=== 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: b331c57d0a21e261d77504111c8b6a615dad1b78 ################# 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, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/01 job id: 48291450 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/02 job id: 48291451 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/03 job id: 48291452 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/04 job id: 48291453 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/05 job id: 48291455 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/06 job id: 48291456 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/07 job id: 48291457 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/08 job id: 48291458 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/09 job id: 48291459 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/10 job id: 48291460 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/11 job id: 48291461 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/12 job id: 48291462 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/13 job id: 48291463 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/14 job id: 48291464 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/15 job id: 48291465 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/16 job id: 48291466 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/17 job id: 48291467 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/18 job id: 48291468 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/19 job id: 48291470 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/20 job id: 48291471 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/21 job id: 48291472 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/22 job id: 48291473 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/23 job id: 48291474 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/24 job id: 48291475 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/25 job id: 48291476 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/26 job id: 48291477 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/27 job id: 48291478 --- 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/b331c57d0a21e261d77504111c8b6a615dad1b78_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 0.022 0.037 133.927 133.928 farming_run 1 2.0 133.379 133.381 133.899 133.903 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.497645E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 115.154 115.154 qs_energies 1 2.0 0.000 0.000 114.956 114.958 mp2_main 1 3.0 0.000 0.000 112.867 112.868 mp2_gpw_main 1 4.0 0.028 0.033 111.834 111.836 mp2_ri_gpw_compute_in 1 5.0 0.175 0.221 92.916 93.330 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 54.991 55.406 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.166 41.370 46.500 get_2c_integrals 1 6.0 0.008 0.009 37.349 37.753 integrate_v_rspace 273 8.0 0.438 0.453 24.743 29.728 pw_transfer 6555 10.6 0.377 0.389 27.367 28.033 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.006 26.481 grid_integrate_task_list 273 9.0 20.548 26.022 20.548 26.022 fft_wrap_pw1pw2_100 2178 12.4 1.174 1.443 23.528 24.009 compute_2c_integrals 1 7.0 0.003 0.003 19.764 19.764 compute_2c_integrals_loop_lm 1 8.0 0.004 0.007 18.838 19.491 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.422 18.834 19.483 rpa_ri_compute_en 1 5.0 0.029 0.032 18.801 19.008 cp_fm_cholesky_decompose 12 8.2 17.513 17.923 17.513 17.923 cholesky_decomp 1 7.0 0.000 0.000 16.426 16.834 fft3d_s 5443 13.4 16.141 16.647 16.163 16.668 ao_to_mo_and_store_B_mult_1 272 7.0 10.775 15.319 10.775 15.319 calculate_wavefunction 272 8.0 5.425 5.583 12.533 13.145 rpa_num_int 1 6.0 0.001 0.011 10.648 10.649 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.622 10.631 calc_mat_Q 8 8.0 0.000 0.000 9.483 9.580 contract_S_to_Q 8 9.0 0.000 0.000 8.905 8.998 calc_potential_gpw 544 9.5 0.005 0.006 8.278 8.735 parallel_gemm_fm 14 9.1 0.000 0.000 8.487 8.587 parallel_gemm_fm_cosma 14 10.1 8.487 8.587 8.487 8.587 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.198 8.577 potential_pw2rs 545 10.0 0.107 0.109 7.668 8.303 collocate_single_gaussian 272 10.0 0.040 0.042 7.447 7.817 create_integ_mat 1 6.0 0.022 0.027 7.634 7.634 array2fm 1 7.0 0.000 0.000 6.730 7.097 pw_scatter_s 2720 13.7 4.433 4.620 4.433 4.620 pw_gather_s 2722 13.2 3.852 4.211 3.852 4.211 array2fm_buffer_send 1 8.0 2.951 3.130 2.951 3.130 pw_poisson_solve 545 10.5 1.097 1.142 2.164 2.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.834447, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 0.028 0.040 420.897 420.898 farming_run 1 2.0 420.146 420.150 420.861 420.862 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.242714E+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 0.015 0.036 217.275 217.276 qs_energies 1 2.0 0.000 0.000 217.019 217.028 scf_env_do_scf 1 3.0 0.000 0.000 114.405 114.405 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.465 113.472 rebuild_ks_matrix 4 6.0 0.000 0.000 113.464 113.471 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.464 113.471 hfx_ks_matrix 4 8.0 0.001 0.001 113.080 113.084 integrate_four_center 4 9.0 0.154 0.471 113.079 113.083 integrate_four_center_main 4 10.0 0.128 0.558 101.670 104.352 integrate_four_center_bin 264 11.0 101.542 104.348 101.542 104.348 mp2_main 1 3.0 0.000 0.000 102.318 102.327 mp2_gpw_main 1 4.0 0.048 0.071 101.417 101.431 init_scf_loop 1 4.0 0.000 0.000 96.231 96.231 mp2_ri_gpw_compute_in 1 5.0 0.064 0.066 74.485 75.576 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.063 55.155 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 41.679 46.722 integrate_v_rspace 95 8.0 0.394 0.550 28.048 32.908 pw_transfer 2240 10.6 0.145 0.168 29.948 30.257 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.960 29.277 mp2_ri_gpw_compute_en 1 5.0 0.058 0.081 26.769 28.871 ao_to_mo_and_store_B_mult_1 91 7.0 10.705 28.406 10.705 28.406 grid_integrate_task_list 95 9.0 23.335 28.298 23.335 28.298 fft_wrap_pw1pw2_100 730 12.4 1.267 1.450 26.685 27.009 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.894 25.009 25.019 get_2c_integrals 1 6.0 0.000 0.000 20.335 20.356 compute_2c_integrals 1 7.0 0.003 0.003 19.310 19.313 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.843 19.185 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.832 18.842 19.185 fft3d_s 1823 13.4 18.451 18.834 18.465 18.847 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.172 18.172 calculate_wavefunction 91 8.0 2.011 2.048 9.745 9.966 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.578 8.828 9.292 potential_pw2rs 186 10.0 0.033 0.038 8.646 9.196 local_gemm 172 8.0 8.273 8.727 8.273 8.727 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.237 8.592 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.515 7.971 8.578 collocate_single_gaussian 91 10.0 0.017 0.021 7.878 8.153 calc_potential_gpw 182 9.5 0.002 0.002 7.907 8.139 integrate_four_center_load 4 10.0 0.000 0.000 6.754 6.758 hfx_load_balance 1 11.0 0.000 0.000 6.754 6.758 mp_sendrecv_dm3 2068 8.0 6.014 6.622 6.014 6.622 mp_sync 37 10.5 3.387 6.503 3.387 6.503 mp2_ri_gpw_compute_en_ener 172 7.0 6.331 6.405 6.331 6.405 pw_gather_s 912 13.2 4.931 5.383 4.931 5.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.413034, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 451.801088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 55.016 55.017 qs_mol_dyn_low 1 2.0 0.003 0.004 54.779 54.787 qs_forces 11 3.9 0.002 0.003 54.718 54.719 qs_energies 11 4.9 0.001 0.001 53.203 53.218 scf_env_do_scf 11 5.9 0.000 0.001 46.914 46.914 scf_env_do_scf_inner_loop 108 6.5 0.003 0.010 44.818 44.818 qs_scf_new_mos 108 7.5 0.000 0.001 34.532 34.837 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.531 34.837 dbcsr_multiply_generic 2286 12.5 0.093 0.098 34.210 34.703 ot_scf_mini 108 9.5 0.002 0.002 32.828 33.008 multiply_cannon 2286 13.5 0.188 0.199 26.835 28.637 multiply_cannon_loop 2286 14.5 1.842 1.947 26.151 27.967 velocity_verlet 10 3.0 0.001 0.002 27.474 27.475 ot_mini 108 10.5 0.001 0.001 19.973 20.231 qs_ot_get_derivative 108 11.5 0.001 0.001 16.858 17.036 mp_waitall_1 245248 16.5 8.712 15.320 8.712 15.320 multiply_cannon_metrocomm3 54864 15.5 0.073 0.081 6.241 13.396 multiply_cannon_multrec 54864 15.5 3.641 5.717 7.902 11.355 qs_ot_get_p 119 10.4 0.001 0.001 8.143 8.425 rebuild_ks_matrix 119 8.3 0.000 0.000 7.966 8.097 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.966 8.097 mp_sum_l 7287 12.8 5.381 7.136 5.381 7.136 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.009 7.133 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.069 6.554 multiply_cannon_sync_h2d 54864 15.5 5.131 6.273 5.131 6.273 dbcsr_mm_accdrv_process 76910 16.1 1.847 2.904 4.174 5.937 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.279 5.317 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.099 5.206 init_scf_run 11 5.9 0.000 0.001 5.029 5.030 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.029 5.029 sum_up_and_integrate 119 10.3 0.012 0.014 4.568 4.576 integrate_v_rspace 119 11.3 0.002 0.003 4.556 4.565 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.515 4.515 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.337 4.476 calculate_rho_elec 119 8.7 0.012 0.017 4.337 4.475 cp_fm_diag_elpa 50 13.0 0.000 0.001 4.298 4.298 cp_fm_redistribute_end 50 14.0 2.192 4.270 2.199 4.273 cp_fm_diag_elpa_base 50 14.0 2.068 4.161 2.072 4.169 calculate_dm_sparse 119 9.5 0.000 0.001 3.052 3.255 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.940 3.140 apply_single 119 13.6 0.000 0.000 2.939 3.140 multiply_cannon_metrocomm1 54864 15.5 0.058 0.065 1.784 2.963 jit_kernel_multiply 13 15.8 2.262 2.941 2.262 2.941 rs_pw_transfer 974 11.9 0.011 0.013 2.775 2.920 acc_transpose_blocks 54864 15.5 0.236 0.260 2.303 2.854 ot_diis_step 108 11.5 0.006 0.006 2.773 2.774 calculate_first_density_matrix 1 7.0 0.000 0.000 2.753 2.757 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.493 2.555 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.446 2.449 density_rs2pw 119 9.7 0.004 0.004 2.244 2.390 wfi_extrapolate 11 7.9 0.001 0.001 2.204 2.204 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.106 2.165 grid_integrate_task_list 119 12.3 2.022 2.129 2.022 2.129 init_scf_loop 11 6.9 0.000 0.000 2.077 2.077 mp_sum_d 4135 12.0 1.430 1.990 1.430 1.990 potential_pw2rs 119 12.3 0.004 0.004 1.906 1.927 pw_transfer 1439 11.6 0.052 0.060 1.775 1.850 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.699 1.777 make_m2s 4572 13.5 0.054 0.056 1.679 1.727 make_images 4572 14.5 0.133 0.138 1.596 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.509 1.550 mp_alltoall_d11v 2130 13.8 1.336 1.528 1.336 1.528 acc_transpose_blocks_sync 164592 16.5 1.222 1.462 1.222 1.462 grid_collocate_task_list 119 9.7 1.357 1.426 1.357 1.426 fft3d_ps 1201 14.6 0.376 0.480 1.342 1.411 fft_wrap_pw1pw2_140 487 13.2 0.184 0.199 1.311 1.389 mp_waitany 12084 13.8 1.247 1.388 1.247 1.388 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.844 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.017000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 487.989248E+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 597017. 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 0.012 0.031 39.668 39.670 qs_mol_dyn_low 1 2.0 0.003 0.003 39.465 39.473 qs_forces 11 3.9 0.002 0.003 39.381 39.382 qs_energies 11 4.9 0.001 0.001 37.662 37.666 scf_env_do_scf 11 5.9 0.000 0.001 32.599 32.599 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.090 30.091 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.118 22.457 qs_scf_new_mos 108 7.5 0.001 0.001 20.828 21.048 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.827 21.048 ot_scf_mini 108 9.5 0.002 0.003 19.907 20.069 velocity_verlet 10 3.0 0.001 0.002 18.933 18.934 multiply_cannon 2286 13.5 0.209 0.216 16.891 18.307 multiply_cannon_loop 2286 14.5 1.193 1.258 15.702 17.259 ot_mini 108 10.5 0.001 0.001 12.279 12.514 mp_waitall_1 200699 16.5 5.546 10.664 5.546 10.664 qs_ot_get_derivative 108 11.5 0.001 0.001 9.905 10.074 multiply_cannon_metrocomm3 27432 15.5 0.073 0.077 4.009 9.395 multiply_cannon_multrec 27432 15.5 1.814 4.165 6.112 8.976 rebuild_ks_matrix 119 8.3 0.000 0.000 7.273 7.422 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.272 7.422 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.409 6.546 dbcsr_mm_accdrv_process 47894 16.0 3.453 5.531 4.218 6.274 qs_ot_get_p 119 10.4 0.001 0.001 4.692 4.910 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.683 4.583 sum_up_and_integrate 119 10.3 0.024 0.028 4.178 4.183 integrate_v_rspace 119 11.3 0.002 0.003 4.154 4.161 mp_sum_l 7287 12.8 2.113 4.111 2.113 4.111 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.875 3.907 calculate_rho_elec 119 8.7 0.021 0.024 3.874 3.907 init_scf_run 11 5.9 0.000 0.001 3.873 3.873 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.872 3.873 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.066 3.734 apply_single 119 13.6 0.000 0.000 3.066 3.734 rs_pw_transfer 974 11.9 0.010 0.011 2.576 3.091 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.031 3.050 make_m2s 4572 13.5 0.052 0.054 2.663 3.008 make_images 4572 14.5 0.206 0.245 2.575 2.923 density_rs2pw 119 9.7 0.004 0.004 2.138 2.671 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.626 2.626 init_scf_loop 11 6.9 0.000 0.000 2.488 2.488 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.376 2.461 ot_diis_step 108 11.5 0.011 0.011 2.323 2.324 calculate_first_density_matrix 1 7.0 0.000 0.000 2.313 2.315 cp_fm_diag_elpa 50 13.0 0.001 0.001 2.297 2.297 cp_fm_redistribute_end 50 14.0 1.164 2.270 1.167 2.273 multiply_cannon_sync_h2d 27432 15.5 1.693 2.237 1.693 2.237 cp_fm_diag_elpa_base 50 14.0 1.074 2.183 1.102 2.217 calculate_dm_sparse 119 9.5 0.000 0.001 2.090 2.170 jit_kernel_multiply 11 16.4 0.705 2.121 0.705 2.121 pw_transfer 1439 11.6 0.065 0.070 1.954 1.994 acc_transpose_blocks 27432 15.5 0.118 0.125 1.649 1.992 grid_integrate_task_list 119 12.3 1.838 1.935 1.838 1.935 potential_pw2rs 119 12.3 0.006 0.006 1.899 1.910 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.862 1.