=== 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: 4955cfc0d24997d34b1dc70cd9db47ebb77d3489 ################# 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, SIRIUS 7.4.3, 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: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/01 job id: 50504858 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/02 job id: 50504862 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/03 job id: 50504863 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/04 job id: 50504865 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/05 job id: 50504867 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/06 job id: 50504869 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/07 job id: 50504870 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/08 job id: 50504871 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/09 job id: 50504873 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/10 job id: 50504875 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/11 job id: 50504876 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/12 job id: 50504877 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/13 job id: 50504878 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/14 job id: 50504879 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/15 job id: 50504882 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/16 job id: 50504883 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/17 job id: 50504884 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/18 job id: 50504885 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/19 job id: 50504886 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/20 job id: 50504888 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/21 job id: 50504890 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/22 job id: 50504891 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/23 job id: 50504892 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/24 job id: 50504893 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/25 job id: 50504894 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/26 job id: 50504895 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/27 job id: 50504897 --- 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/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.050 135.318 135.319 farming_run 1 2.0 134.800 134.803 135.255 135.259 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.504973E+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 0.026 116.207 116.207 qs_energies 1 2.0 0.000 0.000 116.013 116.014 mp2_main 1 3.0 0.000 0.000 113.806 113.808 mp2_gpw_main 1 4.0 0.028 0.034 112.719 112.720 mp2_ri_gpw_compute_in 1 5.0 0.173 0.187 93.368 93.810 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.631 55.074 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.034 46.180 get_2c_integrals 1 6.0 0.008 0.009 37.256 38.565 integrate_v_rspace 273 8.0 0.435 0.453 24.624 29.550 pw_transfer 6555 10.6 0.377 0.396 26.802 27.347 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.427 26.061 grid_integrate_task_list 273 9.0 20.532 25.914 20.532 25.914 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.031 22.972 23.575 compute_2c_integrals 1 7.0 0.002 0.002 19.455 19.457 rpa_ri_compute_en 1 5.0 0.019 0.020 19.245 19.420 compute_2c_integrals_loop_lm 1 8.0 0.001 0.004 18.566 19.174 mp2_eri_2c_integrate_gpw 1 9.0 2.375 2.423 18.564 19.173 cp_fm_cholesky_decompose 12 8.2 17.774 19.115 17.774 19.115 cholesky_decomp 1 7.0 0.000 0.000 16.645 17.976 fft3d_s 5443 13.4 16.094 16.499 16.115 16.519 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.313 10.771 15.313 calculate_wavefunction 272 8.0 5.396 5.528 12.319 12.958 rpa_num_int 1 6.0 0.000 0.001 10.707 10.717 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.688 10.697 calc_mat_Q 8 8.0 0.000 0.000 9.509 9.602 contract_S_to_Q 8 9.0 0.000 0.000 8.929 9.024 calc_potential_gpw 544 9.5 0.004 0.005 8.237 8.746 parallel_gemm_fm 14 9.1 0.000 0.000 8.512 8.607 parallel_gemm_fm_cosma 14 10.1 8.512 8.607 8.512 8.607 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.075 8.468 potential_pw2rs 545 10.0 0.107 0.109 7.485 8.110 create_integ_mat 1 6.0 0.016 0.030 8.005 8.015 collocate_single_gaussian 272 10.0 0.039 0.041 7.312 7.638 array2fm 1 7.0 0.000 0.000 6.883 7.403 pw_scatter_s 2720 13.7 4.328 4.469 4.328 4.469 pw_gather_s 2722 13.2 3.458 3.815 3.458 3.815 array2fm_buffer_send 1 8.0 2.913 3.205 2.913 3.205 pw_poisson_solve 545 10.5 1.120 1.200 2.199 2.491 mp_sync 25 8.8 1.326 2.478 1.326 2.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.720439, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2822.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.035 419.695 419.696 farming_run 1 2.0 418.876 418.879 419.660 419.662 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.240216E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 216.399 216.400 qs_energies 1 2.0 0.000 0.000 216.172 216.178 scf_env_do_scf 1 3.0 0.000 0.000 114.320 114.320 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.384 113.392 rebuild_ks_matrix 4 6.0 0.000 0.000 113.383 113.390 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.383 113.390 hfx_ks_matrix 4 8.0 0.001 0.001 113.034 113.039 integrate_four_center 4 9.0 0.151 0.458 113.033 113.038 integrate_four_center_main 4 10.0 0.113 0.500 101.741 104.365 integrate_four_center_bin 266 11.0 101.628 104.150 101.628 104.150 mp2_main 1 3.0 0.000 0.000 101.563 101.569 mp2_gpw_main 1 4.0 0.056 0.082 100.650 100.657 init_scf_loop 1 4.0 0.000 0.000 96.463 96.463 mp2_ri_gpw_compute_in 1 5.0 0.066 0.068 73.907 74.853 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.723 54.673 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.519 46.380 integrate_v_rspace 95 8.0 0.396 0.561 27.995 32.707 pw_transfer 2240 10.6 0.146 0.167 29.482 29.875 ao_to_mo_and_store_B_mult_1 91 7.0 10.537 28.923 10.537 28.923 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.495 28.908 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 26.588 28.642 grid_integrate_task_list 95 9.0 23.414 28.358 23.414 28.358 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.249 26.603 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.899 24.983 24.993 get_2c_integrals 1 6.0 0.000 0.000 20.067 20.128 compute_2c_integrals 1 7.0 0.002 0.003 19.054 19.070 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.608 18.923 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.931 18.607 18.922 fft3d_s 1823 13.4 18.472 18.858 18.485 18.872 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.855 17.855 calculate_wavefunction 91 8.0 2.014 2.054 9.617 9.833 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.584 8.738 9.225 potential_pw2rs 186 10.0 0.033 0.035 8.398 9.027 local_gemm 172 8.0 8.183 8.659 8.183 8.659 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.142 8.470 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.516 8.012 8.451 calc_potential_gpw 182 9.5 0.002 0.002 7.915 8.193 collocate_single_gaussian 91 10.0 0.017 0.026 7.744 7.925 integrate_four_center_load 4 10.0 0.000 0.000 6.764 6.768 hfx_load_balance 1 11.0 0.000 0.000 6.764 6.768 mp_sync 37 10.5 3.291 6.612 3.291 6.612 mp_sendrecv_dm3 2068 8.0 6.053 6.474 6.053 6.474 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.433 6.344 6.433 pw_gather_s 912 13.2 4.473 4.953 4.473 4.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.653023, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 453.586944E+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 778127. 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 0.020 0.051 54.472 54.473 qs_mol_dyn_low 1 2.0 0.003 0.003 54.203 54.211 qs_forces 11 3.9 0.002 0.002 54.143 54.145 qs_energies 11 4.9 0.002 0.002 52.628 52.663 scf_env_do_scf 11 5.9 0.000 0.001 46.702 46.702 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 44.626 44.626 qs_scf_new_mos 108 7.5 0.000 0.001 34.585 34.899 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.584 34.898 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.068 34.525 ot_scf_mini 108 9.5 0.002 0.002 32.903 33.107 multiply_cannon 2286 13.5 0.188 0.204 26.379 28.176 multiply_cannon_loop 2286 14.5 1.834 1.947 25.714 27.532 velocity_verlet 10 3.0 0.001 0.002 27.098 27.098 ot_mini 108 10.5 0.001 0.001 19.880 20.130 qs_ot_get_derivative 108 11.5 0.001 0.001 16.815 17.034 mp_waitall_1 245248 16.5 8.702 14.792 8.702 14.792 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.193 13.027 multiply_cannon_multrec 54864 15.5 3.625 5.786 7.563 10.972 qs_ot_get_p 119 10.4 0.001 0.002 8.229 8.496 rebuild_ks_matrix 119 8.3 0.000 0.000 7.852 7.998 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.852 7.998 mp_sum_l 7287 12.8 5.711 7.459 5.711 7.459 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.911 7.043 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.017 6.490 multiply_cannon_sync_h2d 54864 15.5 5.131 6.080 5.131 6.080 dbcsr_mm_accdrv_process 76910 16.1 1.850 2.923 3.851 5.521 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.386 5.496 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.318 5.361 init_scf_run 11 5.9 0.000 0.001 4.655 4.655 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.655 4.655 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.580 4.580 sum_up_and_integrate 119 10.3 0.001 0.002 4.449 4.455 integrate_v_rspace 119 11.3 0.002 0.002 4.438 4.445 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.373 4.373 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.154 4.367 calculate_rho_elec 119 8.7 0.011 0.016 4.154 4.366 cp_fm_redistribute_end 50 14.0 2.228 4.346 2.235 4.349 cp_fm_diag_elpa_base 50 14.0 2.108 4.242 2.112 4.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.942 3.136 apply_single 119 13.6 0.000 0.000 2.942 3.136 calculate_dm_sparse 119 9.5 0.000 0.000 2.903 3.071 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.845 3.004 ot_diis_step 108 11.5 0.006 0.006 2.776 2.776 acc_transpose_blocks 54864 15.5 0.224 0.247 2.217 2.752 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.563 2.628 jit_kernel_multiply 13 15.8 1.937 2.512 1.937 2.512 calculate_first_density_matrix 1 7.0 0.000 0.000 2.407 2.419 density_rs2pw 119 9.7 0.004 0.004 2.141 2.266 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.200 2.202 wfi_extrapolate 11 7.9 0.001 0.001 2.170 2.170 grid_integrate_task_list 119 12.3 2.011 2.103 2.011 2.103 init_scf_loop 11 6.9 0.000 0.000 2.058 2.058 mp_sum_d 4137 12.0 1.416 1.971 1.416 1.971 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.907 1.966 potential_pw2rs 119 12.3 0.004 0.004 1.823 1.834 pw_transfer 1439 11.6 0.051 0.059 1.670 1.742 make_m2s 4572 13.5 0.053 0.055 1.663 1.712 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.595 1.669 mp_alltoall_d11v 2130 13.8 1.342 1.645 1.342 1.645 make_images 4572 14.5 0.133 0.139 1.581 1.629 acc_transpose_blocks_sync 164592 16.5 1.212 1.456 1.212 1.456 transfer_rs2pw 487 10.6 0.005 0.007 1.344 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.397 1.449 mp_waitany 12084 13.8 1.247 1.443 1.247 1.443 grid_collocate_task_list 119 9.7 1.352 1.422 1.352 1.422 fft3d_ps 1201 14.6 0.368 0.473 1.306 1.375 transfer_pw2rs 487 13.2 0.006 0.006 1.334 1.344 fft_wrap_pw1pw2_140 487 13.2 0.110 0.118 1.233 1.307 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.790 1.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.473000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 491.520000E+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 1031898. 