6 0.005 0.006 1.301 1.496 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.010 1.372 1.485 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.036 1.478 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.475 mp_waitany 17072 13.8 1.178 1.407 1.178 1.407 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.283 1.290 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 0.332 1.273 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.185 1.250 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.576000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.363636, yerr=9.177767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 732.721152E+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 0.085 0.125 66.199 66.206 qs_mol_dyn_low 1 2.0 0.009 0.055 65.304 65.387 qs_forces 11 3.9 0.003 0.003 64.948 64.949 qs_energies 11 4.9 0.007 0.051 61.559 61.564 scf_env_do_scf 11 5.9 0.001 0.001 52.998 53.000 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.437 41.438 velocity_verlet 10 3.0 0.014 0.033 37.210 37.225 dbcsr_multiply_generic 2055 12.4 0.114 0.120 30.395 30.555 qs_scf_new_mos 99 7.5 0.001 0.001 27.182 27.326 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.181 27.326 ot_scf_mini 99 9.5 0.003 0.003 25.488 25.599 multiply_cannon 2055 13.4 0.238 0.256 23.361 24.148 multiply_cannon_loop 2055 14.4 1.390 1.438 22.012 22.572 ot_mini 99 10.5 0.001 0.001 14.719 14.854 multiply_cannon_multrec 24660 15.4 4.099 6.878 12.998 14.132 rebuild_ks_matrix 110 8.3 0.000 0.000 11.854 11.956 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 11.854 11.956 init_scf_loop 11 6.9 0.000 0.002 11.522 11.522 qs_ot_get_derivative 99 11.5 0.001 0.001 10.534 10.649 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.464 10.558 prepare_preconditioner 11 7.9 0.000 0.000 9.805 9.821 make_preconditioner 11 8.9 0.000 0.001 9.805 9.821 dbcsr_mm_accdrv_process 52304 16.0 7.676 8.997 8.739 9.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.056 9.511 sum_up_and_integrate 110 10.3 0.002 0.003 6.079 6.092 integrate_v_rspace 110 11.3 0.003 0.003 6.053 6.066 mp_waitall_1 121746 16.5 4.146 6.002 4.146 6.002 qs_ot_get_p 110 10.4 0.001 0.001 5.700 5.854 make_m2s 4110 13.4 0.059 0.061 5.398 5.656 init_scf_run 11 5.9 0.000 0.001 5.583 5.584 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.583 5.583 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.571 5.581 calculate_rho_elec 110 8.6 0.077 0.081 5.571 5.580 make_images 4110 14.4 0.574 0.698 5.259 5.513 cp_fm_upper_to_full 70 14.2 3.365 4.861 3.365 4.861 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.074 4.163 apply_single 110 13.6 0.000 0.000 4.074 4.163 ot_diis_step 99 11.5 0.011 0.012 4.146 4.146 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.859 3.876 dbcsr_complete_redistribute 325 12.2 0.430 0.500 2.658 3.776 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.469 3.515 grid_integrate_task_list 110 12.3 3.282 3.443 3.282 3.443 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.307 3.370 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.301 3.301 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.147 3.256 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.411 3.202 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.160 3.183 calculate_dm_sparse 110 9.5 0.001 0.001 2.988 3.024 wfi_extrapolate 11 7.9 0.002 0.013 2.986 2.987 make_images_data 4110 15.4 0.048 0.051 2.666 2.968 density_rs2pw 110 9.6 0.004 0.005 2.698 2.921 hybrid_alltoall_any 4261 16.3 0.122 0.458 2.303 2.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.846 2.847 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.740 2.833 mp_alltoall_i22 605 13.7 1.644 2.813 1.644 2.813 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.774 2.785 cp_fm_diag_elpa_base 48 14.0 2.627 2.686 2.772 2.783 fft_wrap_pw1pw2_140 451 12.1 0.145 0.148 2.705 2.730 cp_fm_cholesky_invert 11 10.9 2.675 2.683 2.675 2.683 multiply_cannon_sync_h2d 24660 15.4 2.381 2.555 2.381 2.555 calculate_first_density_matrix 1 7.0 0.000 0.002 2.501 2.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.433 2.475 fft3d_ps 1111 13.6 1.124 1.163 2.430 2.449 acc_transpose_blocks 24660 15.4 0.107 0.111 2.356 2.443 qs_energies_init_hamiltonians 11 5.9 0.004 0.013 2.422 2.423 grid_collocate_task_list 110 9.6 2.264 2.370 2.264 2.370 potential_pw2rs 110 12.3 0.012 0.013 1.884 1.887 mp_alltoall_d11v 2046 13.8 1.732 1.859 1.732 1.859 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.786 1.818 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.709 cp_fm_cholesky_decompose 22 10.9 1.632 1.696 1.632 1.696 mp_sum_l 6594 12.7 1.004 1.682 1.004 1.682 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.643 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.555 transfer_rs2pw 451 10.6 0.005 0.006 1.282 1.529 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.847 1.455 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.345 1.442 acc_transpose_blocks_sync 73980 16.4 1.352 1.432 1.352 1.432 mp_irecv_dv 62702 16.1 0.744 1.373 0.744 1.373 mp_waitany 13376 13.8 1.104 1.349 1.104 1.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.206000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.545455, yerr=8.700717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 864.456704E+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 0.024 0.050 54.983 54.986 qs_mol_dyn_low 1 2.0 0.003 0.004 54.482 54.499 qs_forces 11 3.9 0.015 0.030 54.085 54.087 qs_energies 11 4.9 0.015 0.029 50.453 50.466 scf_env_do_scf 11 5.9 0.000 0.001 42.300 42.300 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.666 34.667 velocity_verlet 10 3.0 0.008 0.009 31.001 31.025 dbcsr_multiply_generic 2055 12.4 0.119 0.138 23.127 23.256 qs_scf_new_mos 99 7.5 0.001 0.001 20.502 20.561 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.502 20.560 ot_scf_mini 99 9.5 0.002 0.002 19.254 19.275 multiply_cannon 2055 13.4 0.239 0.248 17.596 18.949 multiply_cannon_loop 2055 14.4 0.604 0.632 16.330 16.619 rebuild_ks_matrix 110 8.3 0.000 0.000 11.484 11.499 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 11.484 11.499 ot_mini 99 10.5 0.001 0.001 10.630 10.643 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.218 10.233 multiply_cannon_multrec 8220 15.4 3.214 4.321 7.651 8.529 init_scf_loop 11 6.9 0.000 0.000 7.589 7.591 mp_waitall_1 103326 16.6 5.703 7.177 5.703 7.177 qs_ot_get_derivative 99 11.5 0.001 0.001 6.958 6.975 sum_up_and_integrate 110 10.3 0.001 0.002 6.020 6.038 integrate_v_rspace 110 11.3 0.003 0.003 5.993 6.011 prepare_preconditioner 11 7.9 0.000 0.000 5.958 5.964 make_preconditioner 11 8.9 0.000 0.000 5.958 5.964 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.667 5.678 calculate_rho_elec 110 8.6 0.115 0.116 5.666 5.678 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.537 5.604 dbcsr_mm_accdrv_process 17442 15.9 3.206 4.220 4.298 5.266 init_scf_run 11 5.9 0.000 0.001 5.000 5.000 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.000 5.000 qs_ot_get_p 110 10.4 0.001 0.002 4.676 4.699 make_m2s 4110 13.4 0.037 0.038 4.208 4.481 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.938 4.365 make_images 4110 14.4 0.647 0.706 4.078 4.348 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.717 3.747 apply_single 110 13.6 0.000 0.000 3.717 3.747 ot_diis_step 99 11.5 0.012 0.012 3.650 3.650 grid_integrate_task_list 