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.881 1.883 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.665 1.705 prepare_preconditioner 11 7.9 0.000 0.000 1.558 1.584 make_preconditioner 11 8.9 0.000 0.000 1.558 1.584 make_images_data 4572 15.5 0.048 0.054 1.246 1.576 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.459 1.514 wfi_extrapolate 11 7.9 0.001 0.001 1.497 1.497 fft_wrap_pw1pw2_140 487 13.2 0.203 0.213 1.457 1.494 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.114 1.485 fft3d_ps 1201 14.6 0.519 0.578 1.411 1.445 grid_collocate_task_list 119 9.7 1.280 1.416 1.280 1.416 mp_alltoall_d11v 2130 13.8 1.220 1.355 1.220 1.355 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.277 1.286 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.216 1.267 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.534 1.053 mp_waitany 5720 13.7 0.508 1.047 0.508 1.047 mp_allgather_i34 2286 14.5 0.608 1.016 0.608 1.016 mp_sum_d 4135 12.0 0.540 1.014 0.540 1.014 acc_transpose_blocks_sync 82296 16.5 0.822 0.949 0.822 0.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.920 0.934 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.930 0.931 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.683 0.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.670000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.545455, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 519.127040E+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 639328. 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 0.016 0.036 32.650 32.652 qs_mol_dyn_low 1 2.0 0.003 0.004 32.391 32.399 qs_forces 11 3.9 0.007 0.036 32.331 32.333 qs_energies 11 4.9 0.003 0.005 30.706 30.709 scf_env_do_scf 11 5.9 0.002 0.011 26.008 26.008 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 23.536 23.538 dbcsr_multiply_generic 2286 12.5 0.094 0.095 16.831 16.938 qs_scf_new_mos 108 7.5 0.001 0.001 15.403 15.422 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.403 15.421 velocity_verlet 10 3.0 0.002 0.002 15.392 15.393 ot_scf_mini 108 9.5 0.003 0.004 14.647 14.666 multiply_cannon 2286 13.5 0.193 0.196 13.588 14.430 multiply_cannon_loop 2286 14.5 0.855 0.897 12.838 13.673 ot_mini 108 10.5 0.001 0.002 9.043 9.056 qs_ot_get_derivative 108 11.5 0.001 0.002 7.558 7.576 multiply_cannon_multrec 18288 15.5 1.877 2.994 6.932 7.307 rebuild_ks_matrix 119 8.3 0.000 0.000 6.370 6.389 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.370 6.388 dbcsr_mm_accdrv_process 38222 16.0 4.885 5.769 4.962 5.835 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.618 5.636 sum_up_and_integrate 119 10.3 0.031 0.034 3.936 3.941 integrate_v_rspace 119 11.3 0.003 0.003 3.904 3.914 mp_waitall_1 158411 16.6 2.581 3.739 2.581 3.739 init_scf_run 11 5.9 0.000 0.001 3.491 3.491 scf_env_initial_rho_setup 11 6.9 0.001 0.004 3.490 3.491 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.440 3.445 calculate_rho_elec 119 8.7 0.031 0.032 3.439 3.444 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.844 3.431 qs_ot_get_p 119 10.4 0.001 0.001 3.393 3.421 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.430 2.494 init_scf_loop 11 6.9 0.002 0.003 2.446 2.448 rs_pw_transfer 974 11.9 0.009 0.010 2.210 2.447 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 1.993 2.334 apply_single 119 13.6 0.000 0.000 1.993 2.334 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.263 2.269 calculate_first_density_matrix 1 7.0 0.001 0.002 2.245 2.247 density_rs2pw 119 9.7 0.004 0.006 1.932 2.175 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.964 1.964 make_m2s 4572 13.5 0.044 0.045 1.796 1.915 pw_transfer 1439 11.6 0.065 0.069 1.904 1.913 grid_integrate_task_list 119 12.3 1.803 1.904 1.803 1.904 calculate_dm_sparse 119 9.5 0.000 0.001 1.860 1.873 make_images 4572 14.5 0.193 0.204 1.711 1.830 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.811 1.823 acc_transpose_blocks 18288 15.5 0.081 0.083 1.654 1.760 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.718 1.724 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.714 1.724 cp_fm_diag_elpa_base 50 14.0 1.689 1.700 1.711 1.721 potential_pw2rs 119 12.3 0.007 0.008 1.692 1.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.659 1.661 prepare_preconditioner 11 7.9 0.000 0.000 1.627 1.630 make_preconditioner 11 8.9 0.000 0.001 1.627 1.630 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.488 1.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.501 1.509 fft_wrap_pw1pw2_140 487 13.2 0.256 0.266 1.460 1.471 ot_diis_step 108 11.5 0.012 0.014 1.469 1.469 mp_sum_l 7287 12.8 1.064 1.439 1.064 1.439 grid_collocate_task_list 119 9.7 1.246 1.402 1.246 1.402 fft3d_ps 1201 14.6 0.529 0.548 1.294 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.200 1.205 wfi_extrapolate 11 7.9 0.001 0.005 1.178 1.178 multiply_cannon_sync_h2d 18288 15.5 1.024 1.160 1.024 1.160 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 0.980 0.981 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.943 0.959 make_images_data 4572 15.5 0.047 0.051 0.772 0.914 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.671 0.848 acc_transpose_blocks_sync 54864 16.5 0.743 0.839 0.743 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.825 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.810 0.824 mp_waitany 9880 13.7 0.516 0.789 0.516 0.789 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.123 0.513 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.679 0.746 mp_alltoall_d11v 2130 13.8 0.643 0.736 0.643 0.736 mp_alltoall_z22v 1201 16.6 0.618 0.694 0.618 0.694 cp_fm_cholesky_invert 11 10.9 0.683 0.687 0.683 0.687 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.652000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.818182, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 560.107520E+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 0.017 0.034 37.135 37.136 qs_mol_dyn_low 1 2.0 0.003 0.005 36.860 36.919 qs_forces 11 3.9 0.004 0.015 36.798 36.799 qs_energies 11 4.9 0.002 0.005 35.025 35.030 scf_env_do_scf 11 5.9 0.001 0.001 29.943 29.944 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.509 26.509 dbcsr_multiply_generic 2286 12.5 0.099 0.105 19.833 19.934 velocity_verlet 10 3.0 0.001 0.002 18.854 18.856 qs_scf_new_mos 108 7.5 0.001 0.001 17.938 18.002 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.937 18.001 ot_scf_mini 108 9.5 0.002 0.003 16.903 16.957 multiply_cannon 2286 13.5 0.218 0.227 16.056 16.526 multiply_cannon_loop 2286 14.5 1.532 1.618 15.098 15.428 ot_mini 108 10.5 0.001 0.001 10.362 10.423 multiply_cannon_multrec 27432 15.5 2.458 3.143 9.001 9.290 qs_ot_get_derivative 108 11.5 0.001 0.001 8.481 8.536 dbcsr_mm_accdrv_process 47916 15.9 6.015 7.535 6.440 7.781 rebuild_ks_matrix 119 8.3 0.000 0.000 6.667 6.714 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.666 6.713 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.897 5.932 sum_up_and_integrate 119 10.3 0.035 0.038 3.837 3.849 integrate_v_rspace 119 11.3 0.003 0.003 3.802 3.813 qs_ot_get_p 119 10.4 0.001 0.002 3.699 3.772 init_scf_run 11 5.9 0.000 0.001 3.690 3.691 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.690 3.690 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.616 3.654 calculate_rho_elec 119 8.7 0.040 0.046 3.615 3.654 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.091 3.556 init_scf_loop 11 6.9 0.002 0.007 3.414 3.416 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.178 2.605 apply_single 119 13.6 0.000 0.000 2.178 2.604 prepare_preconditioner 11 7.9 0.000 0.000 2.535 2.544 make_preconditioner 11 8.9 0.000 0.001 2.535 2.544 acc_transpose_blocks 27432 15.5 0.121 0.125 2.416 2.510 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.139 2.468 make_m2s 4572 13.5 0.054 0.057 2.300 2.410 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.353 2.380 make_images 4572 14.5 0.273 0.336 2.191 2.300 calculate_first_density_matrix 1 7.0 0.000 0.002 2.281 2.284 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.230 2.240 rs_pw_transfer 974 11.9 0.009 0.009 2.039 2.236 calculate_dm_sparse 119 9.5 0.000 0.000 2.147 2.204 density_rs2pw 119 9.7 0.004 0.004 1.974 2.162 mp_waitall_1 137007 16.6 1.690 2.101 1.690 2.101 pw_transfer 1439 11.6 0.065 0.070 2.033 2.070 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.940 1.982 grid_integrate_task_list 119 12.3 1.830 1.925 1.830 1.925 ot_diis_step 108 11.5 0.012 0.013 1.836 1.836 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.823 1.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.817 1.818 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.633 1.645 fft_wrap_pw1pw2_140 487 13.2 0.290 0.304 1.596 1.641 potential_pw2rs 119 12.3 0.008 0.009 1.582 1.589 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.576 1.583 cp_fm_diag_elpa_base 50 14.0 1.541 1.556 1.573 1.580 mp_sum_l 7287 12.8 1.058 1.563 1.058 1.563 acc_transpose_blocks_sync 82296 16.5 1.441 1.528 1.441 1.528 fft3d_ps 1201 14.6 0.556 0.607 1.373 1.402 grid_collocate_task_list 119 9.7 1.248 1.370 1.248 1.370 wfi_extrapolate 11 7.9 0.001 0.001 1.351 1.351 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.295 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.244 1.254 cp_fm_upper_to_full 72 14.2 0.810 1.157 0.810 1.157 multiply_cannon_metrocomm3 27432 15.5 0.041 0.043 0.682 1.139 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.117 1.119 jit_kernel_multiply 6 16.1 0.359 1.085 0.359 1.085 dbcsr_complete_redistribute 329 12.2 0.123 0.153 0.786 1.068 make_images_data 4572 15.5 0.047 0.051 0.850 0.956 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.729 0.891 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.799 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.864 0.870 mp_alltoall_d11v 2130 13.8 0.757 0.863 0.757 0.863 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.585 0.857 acc_transpose_blocks_kernels 27432 16.5 0.273 0.284 0.824 0.836 cp_fm_cholesky_invert 11 10.9 0.745 0.748 0.745 0.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.136000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.090909, yerr=3.871916 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 619.413504E+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 0.011 0.027 28.721 28.722 qs_mol_dyn_low 1 2.0 0.003 0.003 28.555 28.563 qs_forces 11 3.9 0.002 0.002 28.499 28.500 qs_energies 11 4.9 0.001 0.001 26.771 26.773 scf_env_do_scf 11 5.9 0.000 0.001 22.082 22.082 scf_env_do_scf_inner_loop 108 6.5 0.006 0.010 19.633 19.634 velocity_verlet 10 3.0 0.002 0.002 14.839 14.842 dbcsr_multiply_generic 2286 12.5 0.091 0.097 12.988 13.050 qs_scf_new_mos 108 7.5 0.001 0.001 11.660 11.686 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.659 11.685 ot_scf_mini 108 9.5 0.002 0.002 10.953 10.978 multiply_cannon 2286 13.5 0.222 0.228 10.415 10.815 multiply_cannon_loop 2286 14.5 0.639 0.670 9.495 9.681 ot_mini 108 10.5 0.001 0.001 6.282 6.313 multiply_cannon_multrec 9144 15.5 1.665 1.859 5.938 6.146 rebuild_ks_matrix 119 8.3 0.000 0.000 5.972 5.996 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.971 5.996 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.309 5.331 qs_ot_get_derivative 108 11.5 0.001 0.001 4.988 5.014 dbcsr_mm_accdrv_process 12550 15.8 3.503 4.159 4.162 4.227 sum_up_and_integrate 119 10.3 0.037 0.041 3.633 3.637 integrate_v_rspace 119 11.3 0.003 0.003 3.596 3.600 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.528 3.534 calculate_rho_elec 119 8.7 0.060 0.061 3.528 3.533 init_scf_run 11 5.9 0.000 0.001 3.237 3.237 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.237 3.237 qs_ot_get_p 119 10.4 0.001 0.001 2.782 2.814 init_scf_loop 11 6.9 0.000 0.000 2.427 2.427 calculate_first_density_matrix 1 7.0 0.000 0.000 2.094 2.094 pw_transfer 1439 11.6 0.066 0.069 2.033 2.042 density_rs2pw 119 9.7 0.004 0.004 1.824 1.970 make_m2s 4572 13.5 0.034 0.035 1.795 1.953 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.939 1.950 grid_integrate_task_list 119 12.3 1.868 1.943 1.868 1.943 mp_waitall_1 115863 16.7 1.366 1.871 1.366 1.871 make_images 4572 14.5 0.269 0.299 1.707 1.863 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.820 1.824 rs_pw_transfer 974 11.9 0.008 0.008 1.645 1.790 calculate_dm_sparse 119 9.5 0.000 0.000 1.730 1.748 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.720 1.740 prepare_preconditioner 11 7.9 0.000 0.000 1.683 1.687 make_preconditioner 11 8.9 0.000 0.000 1.683 1.687 fft_wrap_pw1pw2_140 487 13.2 0.367 0.375 1.603 1.615 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.575 1.600 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.578 1.