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 0.008 0.028 39.426 39.427 qs_mol_dyn_low 1 2.0 0.003 0.003 39.215 39.223 qs_forces 11 3.9 0.002 0.003 39.154 39.155 qs_energies 11 4.9 0.001 0.002 37.470 37.476 scf_env_do_scf 11 5.9 0.000 0.001 32.167 32.168 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.708 29.708 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.365 22.702 qs_scf_new_mos 108 7.5 0.001 0.001 20.922 21.144 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.921 21.144 ot_scf_mini 108 9.5 0.002 0.003 19.993 20.151 multiply_cannon 2286 13.5 0.209 0.217 17.188 18.650 velocity_verlet 10 3.0 0.001 0.002 18.564 18.565 multiply_cannon_loop 2286 14.5 1.192 1.259 16.002 17.472 ot_mini 108 10.5 0.001 0.001 12.313 12.539 mp_waitall_1 200699 16.5 5.772 10.900 5.772 10.900 qs_ot_get_derivative 108 11.5 0.001 0.001 9.943 10.102 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.182 9.600 multiply_cannon_multrec 27432 15.5 1.815 4.097 6.139 8.996 rebuild_ks_matrix 119 8.3 0.000 0.000 6.857 6.988 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.856 6.988 dbcsr_mm_accdrv_process 47894 16.0 3.561 5.765 4.243 6.344 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.028 6.148 qs_ot_get_p 119 10.4 0.001 0.002 4.746 4.963 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.808 4.650 init_scf_run 11 5.9 0.000 0.001 4.090 4.091 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.090 4.090 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.001 4.073 apply_single 119 13.6 0.000 0.000 3.001 4.072 mp_sum_l 7287 12.8 2.051 3.963 2.051 3.963 sum_up_and_integrate 119 10.3 0.001 0.001 3.769 3.776 integrate_v_rspace 119 11.3 0.002 0.002 3.755 3.762 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.710 3.753 calculate_rho_elec 119 8.7 0.021 0.024 3.709 3.752 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.063 3.082 make_m2s 4572 13.5 0.052 0.054 2.676 3.018 make_images 4572 14.5 0.205 0.242 2.587 2.931 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.660 2.660 calculate_first_density_matrix 1 7.0 0.000 0.000 2.545 2.551 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.365 2.451 init_scf_loop 11 6.9 0.000 0.000 2.433 2.434 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.336 2.337 ot_diis_step 108 11.5 0.011 0.011 2.322 2.323 cp_fm_redistribute_end 50 14.0 1.185 2.308 1.190 2.312 cp_fm_diag_elpa_base 50 14.0 1.089 2.207 1.118 2.253 multiply_cannon_sync_h2d 27432 15.5 1.688 2.196 1.688 2.196 calculate_dm_sparse 119 9.5 0.000 0.001 2.086 2.171 acc_transpose_blocks 27432 15.5 0.114 0.120 1.745 2.148 density_rs2pw 119 9.7 0.004 0.004 2.003 2.090 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.069 2.071 grid_integrate_task_list 119 12.3 1.843 1.955 1.843 1.955 pw_transfer 1439 11.6 0.065 0.068 1.907 1.945 jit_kernel_multiply 10 16.5 0.624 1.918 0.624 1.918 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.853 1.892 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.815 1.856 make_images_data 4572 15.5 0.047 0.054 1.253 1.592 prepare_preconditioner 11 7.9 0.000 0.000 1.546 1.572 make_preconditioner 11 8.9 0.000 0.000 1.546 1.572 potential_pw2rs 119 12.3 0.006 0.006 1.510 1.516 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.448 1.503 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.114 1.484 wfi_extrapolate 11 7.9 0.001 0.001 1.481 1.481 fft3d_ps 1201 14.6 0.523 0.588 1.448 1.479 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.420 1.461 mp_alltoall_d11v 2130 13.8 1.232 1.381 1.232 1.381 grid_collocate_task_list 119 9.7 1.288 1.350 1.288 1.350 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.269 1.282 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.212 1.261 transfer_rs2pw 487 10.6 0.005 0.006 1.143 1.226 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.788 1.070 mp_sum_d 4137 12.0 0.549 1.036 0.549 1.036 mp_allgather_i34 2286 14.5 0.609 1.006 0.609 1.006 transfer_pw2rs 487 13.2 0.004 0.005 0.950 0.957 acc_transpose_blocks_sync 82296 16.5 0.818 0.944 0.818 0.944 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 0.933 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.899 make_images_sizes 4572 15.5 0.005 0.005 0.557 0.802 jit_kernel_transpose 5 15.5 0.599 0.798 0.599 0.798 mp_alltoall_i44 4572 16.5 0.552 0.797 0.552 0.797 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.427000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=467.818182, yerr=2.036851 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.671808E+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 0.007 0.027 32.805 32.807 qs_mol_dyn_low 1 2.0 0.003 0.003 32.535 32.545 qs_forces 11 3.9 0.002 0.003 32.475 32.476 qs_energies 11 4.9 0.001 0.002 30.914 30.916 scf_env_do_scf 11 5.9 0.000 0.001 25.919 25.920 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.395 23.396 dbcsr_multiply_generic 2286 12.5 0.093 0.096 17.249 17.322 qs_scf_new_mos 108 7.5 0.001 0.001 15.607 15.624 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.607 15.623 velocity_verlet 10 3.0 0.001 0.002 15.405 15.406 ot_scf_mini 108 9.5 0.002 0.002 14.849 14.864 multiply_cannon 2286 13.5 0.195 0.198 13.987 14.578 multiply_cannon_loop 2286 14.5 0.862 0.892 13.207 13.844 ot_mini 108 10.5 0.001 0.001 9.171 9.188 qs_ot_get_derivative 108 11.5 0.001 0.001 7.668 7.697 multiply_cannon_multrec 18288 15.5 1.876 2.870 7.060 7.360 rebuild_ks_matrix 119 8.3 0.000 0.000 6.036 6.051 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.035 6.050 dbcsr_mm_accdrv_process 38222 16.0 5.006 5.780 5.089 5.846 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.321 5.333 mp_waitall_1 158411 16.6 2.858 3.919 2.858 3.919 init_scf_run 11 5.9 0.000 0.001 3.793 3.793 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.793 3.793 sum_up_and_integrate 119 10.3 0.001 0.001 3.603 3.609 integrate_v_rspace 119 11.3 0.002 0.003 3.590 3.597 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.940 3.506 qs_ot_get_p 119 10.4 0.001 0.001 3.430 3.446 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.334 3.340 calculate_rho_elec 119 8.7 0.031 0.032 3.333 3.339 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.571 2.567 calculate_first_density_matrix 1 7.0 0.000 0.000 2.538 2.539 init_scf_loop 11 6.9 0.000 0.000 2.505 2.505 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.978 2.351 apply_single 119 13.6 0.000 0.000 1.977 2.351 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.297 2.303 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.001 2.002 make_m2s 4572 13.5 0.044 0.045 1.850 1.988 density_rs2pw 119 9.7 0.004 0.004 1.896 1.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.970 1.972 pw_transfer 1439 11.6 0.065 0.068 1.890 1.903 make_images 4572 14.5 0.191 0.203 1.763 1.901 grid_integrate_task_list 119 12.3 1.802 1.877 1.802 1.877 calculate_dm_sparse 119 9.5 0.000 0.001 1.840 1.847 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.798 1.812 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.789 1.800 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.741 1.750 cp_fm_diag_elpa_base 50 14.0 1.717 1.730 1.739 1.749 prepare_preconditioner 11 7.9 0.000 0.000 1.742 1.745 make_preconditioner 11 8.9 0.000 0.000 1.742 1.745 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.726 1.735 acc_transpose_blocks 18288 15.5 0.078 0.079 1.655 1.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.596 1.681 ot_diis_step 108 11.5 0.011 0.012 1.474 1.475 fft_wrap_pw1pw2_140 487 13.2 0.179 0.182 1.440 1.453 mp_sum_l 7287 12.8 1.025 1.446 1.025 1.446 potential_pw2rs 119 12.3 0.007 0.008 1.398 1.402 fft3d_ps 1201 14.6 0.547 0.569 1.352 1.362 grid_collocate_task_list 119 9.7 1.234 1.322 1.234 1.322 wfi_extrapolate 11 7.9 0.001 0.001 1.201 1.201 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.181 multiply_cannon_sync_h2d 18288 15.5 1.047 1.177 1.047 1.177 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.486 1.106 transfer_rs2pw 487 10.6 0.005 0.005 1.019 1.089 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.963 0.981 make_images_data 4572 15.5 0.047 0.051 0.789 0.952 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.939 0.940 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.687 0.877 transfer_pw2rs 487 13.2 0.004 0.004 0.860 0.864 acc_transpose_blocks_kernels 18288 16.5 0.218 0.227 0.821 0.848 mp_alltoall_d11v 2130 13.8 0.705 0.820 0.705 0.820 acc_transpose_blocks_sync 54864 16.5 0.737 0.799 0.737 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.786 0.788 jit_kernel_multiply 4 15.0 0.024 0.766 0.024 0.766 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.673 0.746 cp_fm_cholesky_invert 11 10.9 0.732 0.736 0.732 0.736 mp_alltoall_z22v 1201 16.6 0.651 0.717 0.651 0.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.807000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.363636, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 561.029120E+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 0.009 0.029 40.482 40.484 qs_mol_dyn_low 1 2.0 0.003 0.003 40.237 40.247 qs_forces 11 3.9 0.009 0.021 40.166 40.167 qs_energies 11 4.9 0.001 0.001 38.351 38.361 scf_env_do_scf 11 5.9 0.001 0.002 32.740 32.741 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.764 28.765 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.975 21.074 velocity_verlet 10 3.0 0.002 0.002 20.557 20.559 qs_scf_new_mos 108 7.5 0.001 0.001 19.300 19.363 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.299 19.362 ot_scf_mini 108 9.5 0.002 0.003 18.227 18.277 multiply_cannon 2286 13.5 0.232 0.242 16.428 17.070 multiply_cannon_loop 2286 14.5 1.501 1.596 15.305 15.720 ot_mini 108 10.5 0.001 0.001 11.043 11.102 multiply_cannon_multrec 27432 15.5 2.501 3.171 9.233 9.494 qs_ot_get_derivative 108 11.5 0.001 0.001 9.139 9.190 dbcsr_mm_accdrv_process 47916 15.9 6.394 7.815 6.630 8.046 rebuild_ks_matrix 119 8.3 0.000 0.000 7.160 7.214 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.160 7.213 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.346 6.394 qs_ot_get_p 119 10.4 0.001 0.002 4.231 4.305 sum_up_and_integrate 119 10.3 0.001 0.001 4.218 4.227 integrate_v_rspace 119 11.3 0.003 0.003 4.207 4.216 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.159 4.192 calculate_rho_elec 119 8.7 0.040 0.046 4.158 4.191 init_scf_run 11 5.9 0.000 0.001 4.109 4.110 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.109 4.109 init_scf_loop 11 6.9 0.000 0.000 3.949 3.950 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.227 3.732 mp_waitall_1 137007 16.6 2.586 3.127 2.586 3.127 make_m2s 4572 13.5 0.054 0.056 2.917 3.038 prepare_preconditioner 11 7.9 0.000 0.000 3.019 3.025 make_preconditioner 11 8.9 0.000 0.000 3.019 3.025 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.602 2.946 make_images 4572 14.5 0.273 0.335 2.808 2.928 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.243 2.685 apply_single 119 13.6 0.000 0.000 2.243 2.685 density_rs2pw 119 9.7 0.003 0.004 2.588 2.680 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.611 2.621 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.547 2.574 calculate_first_density_matrix 1 7.0 0.000 0.000 2.482 2.484 acc_transpose_blocks 27432 15.5 0.118 0.122 2.396 2.473 pw_transfer 1439 11.6 0.065 0.068 2.304 2.348 calculate_dm_sparse 119 9.5 0.000 0.000 2.285 2.332 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.213 2.259 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.176 2.177 potential_pw2rs 119 12.3 0.008 0.009 1.975 1.995 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.948 1.949 grid_integrate_task_list 119 12.3 1.823 1.915 1.823 1.915 ot_diis_step 108 11.5 0.012 0.013 1.863 1.863 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.852 1.862 cp_fm_diag_elpa_base 50 14.0 1.813 1.833 1.849 1.860 mp_sum_l 7287 12.8 1.207 1.820 1.207 1.820 fft_wrap_pw1pw2_140 487 13.2 0.228 0.234 1.731 1.780 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.740 1.755 fft3d_ps 1201 14.6 0.594 0.653 1.702 1.734 transfer_rs2pw 487 10.6 0.004 0.005 1.511 1.618 wfi_extrapolate 11 7.9 0.001 0.001 1.566 1.566 acc_transpose_blocks_sync 82296 16.5 1.396 1.466 1.396 1.466 make_images_data 4572 15.5 0.048 0.053 1.158 1.386 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.350 1.359 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.339 1.355 grid_collocate_task_list 119 9.7 1.249 1.346 1.249 1.346 transfer_pw2rs 487 13.2 0.004 0.004 1.300 1.320 cp_fm_upper_to_full 72 14.2 0.854 1.219 0.854 1.219 hybrid_alltoall_any 4725 16.4 0.066 0.154 0.955 1.213 dbcsr_complete_redistribute 329 12.2 0.116 0.151 0.891 1.174 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.728 1.142 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.123 1.130 cp_fm_cholesky_invert 11 10.9 1.033 1.036 1.033 1.036 mp_alltoall_d11v 2130 13.8 0.926 1.034 0.926 1.034 mp_alltoall_z22v 1201 16.6 0.991 1.026 0.991 1.026 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.641 0.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.909 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.877 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.853 0.862 make_images_sizes 4572 15.5 0.005 0.005 0.629 0.854 mp_alltoall_i44 4572 16.5 0.624 0.849 0.624 0.849 mp_sendrecv_dv 8211 12.7 0.818 0.848 0.818 0.848 jit_kernel_multiply 4 16.1 0.170 0.814 0.170 0.814 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=40.484000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.909091, yerr=3.117532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 615.251968E+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 0.010 0.031 33.377 33.379 qs_mol_dyn_low 1 2.0 0.003 0.003 33.140 33.150 qs_forces 11 3.9 0.003 0.003 33.025 33.025 qs_energies 11 4.9 0.001 0.001 31.225 31.229 scf_env_do_scf 11 5.9 0.001 0.002 26.081 26.081 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.009 23.010 velocity_verlet 10 3.0 0.002 0.002 17.249 17.255 dbcsr_multiply_generic 2286 12.5 0.094 0.097 14.561 14.680 qs_scf_new_mos 108 7.5 0.001 0.001 13.918 13.958 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.917 13.957 ot_scf_mini 108 9.5 0.002 0.002 13.115 13.145 multiply_cannon 2286 13.5 0.228 0.235 10.926 11.554 multiply_cannon_loop 2286 14.5 0.647 0.667 9.879 10.175 ot_mini 108 10.5 0.001 0.001 7.391 7.429 rebuild_ks_matrix 119 8.3 0.000 0.000 6.635 6.673 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.634 6.673 multiply_cannon_multrec 9144 15.5 1.669 1.835 5.925 6.109 qs_ot_get_derivative 108 11.5 0.001 0.001 6.020 6.051 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.923 5.958 dbcsr_mm_accdrv_process 12550 15.8 3.482 4.169 4.144 4.220 sum_up_and_integrate 119 10.3 0.001 0.001 4.149 4.158 integrate_v_rspace 119 11.3 0.003 0.003 4.139 4.148 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.135 4.148 calculate_rho_elec 119 8.7 0.060 0.061 4.134 4.147 qs_ot_get_p 119 10.4 0.002 0.002 3.653 3.694 init_scf_run 11 5.9 0.000 0.001 3.608 3.608 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.608 3.608 init_scf_loop 11 6.9 0.000 0.000 3.042 3.044 mp_waitall_1 115863 16.7 2.597 2.988 2.597 2.988 make_m2s 4572 13.5 0.034 0.035 2.544 2.734 make_images 4572 14.5 0.269 0.300 2.453 2.643 density_rs2pw 119 9.7 0.003 0.003 2.450 2.522 pw_transfer 1439 11.6 0.065 0.067 2.453 2.464 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.359 2.371 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.352 2.356 prepare_preconditioner 11 7.9 0.000 0.000 2.229 2.235 make_preconditioner 11 8.9 0.000 0.000 2.229 2.235 calculate_first_density_matrix 1 7.0 0.000 0.000 2.165 2.167 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.095 2.140 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.063 2.064 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.030 2.055 grid_integrate_task_list 119 12.3 1.916 1.976 1.916 1.976 calculate_dm_sparse 119 9.5 0.000 0.000 1.845 1.868 fft_wrap_pw1pw2_140 487 13.2 0.324 0.331 1.839 1.852 potential_pw2rs 119 12.3 0.010 0.011 1.838 1.843 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.728 1.734 cp_fm_diag_elpa_base 50 14.0 1.695 1.713 1.726 1.732 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.714 1.732 fft3d_ps 1201 14.6 0.653 0.667 1.713 1.