=== 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: ac71c107aafb4ece195b773ce8dc175d9d1cdc7c ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # 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: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/01 job id: 49967972 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/02 job id: 49967973 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/03 job id: 49967974 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/04 job id: 49967975 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/05 job id: 49967976 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/06 job id: 49967977 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/07 job id: 49967978 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/08 job id: 49967979 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/09 job id: 49967980 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/10 job id: 49967982 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/11 job id: 49967984 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/12 job id: 49967986 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/13 job id: 49967987 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/14 job id: 49967988 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/15 job id: 49967989 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/16 job id: 49967991 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/17 job id: 49967992 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/18 job id: 49967993 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/19 job id: 49967994 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/20 job id: 49967995 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/21 job id: 49967996 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/22 job id: 49967997 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/23 job id: 49967998 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/24 job id: 49967999 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/25 job id: 49968000 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/26 job id: 49968001 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/27 job id: 49968002 --- 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/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 0.038 134.110 134.112 farming_run 1 2.0 133.366 133.367 134.065 134.069 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.491591E+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 0.010 0.027 114.351 114.352 qs_energies 1 2.0 0.000 0.000 114.078 114.081 mp2_main 1 3.0 0.000 0.000 112.009 112.013 mp2_gpw_main 1 4.0 0.028 0.034 110.990 110.994 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 91.858 92.206 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.596 54.946 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.007 46.103 get_2c_integrals 1 6.0 0.008 0.009 36.456 37.089 integrate_v_rspace 273 8.0 0.443 0.462 24.586 29.565 pw_transfer 6555 10.6 0.378 0.391 26.739 27.175 grid_integrate_task_list 273 9.0 20.499 25.955 20.499 25.955 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.370 25.895 fft_wrap_pw1pw2_100 2178 12.4 1.103 1.243 22.927 23.436 rpa_ri_compute_en 1 5.0 0.022 0.069 19.028 19.180 compute_2c_integrals 1 7.0 0.002 0.002 18.926 18.926 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.546 18.650 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.426 18.544 18.647 cp_fm_cholesky_decompose 12 8.2 17.477 18.093 17.477 18.093 cholesky_decomp 1 7.0 0.000 0.000 16.358 16.975 fft3d_s 5443 13.4 16.093 16.485 16.115 16.506 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.319 10.764 15.319 calculate_wavefunction 272 8.0 5.412 5.539 12.333 13.032 rpa_num_int 1 6.0 0.002 0.028 10.704 10.714 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.660 10.666 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.576 contract_S_to_Q 8 9.0 0.000 0.000 