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.567 1.568 acc_transpose_blocks 9144 15.5 0.043 0.045 1.445 1.507 grid_collocate_task_list 119 9.7 1.296 1.417 1.296 1.417 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.401 1.409 potential_pw2rs 119 12.3 0.010 0.011 1.361 1.363 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.328 1.340 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.287 1.295 cp_fm_diag_elpa_base 50 14.0 1.259 1.277 1.285 1.293 ot_diis_step 108 11.5 0.013 0.013 1.283 1.283 fft3d_ps 1201 14.6 0.559 0.569 1.269 1.277 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.227 1.259 apply_single 119 13.6 0.000 0.000 1.227 1.259 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.238 1.239 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.121 1.125 wfi_extrapolate 11 7.9 0.001 0.001 1.090 1.090 jit_kernel_multiply 6 15.5 0.619 1.048 0.619 1.048 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.789 1.005 make_images_data 4572 15.5 0.042 0.046 0.786 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.928 cp_fm_cholesky_invert 11 10.9 0.843 0.845 0.843 0.845 mp_alltoall_d11v 2130 13.8 0.765 0.843 0.765 0.843 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.814 0.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.786 acc_transpose_blocks_sync 27432 16.5 0.717 0.762 0.717 0.762 qs_env_update_s_mstruct 11 6.9 0.003 0.005 0.685 0.737 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.669 0.688 mp_allgather_i34 2286 14.5 0.235 0.672 0.235 0.672 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.351 0.670 mp_alltoall_z22v 1201 16.6 0.585 0.611 0.585 0.611 mp_waitany 5200 13.7 0.451 0.583 0.451 0.583 qs_create_task_list 11 7.9 0.000 0.001 0.547 0.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.722000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.454545, yerr=6.050750 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 768.344064E+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 0.015 0.034 43.606 43.608 qs_mol_dyn_low 1 2.0 0.003 0.004 43.367 43.383 qs_forces 11 3.9 0.002 0.003 43.307 43.308 qs_energies 11 4.9 0.001 0.002 41.270 41.273 scf_env_do_scf 11 5.9 0.001 0.001 35.384 35.384 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.511 27.512 velocity_verlet 10 3.0 0.002 0.002 24.573 24.578 dbcsr_multiply_generic 2286 12.5 0.100 0.102 19.207 19.446 qs_scf_new_mos 108 7.5 0.001 0.001 17.586 17.684 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.585 17.683 ot_scf_mini 108 9.5 0.002 0.002 16.419 16.518 multiply_cannon 2286 13.5 0.297 0.301 15.204 16.109 multiply_cannon_loop 2286 14.5 0.883 0.906 13.907 14.851 ot_mini 108 10.5 0.001 0.001 10.032 10.145 multiply_cannon_multrec 9144 15.5 3.422 4.728 8.871 8.971 qs_ot_get_derivative 108 11.5 0.001 0.001 7.933 8.033 init_scf_loop 11 6.9 0.000 0.000 7.846 7.850 rebuild_ks_matrix 119 8.3 0.000 0.000 7.427 7.564 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.426 7.563 prepare_preconditioner 11 7.9 0.000 0.000 6.853 6.865 make_preconditioner 11 8.9 0.000 0.000 6.853 6.865 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.672 6.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.438 6.744 dbcsr_mm_accdrv_process 12550 15.8 4.689 6.522 5.312 6.560 cp_fm_upper_to_full 72 14.2 3.169 4.528 3.169 4.528 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.355 4.362 calculate_rho_elec 119 8.7 0.118 0.121 4.355 4.362 sum_up_and_integrate 119 10.3 0.064 0.066 4.086 4.092 integrate_v_rspace 119 11.3 0.003 0.004 4.021 4.028 init_scf_run 11 5.9 0.000 0.001 3.714 3.714 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.714 3.714 qs_ot_get_p 119 10.4 0.001 0.001 3.433 3.564 mp_waitall_1 94719 16.7 2.352 3.301 2.352 3.301 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.857 3.254 dbcsr_complete_redistribute 329 12.2 0.285 0.292 1.965 2.784 pw_transfer 1439 11.6 0.069 0.069 2.703 2.708 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.604 2.610 make_m2s 4572 13.5 0.037 0.038 2.403 2.582 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.241 2.555 apply_single 119 13.6 0.000 0.000 2.241 2.554 make_images 4572 14.5 0.353 0.388 2.283 2.462 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.650 2.452 calculate_dm_sparse 119 9.5 0.000 0.000 2.242 2.261 density_rs2pw 119 9.7 0.004 0.004 2.220 2.254 calculate_first_density_matrix 1 7.0 0.000 0.000 2.208 2.209 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.410 2.206 fft_wrap_pw1pw2_140 487 13.2 0.621 0.624 2.192 2.198 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.337 2.191 mp_alltoall_i22 627 13.8 1.386 2.184 1.386 2.184 grid_integrate_task_list 119 12.3 2.103 2.152 2.103 2.152 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.091 2.142 ot_diis_step 108 11.5 0.014 0.015 2.074 2.074 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.044 2.045 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.888 1.889 acc_transpose_blocks 9144 15.5 0.046 0.047 1.849 1.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.825 1.826 mp_sum_l 7287 12.8 1.051 1.744 1.051 1.744 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.682 fft3d_ps 1201 14.6 0.594 0.606 1.615 1.619 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.575 1.618 grid_collocate_task_list 119 9.7 1.538 1.567 1.538 1.567 rs_pw_transfer 974 11.9 0.009 0.009 1.513 1.554 cp_fm_cholesky_invert 11 10.9 1.471 1.474 1.471 1.474 wfi_extrapolate 11 7.9 0.001 0.001 1.439 1.439 potential_pw2rs 119 12.3 0.014 0.014 1.407 1.409 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.394 1.394 cp_fm_diag_elpa_base 50 14.0 1.248 1.300 1.391 1.391 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.090 1.352 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.322 1.342 make_images_data 4572 15.5 0.046 0.048 1.043 1.274 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.250 qs_env_update_s_mstruct 11 6.9 0.002 0.004 1.210 1.232 acc_transpose_blocks_sync 27432 16.5 1.137 1.163 1.137 1.163 mp_alltoall_d11v 2130 13.8 1.121 1.139 1.121 1.139 qs_create_task_list 11 7.9 0.001 0.001 1.041 1.053 generate_qs_task_list 11 8.9 0.392 0.496 1.041 1.052 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.964 0.977 mp_alltoall_z22v 1201 16.6 0.885 0.905 0.885 0.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.608000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=724.818182, yerr=8.386333 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 500.129792E+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 1350687. 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 0.011 0.032 83.483 83.485 qs_mol_dyn_low 1 2.0 0.003 0.004 83.229 83.240 qs_forces 11 3.9 0.003 0.005 83.159 83.159 qs_energies 11 4.9 0.001 0.002 80.269 80.284 scf_env_do_scf 11 5.9 0.000 0.001 71.223 71.225 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 65.713 65.713 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.922 52.218 qs_scf_new_mos 99 7.5 0.000 0.001 48.239 48.441 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.239 48.440 ot_scf_mini 99 9.5 0.002 0.002 45.797 45.933 multiply_cannon 2055 13.4 0.181 0.187 43.290 44.170 velocity_verlet 10 3.0 0.002 0.002 44.045 44.045 multiply_cannon_loop 2055 14.4 1.813 1.855 42.320 43.242 ot_mini 99 10.5 0.001 0.001 27.297 27.439 qs_ot_get_derivative 99 11.5 0.001 0.001 20.421 20.562 multiply_cannon_multrec 49320 15.4 11.340 12.227 17.441 18.242 rebuild_ks_matrix 110 8.3 0.000 0.000 14.635 14.811 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.031 14.634 14.811 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.808 12.985 mp_waitall_1 220248 16.4 11.353 12.061 11.353 12.061 multiply_cannon_sync_h2d 49320 15.4 9.574 10.186 9.574 10.186 qs_ot_get_p 110 10.4 0.001 0.001 9.855 9.998 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.892 8.434 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.802 8.065 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.232 7.745 apply_single 110 13.6 0.000 0.001 7.232 7.744 sum_up_and_integrate 110 10.3 0.036 0.043 7.121 7.134 integrate_v_rspace 110 11.3 0.003 0.004 7.084 7.107 init_scf_run 11 5.9 0.000 0.001 6.988 6.988 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.988 6.988 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.696 6.889 calculate_rho_elec 110 8.6 0.021 0.026 6.695 6.888 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.682 6.715 ot_diis_step 99 11.5 0.006 0.006 6.609 6.609 dbcsr_mm_accdrv_process 87628 16.1 3.044 3.155 5.971 6.241 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.825 5.826 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.419 5.493 init_scf_loop 11 6.9 0.000 0.000 5.482 5.482 cp_fm_diag_elpa 48 13.0 0.001 0.001 5.252 5.276 cp_fm_diag_elpa_base 48 14.0 5.237 5.263 5.249 5.273 mp_sum_l 6594 12.7 4.304 5.178 4.304 5.178 rs_pw_transfer 902 11.9 0.011 0.013 3.718 4.208 multiply_cannon_metrocomm1 49320 15.4 0.067 0.069 2.732 4.083 wfi_extrapolate 11 7.9 0.001 0.001 4.057 4.058 make_m2s 4110 13.4 0.061 0.066 3.872 3.994 density_rs2pw 110 9.6 0.004 0.005 3.396 3.927 calculate_dm_sparse 110 9.5 0.001 0.001 3.807 3.921 make_images 4110 14.4 0.176 0.189 3.776 3.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.757 3.761 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.403 3.456 grid_integrate_task_list 110 12.3 3.256 3.399 3.256 3.399 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.350 3.391 prepare_preconditioner 11 7.9 0.000 0.000 3.285 3.309 make_preconditioner 11 8.9 0.000 0.000 3.285 3.308 pw_transfer 1331 11.6 0.055 0.067 3.112 3.197 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.083 3.134 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.024 3.111 acc_transpose_blocks 49320 15.4 0.219 0.229 2.625 2.841 calculate_first_density_matrix 1 7.0 0.000 0.001 2.828 2.832 mp_alltoall_d11v 2046 13.8 2.075 2.705 2.075 2.705 fft_wrap_pw1pw2_140 451 13.1 0.452 0.496 2.587 2.678 jit_kernel_multiply 13 15.9 2.649 2.664 2.649 2.664 potential_pw2rs 110 12.3 0.006 0.007 2.635 2.658 mp_waitany 14300 13.8 1.835 2.474 1.835 2.474 grid_collocate_task_list 110 9.6 2.156 2.356 2.156 2.356 fft3d_ps 1111 14.6 0.791 0.877 2.216 2.278 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.955 1.981 mp_sum_d 3889 11.9 1.394 1.945 1.394 1.945 make_images_data 4110 15.4 0.044 0.048 1.746 1.879 cp_fm_cholesky_invert 11 10.9 1.813 1.817 1.813 1.817 hybrid_alltoall_any 4261 16.3 0.085 0.487 1.519 1.794 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.647 1.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.485000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.090909, yerr=2.466302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 593.702912E+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 3632150. 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 0.021 0.035 69.810 69.812 qs_mol_dyn_low 1 2.0 0.003 0.005 69.524 69.535 qs_forces 11 3.9 0.003 0.003 69.459 69.460 qs_energies 11 4.9 0.003 0.007 66.057 66.060 scf_env_do_scf 11 5.9 0.001 0.001 57.457 57.461 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.797 49.799 dbcsr_multiply_generic 2055 12.4 0.116 0.132 37.963 38.100 velocity_verlet 10 3.0 0.002 0.002 36.458 36.459 qs_scf_new_mos 99 7.5 0.001 0.001 33.367 33.489 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.366 33.488 multiply_cannon 2055 13.4 0.224 0.245 31.350 32.290 ot_scf_mini 99 9.5 0.003 0.004 31.693 31.811 multiply_cannon_loop 2055 14.4 1.162 1.196 30.177 31.110 ot_mini 99 10.5 0.001 0.001 18.435 18.550 multiply_cannon_multrec 24660 15.4 6.995 9.198 14.062 16.186 rebuild_ks_matrix 110 8.3 0.000 0.000 13.883 14.017 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.883 14.017 qs_ot_get_derivative 99 11.5 0.001 0.001 12.655 12.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.191 12.310 mp_waitall_1 176588 16.5 7.553 10.189 7.553 10.189 multiply_cannon_metrocomm3 24660 15.4 0.073 0.077 5.182 8.087 init_scf_loop 11 6.9 0.006 0.023 7.624 7.624 multiply_cannon_sync_h2d 24660 15.4 6.450 7.605 6.450 7.605 dbcsr_mm_accdrv_process 52282 16.1 5.538 6.198 6.896 7.210 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.449 7.066 apply_single 110 13.6 0.000 0.001 6.449 7.065 qs_ot_get_p 110 10.4 0.001 0.001 6.621 6.763 sum_up_and_integrate 110 10.3 0.052 0.059 6.684 6.698 integrate_v_rspace 110 11.3 0.002 0.003 6.632 6.647 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.277 6.287 calculate_rho_elec 110 8.6 0.040 0.047 6.277 6.287 init_scf_run 11 5.9 0.000 0.001 6.173 6.173 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.173 6.173 ot_diis_step 99 11.5 0.010 0.010 5.731 5.731 prepare_preconditioner 11 7.9 0.000 0.000 5.539 5.562 make_preconditioner 11 8.9 0.000 0.001 5.539 5.561 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.738 5.