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.627 1.631 make_images_data 4572 15.5 0.042 0.046 1.138 1.464 acc_transpose_blocks 9144 15.5 0.042 0.043 1.429 1.454 hybrid_alltoall_any 4725 16.4 0.065 0.176 1.063 1.440 grid_collocate_task_list 119 9.7 1.344 1.427 1.344 1.427 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.411 1.424 wfi_extrapolate 11 7.9 0.001 0.001 1.380 1.380 transfer_rs2pw 487 10.6 0.004 0.004 1.278 1.345 ot_diis_step 108 11.5 0.013 0.013 1.342 1.343 jit_kernel_multiply 8 16.1 0.623 1.337 0.623 1.337 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.276 1.322 apply_single 119 13.6 0.000 0.000 1.276 1.322 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.242 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.157 1.164 mp_alltoall_d11v 2130 13.8 0.998 1.130 0.998 1.130 cp_fm_cholesky_invert 11 10.9 1.125 1.128 1.125 1.128 transfer_pw2rs 487 13.2 0.003 0.003 1.117 1.120 mp_alltoall_z22v 1201 16.6 0.934 0.966 0.934 0.966 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.938 0.949 make_images_sizes 4572 15.5 0.005 0.005 0.609 0.943 mp_alltoall_i44 4572 16.5 0.604 0.938 0.604 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.924 mp_allgather_i34 2286 14.5 0.356 0.895 0.356 0.895 mp_sum_l 7287 12.8 0.645 0.846 0.645 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.833 0.837 multiply_cannon_metrocomm1 9144 15.5 0.024 0.025 0.686 0.819 acc_transpose_blocks_sync 27432 16.5 0.723 0.745 0.723 0.745 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.691 0.740 dbcsr_complete_redistribute 329 12.2 0.157 0.168 0.647 0.689 cp_fm_cholesky_decompose 22 10.9 0.669 0.676 0.669 0.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.379000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=582.363636, yerr=6.677952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 778.563584E+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 0.019 0.079 49.757 49.770 qs_mol_dyn_low 1 2.0 0.003 0.003 49.475 49.500 qs_forces 11 3.9 0.002 0.002 49.220 49.221 qs_energies 11 4.9 0.001 0.002 47.063 47.070 scf_env_do_scf 11 5.9 0.001 0.001 40.984 40.985 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 31.551 31.552 velocity_verlet 10 3.0 0.002 0.002 28.296 28.326 dbcsr_multiply_generic 2286 12.5 0.105 0.106 21.037 21.405 qs_scf_new_mos 108 7.5 0.001 0.001 20.240 20.338 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.240 20.337 ot_scf_mini 108 9.5 0.002 0.002 18.924 19.016 multiply_cannon 2286 13.5 0.302 0.309 15.706 17.115 multiply_cannon_loop 2286 14.5 0.873 0.880 14.214 15.415 ot_mini 108 10.5 0.001 0.001 11.146 11.282 init_scf_loop 11 6.9 0.000 0.000 9.396 9.399 qs_ot_get_derivative 108 11.5 0.001 0.001 8.897 8.987 multiply_cannon_multrec 9144 15.5 3.444 4.794 8.792 8.882 rebuild_ks_matrix 119 8.3 0.000 0.000 8.291 8.458 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.290 8.458 prepare_preconditioner 11 7.9 0.000 0.000 8.305 8.321 make_preconditioner 11 8.9 0.000 0.000 8.305 8.321 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.786 8.181 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.464 7.616 dbcsr_mm_accdrv_process 12550 15.8 4.527 6.504 5.209 6.566 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.043 5.049 calculate_rho_elec 119 8.7 0.117 0.120 5.042 5.048 mp_waitall_1 94719 16.7 3.705 5.020 3.705 5.020 cp_fm_upper_to_full 72 14.2 3.227 4.686 3.227 4.686 qs_ot_get_p 119 10.4 0.002 0.002 4.462 4.626 sum_up_and_integrate 119 10.3 0.001 0.001 4.603 4.609 integrate_v_rspace 119 11.3 0.003 0.003 4.593 4.599 init_scf_run 11 5.9 0.000 0.001 3.934 3.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.934 3.934 make_m2s 4572 13.5 0.037 0.038 3.309 3.638 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.157 3.618 pw_transfer 1439 11.6 0.068 0.069 3.603 3.614 make_images 4572 14.5 0.352 0.383 3.187 3.516 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.505 3.515 dbcsr_complete_redistribute 329 12.2 0.278 0.286 2.209 3.038 density_rs2pw 119 9.7 0.003 0.004 2.934 2.948 fft_wrap_pw1pw2_140 487 13.2 0.649 0.652 2.881 2.890 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.745 2.885 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.710 2.711 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.815 2.648 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.395 2.617 apply_single 119 13.6 0.000 0.000 2.395 2.617 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.474 2.525 mp_sum_l 7287 12.8 1.511 2.454 1.511 2.454 cp_fm_cholesky_invert 11 10.9 2.443 2.449 2.443 2.449 fft3d_ps 1201 14.6 0.872 0.891 2.426 2.432 calculate_dm_sparse 119 9.5 0.000 0.000 2.406 2.425 mp_alltoall_i22 627 13.8 1.518 2.387 1.518 2.387 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.513 2.332 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.293 2.295 grid_integrate_task_list 119 12.3 2.168 2.189 2.168 2.189 ot_diis_step 108 11.5 0.014 0.015 2.183 2.183 calculate_first_density_matrix 1 7.0 0.000 0.000 2.142 2.144 make_images_data 4572 15.5 0.047 0.049 1.663 2.128 hybrid_alltoall_any 4725 16.4 0.092 0.150 1.609 2.061 potential_pw2rs 119 12.3 0.014 0.014 1.964 1.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.912 1.912 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.893 1.893 cp_fm_diag_elpa_base 50 14.0 1.733 1.792 1.891 1.891 acc_transpose_blocks 9144 15.5 0.045 0.046 1.825 1.838 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.796 1.797 wfi_extrapolate 11 7.9 0.001 0.001 1.711 1.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.600 1.640 grid_collocate_task_list 119 9.7 1.582 1.595 1.582 1.595 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.532 1.584 mp_alltoall_d11v 2130 13.8 1.473 1.504 1.473 1.504 mp_alltoall_z22v 1201 16.6 1.415 1.446 1.415 1.446 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.285 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.122 1.135 acc_transpose_blocks_sync 27432 16.5 1.114 1.129 1.114 1.129 transfer_rs2pw 487 10.6 0.005 0.005 1.108 1.125 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.039 1.054 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.027 jit_kernel_multiply 6 15.8 0.653 1.014 0.653 1.014 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=49.770000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=725.272727, yerr=18.281093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.743040E+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 790692. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 89.367 89.370 qs_mol_dyn_low 1 2.0 0.004 0.011 89.032 89.042 qs_forces 11 3.9 0.005 0.019 88.931 88.935 qs_energies 11 4.9 0.005 0.013 85.830 85.852 scf_env_do_scf 11 5.9 0.000 0.001 76.457 76.460 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 69.742 69.742 dbcsr_multiply_generic 2055 12.4 0.104 0.107 52.682 52.956 qs_scf_new_mos 99 7.5 0.000 0.001 50.969 51.127 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.968 51.126 ot_scf_mini 99 9.5 0.002 0.002 48.471 48.591 velocity_verlet 10 3.0 0.006 0.021 47.018 47.019 multiply_cannon 2055 13.4 0.181 0.186 43.236 44.044 multiply_cannon_loop 2055 14.4 1.805 1.846 41.999 42.868 ot_mini 99 10.5 0.001 0.001 27.864 27.997 qs_ot_get_derivative 99 11.5 0.001 0.002 20.953 21.076 multiply_cannon_multrec 49320 15.4 11.272 11.915 17.384 18.127 rebuild_ks_matrix 110 8.3 0.000 0.000 15.546 15.679 qs_ks_build_kohn_sham_matrix 110 9.3 0.019 0.054 15.545 15.678 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.624 13.750 mp_waitall_1 220248 16.4 12.041 13.033 12.041 13.033 qs_ot_get_p 110 10.4 0.001 0.001 11.851 12.003 multiply_cannon_sync_h2d 49320 15.4 9.480 10.117 9.480 10.117 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.111 8.669 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.688 8.250 qs_ot_p2m_diag 48 11.0 0.015 0.029 8.179 8.200 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.289 7.872 apply_single 110 13.6 0.000 0.000 7.289 7.871 sum_up_and_integrate 110 10.3 0.002 0.004 7.837 7.859 integrate_v_rspace 110 11.3 0.002 0.003 7.811 7.838 qs_rho_update_rho_low 110 7.6 0.001 0.002 7.345 7.493 calculate_rho_elec 110 8.6 0.021 0.025 7.344 7.492 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.256 7.259 init_scf_run 11 5.9 0.000 0.001 7.152 7.152 scf_env_initial_rho_setup 11 6.9 0.001 0.003 7.152 7.152 init_scf_loop 11 6.9 0.000 0.000 6.675 6.676 ot_diis_step 99 11.5 0.006 0.008 6.661 6.662 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.484 6.515 cp_fm_diag_elpa_base 48 14.0 6.460 6.493 6.481 6.513 dbcsr_mm_accdrv_process 87628 16.1 3.084 3.184 5.981 6.226 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.655 5.740 mp_sum_l 6594 12.7 4.317 5.172 4.317 5.172 make_m2s 4110 13.4 0.061 0.065 4.684 4.813 make_images 4110 14.4 0.179 0.193 4.586 4.716 density_rs2pw 110 9.6 0.004 0.004 4.361 4.583 prepare_preconditioner 11 7.9 0.000 0.000 4.390 4.412 make_preconditioner 11 8.9 0.000 0.000 4.390 4.412 wfi_extrapolate 11 7.9 0.001 0.001 4.361 4.361 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.162 4.210 pw_transfer 1331 11.6 0.056 0.067 3.924 4.050 calculate_dm_sparse 110 9.5 0.001 0.001 3.938 4.030 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.835 3.964 multiply_cannon_metrocomm1 49320 15.4 0.067 0.068 2.745 3.884 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.615 3.619 potential_pw2rs 110 12.3 0.005 0.006 3.424 3.497 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.411 3.475 grid_integrate_task_list 110 12.3 3.235 3.466 3.235 3.466 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.216 3.256 fft3d_ps 1111 14.6 0.795 0.891 3.130 3.250 fft_wrap_pw1pw2_140 451 13.1 0.309 0.331 3.069 3.203 mp_alltoall_d11v 2046 13.8 2.251 2.748 2.251 2.748 transfer_rs2pw 451 10.6 0.005 0.006 2.486 2.738 calculate_first_density_matrix 1 7.0 0.000 0.000 2.679 2.683 acc_transpose_blocks 49320 15.4 0.212 0.219 2.583 2.640 jit_kernel_multiply 13 15.9 2.618 2.638 2.618 2.638 cp_fm_cholesky_invert 11 10.9 2.457 2.461 2.457 2.461 transfer_pw2rs 451 13.1 0.006 0.007 2.350 2.415 grid_collocate_task_list 110 9.6 2.155 2.285 2.155 2.285 make_images_data 4110 15.4 0.045 0.048 1.985 2.226 mp_alltoall_z22v 1111 16.6 1.912 2.115 1.912 2.115 mp_waitany 14300 13.8 1.860 2.097 1.860 2.097 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.056 2.074 mp_sum_d 3891 11.9 1.527 2.032 1.527 2.032 hybrid_alltoall_any 4261 16.3 0.085 0.485 1.704 2.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=89.370000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=2.712079 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 592.945152E+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 1244078. 