110 12.3 3.393 3.477 3.393 3.477 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.309 3.313 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.299 3.312 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.973 2.974 density_rs2pw 110 9.6 0.003 0.004 2.645 2.907 fft_wrap_pw1pw2_140 451 12.1 0.193 0.195 2.850 2.868 cp_fm_cholesky_invert 11 10.9 2.837 2.840 2.837 2.840 hybrid_alltoall_any 4261 16.3 0.202 0.865 2.265 2.709 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 2.677 2.684 wfi_extrapolate 11 7.9 0.001 0.001 2.651 2.651 make_images_data 4110 15.4 0.041 0.047 2.255 2.620 grid_collocate_task_list 110 9.6 2.365 2.511 2.365 2.511 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.499 2.500 multiply_cannon_sync_h2d 8220 15.4 2.394 2.494 2.394 2.494 fft3d_ps 1111 13.6 1.273 1.285 2.457 2.471 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.413 2.421 cp_fm_diag_elpa_base 48 14.0 2.358 2.385 2.412 2.419 calculate_first_density_matrix 1 7.0 0.000 0.000 2.258 2.259 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.153 2.160 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.021 2.042 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.995 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.913 1.925 potential_pw2rs 110 12.3 0.015 0.015 1.820 1.824 mp_alltoall_d11v 2046 13.8 1.688 1.803 1.688 1.803 cp_fm_cholesky_decompose 22 10.9 1.674 1.696 1.674 1.696 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.498 1.623 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.597 1.601 dbcsr_complete_redistribute 325 12.2 0.570 0.591 1.422 1.525 mp_allgather_i34 2055 14.4 0.437 1.518 0.437 1.518 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.432 acc_transpose_blocks 8220 15.4 0.039 0.040 1.361 1.408 transfer_rs2pw 451 10.6 0.005 0.005 1.095 1.371 mp_waitany 9240 13.8 1.042 1.336 1.042 1.336 qs_create_task_list 11 7.9 0.001 0.001 1.215 1.315 generate_qs_task_list 11 8.9 0.374 0.443 1.214 1.314 jit_kernel_multiply 7 15.3 0.770 1.219 0.770 1.219 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.184 1.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.986000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=818.000000, yerr=11.053588 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 1.397232E+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 0.087 0.148 88.635 88.644 qs_mol_dyn_low 1 2.0 0.004 0.013 88.017 88.048 qs_forces 11 3.9 0.004 0.004 87.657 87.659 qs_energies 11 4.9 0.002 0.002 83.589 83.592 scf_env_do_scf 11 5.9 0.000 0.001 73.260 73.260 velocity_verlet 10 3.0 0.002 0.002 55.728 55.770 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.927 44.928 dbcsr_multiply_generic 2055 12.4 0.121 0.125 30.847 30.891 init_scf_loop 11 6.9 0.000 0.000 28.262 28.263 qs_scf_new_mos 99 7.5 0.001 0.001 28.010 28.060 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.010 28.059 prepare_preconditioner 11 7.9 0.000 0.000 26.293 26.301 make_preconditioner 11 8.9 0.000 0.000 26.293 26.301 ot_scf_mini 99 9.5 0.002 0.002 26.205 26.252 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.529 25.754 multiply_cannon 2055 13.4 0.331 0.361 23.712 24.570 multiply_cannon_loop 2055 14.4 0.833 0.876 21.945 22.421 cp_fm_upper_to_full 70 14.2 12.787 18.248 12.787 18.248 ot_mini 99 10.5 0.001 0.001 15.249 15.294 rebuild_ks_matrix 110 8.3 0.000 0.001 13.427 13.467 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.426 13.467 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.189 12.226 multiply_cannon_multrec 8220 15.4 4.048 4.222 10.230 10.966 qs_ot_get_derivative 99 11.5 0.001 0.001 10.628 10.676 dbcsr_complete_redistribute 325 12.2 1.015 1.035 7.296 10.463 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.306 9.485 mp_waitall_1 84994 16.7 8.092 9.241 8.092 9.241 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.748 8.903 mp_alltoall_i22 605 13.7 5.370 8.531 5.370 8.531 dbcsr_mm_accdrv_process 11614 15.7 4.238 5.113 6.033 6.876 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.833 6.867 calculate_rho_elec 110 8.6 0.227 0.227 6.833 6.867 sum_up_and_integrate 110 10.3 0.002 0.002 6.588 6.601 integrate_v_rspace 110 11.3 0.004 0.004 6.560 6.573 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.545 6.452 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.888 6.256 apply_single 110 13.6 0.000 0.000 4.888 6.256 init_scf_run 11 5.9 0.000 0.001 6.064 6.064 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.064 6.064 make_m2s 4110 13.4 0.043 0.043 5.253 5.734 qs_ot_get_p 110 10.4 0.001 0.001 5.652 5.703 make_images 4110 14.4 0.884 0.937 5.065 5.546 cp_fm_cholesky_invert 11 10.9 5.240 5.244 5.240 5.244 ot_diis_step 99 11.5 0.015 0.016 4.600 4.600 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.919 4.502 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.478 4.496 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.065 4.071 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.904 3.918 grid_integrate_task_list 110 12.3 3.673 3.734 3.673 3.734 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.663 3.665 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.544 3.545 hybrid_alltoall_any 4261 16.3 0.264 0.569 2.820 3.434 fft3d_ps 1111 13.6 1.829 1.855 3.338 3.357 make_images_data 4110 15.4 0.045 0.047 2.719 3.325 calculate_dm_sparse 110 9.5 0.001 0.001 3.280 3.315 density_rs2pw 110 9.6 0.004 0.004 3.293 3.314 wfi_extrapolate 11 7.9 0.001 0.001 3.286 3.286 multiply_cannon_sync_h2d 8220 15.4 3.127 3.148 3.127 3.148 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.048 3.051 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.952 2.953 cp_fm_diag_elpa_base 48 14.0 2.423 2.613 2.950 2.950 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.656 2.681 grid_collocate_task_list 110 9.6 2.664 2.676 2.664 2.676 calculate_first_density_matrix 1 7.0 0.000 0.000 2.664 2.667 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.549 2.563 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.186 2.251 mp_sum_l 6594 12.7 1.019 2.235 1.019 2.235 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.161 potential_pw2rs 110 12.3 0.021 0.021 2.126 2.128 mp_alltoall_d11v 2046 13.8 2.034 2.081 2.034 2.081 cp_fm_cholesky_decompose 22 10.9 1.982 1.998 1.982 1.998 qs_create_task_list 11 7.9 0.000 0.001 1.895 1.940 generate_qs_task_list 11 8.9 0.732 0.784 1.895 1.940 jit_kernel_multiply 10 15.1 1.596 1.900 1.596 1.900 acc_transpose_blocks 8220 15.4 0.040 0.040 1.844 1.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.841 1.845 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.772 1.787 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.737 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.644000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1259.090909, yerr=60.564698 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.707328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 1284438. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.165 208.006 208.019 qs_mol_dyn_low 1 2.0 0.009 0.054 207.271 207.290 qs_forces 11 3.9 0.005 0.005 207.042 207.045 qs_energies 11 4.9 0.006 0.036 201.436 201.456 scf_env_do_scf 11 5.9 0.001 0.002 184.692 184.697 scf_env_do_scf_inner_loop 117 6.6 0.010 0.062 164.224 164.226 dbcsr_multiply_generic 2507 12.6 0.177 0.181 126.729 127.343 qs_scf_new_mos 117 7.6 0.001 0.001 125.250 125.553 qs_scf_loop_do_ot 117 8.6 0.002 0.012 125.249 125.552 velocity_verlet 10 3.0 0.025 0.076 123.473 123.475 ot_scf_mini 117 9.6 0.003 0.004 118.487 118.828 multiply_cannon 2507 13.6 0.236 0.244 102.201 103.676 multiply_cannon_loop 2507 14.6 2.386 2.431 100.073 101.580 ot_mini 117 10.6 0.001 0.001 67.030 67.347 multiply_cannon_multrec 60168 15.6 31.972 33.856 41.744 43.514 qs_ot_get_derivative 117 11.6 0.001 0.001 42.084 42.392 rebuild_ks_matrix 128 8.3 0.001 0.001 33.606 33.912 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.020 33.605 33.911 mp_waitall_1 267128 16.5 29.440 31.910 29.440 31.910 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.213 30.508 qs_ot_get_p 128 10.4 0.001 0.001 29.034 29.487 multiply_cannon_sync_h2d 60168 15.6 26.267 28.412 26.267 28.412 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.497 25.384 apply_single 128 13.6 0.001 0.001 24.496 25.384 ot_diis_step 117 11.6 0.007 0.008 24.690 24.692 qs_ot_p2m_diag 83 11.4 0.080 0.100 22.402 22.467 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.442 20.667 init_scf_loop 11 6.9 0.002 0.010 20.394 20.395 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.612 19.614 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.562 18.499 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.661 16.692 cp_fm_diag_elpa_base 83 14.4 16.593 16.626 16.656 16.688 prepare_preconditioner 11 7.9 0.000 0.000 15.744 15.791 make_preconditioner 11 8.9 0.000 0.001 15.744 15.791 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.003 15.165 make_m2s 5014 13.6 0.105 0.113 13.828 14.296 make_images 5014 14.6 0.401 0.419 13.643 14.118 sum_up_and_integrate 128 10.3 0.002 0.004 13.637 13.654 integrate_v_rspace 128 11.3 0.003 0.004 13.579 13.597 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.606 12.761 calculate_rho_elec 128 8.7 0.045 0.064 12.606 12.760 init_scf_run 11 5.9 0.000 0.001 12.479 12.480 scf_env_initial_rho_setup 11 6.9 0.006 0.015 12.479 12.480 mp_sum_l 7950 12.9 9.535 11.057 9.535 11.057 dbcsr_mm_accdrv_process 124484 16.2 4.702 4.885 9.341 9.944 wfi_extrapolate 11 7.9 0.003 0.019 9.048 9.048 cp_fm_cholesky_invert 11 10.9 8.993 9.001 8.993 9.001 calculate_dm_sparse 128 9.5 0.001 0.001 8.588 8.699 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.198 8.330 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.196 8.270 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 6.467 7.900 make_images_data 5014 15.6 0.066 0.072 6.765 7.701 grid_integrate_task_list 128 12.3 7.038 7.409 7.038 7.409 hybrid_alltoall_any 5200 16.5 0.292 2.257 5.948 7.262 density_rs2pw 128 9.7 0.005 0.006 6.158 6.751 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.657 6.670 fft_wrap_pw1pw2 1291 11.7 0.018 0.026 6.401 6.667 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.741 5.858 fft_wrap_pw1pw2_140 523 12.2 0.232 0.242 5.558 5.729 fft3d_ps 1291 13.7 2.091 2.594 5.228 5.437 mp_alltoall_d11v 2415 14.1 4.371 5.351 4.371 5.351 grid_collocate_task_list 128 9.7 4.836 5.292 4.836 5.292 cp_fm_cholesky_decompose 22 10.9 4.595 4.609 4.595 4.609 mp_sum_d 4474 12.1 3.505 4.313 3.505 4.313 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.019000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.272727, yerr=6.136195 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 840.556544E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 5616356. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.050 187.845 187.847 qs_mol_dyn_low 1 2.0 0.006 0.032 187.136 187.170 qs_forces 11 3.9 0.022 0.028 187.010 187.012 qs_energies 11 4.9 0.002 0.003 180.240 180.266 scf_env_do_scf 11 5.9 0.001 0.001 163.930 163.940 scf_env_do_scf_inner_loop 116 6.6 0.007 0.016 131.313 131.315 velocity_verlet 10 3.0 0.008 0.030 117.368 117.370 dbcsr_multiply_generic 2485 12.5 0.184 0.190 96.356 97.646 qs_scf_new_mos 116 7.6 0.001 0.001 93.541 94.103 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.540 94.102 ot_scf_mini 116 9.6 0.004 0.004 88.756 89.363 multiply_cannon 2485 13.5 0.498 0.562 76.520 81.540 multiply_cannon_loop 2485 14.5 1.547 1.617 73.174 75.885 ot_mini 116 10.6 0.001 0.001 49.926 50.510 mp_waitall_1 212858 16.6 23.933 37.758 23.933 37.758 multiply_cannon_multrec 29820 15.5 20.530 25.165 31.119 36.185 rebuild_ks_matrix 127 8.3 0.001 0.001 31.955 32.703 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.955 32.702 init_scf_loop 11 6.9 0.000 0.000 32.529 32.530 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.717 29.396 qs_ot_get_derivative 116 11.6 0.001 0.002 27.967 28.571 prepare_preconditioner 11 7.9 0.000 0.000 28.251 28.313 make_preconditioner 11 8.9 0.000 0.000 28.251 28.313 multiply_cannon_metrocomm3 29820 15.5 0.094 0.102 15.796 28.278 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.893 27.406 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.071 23.315 apply_single 127 13.6 0.001 0.001 22.071 23.315 qs_ot_get_p 127 10.4 0.003 0.005 21.131 21.799 ot_diis_step 116 11.6 0.014 0.015 21.779 21.780 multiply_cannon_sync_h2d 29820 15.5 17.590 19.858 17.590 19.858 cp_fm_cholesky_invert 11 10.9 16.459 16.472 16.459 16.472 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.273 16.308 make_m2s 4970 13.5 0.086 0.093 13.893 15.527 make_images 4970 14.5 1.168 1.380 13.679 15.316 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.081 15.081 sum_up_and_integrate 127 10.3 0.002 0.004 13.570 13.601 integrate_v_rspace 127 11.3 0.003 0.004 13.511 13.545 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.829 12.861 calculate_rho_elec 127 8.7 0.087 0.105 12.828 12.860 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.893 11.922 cp_fm_diag_elpa_base 82 14.4 11.644 11.731 11.887 11.912 init_scf_run 11 5.9 0.000 0.001 11.471 11.473 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.471 11.472 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.028 11.440 multiply_cannon_metrocomm4 27335 15.5 0.104 0.117 3.913 11.016 mp_irecv_dv 68888 16.3 3.705 10.613 3.705 10.613 dbcsr_mm_accdrv_process 61726 16.2 5.477 6.099 10.046 10.609 make_images_data 4970 15.5 0.065 0.076 8.127 10.028 hybrid_alltoall_any 5155 16.4 0.347 1.508 7.077 9.624 wfi_extrapolate 11 7.9 0.001 0.001 8.273 8.273 grid_integrate_task_list 127 12.3 7.130 7.470 7.130 7.470 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.305 7.379 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.271 7.050 cp_fm_cholesky_decompose 22 10.9 6.882 6.972 6.882 6.972 density_rs2pw 127 9.7 0.005 0.006 6.472 6.882 calculate_dm_sparse 127 9.5 0.001 0.001 6.476 6.628 fft_wrap_pw1pw2_140 519 12.2 0.246 0.254 6.394 6.463 mp_sum_l 7884 12.9 4.176 6.215 4.176 6.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.118 6.127 fft3d_ps 1281 13.7 2.741 2.921 5.769 5.818 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.370 5.458 grid_collocate_task_list 127 9.7 4.961 5.378 4.961 5.378 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.186 5.361 mp_allgather_i34 2485 14.5 1.875 4.891 1.875 4.891 mp_alltoall_d11v 2401 14.1 4.359 4.751 4.359 4.751 potential_pw2rs 127 12.3 0.015 0.017 4.178 4.192 mp_sum_d 4458 12.1 2.648 4.105 2.648 4.105 dbcsr_complete_redistribute 393 12.7 0.775 0.877 3.181 4.