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.108 5.259 qs_ot_p2m_diag 48 11.0 0.029 0.045 4.687 4.705 make_m2s 4110 13.4 0.057 0.060 4.232 4.584 make_images 4110 14.4 0.411 0.471 4.123 4.471 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.220 4.220 pw_transfer 1331 11.6 0.067 0.073 3.715 3.845 density_rs2pw 110 9.6 0.004 0.005 3.351 3.835 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.607 3.739 cp_fm_diag_elpa 48 13.0 0.001 0.001 3.634 3.645 cp_fm_diag_elpa_base 48 14.0 3.586 3.602 3.631 3.641 rs_pw_transfer 902 11.9 0.012 0.014 3.018 3.554 wfi_extrapolate 11 7.9 0.001 0.001 3.535 3.536 grid_integrate_task_list 110 12.3 3.158 3.388 3.158 3.388 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.307 3.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.252 3.254 fft_wrap_pw1pw2_140 451 13.1 0.526 0.546 3.056 3.192 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.000 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.922 2.959 fft3d_ps 1111 14.6 1.106 1.330 2.573 2.720 make_images_data 4110 15.4 0.049 0.053 2.278 2.654 hybrid_alltoall_any 4261 16.3 0.107 0.459 1.953 2.597 cp_fm_cholesky_invert 11 10.9 2.558 2.565 2.558 2.565 grid_collocate_task_list 110 9.6 2.101 2.534 2.101 2.534 calculate_first_density_matrix 1 7.0 0.005 0.020 2.517 2.519 potential_pw2rs 110 12.3 0.008 0.008 2.451 2.470 mp_sum_l 6594 12.7 1.758 2.392 1.758 2.392 mp_alltoall_d11v 2046 13.8 1.781 2.062 1.781 2.062 acc_transpose_blocks 24660 15.4 0.120 0.124 1.985 2.020 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.988 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.913 1.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.861 1.876 mp_waitany 10164 13.8 1.254 1.824 1.254 1.824 jit_kernel_multiply 9 16.3 0.993 1.725 0.993 1.725 multiply_cannon_metrocomm4 22605 15.4 0.080 0.084 0.749 1.704 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.535 1.544 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.219 1.001 1.523 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.386 1.487 mp_irecv_dv 57340 16.2 0.620 1.482 0.620 1.482 mp_allgather_i34 2055 14.4 0.482 1.463 0.482 1.463 dbcsr_complete_redistribute 325 12.2 0.242 0.305 1.161 1.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.812000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.545455, yerr=6.985226 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 668.098560E+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 0.022 0.035 60.448 60.450 qs_mol_dyn_low 1 2.0 0.010 0.056 60.104 60.114 qs_forces 11 3.9 0.011 0.052 59.973 59.974 qs_energies 11 4.9 0.005 0.020 56.532 56.558 scf_env_do_scf 11 5.9 0.001 0.001 48.566 48.566 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.130 40.130 velocity_verlet 10 3.0 0.003 0.017 32.684 32.686 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.671 28.937 qs_scf_new_mos 99 7.5 0.001 0.001 25.033 25.123 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.033 25.123 ot_scf_mini 99 9.5 0.003 0.003 23.788 23.899 multiply_cannon 2055 13.4 0.214 0.223 22.559 23.741 multiply_cannon_loop 2055 14.4 0.810 0.833 21.382 22.489 ot_mini 99 10.5 0.001 0.002 13.685 13.800 rebuild_ks_matrix 110 8.3 0.000 0.000 12.544 12.697 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.544 12.697 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.927 11.061 multiply_cannon_multrec 16440 15.4 3.560 4.632 9.720 10.763 mp_waitall_1 139946 16.5 7.041 9.678 7.041 9.678 qs_ot_get_derivative 99 11.5 0.001 0.002 9.174 9.285 init_scf_loop 11 6.9 0.002 0.005 8.403 8.405 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 4.540 7.208 sum_up_and_integrate 110 10.3 0.060 0.060 6.629 6.665 prepare_preconditioner 11 7.9 0.000 0.000 6.610 6.627 make_preconditioner 11 8.9 0.000 0.001 6.610 6.627 integrate_v_rspace 110 11.3 0.003 0.004 6.569 6.606 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.933 6.282 dbcsr_mm_accdrv_process 34862 16.1 5.322 5.731 6.005 6.144 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.007 6.014 calculate_rho_elec 110 8.6 0.059 0.060 6.006 6.013 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.965 5.507 apply_single 110 13.6 0.000 0.000 4.965 5.506 init_scf_run 11 5.9 0.000 0.001 5.497 5.497 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.497 5.497 qs_ot_get_p 110 10.4 0.001 0.001 5.283 5.424 make_m2s 4110 13.4 0.050 0.052 4.119 4.530 ot_diis_step 99 11.5 0.011 0.012 4.481 4.481 make_images 4110 14.4 0.400 0.519 4.004 4.413 density_rs2pw 110 9.6 0.004 0.005 3.084 4.372 rs_pw_transfer 902 11.9 0.010 0.011 2.659 3.914 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.193 3.802 pw_transfer 1331 11.6 0.066 0.073 3.714 3.747 multiply_cannon_sync_h2d 16440 15.4 3.226 3.710 3.226 3.710 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.639 3.644 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.607 3.643 grid_integrate_task_list 110 12.3 3.184 3.427 3.184 3.427 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.273 3.274 fft_wrap_pw1pw2_140 451 13.1 0.644 0.653 3.143 3.179 wfi_extrapolate 11 7.9 0.001 0.001 2.960 2.960 make_images_data 4110 15.4 0.046 0.049 2.332 2.840 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.761 2.762 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.751 2.759 cp_fm_diag_elpa_base 48 14.0 2.685 2.716 2.749 2.757 hybrid_alltoall_any 4261 16.3 0.110 0.383 2.062 2.740 cp_fm_cholesky_invert 11 10.9 2.555 2.560 2.555 2.560 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.548 mp_waitany 17072 13.8 1.220 2.536 1.220 2.536 multiply_cannon_metrocomm4 14385 15.4 0.047 0.052 0.915 2.534 grid_collocate_task_list 110 9.6 2.118 2.530 2.118 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.438 2.487 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.399 2.460 fft3d_ps 1111 14.6 1.089 1.102 2.421 2.453 calculate_first_density_matrix 1 7.0 0.001 0.002 2.442 2.444 mp_irecv_dv 48980 15.7 0.841 2.399 0.841 2.399 potential_pw2rs 110 12.3 0.011 0.011 2.346 2.376 mp_alltoall_d11v 2046 13.8 1.732 2.231 1.732 2.231 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.188 0.929 2.179 mp_sum_l 6594 12.7 1.372 2.062 1.372 2.062 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.992 1.994 dbcsr_complete_redistribute 325 12.2 0.332 0.351 1.445 1.897 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.811 1.828 cp_fm_upper_to_full 70 14.2 1.373 1.747 1.373 1.747 acc_transpose_blocks 16440 15.4 0.081 0.084 1.575 1.644 cp_fm_cholesky_decompose 22 10.9 1.539 1.560 1.539 1.560 mp_allgather_i34 2055 14.4 0.488 1.543 0.488 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.379 1.511 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.475 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.972 1.420 rs_gather_matrices 110 12.3 0.233 0.261 0.962 1.399 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.271 1.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.450000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.181818, yerr=8.778260 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 735.653888E+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 0.017 0.029 66.740 66.742 qs_mol_dyn_low 1 2.0 0.003 0.003 66.420 66.430 qs_forces 11 3.9 0.016 0.110 66.355 66.357 qs_energies 11 4.9 0.002 0.010 62.786 62.791 scf_env_do_scf 11 5.9 0.001 0.001 54.131 54.134 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.555 42.556 velocity_verlet 10 3.0 0.002 0.002 37.634 37.637 dbcsr_multiply_generic 2055 12.4 0.113 0.119 30.778 31.007 qs_scf_new_mos 99 7.5 0.001 0.001 27.403 27.500 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.402 27.500 ot_scf_mini 99 9.5 0.003 0.003 25.716 25.820 multiply_cannon 2055 13.4 0.241 0.262 23.737 24.703 multiply_cannon_loop 2055 14.4 1.420 1.480 22.182 22.821 ot_mini 99 10.5 0.001 0.001 14.835 14.964 multiply_cannon_multrec 24660 15.4 4.104 6.746 13.046 14.198 rebuild_ks_matrix 110 8.3 0.000 0.000 12.292 12.418 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.292 12.418 init_scf_loop 11 6.9 0.001 0.003 11.537 11.538 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.820 10.933 qs_ot_get_derivative 99 11.5 0.001 0.001 10.635 10.741 dbcsr_mm_accdrv_process 52304 16.0 7.703 9.384 8.784 9.809 prepare_preconditioner 11 7.9 0.000 0.000 9.771 9.787 make_preconditioner 11 8.9 0.000 0.001 9.771 9.787 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.969 9.439 sum_up_and_integrate 110 10.3 0.068 0.071 6.538 6.552 integrate_v_rspace 110 11.3 0.003 0.004 6.470 6.483 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.181 6.189 calculate_rho_elec 110 8.6 0.078 0.081 6.181 6.188 mp_waitall_1 121746 16.5 4.108 6.017 4.108 6.017 qs_ot_get_p 110 10.4 0.001 0.001 5.738 5.878 make_m2s 4110 13.4 0.059 0.061 5.381 5.875 make_images 4110 14.4 0.579 0.699 5.240 5.730 init_scf_run 11 5.9 0.000 0.001 5.655 5.655 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.654 5.655 cp_fm_upper_to_full 70 14.2 3.338 4.851 3.338 4.851 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.070 4.177 apply_single 110 13.6 0.000 0.000 4.070 4.177 ot_diis_step 99 11.5 0.011 0.011 4.161 4.161 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.870 3.886 dbcsr_complete_redistribute 325 12.2 0.425 0.474 2.692 3.838 pw_transfer 1331 11.6 0.066 0.076 3.757 3.788 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.650 3.685 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.544 3.591 density_rs2pw 110 9.6 0.004 0.004 3.043 3.568 grid_integrate_task_list 110 12.3 3.284 3.457 3.284 3.457 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.354 3.415 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.168 3.301 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.300 3.300 fft_wrap_pw1pw2_140 451 13.1 0.672 0.693 3.134 3.170 make_images_data 4110 15.4 0.048 0.053 2.654 3.153 calculate_dm_sparse 110 9.5 0.001 0.001 3.043 3.085 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.286 3.084 wfi_extrapolate 11 7.9 0.001 0.001 3.022 3.023 rs_pw_transfer 902 11.9 0.010 0.011 2.442 2.985 multiply_cannon_metrocomm3 24660 15.4 0.039 0.040 1.333 2.974 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.793 2.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.841 2.843 mp_alltoall_i22 605 13.7 1.669 2.829 1.669 2.829 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.773 2.783 cp_fm_diag_elpa_base 48 14.0 2.626 2.685 2.770 2.780 cp_fm_cholesky_invert 11 10.9 2.599 2.610 2.599 2.610 multiply_cannon_sync_h2d 24660 15.4 2.400 2.568 2.400 2.568 acc_transpose_blocks 24660 15.4 0.113 0.115 2.451 2.563 calculate_first_density_matrix 1 7.0 0.002 0.012 2.531 2.535 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.472 2.511 grid_collocate_task_list 110 9.6 2.222 2.497 2.222 2.497 fft3d_ps 1111 14.6 1.084 1.116 2.406 2.423 qs_energies_init_hamiltonians 11 5.9 0.005 0.028 2.344 2.346 potential_pw2rs 110 12.3 0.012 0.013 2.163 2.173 mp_alltoall_d11v 2046 13.8 1.772 1.999 1.772 1.999 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.796 1.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.713 1.727 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.013 1.614 1.703 cp_fm_cholesky_decompose 22 10.9 1.626 1.679 1.626 1.679 mp_waitany 13376 13.8 1.119 1.636 1.119 1.636 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.008 0.029 1.603 1.618 mp_sum_l 6594 12.7 0.979 1.530 0.979 1.530 acc_transpose_blocks_sync 73980 16.4 1.418 1.523 1.418 1.523 multiply_cannon_metrocomm4 20550 15.4 0.061 0.064 0.840 1.487 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.174 0.884 1.436 mp_irecv_dv 62702 16.1 0.737 1.406 0.737 1.406 mp_allgather_i34 2055 14.4 0.642 1.400 0.642 1.400 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.265 1.339 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.742000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.363636, yerr=10.011564 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 850.026496E+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 0.031 55.529 55.531 qs_mol_dyn_low 1 2.0 0.003 0.003 55.302 55.311 qs_forces 11 3.9 0.003 0.003 55.226 55.227 qs_energies 11 4.9 0.001 0.001 51.518 51.524 scf_env_do_scf 11 5.9 0.000 0.001 43.297 43.297 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.644 35.645 velocity_verlet 10 3.0 0.006 0.006 31.503 31.507 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.092 23.233 qs_scf_new_mos 99 7.5 0.001 0.001 20.520 20.569 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.520 20.569 ot_scf_mini 99 9.5 0.002 0.002 19.269 19.287 multiply_cannon 2055 13.4 0.236 0.246 17.541 18.952 multiply_cannon_loop 2055 14.4 0.601 0.623 16.273 16.610 rebuild_ks_matrix 110 8.3 0.000 0.000 11.945 11.966 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.944 11.965 ot_mini 99 10.5 0.001 0.001 10.664 10.675 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.586 10.605 multiply_cannon_multrec 8220 15.4 3.173 4.418 7.557 8.445 init_scf_loop 11 6.9 0.000 0.000 7.606 7.608 mp_waitall_1 103326 16.6 5.785 7.255 5.785 7.255 qs_ot_get_derivative 99 11.5 0.001 0.001 6.975 6.992 sum_up_and_integrate 110 10.3 0.079 0.080 6.548 6.560 integrate_v_rspace 110 11.3 0.003 0.003 6.468 6.482 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.359 6.378 calculate_rho_elec 110 8.6 0.114 0.115 6.358 6.377 prepare_preconditioner 11 7.9 0.000 0.000 5.940 5.949 make_preconditioner 11 8.9 0.000 0.000 5.940 5.