8.917 8.993 calc_potential_gpw 544 9.5 0.005 0.005 8.204 8.651 parallel_gemm_fm 14 9.1 0.000 0.000 8.505 8.628 parallel_gemm_fm_cosma 14 10.1 8.505 8.628 8.505 8.628 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.046 8.268 potential_pw2rs 545 10.0 0.106 0.108 7.467 8.084 create_integ_mat 1 6.0 0.013 0.028 7.841 7.850 collocate_single_gaussian 272 10.0 0.039 0.041 7.306 7.566 array2fm 1 7.0 0.000 0.000 6.837 7.264 pw_scatter_s 2720 13.7 4.341 4.558 4.341 4.558 pw_gather_s 2722 13.2 3.438 3.767 3.438 3.767 array2fm_buffer_send 1 8.0 3.017 3.190 3.017 3.190 pw_poisson_solve 545 10.5 1.119 1.172 2.193 2.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.993798, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2808.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 422.757 422.758 farming_run 1 2.0 421.918 421.923 422.720 422.721 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.246740E+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 1644041. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 217.040 217.040 qs_energies 1 2.0 0.000 0.000 216.795 216.807 scf_env_do_scf 1 3.0 0.000 0.000 114.866 114.866 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.921 113.927 rebuild_ks_matrix 4 6.0 0.000 0.000 113.920 113.926 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.919 113.926 hfx_ks_matrix 4 8.0 0.001 0.001 113.573 113.577 integrate_four_center 4 9.0 0.151 0.468 113.572 113.576 integrate_four_center_main 4 10.0 0.144 0.824 101.776 104.895 integrate_four_center_bin 264 11.0 101.631 104.860 101.631 104.860 mp2_main 1 3.0 0.000 0.000 101.643 101.655 mp2_gpw_main 1 4.0 0.049 0.071 100.761 100.776 init_scf_loop 1 4.0 0.000 0.000 96.893 96.894 mp2_ri_gpw_compute_in 1 5.0 0.065 0.089 73.736 74.876 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.769 54.908 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.158 41.487 46.678 integrate_v_rspace 95 8.0 0.397 0.564 27.966 32.981 pw_transfer 2240 10.6 0.147 0.159 29.423 29.884 ao_to_mo_and_store_B_mult_1 91 7.0 10.613 29.076 10.613 29.076 mp2_ri_gpw_compute_en 1 5.0 0.059 0.073 26.878 29.013 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.459 28.952 grid_integrate_task_list 95 9.0 23.387 28.566 23.387 28.566 fft_wrap_pw1pw2_100 730 12.4 1.233 1.392 26.227 26.740 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.893 25.072 25.083 get_2c_integrals 1 6.0 0.000 0.000 19.875 19.903 fft3d_s 1823 13.4 18.473 18.884 18.487 18.898 compute_2c_integrals 1 7.0 0.002 0.003 18.860 18.864 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.478 18.723 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.822 18.477 18.721 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.971 17.971 calculate_wavefunction 91 8.0 2.011 2.046 9.617 9.828 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.582 8.838 9.671 local_gemm 172 8.0 8.282 9.118 8.282 9.118 potential_pw2rs 186 10.0 0.033 0.034 8.370 8.956 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.520 8.031 8.467 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.063 8.431 calc_potential_gpw 182 9.5 0.002 0.002 7.858 8.046 collocate_single_gaussian 91 10.0 0.016 0.019 7.688 7.969 integrate_four_center_load 4 10.0 0.000 0.000 6.764 6.768 hfx_load_balance 1 11.0 0.000 0.000 6.764 6.768 mp_sync 37 10.5 3.765 6.672 3.765 6.672 mp_sendrecv_dm3 2068 8.0 6.068 6.513 6.068 6.513 mp2_ri_gpw_compute_en_ener 172 7.0 6.327 6.417 6.327 6.417 pw_gather_s 912 13.2 4.496 4.978 4.496 4.