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.847000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.181818, yerr=4.549089 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 972.808192E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.087 176.230 176.250 qs_mol_dyn_low 1 2.0 0.006 0.024 175.629 175.662 qs_forces 11 3.9 0.004 0.005 174.569 174.575 qs_energies 11 4.9 0.030 0.062 168.028 168.041 scf_env_do_scf 11 5.9 0.001 0.002 152.335 152.337 scf_env_do_scf_inner_loop 117 6.6 0.008 0.042 117.587 117.588 velocity_verlet 10 3.0 0.008 0.059 112.311 112.344 dbcsr_multiply_generic 2507 12.6 0.184 0.193 83.191 84.482 qs_scf_new_mos 117 7.6 0.001 0.001 81.002 81.388 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.001 81.387 ot_scf_mini 117 9.6 0.003 0.004 76.774 77.232 multiply_cannon 2507 13.6 0.509 0.529 62.770 67.788 multiply_cannon_loop 2507 14.6 1.127 1.184 59.627 63.233 ot_mini 117 10.6 0.001 0.001 43.789 44.236 init_scf_loop 11 6.9 0.001 0.003 34.652 34.653 mp_waitall_1 170520 16.6 24.765 34.077 24.765 34.077 rebuild_ks_matrix 128 8.3 0.001 0.001 30.311 30.872 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.032 30.310 30.871 prepare_preconditioner 11 7.9 0.000 0.000 30.534 30.579 make_preconditioner 11 8.9 0.000 0.001 30.534 30.579 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.167 29.571 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.341 27.859 multiply_cannon_multrec 20056 15.6 13.144 15.987 22.631 26.138 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.445 26.064 qs_ot_get_derivative 117 11.6 0.003 0.010 24.255 24.716 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.350 21.896 apply_single 128 13.6 0.001 0.001 20.350 21.896 ot_diis_step 117 11.6 0.017 0.018 19.426 19.427 qs_ot_get_p 128 10.4 0.001 0.003 18.389 18.839 make_m2s 5014 13.6 0.077 0.083 14.547 15.990 make_images 5014 14.6 1.149 1.244 14.312 15.752 multiply_cannon_sync_h2d 20056 15.6 13.623 15.356 13.623 15.356 cp_fm_cholesky_invert 11 10.9 14.128 14.137 14.128 14.137 qs_ot_p2m_diag 83 11.4 0.264 0.271 14.042 14.049 sum_up_and_integrate 128 10.3 0.002 0.004 13.585 13.611 integrate_v_rspace 128 11.3 0.003 0.004 13.525 13.553 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.977 13.012 calculate_rho_elec 128 8.7 0.132 0.146 12.976 13.011 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.996 12.997 init_scf_run 11 5.9 0.000 0.001 10.565 10.567 scf_env_initial_rho_setup 11 6.9 0.003 0.016 10.564 10.566 make_images_data 5014 15.6 0.061 0.072 8.683 10.563 dbcsr_mm_accdrv_process 41502 16.2 5.756 6.796 8.943 10.151 hybrid_alltoall_any 5200 16.5 0.447 2.021 7.621 10.063 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.844 9.868 cp_fm_diag_elpa_base 83 14.4 9.436 9.603 9.841 9.865 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.132 9.467 multiply_cannon_metrocomm4 17549 15.6 0.066 0.076 3.453 9.135 mp_irecv_dv 50230 16.2 3.322 8.880 3.322 8.880 grid_integrate_task_list 128 12.3 7.284 7.708 7.284 7.708 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.487 7.590 wfi_extrapolate 11 7.9 0.003 0.017 7.385 7.386 cp_fm_cholesky_decompose 22 10.9 7.240 7.292 7.240 7.292 cp_fm_upper_to_full 105 14.8 5.738 7.270 5.738 7.270 fft_wrap_pw1pw2_140 523 12.2 0.283 0.292 6.609 6.725 density_rs2pw 128 9.7 0.005 0.007 6.329 6.695 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.070 6.462 dbcsr_complete_redistribute 395 12.7 1.184 1.246 4.616 6.402 mp_sum_l 7950 12.9 4.086 6.172 4.086 6.172 calculate_dm_sparse 128 9.5 0.001 0.001 5.873 5.990 fft3d_ps 1291 13.7 2.757 2.983 5.797 5.871 grid_collocate_task_list 128 9.7 5.147 5.573 5.147 5.573 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.532 5.539 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.483 5.265 mp_alltoall_d11v 2415 14.1 4.341 4.968 4.341 4.968 mp_allgather_i34 2507 14.6 1.642 4.823 1.642 4.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.580 4.723 transfer_fm_to_dbcsr 11 9.9 0.018 0.023 2.348 4.113 potential_pw2rs 128 12.3 0.020 0.022 4.011 4.025 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.881 3.919 mp_alltoall_i22 716 14.1 1.992 3.881 1.992 3.881 qs_energies_init_hamiltonians 11 5.9 0.013 0.021 3.781 3.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.250000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=923.545455, yerr=8.794251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.150767E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922014800 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509627. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.133 0.156 188.218 188.219 qs_mol_dyn_low 1 2.0 0.003 0.004 187.290 187.311 qs_forces 11 3.9 0.004 0.004 187.177 187.181 qs_energies 11 4.9 0.002 0.002 180.065 180.074 scf_env_do_scf 11 5.9 0.001 0.001 163.359 163.369 velocity_verlet 10 3.0 0.009 0.011 125.013 125.016 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 116.938 116.939 dbcsr_multiply_generic 2529 12.6 0.192 0.196 80.982 81.652 qs_scf_new_mos 118 7.6 0.001 0.001 80.925 81.245 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.924 81.245 ot_scf_mini 118 9.6 0.003 0.003 76.329 76.656 multiply_cannon 2529 13.6 0.555 0.585 56.149 59.644 multiply_cannon_loop 2529 14.6 1.864 1.914 52.472 54.043 init_scf_loop 11 6.9 0.000 0.000 46.301 46.301 ot_mini 118 10.6 0.001 0.001 43.180 43.491 prepare_preconditioner 11 7.9 0.000 0.000 42.228 42.250 make_preconditioner 11 8.9 0.000 0.000 42.228 42.250 make_full_inverse_cholesky 11 9.9 0.014 0.023 35.756 40.786 multiply_cannon_multrec 30348 15.6 13.748 18.983 26.747 31.637 rebuild_ks_matrix 129 8.3 0.001 0.001 29.487 29.811 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.486 29.811 mp_waitall_1 149172 16.7 17.159 27.521 17.159 27.521 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.612 26.898 qs_ot_get_derivative 118 11.6 0.001 0.002 23.285 23.619 make_m2s 5058 13.6 0.093 0.096 20.436 21.526 make_images 5058 14.6 1.984 2.271 20.134 21.230 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.296 19.899 apply_single 129 13.6 0.001 0.001 19.296 19.898 ot_diis_step 118 11.6 0.018 0.020 19.767 19.768 qs_ot_get_p 129 10.4 0.001 0.002 19.134 19.468 cp_fm_upper_to_full 106 14.8 11.470 16.909 11.470 16.909 cp_fm_cholesky_invert 11 10.9 16.042 16.051 16.042 16.051 multiply_cannon_metrocomm3 30348 15.6 0.050 0.053 6.383 15.175 qs_ot_p2m_diag 84 11.4 0.346 0.394 14.991 15.043 sum_up_and_integrate 129 10.3 0.002 0.003 13.766 13.799 integrate_v_rspace 129 11.3 0.003 0.004 13.706 13.741 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.555 13.557 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.242 13.292 calculate_rho_elec 129 8.7 0.176 0.191 13.242 13.292 dbcsr_mm_accdrv_process 62780 16.2 8.451 9.130 12.563 13.051 make_images_data 5058 15.6 0.065 0.071 10.924 12.852 dbcsr_complete_redistribute 397 12.7 1.497 1.598 9.045 12.835 hybrid_alltoall_any 5245 16.5 0.534 2.231 9.917 12.130 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.699 11.479 multiply_cannon_sync_h2d 30348 15.6 10.481 11.444 10.481 11.444 init_scf_run 11 5.9 0.000 0.001 10.757 10.758 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.757 10.758 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.446 10.457 cp_fm_diag_elpa_base 84 14.4 9.474 9.801 10.439 10.454 transfer_fm_to_dbcsr 11 9.9 0.002 0.007 6.450 10.165 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.839 10.074 mp_alltoall_i22 720 14.1 5.619 9.384 5.619 9.384 grid_integrate_task_list 129 12.3 7.568 7.923 7.568 7.923 fft_wrap_pw1pw2 1301 11.7 0.019 0.023 7.796 7.891 wfi_extrapolate 11 7.9 0.001 0.001 7.694 7.694 cp_fm_cholesky_decompose 22 10.9 7.509 7.613 7.509 7.613 multiply_cannon_metrocomm4 25290 15.6 0.079 0.089 2.819 7.286 mp_irecv_dv 76751 16.2 2.664 7.009 2.664 7.009 fft_wrap_pw1pw2_140 527 12.2 0.326 0.333 6.860 6.985 density_rs2pw 129 9.7 0.005 0.005 6.231 6.726 calculate_dm_sparse 129 9.5 0.001 0.001 6.298 6.393 fft3d_ps 1301 13.7 2.955 3.034 5.969 6.022 grid_collocate_task_list 129 9.7 5.329 5.651 5.329 5.651 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.490 5.544 mp_alltoall_d11v 2429 14.1 4.919 5.261 4.919 5.261 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.441 4.545 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.382 4.483 qs_energies_init_hamiltonians 11 5.9 0.002 0.010 4.447 4.448 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.300 4.369 potential_pw2rs 129 12.3 0.022 0.023 3.978 3.