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.510 5.587 dbcsr_mm_accdrv_process 17442 15.9 3.077 4.112 4.244 5.165 init_scf_run 11 5.9 0.000 0.001 5.069 5.069 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.069 5.069 qs_ot_get_p 110 10.4 0.001 0.001 4.658 4.672 make_m2s 4110 13.4 0.038 0.039 4.230 4.520 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.943 4.392 make_images 4110 14.4 0.650 0.710 4.101 4.391 pw_transfer 1331 11.6 0.066 0.071 3.945 3.956 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.838 3.853 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.730 3.785 apply_single 110 13.6 0.000 0.000 3.729 3.784 ot_diis_step 99 11.5 0.012 0.012 3.668 3.668 grid_integrate_task_list 110 12.3 3.364 3.548 3.364 3.548 density_rs2pw 110 9.6 0.004 0.004 3.038 3.412 fft_wrap_pw1pw2_140 451 13.1 0.839 0.850 3.320 3.344 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.300 3.305 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.962 2.962 hybrid_alltoall_any 4261 16.3 0.199 0.847 2.299 2.833 cp_fm_cholesky_invert 11 10.9 2.815 2.819 2.815 2.819 wfi_extrapolate 11 7.9 0.001 0.001 2.737 2.737 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.694 2.694 make_images_data 4110 15.4 0.041 0.046 2.266 2.667 rs_pw_transfer 902 11.9 0.010 0.010 2.191 2.593 grid_collocate_task_list 110 9.6 2.314 2.571 2.314 2.571 calculate_dm_sparse 110 9.5 0.001 0.001 2.473 2.509 multiply_cannon_sync_h2d 8220 15.4 2.379 2.501 2.379 2.501 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.486 2.487 cp_fm_diag_elpa 48 13.0 0.016 0.017 2.399 2.406 fft3d_ps 1111 14.6 1.141 1.176 2.375 2.393 cp_fm_diag_elpa_base 48 14.0 2.327 2.348 2.381 2.388 calculate_first_density_matrix 1 7.0 0.000 0.000 2.230 2.231 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.125 2.140 potential_pw2rs 110 12.3 0.015 0.015 2.081 2.084 mp_alltoall_d11v 2046 13.8 1.760 2.046 1.760 2.046 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.011 2.034 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.987 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.914 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.686 1.689 cp_fm_cholesky_decompose 22 10.9 1.658 1.677 1.658 1.677 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.514 1.640 mp_allgather_i34 2055 14.4 0.439 1.535 0.439 1.535 dbcsr_complete_redistribute 325 12.2 0.568 0.592 1.440 1.529 mp_waitany 9240 13.8 1.076 1.500 1.076 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.427 1.440 acc_transpose_blocks 8220 15.4 0.040 0.041 1.357 1.389 qs_create_task_list 11 7.9 0.000 0.001 1.234 1.341 generate_qs_task_list 11 8.9 0.389 0.457 1.233 1.340 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.752 1.282 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.212 1.231 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.166 0.802 1.205 rs_gather_matrices 110 12.3 0.323 0.362 0.933 1.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.531000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=802.727273, yerr=10.871906 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 1.410363E+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 0.017 0.037 88.190 88.191 qs_mol_dyn_low 1 2.0 0.003 0.008 87.884 87.893 qs_forces 11 3.9 0.018 0.019 87.820 87.822 qs_energies 11 4.9 0.001 0.001 83.574 83.576 scf_env_do_scf 11 5.9 0.001 0.001 73.434 73.434 velocity_verlet 10 3.0 0.002 0.002 56.098 56.103 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.351 45.352 dbcsr_multiply_generic 2055 12.4 0.127 0.133 29.604 29.732 init_scf_loop 11 6.9 0.000 0.000 28.011 28.013 qs_scf_new_mos 99 7.5 0.001 0.001 26.899 26.956 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.899 26.956 prepare_preconditioner 11 7.9 0.000 0.000 25.958 25.967 make_preconditioner 11 8.9 0.000 0.000 25.958 25.967 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.303 25.427 ot_scf_mini 99 9.5 0.002 0.002 25.093 25.144 multiply_cannon 2055 13.4 0.333 0.355 22.568 23.159 multiply_cannon_loop 2055 14.4 0.842 0.873 20.800 21.112 cp_fm_upper_to_full 70 14.2 12.653 18.019 12.653 18.019 rebuild_ks_matrix 110 8.3 0.000 0.000 14.273 14.331 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.273 14.330 ot_mini 99 10.5 0.001 0.001 14.143 14.193 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.902 12.954 dbcsr_complete_redistribute 325 12.2 1.017 1.040 7.192 10.276 multiply_cannon_multrec 8220 15.4 4.051 4.199 9.726 9.885 qs_ot_get_derivative 99 11.5 0.001 0.001 9.504 9.552 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.194 9.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.638 8.692 mp_alltoall_i22 605 13.7 5.281 8.411 5.281 8.411 mp_waitall_1 84994 16.7 7.535 8.394 7.535 8.394 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.811 7.849 calculate_rho_elec 110 8.6 0.226 0.226 7.811 7.848 sum_up_and_integrate 110 10.3 0.150 0.152 7.429 7.446 integrate_v_rspace 110 11.3 0.004 0.004 7.279 7.295 init_scf_run 11 5.9 0.000 0.001 5.829 5.829 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.829 5.829 dbcsr_mm_accdrv_process 11614 15.7 4.104 4.460 5.528 5.770 make_m2s 4110 13.4 0.043 0.044 5.264 5.735 qs_ot_get_p 110 10.4 0.001 0.001 5.649 5.708 make_images 4110 14.4 0.887 0.945 5.076 5.548 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.950 5.253 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.759 5.223 apply_single 110 13.6 0.000 0.000 4.759 5.223 cp_fm_cholesky_invert 11 10.9 5.158 5.162 5.158 5.162 pw_transfer 1331 11.6 0.076 0.076 5.011 5.015 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.893 4.897 ot_diis_step 99 11.5 0.015 0.016 4.612 4.612 fft_wrap_pw1pw2_140 451 13.1 1.341 1.346 4.269 4.276 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.055 4.061 grid_integrate_task_list 110 12.3 3.757 3.870 3.757 3.870 density_rs2pw 110 9.6 0.004 0.004 3.747 3.826 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.704 3.707 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.528 3.529 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.020 3.477 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.796 3.420 wfi_extrapolate 11 7.9 0.001 0.001 3.360 3.360 make_images_data 4110 15.4 0.045 0.047 2.705 3.315 calculate_dm_sparse 110 9.5 0.001 0.001 3.246 3.280 multiply_cannon_sync_h2d 8220 15.4 3.092 3.147 3.092 3.147 cp_fm_diag_elpa 48 13.0 0.035 0.035 2.939 2.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.908 2.911 cp_fm_diag_elpa_base 48 14.0 2.368 2.562 2.902 2.902 fft3d_ps 1111 14.6 1.300 1.308 2.814 2.822 grid_collocate_task_list 110 9.6 2.734 2.790 2.734 2.790 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.670 2.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.401 2.429 rs_pw_transfer 902 11.9 0.010 0.011 2.309 2.411 potential_pw2rs 110 12.3 0.021 0.021 2.400 2.402 calculate_first_density_matrix 1 7.0 0.000 0.000 2.347 2.348 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.207 2.261 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.203 mp_alltoall_d11v 2046 13.8 1.967 2.104 1.967 2.104 cp_fm_cholesky_decompose 22 10.9 1.986 2.003 1.986 2.003 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.968 1.974 qs_create_task_list 11 7.9 0.000 0.001 1.913 1.963 generate_qs_task_list 11 8.9 0.756 0.810 1.912 1.963 acc_transpose_blocks 8220 15.4 0.042 0.042 1.820 1.886 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.746 1.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.191000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1262.363636, yerr=50.056827 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.281344E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2725907. 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 0.011 0.031 207.818 207.819 qs_mol_dyn_low 1 2.0 0.003 0.004 207.377 207.390 qs_forces 11 3.9 0.008 0.009 207.295 207.296 qs_energies 11 4.9 0.001 0.002 201.640 201.658 scf_env_do_scf 11 5.9 0.001 0.001 185.221 185.225 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.708 164.710 dbcsr_multiply_generic 2507 12.6 0.179 0.184 125.867 126.375 qs_scf_new_mos 117 7.6 0.001 0.001 124.622 124.878 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.621 124.877 velocity_verlet 10 3.0 0.001 0.002 123.981 123.982 ot_scf_mini 117 9.6 0.003 0.003 117.924 118.187 multiply_cannon 2507 13.6 0.238 0.245 101.647 103.315 multiply_cannon_loop 2507 14.6 2.420 2.517 99.437 100.974 ot_mini 117 10.6 0.001 0.001 66.955 67.200 multiply_cannon_multrec 60168 15.6 31.833 33.544 41.556 43.038 qs_ot_get_derivative 117 11.6 0.001 0.001 42.113 42.352 rebuild_ks_matrix 128 8.3 0.001 0.001 33.734 34.031 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.733 34.030 mp_waitall_1 267128 16.5 28.963 31.965 28.963 31.965 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.267 30.575 qs_ot_get_p 128 10.4 0.001 0.001 28.858 29.151 multiply_cannon_sync_h2d 60168 15.6 26.194 28.162 26.194 28.162 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.341 25.036 apply_single 128 13.6 0.001 0.001 24.340 25.036 ot_diis_step 117 11.6 0.008 0.008 24.558 24.559 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.332 22.393 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.613 20.835 init_scf_loop 11 6.9 0.000 0.001 20.439 20.441 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.704 19.704 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 16.251 18.501 cp_fm_diag_elpa 83 13.4 0.001 0.001 16.772 16.824 cp_fm_diag_elpa_base 83 14.4 16.704 16.775 16.767 16.821 prepare_preconditioner 11 7.9 0.000 0.000 15.814 15.847 make_preconditioner 11 8.9 0.000 0.000 15.814 15.847 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.067 15.252 sum_up_and_integrate 128 10.3 0.089 0.108 14.139 14.154 make_m2s 5014 13.6 0.104 0.113 13.796 14.129 integrate_v_rspace 128 11.3 0.004 0.004 14.049 14.070 make_images 5014 14.6 0.398 0.418 13.615 13.959 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.766 13.880 calculate_rho_elec 128 8.7 0.046 0.065 13.766 13.879 init_scf_run 11 5.9 0.000 0.001 12.266 12.267 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.266 12.267 mp_sum_l 7950 12.9 9.161 10.441 9.161 10.441 density_rs2pw 128 9.7 0.006 0.007 7.009 10.180 dbcsr_mm_accdrv_process 124484 16.2 4.694 4.954 9.286 9.965 wfi_extrapolate 11 7.9 0.001 0.001 9.099 9.100 cp_fm_cholesky_invert 11 10.9 9.032 9.040 9.032 9.040 rs_pw_transfer 1046 11.9 0.017 0.019 5.619 8.815 calculate_dm_sparse 128 9.5 0.001 0.001 8.503 8.617 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.030 8.163 multiply_cannon_metrocomm1 60168 15.6 0.098 0.102 6.269 8.054 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.973 8.037 pw_transfer 1547 11.6 0.074 0.094 7.734 7.904 make_images_data 5014 15.6 0.069 0.074 6.799 7.753 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.531 7.700 grid_integrate_task_list 128 12.3 7.046 7.623 7.046 7.623 hybrid_alltoall_any 5200 16.5 0.295 2.269 5.956 7.359 fft_wrap_pw1pw2_140 523 13.2 1.274 1.318 6.616 6.790 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.579 6.589 grid_collocate_task_list 128 9.7 4.661 5.971 4.661 5.971 mp_waitany 16020 13.9 2.650 5.852 2.650 5.852 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.730 5.833 fft3d_ps 1291 14.7 2.140 2.777 5.296 5.610 mp_alltoall_d11v 2415 14.1 4.254 5.355 4.254 5.355 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.294 2.141 5.332 cp_fm_cholesky_decompose 22 10.9 4.603 4.616 4.603 4.616 potential_pw2rs 128 12.3 0.009 0.010 4.512 4.537 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.819000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.727273, yerr=5.753906 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 839.864320E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 5313859. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.053 189.444 189.445 qs_mol_dyn_low 1 2.0 0.003 0.003 188.935 188.948 qs_forces 11 3.9 0.007 0.012 188.847 188.848 qs_energies 11 4.9 0.014 0.035 181.983 181.994 scf_env_do_scf 11 5.9 0.001 0.001 165.463 165.473 scf_env_do_scf_inner_loop 116 6.6 0.011 0.036 132.652 132.654 velocity_verlet 10 3.0 0.002 0.002 118.604 118.606 dbcsr_multiply_generic 2485 12.5 0.187 0.197 96.638 97.929 qs_scf_new_mos 116 7.6 0.001 0.001 93.243 93.926 qs_scf_loop_do_ot 116 8.6 0.001 0.002 93.242 93.925 ot_scf_mini 116 9.6 0.004 0.005 88.442 89.224 multiply_cannon 2485 13.5 0.500 0.563 76.669 81.266 multiply_cannon_loop 2485 14.5 1.549 1.625 73.311 76.058 ot_mini 116 10.6 0.001 0.001 49.897 50.569 mp_waitall_1 212858 16.6 23.802 38.601 23.802 38.601 multiply_cannon_multrec 29820 15.5 20.784 25.292 31.282 36.046 rebuild_ks_matrix 127 8.3 0.001 0.001 32.845 33.576 qs_ks_build_kohn_sham_matrix 127 9.3 0.040 0.075 32.844 33.575 init_scf_loop 11 6.9 0.003 0.012 32.724 32.725 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.503 30.167 qs_ot_get_derivative 116 11.6 0.001 0.002 27.917 28.677 multiply_cannon_metrocomm3 29820 15.5 0.096 0.103 15.561 28.604 prepare_preconditioner 11 7.9 0.000 0.000 28.183 28.266 make_preconditioner 11 8.9 0.000 0.001 28.183 28.266 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.786 27.339 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.127 23.331 apply_single 127 13.6 0.001 0.001 22.127 23.331 ot_diis_step 116 11.6 0.014 0.015 21.800 21.801 qs_ot_get_p 127 10.4 0.001 0.001 20.744 21.511 multiply_cannon_sync_h2d 29820 15.5 17.858 20.399 17.858 20.399 cp_fm_cholesky_invert 11 10.9 16.369 16.382 16.369 16.382 qs_ot_p2m_diag 82 11.4 0.186 0.215 15.934 15.968 make_m2s 4970 13.5 0.087 0.092 14.050 15.728 make_images 4970 14.5 1.175 1.364 13.843 15.524 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.749 14.750 sum_up_and_integrate 127 10.3 0.116 0.132 14.513 14.540 integrate_v_rspace 127 11.3 0.004 0.006 14.397 14.429 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.062 14.103 calculate_rho_elec 127 8.7 0.088 0.112 14.061 14.102 cp_fm_diag_elpa 82 13.4 0.001 0.001 11.582 11.615 cp_fm_diag_elpa_base 82 14.4 11.333 11.415 11.576 11.604 init_scf_run 11 5.9 0.000 0.001 11.574 11.576 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.574 11.575 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.029 11.560 multiply_cannon_metrocomm4 27335 15.5 0.107 0.123 3.847 11.245 mp_irecv_dv 68888 16.3 3.640 10.832 3.640 10.832 dbcsr_mm_accdrv_process 61726 16.2 5.378 6.152 9.947 10.506 make_images_data 4970 15.5 0.068 0.077 8.291 10.320 hybrid_alltoall_any 5155 16.4 0.348 1.510 7.209 9.795 density_rs2pw 127 9.7 0.006 0.007 7.317 9.725 pw_transfer 1535 11.6 0.086 0.099 8.674 8.723 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.451 8.500 wfi_extrapolate 11 7.9 0.001 0.001 8.390 8.390 rs_pw_transfer 1038 11.9 0.014 0.016 5.616 8.034 fft_wrap_pw1pw2_140 519 13.2 1.331 1.350 7.491 7.559 grid_integrate_task_list 127 12.3 7.137 7.554 7.137 7.554 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.228 7.068 cp_fm_cholesky_decompose 22 10.9 6.819 6.914 6.819 6.914 calculate_dm_sparse 127 9.5 0.001 0.001 6.464 6.601 mp_sum_l 7884 12.9 4.131 6.127 4.131 6.127 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.113 6.122 grid_collocate_task_list 127 9.7 4.751 5.924 4.751 5.924 fft3d_ps 1281 14.7 2.789 2.980 5.771 5.830 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.361 5.438 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.209 5.397 mp_allgather_i34 2485 14.5 1.889 5.078 1.889 5.078 mp_alltoall_d11v 2401 14.1 4.429 4.978 4.429 4.978 mp_waitany 11660 13.9 2.501 4.903 2.501 4.903 potential_pw2rs 127 12.3 0.015 0.017 4.844 4.875 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.371 2.120 4.521 mp_sum_d 4454 12.1 2.720 4.088 2.720 4.088 dbcsr_complete_redistribute 393 12.7 0.774 0.866 3.198 4.060 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.749 3.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.445000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.363636, yerr=0.481046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 973.332480E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57812. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.224 176.375 176.376 qs_mol_dyn_low 1 2.0 0.003 0.004 175.713 175.741 qs_forces 11 3.9 0.004 0.004 175.591 175.592 qs_energies 11 4.9 0.005 0.021 168.903 168.913 scf_env_do_scf 11 5.9 0.001 0.002 153.240 153.241 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 118.826 118.827 velocity_verlet 10 3.0 0.002 0.002 113.245 113.248 dbcsr_multiply_generic 2527 12.6 0.182 0.191 82.347 83.588 qs_scf_new_mos 118 7.6 0.001 0.001 80.159 80.533 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.158 80.532 ot_scf_mini 118 9.6 0.004 0.004 75.902 76.329 multiply_cannon 2527 13.6 0.518 0.558 62.630 67.562 multiply_cannon_loop 2527 14.6 1.134 1.209 59.550 62.318 ot_mini 118 10.6 0.001 0.001 42.833 43.256 mp_waitall_1 171878 16.6 24.839 34.534 24.839 34.534 init_scf_loop 11 6.9 0.005 0.026 34.315 34.316 rebuild_ks_matrix 129 8.3 0.001 0.001 31.061 31.539 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 31.061 31.539 prepare_preconditioner 11 7.9 0.000 0.000 30.260 30.303 make_preconditioner 11 8.9 0.000 0.001 30.260 30.303 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.878 29.276 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.938 28.372 multiply_cannon_multrec 20216 15.6 13.124 16.573 22.506 25.919 multiply_cannon_metrocomm3 20216 15.6 0.063 0.067 15.527 24.881 qs_ot_get_derivative 118 11.6 0.002 0.002 22.782 23.208 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.155 21.186 apply_single 129 13.6 0.001 0.001 20.154 21.185 ot_diis_step 118 11.6 0.018 0.018 19.944 19.944 qs_ot_get_p 129 10.4 0.001 0.001 18.281 18.841 make_m2s 5054 13.6 0.079 0.083 14.739 16.089 make_images 5054 14.6 1.161 1.260 14.507 15.852 multiply_cannon_sync_h2d 20216 15.6 13.627 15.480 13.627 15.480 sum_up_and_integrate 129 10.3 0.135 0.146 14.688 14.722 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.550 14.581 calculate_rho_elec 129 8.7 0.131 0.146 14.549 14.580 integrate_v_rspace 129 11.3 0.004 0.004 14.553 14.579 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.095 14.104 cp_fm_cholesky_invert 11 10.9 14.043 14.052 14.043 14.052 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.043 13.044 init_scf_run 11 5.9 0.000 0.001 10.563 10.563 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.562 10.563 make_images_data 5054 15.6 0.064 0.074 8.851 10.537 cp_fm_diag_elpa 83 13.4 0.006 0.011 9.915 9.937 cp_fm_diag_elpa_base 83 14.4 9.500 9.664 9.901 9.923 hybrid_alltoall_any 5240 16.5 0.455 2.052 7.635 9.760 multiply_cannon_metrocomm4 17689 15.6 0.066 0.076 3.568 9.643 mp_irecv_dv 50625 16.2 3.439 9.377 3.439 9.377 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.919 9.233 density_rs2pw 129 9.7 0.006 0.006 7.177 9.080 pw_transfer 1559 11.6 0.087 0.102 8.906 9.012 dbcsr_mm_accdrv_process 41830 16.2 5.603 5.968 8.833 9.001 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.680 8.792 fft_wrap_pw1pw2_140 527 13.2 1.432 1.458 7.725 7.848 grid_integrate_task_list 129 12.3 7.311 7.772 7.311 7.772 wfi_extrapolate 11 7.9 0.001 0.001 7.467 7.467 cp_fm_upper_to_full 105 14.8 5.710 7.229 5.710 7.229 cp_fm_cholesky_decompose 22 10.9 7.079 7.145 7.079 7.145 rs_pw_transfer 1054 12.0 0.014 0.014 5.137 7.077 dbcsr_complete_redistribute 395 12.7 1.168 1.201 4.601 6.350 calculate_dm_sparse 129 9.5 0.001 0.001 5.870 5.978 grid_collocate_task_list 129 9.7 5.028 5.915 5.028 5.915 fft3d_ps 1301 14.7 2.761 2.994 5.767 5.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.457 5.468 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.610 5.307 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.450 5.196 mp_alltoall_d11v 2423 14.1 4.426 5.058 4.426 5.058 mp_allgather_i34 2527 14.6 1.559 4.772 1.559 4.772 potential_pw2rs 129 12.3 0.020 0.022 4.708 4.721 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.557 4.717 mp_sum_l 8010 12.9 3.170 4.653 3.170 4.653 mp_waitany 11836 13.9 2.380 4.373 2.380 4.373 transfer_fm_to_dbcsr 11 9.9 0.017 0.021 2.362 4.054 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.979 4.013 rs_pw_transfer_RS2PW_140 140 11.5 0.331 0.349 1.936 3.869 mp_alltoall_i22 718 14.1 1.973 3.843 1.973 3.843 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.800 3.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.624 3.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.376000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=920.727273, yerr=12.764232 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.157378E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58132. MP_Allreduce 11095 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.185 187.812 187.814 qs_mol_dyn_low 1 2.0 0.003 0.004 186.569 186.806 qs_forces 11 3.9 0.007 0.024 186.458 186.461 qs_energies 11 4.9 0.003 0.011 179.211 179.224 scf_env_do_scf 11 5.9 0.001 0.001 162.325 162.332 velocity_verlet 10 3.0 0.002 0.003 123.786 123.789 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 116.483 116.484 dbcsr_multiply_generic 2485 12.5 0.185 0.192 79.791 80.506 qs_scf_new_mos 116 7.6 0.001 0.001 79.336 79.682 qs_scf_loop_do_ot 116 8.6 0.001 0.002 79.335 79.681 ot_scf_mini 116 9.6 0.003 0.004 74.827 75.164 multiply_cannon 2485 13.5 0.554 0.588 55.551 58.369 multiply_cannon_loop 2485 14.5 1.817 1.917 51.714 53.504 init_scf_loop 11 6.9 0.001 0.003 45.723 45.724 ot_mini 116 10.6 0.001 0.001 42.611 42.949 prepare_preconditioner 11 7.9 0.000 0.000 41.579 41.598 make_preconditioner 11 8.9 0.000 0.001 41.578 41.598 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.344 40.299 multiply_cannon_multrec 29820 15.5 13.450 18.948 26.245 31.400 rebuild_ks_matrix 127 8.3 0.001 0.001 29.710 30.030 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.710 30.029 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.698 26.993 mp_waitall_1 146592 16.7 16.861 25.854 16.861 25.854 qs_ot_get_derivative 116 11.6 0.002 0.002 23.117 23.447 make_m2s 4970 13.5 0.094 0.098 19.852 21.049 make_images 4970 14.5 1.943 2.250 19.548 20.745 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.948 19.475 apply_single 127 13.6 0.001 0.001 18.948 19.475 ot_diis_step 116 11.6 0.017 0.018 19.372 19.373 qs_ot_get_p 127 10.4 0.001 0.001 18.572 18.954 cp_fm_upper_to_full 104 14.8 11.351 16.700 11.351 16.700 cp_fm_cholesky_invert 11 10.9 15.863 15.872 15.863 15.872 multiply_cannon_metrocomm3 29820 15.5 0.050 0.053 6.323 14.770 sum_up_and_integrate 127 10.3 0.139 0.150 14.617 14.641 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.491 14.521 calculate_rho_elec 127 8.7 0.172 0.187 14.490 14.520 integrate_v_rspace 127 11.3 0.004 0.005 14.477 14.505 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.379 14.428 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.000 13.001 dbcsr_mm_accdrv_process 61748 16.2 8.406 9.161 12.366 12.882 dbcsr_complete_redistribute 393 12.7 1.509 1.621 8.924 12.746 make_images_data 4970 15.5 0.066 0.072 10.599 12.356 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.585 11.406 multiply_cannon_sync_h2d 29820 15.5 10.315 11.362 10.315 11.362 hybrid_alltoall_any 5155 16.4 0.527 2.215 9.546 10.998 init_scf_run 11 5.9 0.000 0.001 10.930 10.932 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.930 10.931 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.211 9.915 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.645 9.896 cp_fm_diag_elpa 82 13.4 0.001 0.002 9.864 9.882 cp_fm_diag_elpa_base 82 14.4 8.912 9.226 9.857 9.873 mp_alltoall_i22 712 14.1 5.521 9.459 5.521 9.459 pw_transfer 1535 11.6 0.086 0.103 9.060 9.161 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.835 8.942 density_rs2pw 127 9.7 0.006 0.006 6.884 8.266 fft_wrap_pw1pw2_140 519 13.2 1.554 1.587 7.880 8.001 grid_integrate_task_list 127 12.3 7.429 7.898 7.429 7.898 wfi_extrapolate 11 7.9 0.001 0.001 7.777 7.777 cp_fm_cholesky_decompose 22 10.9 7.378 7.469 7.378 7.469 multiply_cannon_metrocomm4 24850 15.5 0.082 0.092 2.820 7.061 mp_irecv_dv 75445 16.2 2.666 6.789 2.666 6.789 calculate_dm_sparse 127 9.5 0.001 0.001 6.194 6.285 rs_pw_transfer 1038 11.9 0.013 0.014 4.554 5.988 fft3d_ps 1281 14.7 2.817 2.913 5.771 5.850 grid_collocate_task_list 127 9.7 5.109 5.821 5.109 5.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.486 5.548 mp_alltoall_d11v 2401 14.1 4.959 5.502 4.959 5.502 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.482 4.579 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.411 4.531 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.495 4.496 potential_pw2rs 127 12.3 0.023 0.024 4.435 4.450 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.160 4.214 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.814000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1099.272727, yerr=14.948949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/19/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 5.865088E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.531404E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.065 170.418 170.426 qs_mol_dyn_low 1 2.0 0.003 0.003 169.990 170.017 qs_forces 11 3.9 0.004 0.004 169.889 169.895 qs_energies 11 4.9 0.002 0.002 162.270 162.277 scf_env_do_scf 11 5.9 0.001 0.001 144.841 144.850 velocity_verlet 10 3.0 0.002 0.002 112.386 112.390 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 109.370 109.371 dbcsr_multiply_generic 2507 12.6 0.180 0.192 71.653 71.931 qs_scf_new_mos 117 7.6 0.001 0.001 70.812 70.893 qs_scf_loop_do_ot 117 8.6 0.001 0.001 70.812 70.892 ot_scf_mini 117 9.6 0.004 0.004 66.404 66.447 multiply_cannon 2507 13.6 0.562 0.590 53.136 56.537 multiply_cannon_loop 2507 14.6 0.804 0.832 50.235 50.869 ot_mini 117 10.6 0.001 0.001 37.266 37.320 init_scf_loop 11 6.9 0.000 0.000 35.324 35.325 prepare_preconditioner 11 7.9 0.000 0.000 31.441 31.466 make_preconditioner 11 8.9 0.000 0.000 31.441 31.466 mp_waitall_1 125778 16.7 24.546 30.684 24.546 30.684 rebuild_ks_matrix 128 8.3 0.001 0.001 29.867 29.959 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.866 29.958 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.340 29.625 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.091 27.183 multiply_cannon_multrec 10028 15.6 10.410 14.706 17.972 21.036 qs_ot_get_derivative 117 11.6 0.002 0.002 20.257 20.312 multiply_cannon_metrocomm3 10028 15.6 0.026 0.028 12.449 19.675 cp_fm_cholesky_invert 11 10.9 17.906 17.912 17.906 17.912 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.972 17.083 apply_single 128 13.6 0.001 0.001 16.972 17.082 ot_diis_step 117 11.6 0.020 0.020 16.936 16.937 qs_ot_get_p 128 10.4 0.001 0.001 16.467 16.525 make_m2s 5014 13.6 0.065 0.070 14.695 15.500 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.382 15.417 calculate_rho_elec 128 8.7 0.256 0.267 15.382 15.416 make_images 5014 14.6 2.176 2.603 14.390 15.192 sum_up_and_integrate 128 10.3 0.179 0.189 14.954 15.001 integrate_v_rspace 128 11.3 0.004 0.004 14.775 14.830 qs_ot_p2m_diag 83 11.4 0.495 0.500 12.756 12.773 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.597 11.597 multiply_cannon_sync_h2d 10028 15.6 10.881 11.132 10.881 11.132 init_scf_run 11 5.9 0.000 0.001 10.555 10.556 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.555 10.555 hybrid_alltoall_any 5200 16.5 0.844 3.777 8.314 9.773 make_images_data 5014 15.6 0.056 0.065 8.462 9.713 pw_transfer 1547 11.6 0.086 0.096 9.433 9.452 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.208 9.233 cp_fm_diag_elpa 83 13.4 0.002 0.002 8.774 8.786 cp_fm_diag_elpa_base 83 14.4 8.535 8.610 8.769 8.782 grid_integrate_task_list 128 12.3 7.727 8.281 7.727 8.281 fft_wrap_pw1pw2_140 523 13.2 1.922 1.960 8.165 8.193 cp_fm_cholesky_decompose 22 10.9 7.882 8.037 7.882 8.037 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.945 7.987 dbcsr_mm_accdrv_process 20762 16.1 3.269 4.456 7.206 7.915 density_rs2pw 128 9.7 0.006 0.006 6.865 7.780 wfi_extrapolate 11 7.9 0.001 0.001 7.554 7.554 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.434 7.441 calculate_dm_sparse 128 9.5 0.001 0.001 6.110 6.182 grid_collocate_task_list 128 9.7 5.468 6.144 5.468 6.144 mp_alltoall_d11v 2415 14.1 4.855 5.890 4.855 5.890 fft3d_ps 1291 14.7 2.761 2.841 5.671 5.715 dbcsr_complete_redistribute 395 12.7 2.120 2.198 5.124 5.526 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.336 5.337 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.154 5.188 rs_pw_transfer 1046 11.9 0.012 0.013 4.108 5.023 mp_allgather_i34 2507 14.6 1.088 4.521 1.088 4.521 potential_pw2rs 128 12.3 0.026 0.026 4.367 4.378 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.051 4.091 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.591 3.895 multiply_cannon_metrocomm4 7521 15.6 0.026 0.028 1.723 3.780 mp_irecv_dv 28860 15.9 1.685 3.708 1.685 3.