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.758627, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.070848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.025 53.927 53.928 qs_mol_dyn_low 1 2.0 0.003 0.003 53.634 53.641 qs_forces 11 3.9 0.004 0.006 53.503 53.504 qs_energies 11 4.9 0.003 0.005 51.995 52.013 scf_env_do_scf 11 5.9 0.000 0.001 45.349 45.350 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.311 43.312 dbcsr_multiply_generic 2286 12.5 0.101 0.116 33.379 33.889 qs_scf_new_mos 108 7.5 0.000 0.001 33.406 33.706 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.405 33.706 ot_scf_mini 108 9.5 0.002 0.002 31.693 31.894 multiply_cannon 2286 13.5 0.207 0.234 26.166 27.711 multiply_cannon_loop 2286 14.5 1.823 1.931 25.501 27.064 velocity_verlet 10 3.0 0.002 0.006 26.085 26.087 ot_mini 108 10.5 0.001 0.001 19.036 19.293 qs_ot_get_derivative 108 11.5 0.001 0.001 16.079 16.265 mp_waitall_1 245248 16.5 8.178 14.519 8.178 14.519 multiply_cannon_metrocomm3 54864 15.5 0.074 0.081 5.927 12.776 multiply_cannon_multrec 54864 15.5 3.653 5.671 7.792 11.160 qs_ot_get_p 119 10.4 0.001 0.001 7.994 8.295 rebuild_ks_matrix 119 8.3 0.000 0.000 7.794 7.981 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.794 7.980 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.850 7.021 mp_sum_l 7287 12.8 5.248 6.967 5.248 6.967 multiply_cannon_sync_h2d 54864 15.5 5.141 6.290 5.141 6.290 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.680 6.171 dbcsr_mm_accdrv_process 76910 16.1 1.841 2.898 4.052 5.775 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.254 5.369 init_scf_run 11 5.9 0.000 0.001 5.367 5.368 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.367 5.367 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.262 5.290 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.474 4.475 sum_up_and_integrate 119 10.3 0.001 0.001 4.396 4.402 integrate_v_rspace 119 11.3 0.002 0.002 4.386 4.392 cp_fm_diag_elpa 50 13.0 0.003 0.013 4.280 4.280 cp_fm_redistribute_end 50 14.0 2.177 4.243 2.183 4.246 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.062 4.170 calculate_rho_elec 119 8.7 0.011 0.017 4.061 4.170 cp_fm_diag_elpa_base 50 14.0 2.057 4.138 2.061 4.143 calculate_dm_sparse 119 9.5 0.000 0.000 3.051 3.211 calculate_first_density_matrix 1 7.0 0.000 0.000 3.175 3.182 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.930 3.142 apply_single 119 13.6 0.000 0.000 2.930 3.142 acc_transpose_blocks 54864 15.5 0.233 0.264 2.270 2.789 jit_kernel_multiply 13 15.8 2.148 2.788 2.148 2.788 ot_diis_step 108 11.5 0.006 0.006 2.720 2.720 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.612 2.720 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.562 2.564 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.445 2.525 density_rs2pw 119 9.7 0.004 0.004 2.090 2.194 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.070 2.124 wfi_extrapolate 11 7.9 0.001 0.001 2.121 2.121 grid_integrate_task_list 119 12.3 2.008 2.102 2.008 2.102 init_scf_loop 11 6.9 0.000 0.000 2.022 2.022 mp_sum_d 4135 12.0 1.275 1.941 1.275 1.941 potential_pw2rs 119 12.3 0.004 0.004 1.782 1.793 make_m2s 4572 13.5 0.053 0.055 1.643 1.694 pw_transfer 1439 11.6 0.052 0.057 1.618 1.691 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.543 1.618 make_images 4572 14.5 0.133 0.139 1.562 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.499 1.529 transfer_rs2pw 487 10.6 0.005 0.006 1.328 1.449 mp_alltoall_d11v 2130 13.8 1.280 1.436 1.280 1.436 acc_transpose_blocks_sync 164592 16.5 1.200 1.436 1.200 1.436 grid_collocate_task_list 119 9.7 1.354 1.423 1.354 1.423 mp_waitany 12084 13.8 1.239 1.401 1.239 1.401 fft3d_ps 1201 14.6 0.371 0.475 1.258 1.336 transfer_pw2rs 487 13.2 0.006 0.006 1.306 1.314 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.192 1.268 parallel_gemm_fm 81 9.0 0.000 0.000 1.238 1.243 parallel_gemm_fm_cosma 81 10.0 1.238 1.243 1.238 1.243 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.698 1.126 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.928000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 490.987520E+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 1308524. 