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 0.069 0.206 69.884 69.886 qs_mol_dyn_low 1 2.0 0.003 0.004 69.048 69.059 qs_forces 11 3.9 0.003 0.003 68.980 68.981 qs_energies 11 4.9 0.002 0.006 65.648 65.651 scf_env_do_scf 11 5.9 0.001 0.002 56.875 56.879 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 49.236 49.239 dbcsr_multiply_generic 2055 12.4 0.118 0.152 38.192 38.382 velocity_verlet 10 3.0 0.001 0.002 36.127 36.128 qs_scf_new_mos 99 7.5 0.001 0.001 33.377 33.527 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.376 33.527 multiply_cannon 2055 13.4 0.225 0.242 31.392 32.575 ot_scf_mini 99 9.5 0.003 0.003 31.700 31.857 multiply_cannon_loop 2055 14.4 1.162 1.193 30.200 31.030 ot_mini 99 10.5 0.001 0.001 18.466 18.634 multiply_cannon_multrec 24660 15.4 6.998 8.436 14.064 15.391 rebuild_ks_matrix 110 8.3 0.000 0.000 13.543 13.654 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.542 13.653 qs_ot_get_derivative 99 11.5 0.001 0.001 12.699 12.860 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.908 12.007 mp_waitall_1 176588 16.5 7.599 9.952 7.599 9.952 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.176 7.883 init_scf_loop 11 6.9 0.001 0.002 7.603 7.613 multiply_cannon_sync_h2d 24660 15.4 6.478 7.611 6.478 7.611 dbcsr_mm_accdrv_process 52282 16.1 5.558 6.481 6.895 7.211 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.505 7.133 apply_single 110 13.6 0.000 0.001 6.505 7.132 qs_ot_get_p 110 10.4 0.001 0.002 6.561 6.711 sum_up_and_integrate 110 10.3 0.002 0.002 6.364 6.378 init_scf_run 11 5.9 0.000 0.001 6.358 6.358 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.357 6.358 integrate_v_rspace 110 11.3 0.002 0.003 6.337 6.350 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.899 5.908 calculate_rho_elec 110 8.6 0.039 0.047 5.899 5.908 ot_diis_step 99 11.5 0.010 0.010 5.714 5.714 prepare_preconditioner 11 7.9 0.000 0.000 5.578 5.601 make_preconditioner 11 8.9 0.000 0.001 5.578 5.600 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.793 5.483 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.137 5.290 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.661 4.681 make_m2s 4110 13.4 0.057 0.060 4.311 4.606 make_images 4110 14.4 0.405 0.463 4.200 4.487 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.190 4.191 pw_transfer 1331 11.6 0.066 0.072 3.578 3.715 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.613 3.627 cp_fm_diag_elpa_base 48 14.0 3.567 3.586 3.611 3.624 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.471 3.611 wfi_extrapolate 11 7.9 0.001 0.001 3.497 3.497 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.427 3.429 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.295 3.382 grid_integrate_task_list 110 12.3 3.160 3.369 3.160 3.369 density_rs2pw 110 9.6 0.004 0.004 3.146 3.297 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.067 3.121 fft_wrap_pw1pw2_140 451 13.1 0.358 0.380 2.971 3.115 calculate_dm_sparse 110 9.5 0.001 0.001 2.983 3.015 calculate_first_density_matrix 1 7.0 0.001 0.006 2.761 2.764 fft3d_ps 1111 14.6 1.115 1.353 2.564 2.725 make_images_data 4110 15.4 0.049 0.053 2.345 2.680 mp_sum_l 6594 12.7 1.875 2.625 1.875 2.625 cp_fm_cholesky_invert 11 10.9 2.589 2.596 2.589 2.596 hybrid_alltoall_any 4261 16.3 0.105 0.459 1.990 2.594 grid_collocate_task_list 110 9.6 2.172 2.337 2.172 2.337 potential_pw2rs 110 12.3 0.008 0.009 2.252 2.259 mp_alltoall_d11v 2046 13.8 1.760 2.076 1.760 2.076 acc_transpose_blocks 24660 15.4 0.115 0.118 1.976 2.006 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.990 2.005 jit_kernel_multiply 10 16.4 0.972 1.932 0.972 1.932 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.899 1.900 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.800 multiply_cannon_metrocomm4 22605 15.4 0.075 0.078 0.762 1.683 transfer_rs2pw 451 10.6 0.006 0.007 1.463 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.552 dbcsr_complete_redistribute 325 12.2 0.251 0.366 1.206 1.535 mp_allgather_i34 2055 14.4 0.498 1.504 0.498 1.504 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.494 mp_irecv_dv 57340 16.2 0.635 1.465 0.635 1.465 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.886000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.636364, yerr=6.357140 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 668.565504E+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 0.143 0.327 60.831 60.832 qs_mol_dyn_low 1 2.0 0.003 0.004 59.963 60.024 qs_forces 11 3.9 0.003 0.003 59.865 59.866 qs_energies 11 4.9 0.006 0.009 56.690 56.694 scf_env_do_scf 11 5.9 0.001 0.002 48.593 48.593 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 40.036 40.036 velocity_verlet 10 3.0 0.001 0.002 32.406 32.408 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.695 28.959 qs_scf_new_mos 99 7.5 0.001 0.001 25.367 25.491 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.366 25.490 ot_scf_mini 99 9.5 0.002 0.003 24.121 24.269 multiply_cannon 2055 13.4 0.213 0.222 22.449 23.805 multiply_cannon_loop 2055 14.4 0.817 0.847 21.291 22.559 ot_mini 99 10.5 0.001 0.001 13.741 13.891 rebuild_ks_matrix 110 8.3 0.000 0.000 12.200 12.372 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.029 12.199 12.372 multiply_cannon_multrec 16440 15.4 3.651 4.824 9.838 11.030 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.751 10.903 mp_waitall_1 139946 16.5 6.850 9.732 6.850 9.732 qs_ot_get_derivative 99 11.5 0.001 0.001 9.321 9.469 init_scf_loop 11 6.9 0.001 0.003 8.524 8.525 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.287 7.318 prepare_preconditioner 11 7.9 0.000 0.000 6.741 6.764 make_preconditioner 11 8.9 0.000 0.002 6.741 6.764 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.032 6.404 sum_up_and_integrate 110 10.3 0.002 0.003 6.264 6.282 integrate_v_rspace 110 11.3 0.003 0.004 6.238 6.256 dbcsr_mm_accdrv_process 34862 16.1 5.365 5.736 6.031 6.152 qs_ot_get_p 110 10.4 0.005 0.008 5.580 5.765 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.738 5.749 calculate_rho_elec 110 8.6 0.059 0.059 5.738 5.749 init_scf_run 11 5.9 0.000 0.001 5.634 5.634 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.634 5.634 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.919 5.423 apply_single 110 13.6 0.000 0.000 4.919 5.423 make_m2s 4110 13.4 0.049 0.051 4.156 4.503 ot_diis_step 99 11.5 0.011 0.011 4.389 4.390 make_images 4110 14.4 0.398 0.525 4.039 4.384 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.921 3.925 multiply_cannon_sync_h2d 16440 15.4 3.259 3.906 3.259 3.906 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.183 3.814 pw_transfer 1331 11.6 0.065 0.072 3.615 3.625 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.544 3.544 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.508 3.521 grid_integrate_task_list 110 12.3 3.222 3.378 3.222 3.378 density_rs2pw 110 9.6 0.004 0.004 2.980 3.271 fft_wrap_pw1pw2_140 451 13.1 0.460 0.470 3.034 3.046 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.999 3.009 cp_fm_diag_elpa_base 48 14.0 2.927 2.959 2.997 3.007 wfi_extrapolate 11 7.9 0.001 0.002 2.974 2.974 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.816 2.818 make_images_data 4110 15.4 0.046 0.051 2.330 2.766 hybrid_alltoall_any 4261 16.3 0.108 0.381 2.067 2.697 cp_fm_cholesky_invert 11 10.9 2.607 2.613 2.607 2.613 calculate_first_density_matrix 1 7.0 0.000 0.003 2.570 2.571 calculate_dm_sparse 110 9.5 0.001 0.001 2.529 2.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.467 2.545 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.450 2.502 fft3d_ps 1111 14.6 1.120 1.134 2.425 2.435 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.884 2.432 grid_collocate_task_list 110 9.6 2.218 2.371 2.218 2.371 mp_irecv_dv 48980 15.7 0.809 2.300 0.809 2.300 mp_alltoall_d11v 2046 13.8 1.745 2.172 1.745 2.172 mp_sum_l 6594 12.7 1.497 2.157 1.497 2.157 potential_pw2rs 110 12.3 0.011 0.011 2.119 2.129 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.010 2.011 dbcsr_complete_redistribute 325 12.2 0.321 0.356 1.471 1.954 cp_fm_upper_to_full 70 14.2 1.456 1.861 1.456 1.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.664 acc_transpose_blocks 16440 15.4 0.078 0.080 1.578 1.640 transfer_rs2pw 451 10.6 0.005 0.006 1.309 1.594 cp_fm_cholesky_decompose 22 10.9 1.538 1.553 1.538 1.553 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.039 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.468 1.481 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.357 1.466 mp_allgather_i34 2055 14.4 0.446 1.464 0.446 1.464 mp_waitany 17072 13.8 1.184 1.449 1.184 1.449 rs_gather_matrices 110 12.3 0.137 0.150 0.847 1.292 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.274 1.289 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.832000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.090909, yerr=8.436442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 736.047104E+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 0.018 0.037 66.400 66.401 qs_mol_dyn_low 1 2.0 0.003 0.003 66.107 66.118 qs_forces 11 3.9 0.003 0.003 66.043 66.044 qs_energies 11 4.9 0.001 0.002 62.657 62.660 scf_env_do_scf 11 5.9 0.000 0.001 54.301 54.304 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.625 42.626 velocity_verlet 10 3.0 0.001 0.002 37.304 37.306 dbcsr_multiply_generic 2055 12.4 0.115 0.120 30.865 31.108 qs_scf_new_mos 99 7.5 0.001 0.001 27.741 27.849 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.741 27.849 ot_scf_mini 99 9.5 0.003 0.003 26.052 26.158 multiply_cannon 2055 13.4 0.275 0.294 23.620 24.744 multiply_cannon_loop 2055 14.4 1.399 1.485 22.125 22.818 ot_mini 99 10.5 0.001 0.001 14.893 15.026 multiply_cannon_multrec 24660 15.4 4.128 6.998 13.113 14.311 rebuild_ks_matrix 110 8.3 0.000 0.000 12.056 12.161 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.055 12.161 init_scf_loop 11 6.9 0.000 0.000 11.638 11.638 qs_ot_get_derivative 99 11.5 0.001 0.001 10.701 10.816 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.661 10.753 prepare_preconditioner 11 7.9 0.000 0.000 9.900 9.916 make_preconditioner 11 8.9 0.000 0.000 9.900 9.916 dbcsr_mm_accdrv_process 52304 16.0 7.710 8.944 8.826 9.793 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.094 9.583 sum_up_and_integrate 110 10.3 0.001 0.002 6.235 6.248 integrate_v_rspace 110 11.3 0.003 0.003 6.209 6.221 qs_ot_get_p 110 10.4 0.001 0.001 5.996 6.151 make_m2s 4110 13.4 0.059 0.062 5.504 5.886 mp_waitall_1 121746 16.5 4.168 5.862 4.168 5.862 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.836 5.847 calculate_rho_elec 110 8.6 0.078 0.081 5.836 5.847 make_images 4110 14.4 0.574 0.692 5.362 5.739 init_scf_run 11 5.9 0.000 0.001 5.635 5.636 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.635 5.635 cp_fm_upper_to_full 70 14.2 3.392 4.947 3.392 4.947 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.055 4.157 apply_single 110 13.6 0.000 0.000 4.054 4.157 ot_diis_step 99 11.5 0.011 0.011 4.149 4.150 qs_ot_p2m_diag 48 11.0 0.055 0.065 4.088 4.106 dbcsr_complete_redistribute 325 12.2 0.419 0.481 2.688 3.840 pw_transfer 1331 11.6 0.065 0.074 3.769 3.801 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.662 3.698 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.494 3.544 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.515 3.515 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.433 3.497 grid_integrate_task_list 110 12.3 3.266 3.408 3.266 3.408 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.164 3.310 fft_wrap_pw1pw2_140 451 13.1 0.532 0.543 3.160 3.197 hybrid_alltoall_any 4261 16.3 0.123 0.459 2.323 3.127 density_rs2pw 110 9.6 0.004 0.004 2.967 3.099 make_images_data 4110 15.4 0.049 0.053 2.731 3.087 calculate_dm_sparse 110 9.5 0.001 0.001 2.984 3.031 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.301 3.006 wfi_extrapolate 11 7.9 0.001 0.001 2.988 2.988 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.970 2.979 cp_fm_diag_elpa_base 48 14.0 2.817 2.881 2.968 2.978 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.797 2.923 mp_alltoall_i22 605 13.7 1.713 2.920 1.713 2.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.861 cp_fm_cholesky_invert 11 10.9 2.628 2.636 2.628 2.636 multiply_cannon_sync_h2d 24660 15.4 2.394 2.588 2.394 2.588 calculate_first_density_matrix 1 7.0 0.000 0.000 2.550 2.553 acc_transpose_blocks 24660 15.4 0.111 0.114 2.396 2.547 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.486 2.522 fft3d_ps 1111 14.6 1.150 1.188 2.488 2.507 grid_collocate_task_list 110 9.6 2.262 2.360 2.262 2.360 qs_energies_init_hamiltonians 11 5.9 0.003 0.012 2.232 2.232 potential_pw2rs 110 12.3 0.013 0.013 2.053 2.057 mp_alltoall_d11v 2046 13.8 1.784 1.972 1.784 1.972 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.816 1.862 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.714 cp_fm_cholesky_decompose 22 10.9 1.631 1.687 1.631 1.687 mp_sum_l 6594 12.7 1.052 1.674 1.052 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.646 mp_allgather_i34 2055 14.4 0.547 1.618 0.547 1.618 multiply_cannon_metrocomm4 20550 15.4 0.060 0.062 0.839 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.556 acc_transpose_blocks_sync 73980 16.4 1.384 1.539 1.384 1.539 jit_kernel_multiply 8 16.1 0.772 1.483 0.772 1.483 mp_irecv_dv 62702 16.1 0.735 1.475 0.735 1.475 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.401000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.818182, yerr=9.842562 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 865.091584E+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 0.012 0.033 56.758 56.760 qs_mol_dyn_low 1 2.0 0.003 0.003 56.390 56.399 qs_forces 11 3.9 0.003 0.003 56.323 56.324 qs_energies 11 4.9 0.001 0.002 52.651 52.657 scf_env_do_scf 11 5.9 0.000 0.001 44.166 44.166 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.241 36.242 velocity_verlet 10 3.0 0.001 0.002 31.672 31.675 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.503 23.653 qs_scf_new_mos 99 7.5 0.001 0.001 21.430 21.485 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.429 21.485 ot_scf_mini 99 9.5 0.002 0.002 20.166 20.214 multiply_cannon 2055 13.4 0.238 0.250 17.821 19.071 multiply_cannon_loop 2055 14.4 0.605 0.624 16.545 16.950 rebuild_ks_matrix 110 8.3 0.000 0.000 11.792 11.832 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 11.791 11.832 ot_mini 99 10.5 0.001 0.001 10.985 11.029 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.493 10.532 multiply_cannon_multrec 8220 15.4 3.172 4.623 7.558 8.680 init_scf_loop 11 6.9 0.000 0.000 7.873 7.875 mp_waitall_1 103326 16.6 6.080 7.708 6.080 7.708 qs_ot_get_derivative 99 11.5 0.001 0.001 7.187 7.236 sum_up_and_integrate 110 10.3 0.001 0.002 6.313 6.323 integrate_v_rspace 110 11.3 0.003 0.003 6.286 6.297 prepare_preconditioner 11 7.9 0.000 0.000 6.209 6.215 make_preconditioner 11 8.9 0.000 0.000 6.209 6.215 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.123 6.143 calculate_rho_elec 110 8.6 0.113 0.113 6.122 6.142 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.784 5.858 init_scf_run 11 5.9 0.000 0.001 5.290 5.290 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.289 5.289 qs_ot_get_p 110 10.4 0.002 0.002 5.176 5.226 dbcsr_mm_accdrv_process 17442 15.9 3.135 4.117 4.245 5.166 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.216 4.595 make_m2s 4110 13.4 0.038 0.039 4.265 4.477 make_images 4110 14.4 0.646 0.708 4.135 4.346 pw_transfer 1331 11.6 0.066 0.070 4.203 4.209 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.096 4.104 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.823 3.896 apply_single 110 13.6 0.000 0.000 3.823 3.896 ot_diis_step 99 11.5 0.012 0.012 3.774 3.774 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.744 3.749 fft_wrap_pw1pw2_140 451 13.1 0.722 0.734 3.540 3.554 grid_integrate_task_list 110 12.3 3.389 3.502 3.389 3.502 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.392 3.393 density_rs2pw 110 9.6 0.004 0.004 3.115 3.246 cp_fm_cholesky_invert 11 10.9 3.057 3.060 3.057 3.060 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.837 2.844 cp_fm_diag_elpa_base 48 14.0 2.779 2.801 2.835 2.843 wfi_extrapolate 11 7.9 0.001 0.001 2.762 2.762 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.692 2.693 fft3d_ps 1111 14.6 1.295 1.300 2.637 2.653 hybrid_alltoall_any 4261 16.3 0.199 0.849 2.209 2.629 make_images_data 4110 15.4 0.041 0.047 2.270 2.592 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.565 2.567 calculate_dm_sparse 110 9.5 0.001 0.001 2.483 2.528 multiply_cannon_sync_h2d 8220 15.4 2.358 2.493 2.358 2.493 grid_collocate_task_list 110 9.6 2.363 2.483 2.363 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.429 2.