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.677 3.688 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.357 3.663 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.566 3.635 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.539 3.579 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.372 3.436 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.426000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1459.545455, yerr=1.876342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_performance_tests/20/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 11.786061E+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 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.115057E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 301.341 301.342 qs_mol_dyn_low 1 2.0 0.003 0.007 300.803 300.815 qs_forces 11 3.9 0.005 0.005 300.710 300.714 qs_energies 11 4.9 0.002 0.002 291.239 291.246 scf_env_do_scf 11 5.9 0.001 0.001 268.291 268.298 velocity_verlet 10 3.0 0.002 0.002 217.759 217.767 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 142.389 142.391 init_scf_loop 11 6.9 0.000 0.000 125.623 125.625 prepare_preconditioner 11 7.9 0.000 0.000 120.480 120.502 make_preconditioner 11 8.9 0.000 0.000 120.480 120.502 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.650 117.626 qs_scf_new_mos 118 7.6 0.001 0.001 91.947 92.085 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.946 92.084 ot_scf_mini 118 9.6 0.004 0.004 86.946 87.012 dbcsr_multiply_generic 2529 12.6 0.213 0.226 84.378 84.917 cp_fm_upper_to_full 106 14.8 53.599 77.462 53.599 77.462 multiply_cannon 2529 13.6 0.683 0.727 60.430 61.541 multiply_cannon_loop 2529 14.6 1.078 1.106 56.230 57.802 ot_mini 118 10.6 0.001 0.001 45.509 45.606 dbcsr_complete_redistribute 397 12.7 4.017 4.052 30.845 44.450 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.322 40.908 rebuild_ks_matrix 129 8.3 0.001 0.001 38.097 38.191 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.096 38.191 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.787 38.141 mp_alltoall_i22 720 14.1 22.567 36.290 22.567 36.290 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.029 35.130 mp_waitall_1 104580 16.8 29.605 34.654 29.605 34.654 cp_fm_cholesky_invert 11 10.9 33.165 33.172 33.165 33.172 qs_ot_get_p 129 10.4 0.001 0.001 25.963 26.011 qs_ot_get_derivative 118 11.6 0.002 0.002 25.110 25.180 qs_ot_p2m_diag 84 11.4 0.891 0.897 21.773 21.802 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 19.833 21.432 make_m2s 5058 13.6 0.076 0.080 19.168 20.752 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.584 20.595 calculate_rho_elec 129 8.7 0.485 0.486 20.584 20.595 ot_diis_step 118 11.6 0.022 0.023 20.354 20.354 make_images 5058 14.6 3.096 3.305 18.685 20.272 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.950 19.953 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.600 19.798 apply_single 129 13.6 0.001 0.001 19.599 19.797 sum_up_and_integrate 129 10.3 0.323 0.325 19.657 19.743 integrate_v_rspace 129 11.3 0.004 0.004 19.334 19.419 multiply_cannon_multrec 10116 15.6 10.402 12.229 18.211 18.279 cp_fm_diag_elpa 84 13.4 0.001 0.002 16.738 16.739 cp_fm_diag_elpa_base 84 14.4 12.216 13.889 16.734 16.734 multiply_cannon_sync_h2d 10116 15.6 14.373 14.429 14.373 14.429 hybrid_alltoall_any 5245 16.5 1.327 3.063 11.227 13.584 make_images_data 5058 15.6 0.064 0.068 10.958 13.279 init_scf_run 11 5.9 0.000 0.001 12.503 12.504 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.503 12.503 pw_transfer 1559 11.6 0.094 0.095 11.470 11.476 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.232 11.238 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.025 10.084 fft_wrap_pw1pw2_140 527 13.2 3.139 3.192 10.018 10.027 dbcsr_mm_accdrv_process 20934 16.1 4.267 6.205 7.559 9.474 wfi_extrapolate 11 7.9 0.001 0.001 9.379 9.379 grid_integrate_task_list 129 12.3 8.653 8.838 8.653 8.838 cp_fm_cholesky_decompose 22 10.9 8.804 8.829 8.804 8.829 mp_alltoall_d11v 2429 14.1 8.048 8.717 8.048 8.717 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.244 8.245 density_rs2pw 129 9.7 0.005 0.005 7.996 8.148 calculate_dm_sparse 129 9.5 0.001 0.001 6.879 6.965 grid_collocate_task_list 129 9.7 6.486 6.566 6.486 6.566 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.480 6.516 rs_scatter_matrices 140 9.7 3.985 4.855 6.076 6.298 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.203 6.288 fft3d_ps 1301 14.7 2.874 2.882 6.221 6.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.342000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2866.636364, yerr=139.664060 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 1.259741E+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 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 0.023 0.075 85.158 85.159 qs_energies 1 2.0 0.000 0.000 84.590 84.602 ls_scf 1 3.0 0.000 0.000 83.690 83.703 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.490 72.627 multiply_cannon 111 7.7 0.017 0.020 55.762 56.913 multiply_cannon_loop 111 8.7 0.228 0.246 52.391 53.690 ls_scf_main 1 4.0 0.000 0.000 52.550 52.550 density_matrix_trs4 2 5.0 0.002 0.003 47.017 47.092 ls_scf_init_scf 1 4.0 0.000 0.000 28.093 28.094 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.929 26.987 mp_waitall_1 11031 10.9 22.454 25.501 22.454 25.501 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.831 24.847 multiply_cannon_multrec 2664 9.7 8.153 8.903 15.596 17.386 multiply_cannon_sync_h2d 2664 9.7 13.634 15.655 13.634 15.655 make_m2s 222 7.7 0.009 0.011 13.200 13.607 make_images 222 8.7 0.098 0.108 13.178 13.587 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.547 13.174 make_images_data 222 9.7 0.004 0.005 7.822 8.369 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.422 8.126 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.696 7.061 8.053 hybrid_alltoall_any 227 10.6 0.216 1.822 6.704 7.959 dbcsr_mm_accdrv_process_sort 4760 11.4 6.267 7.168 6.267 7.168 calculate_norms 4752 9.8 5.526 6.209 5.526 6.209 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.138 mp_sum_l 887 5.1 3.010 4.412 3.010 4.412 make_images_sizes 222 9.7 0.000 0.000 0.690 3.699 mp_alltoall_i44 222 10.7 0.689 3.699 0.689 3.699 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.014 3.435 mp_irecv_dv 6231 10.9 1.997 3.427 1.997 3.427 arnoldi_extremal 4 6.8 0.000 0.000 3.253 3.281 arnoldi_normal_ev 4 7.8 0.001 0.003 3.253 3.281 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.214 3.252 build_subspace 16 8.4 0.009 0.012 3.156 3.158 ls_scf_post 1 4.0 0.000 0.000 3.047 3.060 ls_scf_store_result 1 5.0 0.000 0.000 2.866 2.909 dbcsr_special_finalize 555 9.7 0.005 0.006 2.336 2.672 dbcsr_merge_single_wm 555 10.7 0.454 0.576 2.327 2.662 make_images_pack 222 9.7 2.206 2.621 2.208 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.356 2.596 dbcsr_matrix_vector_mult_local 304 10.0 2.079 2.490 2.081 2.492 dbcsr_sort_data 658 11.4 2.133 2.422 2.133 2.422 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.323 2.400 buffer_matrices_ensure_size 222 8.7 1.743 2.053 1.743 2.053 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.776 1.777 rebuild_ks_matrix 3 7.3 0.000 0.000 1.766 1.767 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.766 1.767 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.159000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 2.172269E+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 0.042 0.100 89.636 89.637 qs_energies 1 2.0 0.000 0.000 89.068 89.072 ls_scf 1 3.0 0.000 0.002 87.714 87.717 dbcsr_multiply_generic 111 6.7 0.017 0.031 73.796 74.188 multiply_cannon 111 7.7 0.027 0.040 52.512 56.373 ls_scf_main 1 4.0 0.000 0.010 54.230 54.230 multiply_cannon_loop 111 8.7 0.135 0.147 49.793 53.141 density_matrix_trs4 2 5.0 0.002 0.004 48.472 48.659 mp_waitall_1 9105 10.9 20.552 30.044 20.552 30.044 ls_scf_init_scf 1 4.0 0.000 0.001 29.749 29.751 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.551 28.635 multiply_cannon_multrec 1332 9.7 13.238 17.388 22.572 27.880 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.194 26.208 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.366 20.792 make_m2s 222 7.7 0.007 0.008 14.927 15.492 make_images 222 8.7 1.374 1.703 14.896 15.462 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.532 8.930 10.508 dbcsr_mm_accdrv_process_sort 4041 11.4 8.445 9.976 8.445 9.976 make_images_data 222 9.7 0.004 0.005 8.648 9.517 hybrid_alltoall_any 227 10.6 0.543 2.567 8.012 9.398 mp_sum_l 887 5.1 4.883 7.980 4.883 7.980 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.199 7.810 mp_irecv_dv 3311 11.0 3.179 7.751 3.179 7.751 calculate_norms 2376 9.8 6.072 6.832 6.072 6.832 multiply_cannon_sync_h2d 1332 9.7 4.870 6.434 4.870 6.434 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.648 6.165 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.926 5.178 arnoldi_extremal 4 6.8 0.000 0.000 4.645 4.666 arnoldi_normal_ev 4 7.8 0.001 0.005 4.644 4.666 build_subspace 16 8.4 0.014 0.021 4.385 4.389 ls_scf_post 1 4.0 0.000 0.001 3.735 3.739 ls_scf_store_result 1 5.0 0.000 0.000 3.424 3.557 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.156 3.397 dbcsr_matrix_vector_mult_local 304 10.0 2.754 3.249 2.756 3.251 mp_allgather_i34 111 8.7 0.711 2.879 0.711 2.879 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.664 2.785 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.186 2.457 dbcsr_data_new 4174 10.1 2.107 2.406 2.107 2.406 make_images_pack 222 9.7 1.823 2.142 1.825 2.144 dbcsr_sort_data 436 11.2 1.839 2.071 1.839 2.071 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.876 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.858 1.861 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.858 1.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.637000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1773.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 2.864726E+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 0.041 0.086 92.099 92.100 qs_energies 1 2.0 0.000 0.000 91.447 91.450 ls_scf 1 3.0 0.000 0.001 90.017 90.021 dbcsr_multiply_generic 111 6.7 0.016 0.022 74.636 74.889 ls_scf_main 1 4.0 0.000 0.002 56.047 56.051 multiply_cannon 111 7.7 0.031 0.075 51.656 55.484 multiply_cannon_loop 111 8.7 0.116 0.129 48.907 53.178 density_matrix_trs4 2 5.0 0.004 0.008 50.073 50.230 mp_waitall_1 7281 11.0 22.943 33.872 22.943 33.872 ls_scf_init_scf 1 4.0 0.000 0.001 30.247 30.250 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.034 29.098 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.664 26.680 multiply_cannon_multrec 888 9.7 12.686 15.464 21.320 24.687 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.744 23.473 make_m2s 222 7.7 0.006 0.007 16.062 16.836 make_images 222 8.7 1.584 1.856 16.024 16.796 make_images_data 222 9.7 0.004 0.005 9.295 10.344 hybrid_alltoall_any 227 10.6 0.641 2.966 8.925 10.079 dbcsr_mm_accdrv_process 3754 10.4 0.299 0.491 8.160 9.384 dbcsr_mm_accdrv_process_sort 3754 11.4 7.724 8.893 7.724 8.893 mp_sum_l 887 5.1 5.228 8.423 5.228 8.423 multiply_cannon_sync_h2d 888 9.7 6.008 7.305 6.008 7.305 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.496 6.942 mp_irecv_dv 2335 11.1 2.481 6.888 2.481 6.888 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.549 6.874 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.024 6.477 arnoldi_extremal 4 6.8 0.000 0.000 5.105 5.126 arnoldi_normal_ev 4 7.8 0.001 0.005 5.105 5.125 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.752 4.998 build_subspace 16 8.4 0.014 0.020 4.798 4.804 calculate_norms 1584 9.8 4.331 4.674 4.331 4.674 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.452 3.775 ls_scf_post 1 4.0 0.000 0.001 3.724 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.036 3.614 3.038 3.616 mp_allgather_i34 111 8.7 0.865 3.603 0.865 3.603 ls_scf_store_result 1 5.0 0.000 0.000 3.456 3.548 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.907 3.038 dbcsr_data_new 4116 9.9 2.110 2.459 2.110 2.459 dbcsr_sort_data 325 11.1 1.880 2.110 1.880 2.110 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.940 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.925 1.927 make_images_sizes 222 9.7 0.000 0.000 0.930 1.926 mp_alltoall_i44 222 10.7 0.930 1.925 0.930 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.904 1.907 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.904 1.907 make_images_pack 222 9.7 1.625 1.882 1.628 1.885 dbcsr_merge_all 275 8.9 0.247 0.313 1.532 1.