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 0.046 0.136 39.130 39.135 qs_mol_dyn_low 1 2.0 0.015 0.038 38.742 38.751 qs_forces 11 3.9 0.006 0.030 38.643 38.644 qs_energies 11 4.9 0.034 0.094 36.953 36.959 scf_env_do_scf 11 5.9 0.001 0.007 31.705 31.735 scf_env_do_scf_inner_loop 108 6.5 0.016 0.025 29.054 29.054 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.915 22.248 qs_scf_new_mos 108 7.5 0.001 0.001 20.532 20.756 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.531 20.755 ot_scf_mini 108 9.5 0.003 0.003 19.616 19.783 multiply_cannon 2286 13.5 0.209 0.217 16.867 18.463 velocity_verlet 10 3.0 0.002 0.002 18.089 18.090 multiply_cannon_loop 2286 14.5 1.193 1.256 15.717 17.210 ot_mini 108 10.5 0.001 0.001 12.089 12.331 mp_waitall_1 200699 16.5 5.487 10.624 5.487 10.624 qs_ot_get_derivative 108 11.5 0.001 0.002 9.735 9.908 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.036 9.415 multiply_cannon_multrec 27432 15.5 1.838 4.247 6.149 8.994 rebuild_ks_matrix 119 8.3 0.000 0.000 6.891 7.034 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.022 6.890 7.034 dbcsr_mm_accdrv_process 47894 16.0 3.509 5.675 4.230 6.296 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.079 6.206 qs_ot_get_p 119 10.4 0.001 0.001 4.596 4.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.709 4.569 mp_sum_l 7287 12.8 2.113 4.186 2.113 4.186 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.999 4.060 apply_single 119 13.6 0.000 0.000 2.999 4.060 init_scf_run 11 5.9 0.000 0.001 3.907 3.908 scf_env_initial_rho_setup 11 6.9 0.002 0.003 3.907 3.908 sum_up_and_integrate 119 10.3 0.005 0.016 3.652 3.659 integrate_v_rspace 119 11.3 0.002 0.003 3.634 3.644 qs_rho_update_rho_low 119 7.7 0.002 0.016 3.597 3.625 calculate_rho_elec 119 8.7 0.023 0.040 3.595 3.624 qs_ot_p2m_diag 50 11.0 0.009 0.041 2.972 3.009 make_m2s 4572 13.5 0.051 0.053 2.548 2.879 make_images 4572 14.5 0.206 0.245 2.458 2.793 init_scf_loop 11 6.9 0.003 0.009 2.627 2.656 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.577 2.577 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.299 2.386 calculate_first_density_matrix 1 7.0 0.002 0.004 2.379 2.382 ot_diis_step 108 11.5 0.011 0.011 2.307 2.307 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.263 2.264 cp_fm_redistribute_end 50 14.0 1.148 2.239 1.151 2.242 cp_fm_diag_elpa_base 50 14.0 1.060 2.151 1.087 2.185 multiply_cannon_sync_h2d 27432 15.5 1.673 2.166 1.673 2.166 calculate_dm_sparse 119 9.5 0.000 0.001 2.053 2.130 density_rs2pw 119 9.7 0.004 0.004 1.863 1.966 acc_transpose_blocks 27432 15.5 0.116 0.122 1.622 1.964 jit_kernel_multiply 10 16.3 0.663 1.951 0.663 1.951 grid_integrate_task_list 119 12.3 1.838 1.932 1.838 1.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.897 1.899 pw_transfer 1439 11.6 0.065 0.070 1.815 1.848 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.723 1.759 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.662 1.704 prepare_preconditioner 11 7.9 0.000 0.000 1.515 1.541 make_preconditioner 11 8.9 0.014 0.042 1.515 1.541 make_images_data 4572 15.5 0.048 0.054 1.216 1.530 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.408 1.478 wfi_extrapolate 11 7.9 0.001 0.001 1.427 1.427 fft_wrap_pw1pw2_140 487 13.2 0.159 0.167 1.380 1.419 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.079 1.409 potential_pw2rs 119 12.3 0.006 0.006 1.397 1.407 fft3d_ps 1201 14.6 0.520 0.576 1.360 1.387 grid_collocate_task_list 119 9.7 1.293 1.350 1.293 1.350 mp_alltoall_d11v 2130 13.8 1.200 1.308 1.200 1.308 transfer_rs2pw 487 10.6 0.005 0.006 1.108 1.268 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.222 1.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.254 mp_sum_d 4135 12.0 0.699 1.213 0.699 1.213 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.024 1.027 mp_allgather_i34 2286 14.5 0.574 0.961 0.574 0.961 acc_transpose_blocks_sync 82296 16.5 0.819 0.949 0.819 0.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.871 0.884 transfer_pw2rs 487 13.2 0.005 0.005 0.868 0.874 acc_transpose_blocks_kernels 27432 16.5 0.190 0.279 0.661 0.