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.149 2.167 potential_pw2rs 110 12.3 0.015 0.016 2.097 2.100 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.063 2.095 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.995 2.025 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 2.002 mp_alltoall_d11v 2046 13.8 1.703 1.878 1.703 1.878 cp_fm_cholesky_decompose 22 10.9 1.705 1.717 1.705 1.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.629 1.631 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.520 1.630 dbcsr_complete_redistribute 325 12.2 0.558 0.600 1.534 1.623 mp_allgather_i34 2055 14.4 0.446 1.527 0.446 1.527 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.429 1.441 acc_transpose_blocks 8220 15.4 0.040 0.042 1.382 1.411 qs_create_task_list 11 7.9 0.000 0.000 1.209 1.306 generate_qs_task_list 11 8.9 0.375 0.440 1.208 1.306 transfer_rs2pw 451 10.6 0.005 0.005 1.159 1.302 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.474 1.231 jit_kernel_multiply 7 15.4 0.796 1.206 0.796 1.206 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.175 1.198 mp_irecv_dv 24056 15.7 0.447 1.191 0.447 1.191 mp_waitany 9240 13.8 1.047 1.179 1.047 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.760000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=819.636364, yerr=11.267594 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 1.389941E+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 0.158 0.224 90.986 90.987 qs_mol_dyn_low 1 2.0 0.003 0.003 90.169 90.229 qs_forces 11 3.9 0.031 0.050 90.091 90.093 qs_energies 11 4.9 0.002 0.002 85.857 85.866 scf_env_do_scf 11 5.9 0.000 0.001 75.518 75.519 velocity_verlet 10 3.0 0.010 0.011 57.389 57.394 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.303 46.304 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.283 30.332 init_scf_loop 11 6.9 0.000 0.000 29.141 29.144 qs_scf_new_mos 99 7.5 0.001 0.001 28.039 28.072 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.038 28.071 prepare_preconditioner 11 7.9 0.000 0.000 27.041 27.051 make_preconditioner 11 8.9 0.000 0.000 27.041 27.051 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.122 26.508 ot_scf_mini 99 9.5 0.002 0.003 26.219 26.249 multiply_cannon 2055 13.4 0.331 0.349 22.776 23.473 multiply_cannon_loop 2055 14.4 0.825 0.852 20.934 21.230 cp_fm_upper_to_full 70 14.2 12.956 18.587 12.956 18.587 ot_mini 99 10.5 0.001 0.001 14.611 14.640 rebuild_ks_matrix 110 8.3 0.000 0.001 14.204 14.254 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.203 14.254 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.890 12.935 dbcsr_complete_redistribute 325 12.2 1.004 1.026 7.537 10.952 multiply_cannon_multrec 8220 15.4 4.051 4.206 9.812 9.930 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.486 9.914 qs_ot_get_derivative 99 11.5 0.001 0.001 9.821 9.849 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.903 9.294 mp_alltoall_i22 605 13.7 5.529 8.978 5.529 8.978 mp_waitall_1 84994 16.7 8.003 8.832 8.003 8.832 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.632 7.669 calculate_rho_elec 110 8.6 0.222 0.223 7.632 7.669 sum_up_and_integrate 110 10.3 0.002 0.002 7.046 7.061 integrate_v_rspace 110 11.3 0.004 0.004 7.018 7.033 qs_ot_get_p 110 10.4 0.009 0.016 6.239 6.266 make_m2s 4110 13.4 0.043 0.043 5.667 6.194 init_scf_run 11 5.9 0.000 0.001 6.001 6.001 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.001 6.001 make_images 4110 14.4 0.877 0.922 5.476 6.000 pw_transfer 1331 11.6 0.075 0.075 5.981 5.989 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.864 5.873 dbcsr_mm_accdrv_process 11614 15.7 3.896 4.157 5.612 5.845 cp_fm_cholesky_invert 11 10.9 5.609 5.613 5.609 5.613 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.038 5.430 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.932 5.371 apply_single 110 13.6 0.000 0.000 4.932 5.370 fft_wrap_pw1pw2_140 451 13.1 1.366 1.371 5.131 5.138 ot_diis_step 99 11.5 0.015 0.016 4.766 4.766 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.500 4.507 density_rs2pw 110 9.6 0.004 0.004 4.075 4.098 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.970 3.971 hybrid_alltoall_any 4261 16.3 0.265 0.565 3.074 3.866 make_images_data 4110 15.4 0.045 0.048 3.087 3.780 grid_integrate_task_list 110 12.3 3.687 3.744 3.687 3.744 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.722 3.732 fft3d_ps 1111 14.6 1.882 1.887 3.570 3.578 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.102 3.561 wfi_extrapolate 11 7.9 0.001 0.001 3.429 3.429 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.330 3.330 cp_fm_diag_elpa_base 48 14.0 2.785 2.989 3.328 3.328 calculate_dm_sparse 110 9.5 0.001 0.001 3.300 3.327 multiply_cannon_sync_h2d 8220 15.4 3.125 3.151 3.125 3.151 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.923 2.925 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.753 2.768 grid_collocate_task_list 110 9.6 2.676 2.698 2.676 2.698 potential_pw2rs 110 12.3 0.021 0.022 2.555 2.563 calculate_first_density_matrix 1 7.0 0.000 0.000 2.451 2.451 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.409 2.427 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.243 2.292 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.112 2.205 mp_alltoall_d11v 2046 13.8 2.119 2.176 2.119 2.176 cp_fm_cholesky_decompose 22 10.9 2.035 2.049 2.035 2.049 qs_create_task_list 11 7.9 0.000 0.000 1.886 1.929 generate_qs_task_list 11 8.9 0.733 0.787 1.886 1.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.920 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.801 1.832 acc_transpose_blocks 8220 15.4 0.041 0.042 1.772 1.827 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.987000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1249.000000, yerr=60.539994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 636.411904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3994 57876. MP_Allreduce 11064 797. MP_Sync 87 MP_Alltoall 2226 3473517. 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 0.061 0.095 210.949 210.951 qs_mol_dyn_low 1 2.0 0.004 0.009 210.275 210.289 qs_forces 11 3.9 0.006 0.007 210.181 210.182 qs_energies 11 4.9 0.002 0.003 204.561 204.573 scf_env_do_scf 11 5.9 0.001 0.001 187.992 187.996 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 167.184 167.186 qs_scf_new_mos 117 7.6 0.001 0.001 127.139 127.449 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.138 127.448 dbcsr_multiply_generic 2507 12.6 0.175 0.181 126.354 127.150 velocity_verlet 10 3.0 0.001 0.002 124.682 124.683 ot_scf_mini 117 9.6 0.003 0.003 120.402 120.694 multiply_cannon 2507 13.6 0.237 0.246 101.785 103.448 multiply_cannon_loop 2507 14.6 2.414 2.465 99.548 101.120 ot_mini 117 10.6 0.001 0.001 67.017 67.315 multiply_cannon_multrec 60168 15.6 31.799 34.018 41.554 43.247 qs_ot_get_derivative 117 11.6 0.001 0.003 42.000 42.302 rebuild_ks_matrix 128 8.3 0.001 0.001 33.943 34.178 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.942 34.177 mp_waitall_1 267128 16.5 29.280 32.512 29.280 32.512 qs_ot_get_p 128 10.4 0.001 0.002 31.053 31.349 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.484 30.714 multiply_cannon_sync_h2d 60168 15.6 26.300 28.432 26.300 28.432 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.416 24.937 apply_single 128 13.6 0.001 0.001 24.415 24.936 ot_diis_step 117 11.6 0.008 0.008 24.756 24.757 qs_ot_p2m_diag 83 11.4 0.079 0.092 24.426 24.515 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.577 21.578 init_scf_loop 11 6.9 0.000 0.000 20.730 20.731 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.274 20.459 multiply_cannon_metrocomm3 60168 15.6 0.122 0.130 16.163 18.610 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.459 18.490 cp_fm_diag_elpa_base 83 14.4 18.377 18.414 18.454 18.486 prepare_preconditioner 11 7.9 0.000 0.000 16.092 16.130 make_preconditioner 11 8.9 0.000 0.000 16.092 16.130 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.318 15.513 make_m2s 5014 13.6 0.104 0.112 13.989 14.429 sum_up_and_integrate 128 10.3 0.002 0.005 14.267 14.286 make_images 5014 14.6 0.403 0.421 13.809 14.260 integrate_v_rspace 128 11.3 0.003 0.004 14.208 14.229 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.451 13.584 calculate_rho_elec 128 8.7 0.045 0.064 13.450 13.584 init_scf_run 11 5.9 0.000 0.001 12.374 12.374 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.374 12.374 mp_sum_l 7950 12.9 9.328 10.538 9.328 10.538 dbcsr_mm_accdrv_process 124484 16.2 4.734 4.867 9.321 9.827 cp_fm_cholesky_invert 11 10.9 9.236 9.244 9.236 9.244 wfi_extrapolate 11 7.9 0.001 0.001 9.067 9.067 calculate_dm_sparse 128 9.5 0.001 0.001 8.541 8.610 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.140 8.294 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.166 8.236 pw_transfer 1547 11.6 0.074 0.086 7.844 8.048 multiply_cannon_metrocomm1 60168 15.6 0.096 0.102 6.388 8.023 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.642 7.845 make_images_data 5014 15.6 0.069 0.074 6.854 7.822 density_rs2pw 128 9.7 0.006 0.007 7.065 7.655 grid_integrate_task_list 128 12.3 7.025 7.486 7.025 7.486 hybrid_alltoall_any 5200 16.5 0.296 2.288 5.990 7.237 fft_wrap_pw1pw2_140 523 13.2 0.862 0.901 6.620 6.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.698 6.709 fft3d_ps 1291 14.7 2.197 2.858 5.642 5.973 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.754 5.834 mp_alltoall_d11v 2415 14.1 4.454 5.702 4.454 5.702 grid_collocate_task_list 128 9.7 4.843 5.280 4.843 5.280 potential_pw2rs 128 12.3 0.009 0.010 4.738 4.781 cp_fm_cholesky_decompose 22 10.9 4.655 4.667 4.655 4.667 mp_sum_d 4468 12.1 3.545 4.279 3.545 4.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.951000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=600.636364, yerr=7.957532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.090176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5278145. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.044 192.641 192.643 qs_mol_dyn_low 1 2.0 0.004 0.006 192.091 192.121 qs_forces 11 3.9 0.004 0.006 191.974 191.975 qs_energies 11 4.9 0.004 0.012 185.187 185.200 scf_env_do_scf 11 5.9 0.001 0.004 168.442 168.452 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.261 135.264 velocity_verlet 10 3.0 0.001 0.002 120.336 120.339 dbcsr_multiply_generic 2507 12.6 0.186 0.193 97.548 98.872 qs_scf_new_mos 117 7.6 0.001 0.001 96.555 97.179 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.555 97.178 ot_scf_mini 117 9.6 0.004 0.006 91.742 92.427 multiply_cannon 2507 13.6 0.502 0.562 76.873 81.505 multiply_cannon_loop 2507 14.6 1.561 1.630 73.878 77.182 ot_mini 117 10.6 0.001 0.002 50.626 51.228 multiply_cannon_multrec 30084 15.6 20.873 26.204 31.509 37.551 mp_waitall_1 214728 16.6 24.346 37.516 24.346 37.516 init_scf_loop 11 6.9 0.001 0.004 33.091 33.092 rebuild_ks_matrix 128 8.3 0.001 0.001 32.360 33.018 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.024 32.360 33.017 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.089 29.697 qs_ot_get_derivative 117 11.6 0.001 0.002 28.611 29.287 prepare_preconditioner 11 7.9 0.000 0.000 28.739 28.808 make_preconditioner 11 8.9 0.000 0.003 28.739 28.808 multiply_cannon_metrocomm3 30084 15.6 0.101 0.108 15.812 28.306 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.449 27.999 qs_ot_get_p 128 10.4 0.002 0.002 23.187 23.836 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 22.466 23.705 apply_single 128 13.6 0.001 0.001 22.466 23.705 ot_diis_step 117 11.6 0.015 0.017 21.838 21.840 multiply_cannon_sync_h2d 30084 15.6 17.909 20.113 17.909 20.113 qs_ot_p2m_diag 83 11.4 0.189 0.217 18.202 18.243 cp_dbcsr_syevd 83 12.4 0.006 0.007 16.970 16.971 cp_fm_cholesky_invert 11 10.9 16.808 16.820 16.808 16.820 make_m2s 5014 13.6 0.087 0.091 14.280 15.521 make_images 5014 14.6 1.161 1.359 14.066 15.305 sum_up_and_integrate 128 10.3 0.002 0.005 14.125 14.157 integrate_v_rspace 128 11.3 0.003 0.003 14.065 14.101 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.699 13.738 cp_fm_diag_elpa_base 83 14.4 13.440 13.540 13.692 13.726 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.548 13.583 calculate_rho_elec 128 8.7 0.087 0.104 13.547 13.582 init_scf_run 11 5.9 0.000 0.001 11.728 11.730 scf_env_initial_rho_setup 11 6.9 0.006 0.021 11.728 11.730 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.239 11.716 multiply_cannon_metrocomm4 27577 15.6 0.100 0.114 3.882 10.912 dbcsr_mm_accdrv_process 62242 16.2 5.543 6.499 10.093 10.820 mp_irecv_dv 69486 16.3 3.680 10.507 3.680 10.507 make_images_data 5014 15.6 0.067 0.076 8.441 9.952 hybrid_alltoall_any 5200 16.5 0.349 1.540 7.172 9.330 pw_transfer 1547 11.6 0.085 0.101 8.688 8.744 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.465 8.526 wfi_extrapolate 11 7.9 0.001 0.004 8.390 8.390 grid_integrate_task_list 128 12.3 7.180 7.585 7.180 7.585 density_rs2pw 128 9.7 0.006 0.007 7.219 7.526 fft_wrap_pw1pw2_140 523 13.2 0.932 0.954 7.449 7.519 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.288 7.396 cp_fm_cholesky_decompose 22 10.9 7.113 7.204 7.113 7.204 mp_sum_l 7950 12.9 4.606 6.797 4.606 6.797 calculate_dm_sparse 128 9.5 0.001 0.001 6.513 6.649 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.069 6.080 fft3d_ps 1291 14.7 2.828 3.016 6.025 6.076 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.395 5.472 grid_collocate_task_list 128 9.7 4.997 5.426 4.997 5.426 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.156 5.345 mp_alltoall_d11v 2415 14.1 4.250 4.839 4.250 4.839 mp_allgather_i34 2507 14.6 1.512 4.653 1.512 4.653 potential_pw2rs 128 12.3 0.015 0.017 4.593 4.614 mp_sum_d 4481 12.1 2.687 4.044 2.687 4.044 dbcsr_complete_redistribute 395 12.7 0.762 0.874 3.178 4.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.643000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.727273, yerr=4.024717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 956.968960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 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 0.113 0.351 178.001 178.002 qs_mol_dyn_low 1 2.0 0.003 0.004 176.958 176.981 qs_forces 11 3.9 0.006 0.026 176.865 176.870 qs_energies 11 4.9 0.018 0.034 170.297 170.306 scf_env_do_scf 11 5.9 0.002 0.010 154.581 154.582 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 119.016 119.016 velocity_verlet 10 3.0 0.001 0.002 113.546 113.548 dbcsr_multiply_generic 2507 12.6 0.183 0.188 81.970 83.048 qs_scf_new_mos 117 7.6 0.001 0.001 81.727 82.104 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.727 82.103 ot_scf_mini 117 9.6 0.003 0.004 77.489 77.870 multiply_cannon 2507 13.6 0.506 0.528 62.084 66.917 multiply_cannon_loop 2507 14.6 1.134 1.200 59.346 61.832 ot_mini 117 10.6 0.001 0.002 42.563 42.933 init_scf_loop 11 6.9 0.001 0.005 35.457 35.458 mp_waitall_1 170520 16.6 25.173 34.880 25.173 34.880 prepare_preconditioner 11 7.9 0.000 0.000 31.424 31.471 make_preconditioner 11 8.9 0.000 0.001 31.424 31.471 rebuild_ks_matrix 128 8.3 0.001 0.001 30.357 30.843 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.028 30.357 30.842 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.931 30.450 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.358 27.808 multiply_cannon_multrec 20056 15.6 12.995 16.129 22.279 25.466 multiply_cannon_metrocomm3 20056 15.6 0.062 0.068 15.736 25.128 qs_ot_get_derivative 117 11.6 0.001 0.002 22.910 23.294 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.698 20.674 apply_single 128 13.6 0.001 0.001 19.698 20.674 qs_ot_get_p 128 10.4 0.001 0.003 20.239 20.672 ot_diis_step 117 11.6 0.018 0.022 19.541 19.541 qs_ot_p2m_diag 83 11.4 0.266 0.274 15.973 15.984 make_m2s 5014 13.6 0.078 0.083 14.946 15.848 make_images 5014 14.6 1.145 1.257 14.712 15.605 multiply_cannon_sync_h2d 20056 15.6 13.517 15.325 13.517 15.325 cp_dbcsr_syevd 83 12.4 0.006 0.007 14.902 14.903 cp_fm_cholesky_invert 11 10.9 14.495 14.503 14.495 14.503 sum_up_and_integrate 128 10.3 0.002 0.003 14.030 14.057 integrate_v_rspace 128 11.3 0.003 0.004 13.971 13.999 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.667 13.702 calculate_rho_elec 128 8.7 0.129 0.143 13.667 13.702 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.623 11.642 cp_fm_diag_elpa_base 83 14.4 11.193 11.349 11.618 11.637 init_scf_run 11 5.9 0.000 0.001 10.614 10.614 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.613 10.614 make_images_data 5014 15.6 0.064 0.077 8.913 10.162 multiply_cannon_metrocomm4 17549 15.6 0.065 0.075 3.527 9.630 hybrid_alltoall_any 5200 16.5 0.450 2.056 7.650 9.508 mp_irecv_dv 50230 16.2 3.396 9.367 3.396 9.367 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.082 9.353 pw_transfer 1547 11.6 0.085 0.103 8.890 8.987 dbcsr_mm_accdrv_process 41502 16.2 5.607 5.917 8.749 8.874 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.667 8.770 fft_wrap_pw1pw2_140 523 13.2 1.069 1.110 7.663 7.777 grid_integrate_task_list 128 12.3 7.307 7.679 7.307 7.679 cp_fm_upper_to_full 105 14.8 6.045 7.646 6.045 7.646 wfi_extrapolate 11 7.9 0.001 0.002 7.453 7.453 cp_fm_cholesky_decompose 22 10.9 7.343 7.427 7.343 7.427 density_rs2pw 128 9.7 0.006 0.006 7.046 7.415 dbcsr_complete_redistribute 395 12.7 1.184 1.213 4.784 6.723 fft3d_ps 1291 14.7 2.819 3.055 5.974 6.062 calculate_dm_sparse 128 9.5 0.001 0.001 5.829 5.921 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.588 5.596 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.598 5.530 grid_collocate_task_list 128 9.7 5.165 5.526 5.165 5.526 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.609 5.240 mp_alltoall_d11v 2415 14.1 4.358 5.188 4.358 5.188 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.649 4.810 mp_allgather_i34 2507 14.6 1.230 4.554 1.230 4.554 mp_sum_l 7950 12.9 3.158 4.487 3.158 4.487 potential_pw2rs 128 12.3 0.020 0.022 4.461 4.480 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.474 4.310 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.985 4.013 mp_alltoall_i22 716 14.1 2.015 3.952 2.015 3.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.755 3.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.002000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=909.272727, yerr=11.786671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.183617E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57894. MP_Allreduce 11137 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.129 0.153 191.328 191.329 qs_mol_dyn_low 1 2.0 0.003 0.003 190.679 190.692 qs_forces 11 3.9 0.064 0.085 190.567 190.578 qs_energies 11 4.9 0.001 0.002 183.433 183.501 scf_env_do_scf 11 5.9 0.001 0.002 166.377 166.385 velocity_verlet 10 3.0 0.002 0.002 126.339 126.343 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 119.738 119.741 qs_scf_new_mos 118 7.6 0.001 0.001 82.852 83.192 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.852 83.192 dbcsr_multiply_generic 2527 12.6 0.192 0.198 81.099 81.712 ot_scf_mini 118 9.6 0.003 0.004 78.111 78.422 multiply_cannon 2527 13.6 0.557 0.592 56.030 58.613 multiply_cannon_loop 2527 14.6 1.807 1.935 52.321 53.843 init_scf_loop 11 6.9 0.000 0.000 46.513 46.520 ot_mini 118 10.6 0.001 0.001 43.232 43.524 prepare_preconditioner 11 7.9 0.000 0.000 42.391 42.422 make_preconditioner 11 8.9 0.000 0.000 42.391 42.422 make_full_inverse_cholesky 11 9.9 0.010 0.022 35.983 40.985 multiply_cannon_multrec 30324 15.6 13.839 19.084 26.836 31.766 rebuild_ks_matrix 129 8.3 0.001 0.001 29.658 30.009 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 29.657 30.008 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.827 27.143 mp_waitall_1 149060 16.7 17.115 26.389 17.115 26.389 qs_ot_get_derivative 118 11.6 0.001 0.002 23.529 23.842 make_m2s 5054 13.6 0.093 0.098 20.632 21.731 make_images 5054 14.6 1.949 2.259 20.324 21.426 qs_ot_get_p 129 10.4 0.002 0.002 20.899 21.264 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.101 19.656 apply_single 129 13.6 0.001 0.001 19.101 19.655 ot_diis_step 118 11.6 0.018 0.018 19.570 19.572 cp_fm_upper_to_full 105 14.8 11.514 16.915 11.514 16.915 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.670 16.722 cp_fm_cholesky_invert 11 10.9 16.242 16.251 16.242 16.251 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.255 15.257 multiply_cannon_metrocomm3 30324 15.6 0.049 0.052 6.180 14.594 sum_up_and_integrate 129 10.3 0.002 0.003 14.148 14.173 integrate_v_rspace 129 11.3 0.003 0.004 14.088 14.114 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.008 14.059 calculate_rho_elec 129 8.7 0.171 0.186 14.007 14.058 dbcsr_mm_accdrv_process 62756 16.2 8.477 9.313 12.560 13.032 dbcsr_complete_redistribute 395 12.7 1.511 1.639 9.232 13.009 make_images_data 5054 15.6 0.066 0.073 11.058 12.668 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.950 11.971 cp_fm_diag_elpa_base 83 14.4 10.963 11.280 11.942 11.961 hybrid_alltoall_any 5240 16.5 0.534 2.239 9.804 11.640 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.832 11.631 multiply_cannon_sync_h2d 30324 15.6 10.634 11.376 10.634 11.376 init_scf_run 11 5.9 0.000 0.001 11.019 11.021 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.019 11.020 transfer_fm_to_dbcsr 11 9.9 0.001 0.014 6.384 10.078 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.842 10.067 mp_alltoall_i22 718 14.1 5.596 9.475 5.596 9.475 pw_transfer 1559 11.6 0.087 0.102 9.326 9.387 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.098 9.165 fft_wrap_pw1pw2_140 527 13.2 1.265 1.301 8.080 8.163 grid_integrate_task_list 129 12.3 7.534 7.984 7.534 7.984 multiply_cannon_metrocomm4 25270 15.6 0.081 0.093 2.805 7.806 wfi_extrapolate 11 7.9 0.001 0.001 7.687 7.687 cp_fm_cholesky_decompose 22 10.9 7.558 7.641 7.558 7.641 mp_irecv_dv 76685 16.2 2.650 7.520 2.650 7.520 density_rs2pw 129 9.7 0.005 0.006 6.987 7.402 calculate_dm_sparse 129 9.5 0.001 0.001 6.443 6.515 fft3d_ps 1301 14.7 3.013 3.055 6.115 6.147 grid_collocate_task_list 129 9.7 5.321 5.713 5.321 5.713 mp_alltoall_d11v 2423 14.1 5.144 5.680 5.144 5.680 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.586 5.629 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.620 4.708 qs_energies_init_hamiltonians 11 5.9 0.084 0.110 4.509 4.570 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.446 4.525 potential_pw2rs 129 12.3 0.024 0.024 4.380 4.392 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.299 4.381 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.329000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1101.545455, yerr=26.331328 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 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 1986255912960 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 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 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 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.553854E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57681. MP_Allreduce 11174 1163. MP_Sync 88 MP_Alltoall 1724 18848034. 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 0.205 0.242 174.050 174.059 qs_mol_dyn_low 1 2.0 0.003 0.003 173.170 173.203 qs_forces 11 3.9 0.025 0.030 173.072 173.076 qs_energies 11 4.9 0.023 0.043 165.608 165.633 scf_env_do_scf 11 5.9 0.001 0.012 148.095 148.105 velocity_verlet 10 3.0 0.002 0.002 113.987 113.991 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.984 111.985 qs_scf_new_mos 118 7.6 0.001 0.001 74.299 74.432 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.298 74.431 dbcsr_multiply_generic 2529 12.6 0.184 0.191 73.108 73.562 ot_scf_mini 118 9.6 0.003 0.004 69.824 69.944 multiply_cannon 2529 13.6 0.604 0.632 54.280 57.627 multiply_cannon_loop 2529 14.6 0.817 0.853 51.094 52.148 ot_mini 118 10.6 0.001 0.001 38.296 38.419 init_scf_loop 11 6.9 0.000 0.000 35.949 35.951 prepare_preconditioner 11 7.9 0.000 0.000 32.086 32.118 make_preconditioner 11 8.9 0.000 0.000 32.086 32.118 mp_waitall_1 126876 16.7 25.333 32.059 25.333 32.059 make_full_inverse_cholesky 11 9.9 0.015 0.028 29.999 30.275 rebuild_ks_matrix 129 8.3 0.001 0.001 29.608 29.744 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.608 29.743 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.015 27.136 multiply_cannon_multrec 10116 15.6 10.442 15.924 18.093 22.611 qs_ot_get_derivative 118 11.6 0.002 0.002 20.813 20.933 multiply_cannon_metrocomm3 10116 15.6 0.024 0.026 13.136 20.677 qs_ot_get_p 129 10.4 0.002 0.002 18.636 18.786 cp_fm_cholesky_invert 11 10.9 18.438 18.445 18.438 18.445 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.506 17.708 apply_single 129 13.6 0.001 0.001 17.506 17.708 ot_diis_step 118 11.6 0.020 0.021 17.412 17.412 make_m2s 5058 13.6 0.065 0.070 14.892 15.814 make_images 5058 14.6 2.173 2.605 14.581 15.500 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.763 14.780 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.450 14.475 calculate_rho_elec 129 8.7 0.255 0.266 14.449 14.474 sum_up_and_integrate 129 10.3 0.002 0.003 14.398 14.448 integrate_v_rspace 129 11.3 0.004 0.004 14.337 14.386 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.569 13.570 multiply_cannon_sync_h2d 10116 15.6 10.792 11.349 10.792 11.349 init_scf_run 11 5.9 0.000 0.001 10.656 10.657 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.656 10.656 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.568 10.579 cp_fm_diag_elpa_base 84 14.4 10.323 10.401 10.563 10.574 pw_transfer 1559 11.6 0.086 0.096 10.044 10.082 make_images_data 5058 15.6 0.056 0.064 8.548 9.987 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.820 9.864 hybrid_alltoall_any 5245 16.5 0.846 3.807 8.371 9.584 fft_wrap_pw1pw2_140 527 13.2 1.686 1.716 8.680 8.725 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.236 8.314 grid_integrate_task_list 129 12.3 7.812 8.229 7.812 8.229 cp_fm_cholesky_decompose 22 10.9 8.031 8.169 8.031 8.169 dbcsr_mm_accdrv_process 20934 16.1 3.291 4.555 7.295 7.976 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.391 7.462 density_rs2pw 129 9.7 0.005 0.006 7.099 7.444 wfi_extrapolate 11 7.9 0.001 0.001 7.366 7.366 calculate_dm_sparse 129 9.5 0.001 0.001 6.190 6.278 fft3d_ps 1301 14.7 3.161 3.261 6.246 6.269 grid_collocate_task_list 129 9.7 5.619 5.871 5.619 5.871 dbcsr_complete_redistribute 397 12.7 2.105 2.175 5.235 5.559 multiply_cannon_metrocomm4 7587 15.6 0.026 0.029 1.883 5.507 mp_irecv_dv 29102 15.9 1.843 5.420 1.843 5.420 mp_alltoall_d11v 2429 14.1 4.651 5.401 4.651 5.401 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.248 5.270 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.244 5.263 mp_allgather_i34 2529 14.6 1.290 4.675 1.290 4.675 potential_pw2rs 129 12.3 0.027 0.028 4.471 4.