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.219000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.727273, yerr=23.981398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.549140E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.050 168.627 168.633 qs_mol_dyn_low 1 2.0 0.003 0.004 168.198 168.211 qs_forces 11 3.9 0.004 0.005 168.100 168.104 qs_energies 11 4.9 0.023 0.024 160.727 160.735 scf_env_do_scf 11 5.9 0.001 0.001 143.583 143.594 velocity_verlet 10 3.0 0.012 0.013 111.135 111.139 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 108.179 108.180 dbcsr_multiply_generic 2527 12.6 0.185 0.193 72.476 73.015 qs_scf_new_mos 118 7.6 0.001 0.001 71.731 71.832 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.731 71.831 ot_scf_mini 118 9.6 0.003 0.004 67.289 67.388 multiply_cannon 2527 13.6 0.563 0.591 53.920 56.976 multiply_cannon_loop 2527 14.6 0.818 0.856 50.760 51.932 ot_mini 118 10.6 0.001 0.001 37.941 38.038 init_scf_loop 11 6.9 0.000 0.000 35.256 35.257 mp_waitall_1 126780 16.7 24.862 31.607 24.862 31.607 prepare_preconditioner 11 7.9 0.000 0.000 31.428 31.452 make_preconditioner 11 8.9 0.000 0.000 31.428 31.452 make_full_inverse_cholesky 11 9.9 0.015 0.028 29.237 29.489 rebuild_ks_matrix 129 8.3 0.001 0.001 29.110 29.274 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.110 29.274 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.569 26.719 multiply_cannon_multrec 10108 15.6 10.370 16.235 18.007 22.818 multiply_cannon_metrocomm3 10108 15.6 0.025 0.026 13.070 20.671 qs_ot_get_derivative 118 11.6 0.002 0.002 20.503 20.599 cp_fm_cholesky_invert 11 10.9 17.880 17.886 17.880 17.886 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.473 17.685 apply_single 129 13.6 0.001 0.001 17.473 17.685 ot_diis_step 118 11.6 0.020 0.025 17.365 17.365 qs_ot_get_p 129 10.4 0.002 0.002 16.548 16.692 make_m2s 5054 13.6 0.064 0.070 14.655 15.668 make_images 5054 14.6 2.177 2.602 14.345 15.351 sum_up_and_integrate 129 10.3 0.002 0.002 13.816 13.862 integrate_v_rspace 129 11.3 0.004 0.004 13.755 13.803 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.536 13.565 calculate_rho_elec 129 8.7 0.259 0.270 13.536 13.564 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.811 12.827 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.661 11.661 multiply_cannon_sync_h2d 10108 15.6 10.795 11.352 10.795 11.352 init_scf_run 11 5.9 0.000 0.001 10.295 10.296 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.295 10.295 make_images_data 5054 15.6 0.056 0.066 8.379 9.807 hybrid_alltoall_any 5240 16.5 0.851 3.813 8.235 9.406 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.794 8.805 cp_fm_diag_elpa_base 83 14.4 8.561 8.637 8.789 8.800 grid_integrate_task_list 129 12.3 7.850 8.133 7.850 8.133 fft_wrap_pw1pw2 1301 11.7 0.018 0.020 8.082 8.103 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.968 8.028 dbcsr_mm_accdrv_process 20926 16.1 3.232 4.253 7.289 8.004 cp_fm_cholesky_decompose 22 10.9 7.771 7.871 7.771 7.871 multiply_cannon_metrocomm1 10108 15.6 0.030 0.030 4.213 7.524 wfi_extrapolate 11 7.9 0.001 0.001 7.273 7.273 fft_wrap_pw1pw2_140 527 12.2 0.437 0.444 7.129 7.158 density_rs2pw 129 9.7 0.005 0.005 6.126 6.527 calculate_dm_sparse 129 9.5 0.001 0.001 6.090 6.177 fft3d_ps 1301 13.7 3.133 3.241 6.042 6.072 multiply_cannon_metrocomm4 7581 15.6 0.025 0.029 1.883 5.977 mp_irecv_dv 29086 15.9 1.842 5.877 1.842 5.877 grid_collocate_task_list 129 9.7 5.616 5.873 5.616 5.873 dbcsr_complete_redistribute 395 12.7 2.113 2.205 5.306 5.669 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.247 5.248 mp_alltoall_d11v 2423 14.1 4.724 5.212 4.724 5.212 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.143 5.164 mp_allgather_i34 2527 14.6 1.334 4.644 1.334 4.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.052 4.105 potential_pw2rs 129 12.3 0.026 0.026 3.929 3.934 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.589 3.880 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.685 3.783 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.452 3.766 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.664 3.714 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.432 3.500 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.453 3.468 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.633000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1456.727273, yerr=21.443757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.214131E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 59207. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954382. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.104 286.599 286.606 qs_mol_dyn_low 1 2.0 0.003 0.004 285.828 285.861 qs_forces 11 3.9 0.015 0.024 285.724 285.727 qs_energies 11 4.9 0.032 0.036 276.980 276.991 scf_env_do_scf 11 5.9 0.001 0.001 254.391 254.400 velocity_verlet 10 3.0 0.015 0.017 207.099 207.108 scf_env_do_scf_inner_loop 116 6.6 0.010 0.015 130.679 130.680 init_scf_loop 11 6.9 0.000 0.000 123.433 123.435 prepare_preconditioner 11 7.9 0.000 0.000 118.687 118.705 make_preconditioner 11 8.9 0.000 0.000 118.687 118.705 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.034 115.830 qs_scf_new_mos 116 7.6 0.001 0.001 88.456 88.493 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.455 88.492 ot_scf_mini 116 9.6 0.004 0.004 83.655 83.660 dbcsr_multiply_generic 2485 12.5 0.213 0.223 81.313 81.745 cp_fm_upper_to_full 104 14.8 53.003 75.617 53.003 75.617 multiply_cannon 2485 13.5 0.674 0.701 58.565 60.225 multiply_cannon_loop 2485 14.5 1.035 1.061 54.567 55.685 ot_mini 116 10.6 0.001 0.001 43.830 43.839 dbcsr_complete_redistribute 393 12.7 3.990 4.010 29.548 42.801 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.065 39.317 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.610 36.811 mp_alltoall_i22 712 14.1 21.342 34.531 21.342 34.531 rebuild_ks_matrix 127 8.3 0.001 0.001 33.316 33.326 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.315 33.325 cp_fm_cholesky_invert 11 10.9 32.996 33.002 32.996 33.002 mp_waitall_1 102768 16.8 27.599 31.826 27.599 31.826 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.979 30.996 qs_ot_get_p 127 10.4 0.007 0.007 24.859 24.887 qs_ot_get_derivative 116 11.6 0.002 0.002 24.663 24.669 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.867 20.897 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.669 19.916 make_m2s 4970 13.5 0.073 0.075 18.289 19.781 make_images 4970 14.5 3.058 3.262 17.810 19.304 ot_diis_step 116 11.6 0.021 