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.100000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2225.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 3.379470E+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 0.031 0.048 96.858 96.860 qs_energies 1 2.0 0.000 0.000 96.248 96.254 ls_scf 1 3.0 0.000 0.000 94.599 94.607 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.420 78.668 ls_scf_main 1 4.0 0.000 0.000 58.901 58.902 multiply_cannon 111 7.7 0.046 0.085 51.724 55.585 density_matrix_trs4 2 5.0 0.002 0.004 52.772 52.884 multiply_cannon_loop 111 8.7 0.152 0.164 46.733 49.511 ls_scf_init_scf 1 4.0 0.000 0.000 32.477 32.479 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.267 31.322 mp_waitall_1 6369 11.0 22.779 29.997 22.779 29.997 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.810 28.825 multiply_cannon_multrec 1332 9.7 14.172 17.264 22.129 25.158 make_m2s 222 7.7 0.007 0.008 21.215 22.596 make_images 222 8.7 3.146 3.609 21.165 22.548 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.278 16.488 make_images_data 222 9.7 0.004 0.004 11.844 13.437 hybrid_alltoall_any 227 10.6 0.797 3.742 11.222 12.857 dbcsr_mm_accdrv_process 3641 10.4 0.285 0.481 7.594 9.126 dbcsr_mm_accdrv_process_sort 3641 11.4 7.155 8.643 7.155 8.643 mp_sum_l 887 5.1 3.886 6.680 3.886 6.680 multiply_cannon_sync_h2d 1332 9.7 5.488 6.063 5.488 6.063 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.078 6.038 mp_irecv_dv 3229 10.9 2.054 5.977 2.054 5.977 arnoldi_extremal 4 6.8 0.000 0.000 5.197 5.212 arnoldi_normal_ev 4 7.8 0.001 0.005 5.197 5.212 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.966 5.194 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.492 5.016 build_subspace 16 8.4 0.014 0.021 4.861 4.871 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.528 4.702 calculate_norms 2376 9.8 4.175 4.547 4.175 4.547 mp_allgather_i34 111 8.7 2.108 4.297 2.108 4.297 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.590 3.878 dbcsr_matrix_vector_mult_local 304 10.0 3.210 3.699 3.212 3.701 dbcsr_sort_data 658 11.4 3.073 3.426 3.073 3.426 ls_scf_post 1 4.0 0.000 0.000 3.221 3.227 dbcsr_special_finalize 555 9.7 0.006 0.007 2.819 3.212 dbcsr_merge_single_wm 555 10.7 0.539 0.664 2.810 3.204 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.026 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.975 3.026 dbcsr_data_release 10477 10.7 1.593 2.447 1.593 2.447 dbcsr_finalize 304 7.8 0.049 0.061 1.809 1.957 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.938 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.860000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 4.728730E+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 0.040 0.056 91.940 91.942 qs_energies 1 2.0 0.000 0.000 91.184 91.189 ls_scf 1 3.0 0.000 0.000 89.250 89.254 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.720 70.883 ls_scf_main 1 4.0 0.000 0.000 56.632 56.632 multiply_cannon 111 7.7 0.066 0.116 52.765 56.092 multiply_cannon_loop 111 8.7 0.088 0.095 50.203 51.367 density_matrix_trs4 2 5.0 0.002 0.003 49.613 49.680 ls_scf_init_scf 1 4.0 0.000 0.000 29.226 29.229 mp_waitall_1 5436 11.0 24.494 28.467 24.494 28.467 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.944 27.968 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.860 25.869 multiply_cannon_multrec 444 9.7 13.657 16.627 20.781 22.197 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.815 15.734 make_m2s 222 7.7 0.005 0.005 13.411 14.340 make_images 222 8.7 2.043 2.482 13.344 14.271 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.334 13.565 make_images_data 222 9.7 0.003 0.004 8.165 9.680 hybrid_alltoall_any 227 10.6 0.802 3.792 8.095 9.609 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.409 6.827 7.952 multiply_cannon_sync_h2d 444 9.7 6.762 7.896 6.762 7.896 dbcsr_mm_accdrv_process_sort 3003 11.4 6.464 7.551 6.464 7.551 arnoldi_extremal 4 6.8 0.000 0.000 5.870 5.879 arnoldi_normal_ev 4 7.8 0.002 0.005 5.870 5.879 build_subspace 16 8.4 0.015 0.020 5.477 5.488 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.371 4.528 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.178 4.393 mp_sum_l 887 5.1 2.544 4.216 2.544 4.216 dbcsr_matrix_vector_mult_local 304 10.0 3.711 4.196 3.713 4.198 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.563 3.988 mp_irecv_dv 1241 11.2 1.550 3.949 1.550 3.949 mp_allgather_i34 111 8.7 1.141 3.698 1.141 3.698 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.603 3.679 calculate_norms 792 9.8 3.613 3.674 3.613 3.674 ls_scf_post 1 4.0 0.000 0.000 3.392 3.397 ls_scf_store_result 1 5.0 0.000 0.000 3.188 3.220 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.780 3.208 make_images_sizes 222 9.7 0.000 0.000 0.858 3.133 mp_alltoall_i44 222 10.7 0.858 3.132 0.858 3.132 dbcsr_finalize 304 7.8 0.062 0.077 2.185 2.273 dbcsr_data_new 4608 9.7 1.816 2.270 1.816 2.270 dbcsr_merge_all 275 8.9 0.477 0.523 2.043 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.087 2.088 rebuild_ks_matrix 3 7.3 0.000 0.000 2.054 2.055 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 2.054 2.055 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.918 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.942000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 8.850039E+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 0.076 0.100 105.092 105.093 qs_energies 1 2.0 0.000 0.000 103.706 103.714 ls_scf 1 3.0 0.000 0.000 100.767 100.776 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.413 74.556 ls_scf_main 1 4.0 0.000 0.000 63.474 63.474 density_matrix_trs4 2 5.0 0.002 0.003 54.455 54.518 multiply_cannon 111 7.7 0.144 0.272 48.047 50.007 multiply_cannon_loop 111 8.7 0.099 0.101 45.076 45.703 ls_scf_init_scf 1 4.0 0.000 0.000 33.522 33.523 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.945 31.973 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.265 29.275 mp_waitall_1 4527 11.1 21.636 25.277 21.636 25.277 make_m2s 222 7.7 0.005 0.005 22.746 23.804 make_images 222 8.7 3.575 3.879 22.638 23.695 multiply_cannon_multrec 444 9.7 17.786 18.360 22.442 22.959 hybrid_alltoall_any 227 10.6 1.653 3.604 12.863 15.846 make_images_data 222 9.7 0.003 0.004 13.124 15.226 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.913 10.213 multiply_cannon_sync_h2d 444 9.7 8.788 8.842 8.788 8.842 arnoldi_extremal 4 6.8 0.000 0.000 7.368 7.382 arnoldi_normal_ev 4 7.8 0.003 0.009 7.368 7.382 build_subspace 16 8.4 0.026 0.036 6.822 6.833 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.475 5.627 dbcsr_matrix_vector_mult_local 304 10.0 5.068 5.363 5.070 5.365 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.139 5.231 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.847 5.109 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.358 4.484 4.621 dbcsr_mm_accdrv_process_sort 1814 11.4 4.149 4.288 4.149 4.288 ls_scf_post 1 4.0 0.000 0.000 3.771 3.779 make_images_sizes 222 9.7 0.000 0.000 1.478 3.636 mp_alltoall_i44 222 10.7 1.477 3.636 1.477 3.636 ls_scf_store_result 1 5.0 0.000 0.000 3.494 3.529 mp_allgather_i34 111 8.7 1.074 3.437 1.074 3.437 calculate_norms 792 9.8 3.230 3.271 3.230 3.271 dbcsr_finalize 304 7.8 0.082 0.089 3.070 3.165 dbcsr_merge_all 275 8.9 0.884 0.917 2.855 2.941 qs_energies_init_hamiltonians 1 3.0 0.003 0.006 2.908 2.908 dbcsr_complete_redistribute 5 7.6 1.427 1.461 2.739 2.844 dbcsr_sort_data 325 11.1 2.437 2.494 2.437 2.494 matrix_ls_to_qs 2 6.0 0.000 0.000 2.385 2.490 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.467 2.469 rebuild_ks_matrix 3 7.3 0.000 0.000 2.398 2.400 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.398 2.400 dbcsr_data_new 6591 9.6 1.885 2.355 1.885 2.355 dbcsr_new_transposed 4 7.5 0.241 0.251 2.283 2.300 dbcsr_frobenius_norm 74 6.6 2.058 2.141 2.206 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.201 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.201 dbcsr_data_release 12724 10.6 1.971 2.166 1.971 2.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.093000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6979.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b331c57d0a21e261d77504111c8b6a615dad1b78_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 593.022976E+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 885708282. 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 0.052 0.197 236.277 236.281 qs_mol_dyn_low 1 2.0 0.004 0.045 234.926 234.961 qs_forces 5 3.8 0.006 0.030 234.786 234.787 qs_energies 5 4.8 0.004 0.043 231.606 231.632 scf_env_do_scf 5 5.8 0.000 0.002 217.390 217.393 scf_env_do_scf_inner_loop 105 6.6 0.004 0.017 189.477 189.487 qs_scf_new_mos 105 7.6 0.000 0.001 147.569 147.764 qs_scf_loop_do_ot 105 8.6 0.001 0.001 147.569 147.764 ot_scf_mini 105 9.6 0.003 0.004 137.706 137.864 dbcsr_multiply_generic 1445 12.2 0.126 0.138 132.497 133.012 multiply_cannon 1445 13.2 0.274 0.286 113.510 115.582 multiply_cannon_loop 1445 14.2 2.852 2.996 111.850 113.387 velocity_verlet 4 3.0 0.003 0.010 107.899 107.900 ot_mini 105 10.6 0.001 0.001 59.717 59.852 qs_ot_get_p 112 10.4 0.001 0.002 48.010 48.272 multiply_cannon_multrec 69360 15.2 29.685 35.138 39.629 44.694 mp_waitall_1 488190 16.1 34.676 42.097 34.676 42.097 qs_ot_get_derivative 55 11.6 0.001 0.001 38.027 38.158 qs_ot_p2m_diag 40 11.0 0.020 0.031 36.856 36.976 cp_dbcsr_syevd 40 12.0 0.002 0.004 33.612 33.612 multiply_cannon_sync_h2d 69360 15.2 29.110 33.547 29.110 33.547 multiply_cannon_metrocomm3 69360 15.2 0.204 0.215 25.559 33.087 rebuild_ks_matrix 110 8.4 0.000 0.001 29.976 30.149 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.062 29.976 30.148 cp_fm_syevd 40 13.0 0.001 0.002 28.459 28.610 init_scf_loop 7 6.6 0.001 0.011 27.880 27.883 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.419 27.581 cp_fm_redistribute_end 40 14.0 11.871 23.699 11.877 23.700 cp_fm_syevd_base 40 14.0 11.815 23.643 11.815 23.643 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.963 23.193 apply_single 62 13.6 0.000 0.000 22.963 23.193 prepare_preconditioner 7 7.6 0.000 0.000 22.769 22.796 make_preconditioner 7 8.6 0.000 0.003 22.769 22.796 ot_new_cg_direction 55 11.6 0.001 0.001 20.989 20.990 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.173 18.505 calculate_rho_elec 110 8.6 0.030 0.033 18.173 18.505 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.347 15.425 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.801 14.974 rs_pw_transfer 690 11.5 0.011 0.013 12.463 13.784 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.619 13.724 density_rs2pw 110 9.6 0.006 0.007 12.195 13.577 pw_transfer 1645 12.4 0.082 0.102 12.748 12.956 mp_sum_l 4764 12.2 12.142 12.893 12.142 12.893 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.606 12.820 calculate_dm_sparse 110 9.5 0.000 0.001 11.474 11.755 init_scf_run 5 5.8 0.000 0.001 11.462 11.463 scf_env_initial_rho_setup 5 6.8 0.001 0.003 11.462 11.463 fft_wrap_pw1pw2_240 915 15.0 1.188 1.256 10.766 10.969 qs_vxc_create 110 10.4 0.003 0.005 10.723 10.760 dbcsr_mm_accdrv_process 154766 15.8 6.191 6.373 9.811 10.656 cp_fm_cholesky_invert 7 10.6 10.542 10.549 10.542 10.549 qs_ot_get_derivative_diag 18 12.0 0.001 0.017 10.304 10.369 check_diag 80 13.5 8.596 8.886 9.508 9.662 sum_up_and_integrate 60 10.3 0.028 0.030 9.165 9.211 integrate_v_rspace 60 11.3 0.002 0.003 9.137 9.183 fft3d_pb 915 16.0 2.370 2.556 8.849 9.037 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.543 7.782 acc_transpose_blocks 69360 15.2 0.367 0.389 6.948 7.590 xc_vxc_pw_create 60 11.3 0.039 0.051 7.204 7.241 make_full_single_inverse 7 9.6 0.001 0.012 7.100 7.128 make_m2s 2890 13.2 0.078 0.087 6.496 7.089 make_images 2890 14.2 0.240 0.263 6.390 6.985 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 6.926 6.946 multiply_cannon_metrocomm1 69360 15.2 0.098 0.105 4.623 6.826 xc_pw_derive 510 13.4 0.005 0.006 6.485 6.559 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 6.491 6.551 calculate_first_density_matrix 1 7.0 0.000 0.004 6.508 6.521 mp_alltoall_z22v 2340 17.7 5.931 6.290 5.931 6.290 mp_waitany 7680 13.5 4.447 5.626 4.447 5.626 potential_pw2rs 60 12.3 0.003 0.003 5.260 5.353 multiply_cannon_metrocomm4 67915 15.2 0.185 0.206 2.027 4.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=236.281000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b331c57d0a21e261d77504111c8b6a615dad1b78 Summary: empty Status: OK