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.135000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=467.090909, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 521.814016E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.177 34.819 34.824 qs_mol_dyn_low 1 2.0 0.003 0.003 33.407 33.541 qs_forces 11 3.9 0.002 0.003 33.338 33.339 qs_energies 11 4.9 0.003 0.009 31.761 31.763 scf_env_do_scf 11 5.9 0.001 0.002 26.607 26.608 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 23.969 23.970 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.411 17.501 qs_scf_new_mos 108 7.5 0.001 0.001 16.052 16.072 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.051 16.071 velocity_verlet 10 3.0 0.001 0.002 15.855 15.856 ot_scf_mini 108 9.5 0.002 0.002 15.275 15.292 multiply_cannon 2286 13.5 0.201 0.207 13.718 14.470 multiply_cannon_loop 2286 14.5 0.858 0.887 12.875 13.668 ot_mini 108 10.5 0.001 0.001 9.355 9.372 qs_ot_get_derivative 108 11.5 0.001 0.001 7.830 7.846 multiply_cannon_multrec 18288 15.5 1.883 2.884 6.966 7.285 rebuild_ks_matrix 119 8.3 0.000 0.000 6.156 6.180 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.156 6.180 dbcsr_mm_accdrv_process 38222 16.0 4.913 5.748 4.989 5.814 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.427 5.450 mp_waitall_1 158411 16.6 2.843 4.007 2.843 4.007 init_scf_run 11 5.9 0.000 0.001 3.881 3.881 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.880 3.881 sum_up_and_integrate 119 10.3 0.001 0.002 3.713 3.720 integrate_v_rspace 119 11.3 0.002 0.003 3.701 3.709 qs_ot_get_p 119 10.4 0.001 0.002 3.638 3.675 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.966 3.567 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.391 3.398 calculate_rho_elec 119 8.7 0.031 0.031 3.390 3.397 init_scf_loop 11 6.9 0.000 0.002 2.616 2.619 calculate_first_density_matrix 1 7.0 0.003 0.026 2.581 2.583 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.417 2.423 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.454 2.415 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.030 2.356 apply_single 119 13.6 0.000 0.000 2.029 2.355 make_m2s 4572 13.5 0.043 0.044 2.109 2.227 make_images 4572 14.5 0.192 0.206 2.023 2.140 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.106 2.107 density_rs2pw 119 9.7 0.004 0.004 1.950 2.096 pw_transfer 1439 11.6 0.065 0.068 1.938 1.956 grid_integrate_task_list 119 12.3 1.801 1.904 1.801 1.904 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.882 1.883 calculate_dm_sparse 119 9.5 0.000 0.001 1.857 1.865 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.845 1.865 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.815 1.824 cp_fm_diag_elpa_base 50 14.0 1.790 1.806 1.813 1.822 prepare_preconditioner 11 7.9 0.000 0.000 1.796 1.799 make_preconditioner 11 8.9 0.000 0.001 1.796 1.799 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.780 1.790 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.652 1.736 acc_transpose_blocks 18288 15.5 0.082 0.084 1.642 1.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.679 1.689 mp_sum_l 7287 12.8 1.274 1.642 1.274 1.642 potential_pw2rs 119 12.3 0.007 0.008 1.495 1.503 ot_diis_step 108 11.5 0.011 0.012 1.501 1.501 fft_wrap_pw1pw2_140 487 13.2 0.210 0.216 1.444 1.460 fft3d_ps 1201 14.6 0.528 0.548 1.416 1.436 grid_collocate_task_list 119 9.7 1.233 1.319 1.233 1.319 wfi_extrapolate 11 7.9 0.001 0.001 1.240 1.240 transfer_rs2pw 487 10.6 0.005 0.005 1.118 1.206 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.186 multiply_cannon_sync_h2d 18288 15.5 1.002 1.127 1.002 1.127 make_images_data 4572 15.5 0.048 0.052 0.908 1.118 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.022 1.022 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.801 1.021 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.985 1.008 transfer_pw2rs 487 13.2 0.004 0.004 0.915 0.922 mp_alltoall_d11v 2130 13.8 0.758 0.865 0.758 0.865 acc_transpose_blocks_sync 54864 16.5 0.750 0.819 0.750 0.