0.022 19.143 19.143 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.104 19.105 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.484 18.572 apply_single 127 13.6 0.001 0.001 18.484 18.572 multiply_cannon_multrec 9940 15.5 10.237 12.040 18.138 18.390 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.987 15.989 cp_fm_diag_elpa_base 82 14.4 11.609 13.164 15.983 15.984 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.656 15.667 calculate_rho_elec 127 8.7 0.479 0.479 15.655 15.667 sum_up_and_integrate 127 10.3 0.002 0.002 15.528 15.619 integrate_v_rspace 127 11.3 0.004 0.005 15.465 15.557 multiply_cannon_sync_h2d 9940 15.5 14.202 14.262 14.202 14.262 hybrid_alltoall_any 5155 16.4 1.302 3.005 10.552 12.899 make_images_data 4970 15.5 0.062 0.067 10.253 12.444 init_scf_run 11 5.9 0.000 0.001 12.035 12.035 scf_env_initial_rho_setup 11 6.9 0.009 0.013 12.035 12.035 fft_wrap_pw1pw2 1281 11.7 0.020 0.021 9.856 9.879 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.530 9.535 dbcsr_mm_accdrv_process 20590 16.0 4.351 6.282 7.653 9.522 wfi_extrapolate 11 7.9 0.001 0.001 8.916 8.916 cp_fm_cholesky_decompose 22 10.9 8.796 8.809 8.796 8.809 fft_wrap_pw1pw2_140 519 12.2 0.819 0.821 8.742 8.761 grid_integrate_task_list 127 12.3 8.503 8.700 8.503 8.700 qs_energies_init_hamiltonians 11 5.9 0.034 0.047 8.399 8.410 fft3d_ps 1281 13.7 3.963 3.995 7.166 7.198 mp_alltoall_d11v 2401 14.1 6.815 6.910 6.815 6.910 density_rs2pw 127 9.7 0.005 0.005 6.747 6.806 calculate_dm_sparse 127 9.5 0.001 0.001 6.656 6.710 grid_collocate_task_list 127 9.7 6.367 6.449 6.367 6.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.384 6.439 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.129 6.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.606000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2885.454545, yerr=176.990171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 1.263755E+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 19057443. 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 0.034 0.093 85.493 85.496 qs_energies 1 2.0 0.000 0.001 84.686 84.696 ls_scf 1 3.0 0.000 0.000 83.790 83.799 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.685 72.849 multiply_cannon 111 7.7 0.017 0.020 55.968 56.928 multiply_cannon_loop 111 8.7 0.226 0.248 52.592 53.742 ls_scf_main 1 4.0 0.000 0.000 52.373 52.374 density_matrix_trs4 2 5.0 0.002 0.003 46.786 46.866 ls_scf_init_scf 1 4.0 0.000 0.000 28.359 28.361 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.199 27.246 mp_waitall_1 11031 10.9 22.433 25.773 22.433 25.773 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.010 25.025 multiply_cannon_multrec 2664 9.7 8.137 8.967 15.609 17.453 multiply_cannon_sync_h2d 2664 9.7 13.557 15.484 13.557 15.484 make_m2s 222 7.7 0.008 0.011 13.107 13.591 make_images 222 8.7 0.099 0.109 13.085 13.571 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.637 13.302 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.484 8.821 make_images_data 222 9.7 0.004 0.005 7.656 8.186 dbcsr_mm_accdrv_process 4760 10.4 0.586 0.704 7.089 8.048 hybrid_alltoall_any 227 10.6 0.216 1.843 6.729 7.910 dbcsr_mm_accdrv_process_sort 4760 11.4 6.304 7.188 6.304 7.188 calculate_norms 4752 9.8 5.513 6.194 5.513 6.194 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.047 5.202 mp_sum_l 887 5.1 3.094 4.448 3.094 4.448 make_images_sizes 222 9.7 0.000 0.000 0.697 3.749 mp_alltoall_i44 222 10.7 0.697 3.748 0.697 3.748 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.306 3.357 arnoldi_extremal 4 6.8 0.000 0.000 3.174 3.201 arnoldi_normal_ev 4 7.8 0.001 0.003 3.173 3.201 multiply_cannon_metrocomm4 2442 9.7 0.011 0.013 2.067 3.179 mp_irecv_dv 6231 10.9 2.051 3.158 2.051 3.158 build_subspace 16 8.4 0.009 0.012 3.083 3.085 ls_scf_post 1 4.0 0.000 0.000 3.057 3.067 ls_scf_store_result 1 5.0 0.000 0.000 2.873 2.916 dbcsr_special_finalize 555 9.7 0.005 0.006 2.392 2.737 dbcsr_merge_single_wm 555 10.7 0.458 0.580 2.384 2.729 make_images_pack 222 9.7 2.214 2.629 2.216 2.631 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.311 2.551 dbcsr_sort_data 658 11.4 2.185 2.502 2.185 2.502 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.459 2.071 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.321 2.430 buffer_matrices_ensure_size 222 8.7 1.757 2.048 1.757 2.048 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.841 1.843 rebuild_ks_matrix 3 7.3 0.000 0.000 1.822 1.824 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.020 1.822 1.824 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.496000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1135.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 2.244563E+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 0.071 0.172 91.507 91.577 qs_energies 1 2.0 0.000 0.000 90.493 90.516 ls_scf 1 3.0 0.002 0.015 88.537 88.561 dbcsr_multiply_generic 111 6.7 0.016 0.021 74.585 74.914 multiply_cannon 111 7.7 0.027 0.043 52.601 57.713 ls_scf_main 1 4.0 0.001 0.010 54.840 54.846 multiply_cannon_loop 111 8.7 0.135 0.147 49.810 53.677 density_matrix_trs4 2 5.0 0.002 0.003 49.154 49.356 ls_scf_init_scf 1 4.0 0.000 0.003 29.979 29.981 mp_waitall_1 9105 10.9 20.515 29.813 20.515 29.813 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.777 28.858 multiply_cannon_multrec 1332 9.7 13.373 16.756 22.748 27.228 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 26.367 26.380 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.226 20.459 make_m2s 222 7.7 0.006 0.008 15.203 15.739 make_images 222 8.7 1.371 1.695 15.172 15.709 dbcsr_mm_accdrv_process 4041 10.4 0.330 0.528 8.970 10.539 dbcsr_mm_accdrv_process_sort 4041 11.4 8.489 10.013 8.489 10.013 make_images_data 222 9.7 0.004 0.005 8.817 9.739 hybrid_alltoall_any 227 10.6 0.541 2.558 8.169 9.647 mp_sum_l 887 5.1 5.311 8.561 5.311 8.561 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.204 7.923 mp_irecv_dv 3311 11.0 3.183 7.865 3.183 7.865 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.039 6.774 calculate_norms 2376 9.8 6.070 6.649 6.070 6.649 multiply_cannon_sync_h2d 1332 9.7 4.940 6.271 4.940 6.271 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.934 5.167 arnoldi_extremal 4 6.8 0.000 0.000 4.683 4.700 arnoldi_normal_ev 4 7.8 0.001 0.005 4.683 4.700 build_subspace 16 8.4 0.014 0.021 4.407 4.412 ls_scf_post 1 4.0 0.001 0.010 3.717 3.729 ls_scf_store_result 1 5.0 0.000 0.000 3.432 3.538 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.179 3.399 dbcsr_matrix_vector_mult_local 304 10.0 2.770 3.254 2.772 3.256 mp_allgather_i34 111 8.7 0.787 3.062 0.787 3.062 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.118 2.913 ls_scf_dm_to_ks 2 5.0 0.001 0.013 2.594 2.680 dbcsr_data_new 4174 10.1 2.115 2.408 2.115 2.408 make_images_pack 222 9.7 1.823 2.124 1.826 2.126 dbcsr_sort_data 436 11.2 1.860 2.114 1.860 2.114 qs_energies_init_hamiltonians 1 3.0 0.007 0.058 1.950 1.950 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.904 1.906 rebuild_ks_matrix 3 7.3 0.000 0.000 1.891 1.893 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.024 1.891 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.577000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 2.968691E+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 0.040 0.139 95.580 95.582 qs_energies 1 2.0 0.001 0.012 94.084 94.095 ls_scf 1 3.0 0.002 0.014 92.596 