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.102 4.141 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.589 3.858 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.692 3.735 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.413 3.727 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.616 3.722 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.503 3.517 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.441 3.507 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.059000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1470.545455, yerr=24.298760 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/20/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 11.696234E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.117277E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 59136. MP_Allreduce 11082 1503. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.072 295.334 295.345 qs_mol_dyn_low 1 2.0 0.003 0.003 294.603 294.632 qs_forces 11 3.9 0.004 0.005 294.495 294.501 qs_energies 11 4.9 0.003 0.004 285.546 285.556 scf_env_do_scf 11 5.9 0.001 0.002 262.920 262.935 velocity_verlet 10 3.0 0.002 0.004 213.136 213.144 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 137.066 137.068 init_scf_loop 11 6.9 0.000 0.000 125.579 125.581 prepare_preconditioner 11 7.9 0.000 0.000 120.721 120.742 make_preconditioner 11 8.9 0.000 0.000 120.721 120.742 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.534 117.835 qs_scf_new_mos 117 7.6 0.001 0.001 91.776 91.861 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.775 91.860 ot_scf_mini 117 9.6 0.004 0.004 86.877 86.918 dbcsr_multiply_generic 2507 12.6 0.216 0.226 82.747 83.319 cp_fm_upper_to_full 105 14.8 53.845 76.986 53.845 76.986 multiply_cannon 2507 13.6 0.696 0.730 59.134 60.490 multiply_cannon_loop 2507 14.6 1.046 1.060 55.000 56.442 ot_mini 117 10.6 0.001 0.001 45.123 45.181 dbcsr_complete_redistribute 395 12.7 4.006 4.072 30.186 43.315 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.678 39.759 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.145 37.106 rebuild_ks_matrix 128 8.3 0.001 0.001 35.038 35.089 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 35.037 35.088 mp_alltoall_i22 716 14.1 21.908 34.859 21.908 34.859 cp_fm_cholesky_invert 11 10.9 33.559 33.566 33.559 33.566 mp_waitall_1 103674 16.8 28.466 32.683 28.466 32.683 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.517 32.575 qs_ot_get_p 128 10.4 0.002 0.002 26.786 26.881 qs_ot_get_derivative 117 11.6 0.002 0.002 25.007 25.045 qs_ot_p2m_diag 83 11.4 0.880 0.885 22.593 22.624 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.785 20.787 make_m2s 5014 13.6 0.074 0.077 18.903 20.442 multiply_cannon_metrocomm3 10028 15.6 0.025 0.026 18.974 20.403 ot_diis_step 117 11.6 0.022 0.022 20.080 20.080 make_images 5014 14.6 3.060 3.256 18.419 19.957 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.419 19.563 apply_single 128 13.6 0.001 0.001 19.419 19.563 multiply_cannon_multrec 10028 15.6 10.342 12.152 18.109 18.165 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.534 17.534 cp_fm_diag_elpa_base 83 14.4 13.064 14.697 17.529 17.529 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.502 17.516 calculate_rho_elec 128 8.7 0.478 0.479 17.501 17.515 sum_up_and_integrate 128 10.3 0.002 0.002 16.625 16.713 integrate_v_rspace 128 11.3 0.004 0.005 16.562 16.651 multiply_cannon_sync_h2d 10028 15.6 14.310 14.341 14.310 14.341 pw_transfer 1547 11.6 0.096 0.097 13.288 13.300 hybrid_alltoall_any 5200 16.5 1.312 3.083 10.948 13.189 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 13.050 13.061 make_images_data 5014 15.6 0.064 0.069 10.777 13.059 init_scf_run 11 5.9 0.000 0.001 12.169 12.170 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.169 12.170 fft_wrap_pw1pw2_140 523 13.2 3.214 3.295 11.589 11.597 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.860 9.886 dbcsr_mm_accdrv_process 20762 16.1 4.259 6.174 7.516 9.419 cp_fm_cholesky_decompose 22 10.9 9.009 9.024 9.009 9.024 wfi_extrapolate 11 7.9 0.001 0.001 8.991 8.992 grid_integrate_task_list 128 12.3 8.561 8.738 8.561 8.738 density_rs2pw 128 9.7 0.005 0.005 8.501 8.546 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 8.285 8.286 fft3d_ps 1291 14.7 3.977 3.991 7.518 7.563 mp_alltoall_d11v 2415 14.1 7.048 7.194 7.048 7.194 calculate_dm_sparse 128 9.5 0.001 0.001 6.772 6.809 grid_collocate_task_list 128 9.7 6.419 6.485 6.419 6.485 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.339 6.388 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.160 6.231 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.345000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2832.818182, yerr=138.760630 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 1.260880E+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 2275295. 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 0.094 0.117 85.168 85.169 qs_energies 1 2.0 0.000 0.000 84.375 84.385 ls_scf 1 3.0 0.000 0.000 83.459 83.469 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.323 72.513 multiply_cannon 111 7.7 0.018 0.021 55.767 56.920 multiply_cannon_loop 111 8.7 0.230 0.243 52.327 53.645 ls_scf_main 1 4.0 0.000 0.000 52.129 52.130 density_matrix_trs4 2 5.0 0.002 0.003 46.663 46.729 ls_scf_init_scf 1 4.0 0.000 0.000 28.297 28.298 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.166 27.220 mp_waitall_1 11031 10.9 22.231 25.969 22.231 25.969 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.006 24.939 24.955 multiply_cannon_multrec 2664 9.7 8.145 8.846 15.575 17.268 multiply_cannon_sync_h2d 2664 9.7 13.449 15.201 13.449 15.201 make_m2s 222 7.7 0.009 0.012 12.914 13.485 make_images 222 8.7 0.100 0.110 12.891 13.465 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 9.549 12.156 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.526 8.528 dbcsr_mm_accdrv_process 4760 10.4 0.601 0.713 7.049 8.057 make_images_data 222 9.7 0.005 0.006 7.488 7.968 hybrid_alltoall_any 227 10.6 0.218 1.841 6.440 7.541 dbcsr_mm_accdrv_process_sort 4760 11.4 6.247 7.160 6.247 7.160 calculate_norms 4752 9.8 5.526 6.170 5.526 6.170 mp_sum_l 887 5.1 3.112 5.193 3.112 5.193 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.970 5.063 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.361 4.036 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.043 3.793 mp_irecv_dv 6231 10.9 2.026 3.770 2.026 3.770 arnoldi_extremal 4 6.8 0.000 0.000 3.287 3.312 arnoldi_normal_ev 4 7.8 0.001 0.004 3.287 3.311 make_images_sizes 222 9.7 0.000 0.000 0.670 3.268 mp_alltoall_i44 222 10.7 0.669 3.267 0.669 3.267 build_subspace 16 8.4 0.009 0.012 3.196 3.198 ls_scf_post 1 4.0 0.000 0.000 3.032 3.042 ls_scf_store_result 1 5.0 0.000 0.000 2.834 2.888 dbcsr_special_finalize 555 9.7 0.006 0.006 2.398 2.813 dbcsr_merge_single_wm 555 10.7 0.462 0.594 2.389 2.805 make_images_pack 222 9.7 2.208 2.626 2.210 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.349 2.572 dbcsr_sort_data 658 11.4 2.185 2.534 2.185 2.534 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.454 2.062 2.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.280 2.347 buffer_matrices_ensure_size 222 8.7 1.756 2.112 1.756 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.764 1.765 rebuild_ks_matrix 3 7.3 0.000 0.000 1.754 1.756 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.754 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.169000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 2.171453E+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 0.027 0.045 90.060 90.062 qs_energies 1 2.0 0.000 0.001 89.405 89.409 ls_scf 1 3.0 0.000 0.002 88.059 88.064 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.066 74.370 multiply_cannon 111 7.7 0.029 0.043 52.678 56.964 ls_scf_main 1 4.0 0.000 0.002 54.499 54.504 multiply_cannon_loop 111 8.7 0.136 0.147 49.834 52.980 density_matrix_trs4 2 5.0 0.003 0.008 48.793 48.989 ls_scf_init_scf 1 4.0 0.000 0.001 29.871 29.874 mp_waitall_1 9105 10.9 20.601 29.453 20.601 29.453 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.669 28.762 multiply_cannon_multrec 1332 9.7 13.295 17.236 22.599 27.674 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.300 26.310 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.441 20.367 make_m2s 222 7.7 0.007 0.008 15.029 15.623 make_images 222 8.7 1.375 1.699 14.998 15.594 dbcsr_mm_accdrv_process 4041 10.4 0.330 0.535 8.902 10.500 dbcsr_mm_accdrv_process_sort 4041 11.4 8.420 9.964 8.420 9.964 hybrid_alltoall_any 227 10.6 0.542 2.569 8.011 9.581 make_images_data 222 9.7 0.004 0.005 8.647 9.484 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.227 7.757 mp_irecv_dv 3311 11.0 3.206 7.703 3.206 7.703 mp_sum_l 887 5.1 4.883 7.651 4.883 7.651 calculate_norms 2376 9.8 6.078 6.841 6.078 6.841 multiply_cannon_sync_h2d 1332 9.7 4.835 6.277 4.835 6.277 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.698 6.113 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.937 5.208 arnoldi_extremal 4 6.8 0.000 0.000 4.714 4.725 arnoldi_normal_ev 4 7.8 0.001 0.005 4.714 4.725 build_subspace 16 8.4 0.014 0.021 4.464 4.467 ls_scf_post 1 4.0 0.000 0.002 3.689 3.693 ls_scf_store_result 1 5.0 0.000 0.000 3.380 3.513 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.172 3.416 dbcsr_matrix_vector_mult_local 304 10.0 2.767 3.252 2.769 3.254 mp_allgather_i34 111 8.7 0.820 3.034 0.820 3.034 ls_scf_dm_to_ks 2 5.0 0.000 0.005 2.597 2.699 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.159 2.625 dbcsr_data_new 4174 10.1 2.114 2.403 2.114 2.403 make_images_pack 222 9.7 1.815 2.127 1.817 2.129 dbcsr_sort_data 436 11.2 1.825 2.063 1.825 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.863 1.866 rebuild_ks_matrix 3 7.3 0.000 0.000 1.850 1.853 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.850 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.062000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1770.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 2.835628E+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 0.031 0.049 91.970 91.972 qs_energies 1 2.0 0.000 0.001 91.317 91.321 ls_scf 1 3.0 0.000 0.002 89.879 89.882 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.288 74.666 multiply_cannon 111 7.7 0.037 0.098 51.651 56.114 ls_scf_main 1 4.0 0.000 0.002 55.732 55.736 multiply_cannon_loop 111 8.7 0.117 0.130 48.920 53.348 density_matrix_trs4 2 5.0 0.002 0.005 49.844 50.047 mp_waitall_1 7281 11.0 22.447 32.447 22.447 32.447 ls_scf_init_scf 1 4.0 0.000 0.001 30.401 30.403 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.181 29.282 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.821 26.832 multiply_cannon_multrec 888 9.7 12.752 15.230 21.368 24.424 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.496 22.557 make_m2s 222 7.7 0.006 0.008 15.784 16.445 make_images 222 8.7 1.590 1.849 15.746 16.408 make_images_data 222 9.7 0.004 0.005 8.894 9.865 dbcsr_mm_accdrv_process 3754 10.4 0.318 0.495 8.130 9.414 hybrid_alltoall_any 227 10.6 0.640 2.967 8.424 9.394 dbcsr_mm_accdrv_process_sort 3754 11.4 7.695 8.919 7.695 8.919 mp_sum_l 887 5.1 5.150 8.589 5.150 8.589 multiply_cannon_sync_h2d 888 9.7 6.039 7.346 6.039 7.346 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.490 7.308 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.700 7.275 mp_irecv_dv 2335 11.1 2.474 7.235 2.474 7.235 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.003 6.728 arnoldi_extremal 4 6.8 0.000 0.000 5.318 5.330 arnoldi_normal_ev 4 7.8 0.001 0.005 5.318 5.330 build_subspace 16 8.4 0.014 0.020 5.012 5.019 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.760 4.996 calculate_norms 1584 9.8 4.349 4.742 4.349 4.742 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.577 3.863 ls_scf_post 1 4.0 0.000 0.002 3.746 3.749 dbcsr_matrix_vector_mult_local 304 10.0 3.046 3.630 3.048 3.632 ls_scf_store_result 1 5.0 0.000 0.000 3.449 3.521 mp_allgather_i34 111 8.7 0.827 3.463 0.827 3.463 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.825 2.915 dbcsr_data_new 4116 9.9 2.112 2.450 2.112 2.450 make_images_sizes 222 9.7 0.000 0.001 1.081 2.184 mp_alltoall_i44 222 10.7 1.080 2.183 1.080 2.183 dbcsr_sort_data 325 11.1 1.839 2.153 1.839 2.153 dbcsr_finalize 304 7.8 0.026 0.032 1.617 1.904 make_images_pack 222 9.7 1.621 1.890 1.624 1.892 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.880 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.863 1.867 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.862 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.972000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2228.