92.601 dbcsr_multiply_generic 111 6.7 0.018 0.036 75.795 76.109 multiply_cannon 111 7.7 0.040 0.102 52.331 56.942 ls_scf_main 1 4.0 0.000 0.001 56.870 56.876 multiply_cannon_loop 111 8.7 0.116 0.130 49.542 53.667 density_matrix_trs4 2 5.0 0.007 0.078 50.990 51.170 mp_waitall_1 7281 11.0 23.719 33.277 23.719 33.277 ls_scf_init_scf 1 4.0 0.000 0.002 32.051 32.053 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.643 29.732 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.217 27.278 multiply_cannon_multrec 888 9.7 12.658 15.059 21.346 24.585 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.999 22.927 make_m2s 222 7.7 0.006 0.007 16.527 17.291 make_images 222 8.7 1.583 1.854 16.489 17.251 make_images_data 222 9.7 0.005 0.012 9.584 10.697 hybrid_alltoall_any 227 10.6 0.638 2.947 9.077 10.387 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.489 8.193 9.431 mp_sum_l 887 5.1 5.263 9.286 5.263 9.286 dbcsr_mm_accdrv_process_sort 3754 11.4 7.755 8.942 7.755 8.942 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.038 7.533 multiply_cannon_sync_h2d 888 9.7 5.985 7.175 5.985 7.175 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.451 7.053 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.779 7.049 mp_irecv_dv 2335 11.1 2.435 6.992 2.435 6.992 arnoldi_extremal 4 6.8 0.000 0.000 5.428 5.493 arnoldi_normal_ev 4 7.8 0.001 0.005 5.428 5.493 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.735 4.974 calculate_norms 1584 9.8 4.408 4.861 4.408 4.861 build_subspace 16 8.4 0.014 0.020 4.790 4.796 mp_allgather_i34 111 8.7 0.908 3.870 0.908 3.870 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.471 3.804 ls_scf_post 1 4.0 0.000 0.001 3.673 3.679 dbcsr_matrix_vector_mult_local 304 10.0 3.056 3.639 3.058 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.501 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.827 2.928 ls_scf_initial_guess 1 5.0 0.000 0.000 2.407 2.531 ls_scf_qs_atomic_guess 1 6.0 0.023 0.238 2.407 2.531 dbcsr_data_new 4116 9.9 2.107 2.425 2.107 2.425 dbcsr_sort_data 325 11.1 1.894 2.177 1.894 2.177 make_images_sizes 222 9.7 0.000 0.000 1.082 2.147 mp_alltoall_i44 222 10.7 1.082 2.147 1.082 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.096 2.099 rebuild_ks_matrix 3 7.3 0.000 0.000 2.078 2.081 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.018 2.078 2.081 mp_sum_d 1065 5.9 1.486 2.068 1.486 2.068 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.582000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2280.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 3.380519E+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 0.055 0.102 97.977 97.978 qs_energies 1 2.0 0.000 0.000 97.139 97.145 ls_scf 1 3.0 0.000 0.000 95.455 95.461 dbcsr_multiply_generic 111 6.7 0.016 0.017 79.112 79.417 ls_scf_main 1 4.0 0.000 0.000 59.356 59.357 multiply_cannon 111 7.7 0.048 0.109 51.857 56.704 density_matrix_trs4 2 5.0 0.002 0.003 53.262 53.362 multiply_cannon_loop 111 8.7 0.151 0.167 46.889 50.898 ls_scf_init_scf 1 4.0 0.000 0.000 32.837 32.840 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.607 31.703 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.145 29.158 mp_waitall_1 6369 11.0 22.785 28.694 22.785 28.694 multiply_cannon_multrec 1332 9.7 14.125 17.271 22.130 24.963 make_m2s 222 7.7 0.006 0.008 21.140 22.497 make_images 222 8.7 3.155 3.619 21.090 22.449 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.150 17.712 make_images_data 222 9.7 0.004 0.004 11.716 13.243 hybrid_alltoall_any 227 10.6 0.798 3.841 11.137 12.933 dbcsr_mm_accdrv_process 3641 10.4 0.301 0.509 7.648 9.173 mp_sum_l 887 5.1 4.592 9.167 4.592 9.167 dbcsr_mm_accdrv_process_sort 3641 11.4 7.199 8.683 7.199 8.683 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.541 7.365 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.751 6.133 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.098 6.094 multiply_cannon_sync_h2d 1332 9.7 5.414 6.005 5.414 6.005 mp_irecv_dv 3229 10.9 2.074 6.002 2.074 6.002 arnoldi_extremal 4 6.8 0.000 0.000 5.255 5.267 arnoldi_normal_ev 4 7.8 0.001 0.005 5.255 5.267 build_subspace 16 8.4 0.014 0.021 4.920 4.928 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.551 4.726 calculate_norms 2376 9.8 4.223 4.603 4.223 4.603 mp_allgather_i34 111 8.7 2.090 4.470 2.090 4.470 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.625 3.932 dbcsr_matrix_vector_mult_local 304 10.0 3.238 3.742 3.240 3.744 dbcsr_sort_data 658 11.4 3.114 3.531 3.114 3.531 dbcsr_special_finalize 555 9.7 0.006 0.007 2.881 3.283 dbcsr_merge_single_wm 555 10.7 0.543 0.672 2.873 3.276 ls_scf_post 1 4.0 0.000 0.000 3.261 3.267 ls_scf_store_result 1 5.0 0.000 0.000 2.990 3.058 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.987 3.025 dbcsr_data_release 10477 10.7 1.572 2.407 1.572 2.407 dbcsr_finalize 304 7.8 0.049 0.061 1.790 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.978000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 4.838728E+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 0.041 0.085 92.833 92.833 qs_energies 1 2.0 0.000 0.000 92.008 92.009 ls_scf 1 3.0 0.000 0.000 90.065 90.070 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.612 71.936 ls_scf_main 1 4.0 0.000 0.000 57.255 57.255 multiply_cannon 111 7.7 0.094 0.185 52.981 56.004 multiply_cannon_loop 111 8.7 0.087 0.093 50.344 52.092 density_matrix_trs4 2 5.0 0.002 0.003 50.283 50.359 ls_scf_init_scf 1 4.0 0.000 0.000 29.393 29.397 mp_waitall_1 5436 11.0 24.930 28.940 24.930 28.940 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.141 28.173 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.050 26.063 multiply_cannon_multrec 444 9.7 13.686 16.586 20.852 22.233 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.839 15.786 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.367 15.164 make_m2s 222 7.7 0.005 0.005 13.788 14.767 make_images 222 8.7 2.039 2.473 13.720 14.698 hybrid_alltoall_any 227 10.6 0.804 3.858 8.346 10.034 make_images_data 222 9.7 0.003 0.004 8.545 9.947 multiply_cannon_sync_h2d 444 9.7 6.708 8.138 6.708 8.138 dbcsr_mm_accdrv_process 3003 10.4 0.331 0.403 6.860 7.999 dbcsr_mm_accdrv_process_sort 3003 11.4 6.493 7.599 6.493 7.599 arnoldi_extremal 4 6.8 0.000 0.000 5.858 5.876 arnoldi_normal_ev 4 7.8 0.002 0.005 5.858 5.876 build_subspace 16 8.4 0.015 0.019 5.447 5.457 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.433 4.609 mp_sum_l 887 5.1 2.859 4.507 2.859 4.507 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.217 4.396 dbcsr_matrix_vector_mult_local 304 10.0 3.749 4.179 3.751 4.181 mp_allgather_i34 111 8.7 1.216 3.866 1.216 3.866 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.564 3.793 mp_irecv_dv 1241 11.2 1.550 3.760 1.550 3.760 calculate_norms 792 9.8 3.595 3.706 3.595 3.706 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.491 3.573 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.041 3.559 ls_scf_post 1 4.0 0.000 0.000 3.417 3.419 make_images_sizes 222 9.7 0.000 0.000 0.853 3.346 mp_alltoall_i44 222 10.7 0.852 3.346 0.852 3.346 ls_scf_store_result 1 5.0 0.000 0.000 3.193 3.232 dbcsr_finalize 304 7.8 0.062 0.077 2.205 2.293 dbcsr_data_new 4608 9.7 1.775 2.257 1.775 2.257 dbcsr_merge_all 275 8.9 0.480 0.522 2.061 2.128 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.057 2.058 rebuild_ks_matrix 3 7.3 0.000 0.000 2.024 2.025 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.024 2.025 qs_energies_init_hamiltonians 1 3.0 0.009 0.026 1.927 1.927 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.833000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 8.998531E+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 0.080 0.106 105.804 105.804 qs_energies 1 2.0 0.000 0.000 104.375 104.378 ls_scf 1 3.0 0.000 0.000 101.395 101.398 dbcsr_multiply_generic 111 6.7 0.024 0.026 75.241 75.367 ls_scf_main 1 4.0 0.000 0.000 63.744 63.745 density_matrix_trs4 2 5.0 0.002 0.003 54.953 55.000 multiply_cannon 111 7.7 0.172 0.295 48.936 51.307 multiply_cannon_loop 111 8.7 0.098 0.101 45.857 46.218 ls_scf_init_scf 1 4.0 0.000 0.000 33.837 33.838 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.424 32.447 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.010 29.606 29.611 mp_waitall_1 4527 11.1 22.020 25.754 22.020 25.754 make_m2s 222 7.7 0.005 0.005 22.577 23.745 make_images 222 8.7 3.582 3.897 22.469 23.634 multiply_cannon_multrec 444 9.7 17.836 18.393 22.539 23.005 hybrid_alltoall_any 227 10.6 1.654 3.628 12.724 15.625 make_images_data 222 9.7 0.003 0.004 12.907 15.146 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.513 10.755 multiply_cannon_sync_h2d 444 9.7 8.792 8.844 8.792 8.844 arnoldi_extremal 4 6.8 0.000 0.000 7.483 7.492 arnoldi_normal_ev 4 7.8 0.004 0.012 7.483 7.492 build_subspace 16 8.4 0.026 0.036 6.938 6.949 dbcsr_matrix_vector_mult 304 9.0 0.018 0.035 5.593 5.723 dbcsr_matrix_vector_mult_local 304 10.0 5.160 5.430 5.162 5.433 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.974 5.221 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.902 4.984 dbcsr_mm_accdrv_process 1814 10.4 0.246 0.373 4.519 4.642 dbcsr_mm_accdrv_process_sort 1814 11.4 4.180 4.314 4.180 4.314 ls_scf_post 1 4.0 0.000 0.000 3.813 3.816 mp_allgather_i34 111 8.7 1.183 3.733 1.183 3.733 make_images_sizes 222 9.7 0.000 0.000 1.494 3.641 mp_alltoall_i44 222 10.7 1.494 3.641 1.494 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.525 3.562 calculate_norms 792 9.8 3.228 3.270 3.228 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.178 dbcsr_merge_all 275 8.9 0.884 0.917 2.864 2.957 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.950 2.950 dbcsr_complete_redistribute 5 7.6 1.428 1.469 2.738 2.846 matrix_ls_to_qs 2 6.0 0.000 0.000 2.398 2.518 dbcsr_sort_data 325 11.1 2.440 2.511 2.440 2.511 dbcsr_new_transposed 4 7.5 0.245 0.265 2.330 2.354 dbcsr_data_new 6591 9.6 1.823 2.296 1.823 2.296 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.278 2.280 dbcsr_frobenius_norm 74 6.6 2.054 2.134 2.195 2.226 rebuild_ks_matrix 3 7.3 0.000 0.000 2.211 2.212 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.211 2.212 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.204 dbcsr_add_anytype 103 7.2 0.859 0.892 2.126 2.204 dbcsr_data_release 12724 10.6 1.985 2.202 1.985 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.804000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7eeddba65004ce63710e802b2136684b4a3d7a4d_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 585.527296E+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 993460487. 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 0.062 0.160 225.796 225.811 qs_mol_dyn_low 1 2.0 0.009 0.064 224.167 224.200 qs_forces 5 3.8 0.004 0.008 223.876 223.886 qs_energies 5 4.8 0.005 0.024 220.944 220.969 scf_env_do_scf 5 5.8 0.000 0.001 205.472 205.477 scf_env_do_scf_inner_loop 105 6.6 0.022 0.101 179.266 179.276 qs_scf_new_mos 105 7.6 0.000 0.001 140.447 140.658 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.446 140.658 dbcsr_multiply_generic 1445 12.2 0.128 0.136 133.801 134.133 ot_scf_mini 105 9.6 0.003 0.004 130.500 130.649 multiply_cannon 1445 13.2 0.272 0.287 114.878 116.894 multiply_cannon_loop 1445 14.2 2.843 2.985 113.190 114.258 velocity_verlet 4 3.0 0.014 0.063 104.978 104.983 ot_mini 105 10.6 0.001 0.001 59.941 60.071 multiply_cannon_multrec 69360 15.2 29.809 35.049 39.772 45.459 mp_waitall_1 488190 16.1 34.813 41.883 34.813 41.883 qs_ot_get_p 112 10.4 0.001 0.001 40.594 40.841 qs_ot_get_derivative 55 11.6 0.001 0.012 38.164 38.285 multiply_cannon_sync_h2d 69360 15.2 29.108 33.383 29.108 33.383 multiply_cannon_metrocomm3 69360 15.2 0.197 0.206 25.632 33.128 qs_ot_p2m_diag 40 11.0 0.020 0.041 29.698 29.803 rebuild_ks_matrix 110 8.4 0.000 0.000 28.183 28.387 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.060 28.182 28.387 cp_dbcsr_syevd 40 12.0 0.002 0.003 26.507 26.509 init_scf_loop 7 6.6 0.000 0.000 26.176 26.178 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.903 26.083 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.058 23.271 apply_single 62 13.6 0.000 0.000 23.058 23.271 cp_fm_syevd 40 13.0 0.000 0.001 21.355 21.485 prepare_preconditioner 7 7.6 0.000 0.000 21.319 21.362 make_preconditioner 7 8.6 0.000 0.000 21.319 21.362 ot_new_cg_direction 55 11.6 0.001 0.001 21.049 21.049 cp_fm_redistribute_end 40 14.0 8.333 16.622 8.337 16.623 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.190 16.610 calculate_rho_elec 110 8.6 0.029 0.038 16.189 16.610 cp_fm_syevd_base 40 14.0 8.278 16.568 8.278 16.568 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.765 14.890 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.152 14.224 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.718 13.811 mp_sum_l 4764 12.2 12.231 12.990 12.231 12.990 init_scf_run 5 5.8 0.000 0.001 12.744 12.745 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.744 12.745 calculate_dm_sparse 110 9.5 0.000 0.001 11.592 11.729 density_rs2pw 110 9.6 0.004 0.005 10.431 10.818 dbcsr_mm_accdrv_process 154766 15.8 6.206 6.376 9.831 10.688 fft_wrap_pw1pw2 1425 12.5 0.019 0.026 10.284 10.579 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.281 10.398 cp_fm_cholesky_invert 7 10.6 9.937 9.945 9.937 9.945 qs_vxc_create 110 10.4 0.002 0.002 9.733 9.757 check_diag 80 13.5 8.684 8.935 9.484 9.614 fft_wrap_pw1pw2_240 915 14.0 0.200 0.228 9.022 9.207 multiply_cannon_metrocomm1 69360 15.2 0.093 0.099 4.776 8.764 acc_transpose_blocks 69360 15.2 0.344 0.360 7.964 8.733 sum_up_and_integrate 60 10.3 0.001 0.019 8.227 8.242 integrate_v_rspace 60 11.3 0.002 0.002 8.209 8.226 fft3d_pb 915 15.0 2.254 2.466 7.987 8.174 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 8.018 8.035 calculate_first_density_matrix 1 7.0 0.000 0.002 7.957 7.974 transfer_rs2pw 445 10.6 0.008 0.009 7.537 7.970 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.657 7.714 xc_rho_set_and_dset_create 110 12.4 0.076 0.095 6.875 7.119 make_full_single_inverse 7 9.6 0.001 0.002 6.899 6.941 make_m2s 2890 13.2 0.078 0.086 6.296 6.921 make_images 2890 14.2 0.238 0.259 6.189 6.813 xc_vxc_pw_create 60 11.3 0.036 0.041 6.493 6.520 acc_transpose_blocks_kernels 69360 16.2 0.843 0.878 5.220 5.880 xc_pw_derive 510 13.4 0.005 0.006 5.577 5.657 mp_alltoall_z22v 2340 16.7 5.160 5.361 5.160 5.361 jit_kernel_transpose 5 15.0 4.377 5.020 4.377 5.020 mp_waitany 7680 13.5 4.526 4.973 4.526 4.973 multiply_cannon_metrocomm4 67915 15.2 0.181 0.197 2.022 4.694 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=225.811000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.600000, yerr=2.332381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7eeddba65004ce63710e802b2136684b4a3d7a4d Summary: empty Status: OK