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 3.350131E+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 0.032 0.070 97.525 97.534 qs_energies 1 2.0 0.000 0.005 96.884 96.893 ls_scf 1 3.0 0.000 0.001 95.193 95.200 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.588 78.843 ls_scf_main 1 4.0 0.003 0.027 58.971 58.973 multiply_cannon 111 7.7 0.047 0.098 51.800 56.485 density_matrix_trs4 2 5.0 0.002 0.004 52.782 52.925 multiply_cannon_loop 111 8.7 0.152 0.166 46.686 49.112 ls_scf_init_scf 1 4.0 0.000 0.001 33.002 33.005 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.554 31.624 mp_waitall_1 6369 11.0 22.771 29.237 22.771 29.237 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.081 29.092 multiply_cannon_multrec 1332 9.7 14.252 17.620 22.178 25.166 make_m2s 222 7.7 0.008 0.008 21.229 22.687 make_images 222 8.7 3.148 3.621 21.178 22.638 multiply_cannon_metrocomm3 1332 9.7 0.004 0.004 9.168 17.140 make_images_data 222 9.7 0.004 0.005 11.870 13.511 hybrid_alltoall_any 227 10.6 0.799 3.790 11.178 12.990 dbcsr_mm_accdrv_process 3641 10.4 0.303 0.484 7.572 9.103 dbcsr_mm_accdrv_process_sort 3641 11.4 7.132 8.617 7.132 8.617 mp_sum_l 887 5.1 3.959 7.636 3.959 7.636 multiply_cannon_sync_h2d 1332 9.7 5.413 5.982 5.413 5.982 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.025 5.974 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.084 5.956 mp_irecv_dv 3229 10.9 2.059 5.873 2.059 5.873 arnoldi_extremal 4 6.8 0.000 0.000 5.300 5.313 arnoldi_normal_ev 4 7.8 0.001 0.005 5.300 5.313 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.567 5.120 build_subspace 16 8.4 0.014 0.021 4.964 4.971 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.560 4.741 mp_allgather_i34 111 8.7 2.218 4.677 2.218 4.677 calculate_norms 2376 9.8 4.176 4.468 4.176 4.468 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.639 3.942 dbcsr_matrix_vector_mult_local 304 10.0 3.227 3.717 3.229 3.719 dbcsr_sort_data 658 11.4 3.063 3.410 3.063 3.410 ls_scf_post 1 4.0 0.000 0.003 3.219 3.226 dbcsr_special_finalize 555 9.7 0.006 0.007 2.796 3.191 dbcsr_merge_single_wm 555 10.7 0.536 0.665 2.787 3.183 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.029 3.085 ls_scf_store_result 1 5.0 0.000 0.000 2.960 3.014 dbcsr_data_release 10477 10.7 1.598 2.417 1.598 2.417 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.121 2.122 rebuild_ks_matrix 3 7.3 0.000 0.000 2.097 2.099 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.016 2.097 2.099 dbcsr_finalize 304 7.8 0.049 0.061 1.820 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.534000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 4.759085E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.073 92.051 92.052 qs_energies 1 2.0 0.000 0.000 91.238 91.243 ls_scf 1 3.0 0.000 0.000 89.233 89.237 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.738 70.931 ls_scf_main 1 4.0 0.000 0.000 56.351 56.351 multiply_cannon 111 7.7 0.073 0.128 52.685 56.092 multiply_cannon_loop 111 8.7 0.089 0.098 50.108 51.350 density_matrix_trs4 2 5.0 0.002 0.003 49.336 49.408 ls_scf_init_scf 1 4.0 0.000 0.000 29.449 29.451 mp_waitall_1 5436 11.0 24.360 28.849 24.360 28.849 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.185 28.221 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.064 26.078 multiply_cannon_multrec 444 9.7 13.802 16.614 20.897 22.177 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.853 15.838 make_m2s 222 7.7 0.005 0.005 13.411 14.324 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.139 14.303 make_images 222 8.7 2.039 2.483 13.343 14.255 make_images_data 222 9.7 0.003 0.004 8.182 9.586 hybrid_alltoall_any 227 10.6 0.801 3.824 8.147 9.390 multiply_cannon_sync_h2d 444 9.7 6.709 8.105 6.709 8.105 dbcsr_mm_accdrv_process 3003 10.4 0.343 0.420 6.785 7.878 dbcsr_mm_accdrv_process_sort 3003 11.4 6.415 7.479 6.415 7.479 arnoldi_extremal 4 6.8 0.000 0.000 5.837 5.852 arnoldi_normal_ev 4 7.8 0.002 0.005 5.836 5.851 build_subspace 16 8.4 0.015 0.020 5.435 5.444 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.463 4.678 mp_sum_l 887 5.1 2.654 4.363 2.654 4.363 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.167 4.336 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.137 3.699 4.140 calculate_norms 792 9.8 3.642 3.780 3.642 3.780 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.521 3.668 mp_irecv_dv 1241 11.2 1.510 3.642 1.510 3.642 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.563 3.624 mp_allgather_i34 111 8.7 1.145 3.560 1.145 3.560 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.875 3.523 ls_scf_post 1 4.0 0.000 0.000 3.432 3.437 ls_scf_store_result 1 5.0 0.000 0.000 3.223 3.261 make_images_sizes 222 9.7 0.000 0.000 0.828 3.230 mp_alltoall_i44 222 10.7 0.828 3.229 0.828 3.229 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.264 dbcsr_data_new 4608 9.7 1.809 2.248 1.809 2.248 dbcsr_merge_all 275 8.9 0.480 0.523 2.056 2.105 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.040 2.041 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.008 2.009 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.989 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.052000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3738.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_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 8.976630E+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 0.094 0.099 106.119 106.119 qs_energies 1 2.0 0.000 0.000 104.737 104.743 ls_scf 1 3.0 0.000 0.000 101.730 101.735 dbcsr_multiply_generic 111 6.7 0.023 0.027 75.139 75.288 ls_scf_main 1 4.0 0.000 0.000 63.932 63.933 density_matrix_trs4 2 5.0 0.002 0.003 54.827 54.896 multiply_cannon 111 7.7 0.148 0.229 48.630 50.689 multiply_cannon_loop 111 8.7 0.098 0.099 45.685 46.335 ls_scf_init_scf 1 4.0 0.000 0.000 34.006 34.007 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.435 32.464 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.747 29.754 mp_waitall_1 4527 11.1 22.153 25.825 22.153 25.825 make_m2s 222 7.7 0.005 0.005 22.770 23.801 make_images 222 8.7 3.574 3.892 22.659 23.689 multiply_cannon_multrec 444 9.7 17.821 18.394 22.562 23.300 hybrid_alltoall_any 227 10.6 1.652 3.611 12.915 15.869 make_images_data 222 9.7 0.004 0.004 13.159 15.241 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.305 10.955 multiply_cannon_sync_h2d 444 9.7 8.775 8.830 8.775 8.830 arnoldi_extremal 4 6.8 0.000 0.000 7.428 7.438 arnoldi_normal_ev 4 7.8 0.003 0.009 7.428 7.438 build_subspace 16 8.4 0.026 0.036 6.875 6.886 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.505 5.622 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.310 5.045 5.313 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.214 5.306 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.869 5.130 dbcsr_mm_accdrv_process 1814 10.4 0.314 0.607 4.560 4.823 dbcsr_mm_accdrv_process_sort 1814 11.4 4.132 4.261 4.132 4.261 ls_scf_post 1 4.0 0.000 0.000 3.792 3.797 make_images_sizes 222 9.7 0.000 0.000 1.466 3.607 mp_alltoall_i44 222 10.7 1.465 3.607 1.465 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.515 3.553 calculate_norms 792 9.8 3.244 3.324 3.244 3.324 mp_allgather_i34 111 8.7 1.043 3.311 1.043 3.311 dbcsr_finalize 304 7.8 0.082 0.090 3.074 3.187 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.977 2.977 dbcsr_merge_all 275 8.9 0.884 0.915 2.858 2.963 dbcsr_complete_redistribute 5 7.6 1.427 1.470 2.739 2.862 matrix_ls_to_qs 2 6.0 0.000 0.000 2.411 2.544 dbcsr_sort_data 325 11.1 2.437 2.509 2.437 2.509 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.486 2.488 rebuild_ks_matrix 3 7.3 0.000 0.000 2.412 2.414 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.412 2.414 dbcsr_new_transposed 4 7.5 0.241 0.252 2.336 2.360 dbcsr_data_new 6591 9.6 1.864 2.317 1.864 2.317 dbcsr_frobenius_norm 74 6.6 2.061 2.143 2.210 2.240 dbcsr_add_d 103 6.2 0.000 0.000 2.139 2.224 dbcsr_add_anytype 103 7.2 0.858 0.891 2.139 2.223 dbcsr_data_release 12724 10.6 1.976 2.195 1.976 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.119000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7031.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4955cfc0d24997d34b1dc70cd9db47ebb77d3489_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.907712E+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 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 0.032 0.156 249.517 249.518 qs_mol_dyn_low 1 2.0 0.003 0.005 248.231 248.327 qs_forces 5 3.8 0.006 0.048 248.144 248.145 qs_energies 5 4.8 0.003 0.023 244.987 245.010 scf_env_do_scf 5 5.8 0.000 0.002 230.701 230.704 scf_env_do_scf_inner_loop 105 6.6 0.003 0.019 201.532 201.548 qs_scf_new_mos 105 7.6 0.000 0.001 158.155 158.296 qs_scf_loop_do_ot 105 8.6 0.001 0.001 158.154 158.295 ot_scf_mini 105 9.6 0.003 0.005 148.308 148.417 dbcsr_multiply_generic 1445 12.2 0.125 0.132 133.708 134.100 multiply_cannon 1445 13.2 0.275 0.285 114.237 116.223 multiply_cannon_loop 1445 14.2 2.839 3.003 112.339 113.554 velocity_verlet 4 3.0 0.003 0.011 109.381 109.381 ot_mini 105 10.6 0.001 0.022 60.903 60.994 qs_ot_get_p 112 10.4 0.001 0.003 57.355 57.669 qs_ot_p2m_diag 40 11.0 0.020 0.031 46.016 46.077 multiply_cannon_multrec 69360 15.2 29.644 34.509 39.680 44.348 cp_dbcsr_syevd 40 12.0 0.002 0.003 42.618 42.620 mp_waitall_1 488190 16.1 35.836 41.601 35.836 41.601 qs_ot_get_derivative 55 11.6 0.001 0.001 39.111 39.217 cp_fm_syevd 40 13.0 0.000 0.001 37.063 37.198 multiply_cannon_sync_h2d 69360 15.2 29.163 33.342 29.163 33.342 cp_fm_redistribute_end 40 14.0 16.050 32.029 16.057 32.032 cp_fm_syevd_base 40 14.0 15.967 31.947 15.967 31.947 multiply_cannon_metrocomm3 69360 15.2 0.204 0.215 25.803 31.781 rebuild_ks_matrix 110 8.4 0.000 0.000 30.241 30.403 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.029 30.240 30.402 init_scf_loop 7 6.6 0.001 0.011 29.131 29.133 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.730 27.889 prepare_preconditioner 7 7.6 0.000 0.001 24.007 24.035 make_preconditioner 7 8.6 0.000 0.003 24.007 24.035 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.072 23.394 apply_single 62 13.6 0.000 0.000 23.071 23.394 ot_new_cg_direction 55 11.6 0.001 0.002 21.066 21.067 qs_rho_update_rho_low 110 7.6 0.001 0.005 19.291 19.597 calculate_rho_elec 110 8.6 0.030 0.032 19.290 19.597 make_full_inverse_cholesky 7 9.6 0.000 0.000 16.197 16.265 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.773 14.945 density_rs2pw 110 9.6 0.005 0.006 13.512 14.033 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.905 13.978 mp_sum_l 4764 12.2 12.304 12.872 12.304 12.872 pw_transfer 1645 12.4 0.080 0.109 11.982 12.235 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 11.843 12.099 calculate_dm_sparse 110 9.5 0.000 0.001 11.508 11.677 init_scf_run 5 5.8 0.000 0.001 11.561 11.562 scf_env_initial_rho_setup 5 6.8 0.001 0.007 11.561 11.562 cp_fm_cholesky_invert 7 10.6 11.262 11.272 11.262 11.272 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.741 10.798 dbcsr_mm_accdrv_process 154766 15.8 6.219 6.422 9.903 10.733 fft_wrap_pw1pw2_240 915 15.0 0.848 0.925 10.380 10.643 transfer_rs2pw 445 10.6 0.007 0.009 10.070 10.598 qs_vxc_create 110 10.4 0.002 0.010 10.525 10.570 check_diag 80 13.5 8.686 8.935 10.060 10.195 sum_up_and_integrate 60 10.3 0.001 0.004 9.665 9.686 integrate_v_rspace 60 11.3 0.002 0.002 9.648 9.670 fft3d_pb 915 16.0 2.403 2.674 8.530 8.831 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.356 7.606 make_full_single_inverse 7 9.6 0.001 0.007 7.518 7.546 acc_transpose_blocks 69360 15.2 0.358 0.377 7.191 7.542 make_m2s 2890 13.2 0.079 0.088 6.808 7.391 make_images 2890 14.2 0.241 0.261 6.701 7.282 xc_vxc_pw_create 60 11.3 0.039 0.052 7.060 7.105 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.981 6.999 multiply_cannon_metrocomm1 69360 15.2 0.098 0.104 4.554 6.897 calculate_first_density_matrix 1 7.0 0.000 0.007 6.622 6.635 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.507 6.564 xc_pw_derive 510 13.4 0.005 0.007 6.276 6.337 mp_alltoall_z22v 2340 17.7 5.535 5.946 5.535 5.946 potential_pw2rs 60 12.3 0.002 0.003 5.826 5.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=249.518000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=564.400000, yerr=2.154066 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 4955cfc0d24997d34b1dc70cd9db47ebb77d3489 Summary: empty Status: OK