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.813 mp_alltoall_z22v 1201 16.6 0.723 0.808 0.723 0.808 acc_transpose_blocks_kernels 18288 16.5 0.221 0.226 0.790 0.802 cp_fm_cholesky_invert 11 10.9 0.783 0.787 0.783 0.787 mp_allgather_i34 2286 14.5 0.275 0.781 0.275 0.781 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.681 0.741 create_qs_kind_set 1 2.0 0.000 0.000 0.623 0.697 read_qs_kind 2 3.0 0.047 0.179 0.623 0.697 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.824000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 558.350336E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.030 36.513 36.517 qs_mol_dyn_low 1 2.0 0.003 0.004 36.310 36.317 qs_forces 11 3.9 0.009 0.015 36.250 36.254 qs_energies 11 4.9 0.002 0.004 34.538 34.549 scf_env_do_scf 11 5.9 0.000 0.001 29.335 29.338 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.987 25.991 dbcsr_multiply_generic 2286 12.5 0.099 0.102 19.979 20.084 velocity_verlet 10 3.0 0.002 0.003 18.352 18.356 qs_scf_new_mos 108 7.5 0.001 0.001 17.951 18.004 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.951 18.003 ot_scf_mini 108 9.5 0.002 0.003 16.912 16.964 multiply_cannon 2286 13.5 0.219 0.227 16.278 16.687 multiply_cannon_loop 2286 14.5 1.561 1.610 15.324 15.650 ot_mini 108 10.5 0.001 0.001 10.422 10.482 multiply_cannon_multrec 27432 15.5 2.447 3.113 9.001 9.306 qs_ot_get_derivative 108 11.5 0.002 0.013 8.547 8.600 dbcsr_mm_accdrv_process 47916 15.9 6.035 7.756 6.450 7.819 rebuild_ks_matrix 119 8.3 0.000 0.000 6.264 6.312 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.263 6.312 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.544 5.586 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 3.873 3.873 qs_ot_get_p 119 10.4 0.001 0.001 3.635 3.713 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.128 3.528 sum_up_and_integrate 119 10.3 0.001 0.001 3.457 3.465 integrate_v_rspace 119 11.3 0.003 0.003 3.446 3.454 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.320 3.354 calculate_rho_elec 119 8.7 0.040 0.046 3.320 3.354 init_scf_loop 11 6.9 0.000 0.000 3.328 3.329 acc_transpose_blocks 27432 15.5 0.124 0.129 2.467 2.531 prepare_preconditioner 11 7.9 0.000 0.000 2.505 2.513 make_preconditioner 11 8.9 0.000 0.001 2.505 2.513 calculate_first_density_matrix 1 7.0 0.000 0.001 2.498 2.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.114 2.440 make_m2s 4572 13.5 0.053 0.056 2.298 2.399 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.340 2.368 mp_waitall_1 137007 16.6 1.825 2.290 1.825 2.290 make_images 4572 14.5 0.273 0.336 2.191 2.289 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.112 2.235 apply_single 119 13.6 0.000 0.000 2.112 2.235 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.214 2.222 calculate_dm_sparse 119 9.5 0.000 0.000 2.146 2.201 pw_transfer 1439 11.6 0.065 0.069 1.889 1.925 grid_integrate_task_list 119 12.3 1.836 1.917 1.836 1.917 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.870 1.871 density_rs2pw 119 9.7 0.004 0.004 1.755 1.856 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.797 1.836 ot_diis_step 108 11.5 0.012 0.013 1.834 1.834 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.809 1.810 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.696 1.709 acc_transpose_blocks_sync 82296 16.5 1.485 1.550 1.485 1.550 cp_fm_diag_elpa 50 13.0 0.011 0.015 1.539 1.547 cp_fm_diag_elpa_base 50 14.0 1.490 1.507 1.521 1.530 fft_wrap_pw1pw2_140 487 13.2 0.245 0.256 1.490 1.529 grid_collocate_task_list 119 9.7 1.249 1.351 1.249 1.351 fft3d_ps 1201 14.6 0.554 0.603 1.318 1.344 wfi_extrapolate 11 7.9 0.001 0.001 1.323 1.323 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.318 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.808 1.285 potential_pw2rs 119 12.3 0.008 0.009 1.260 1.263 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.236 1.246 mp_sum_l 7287 12.8 0.979 1.223 0.979 1.223 cp_fm_upper_to_full 72 14.2 0.819 1.165 0.819 1.165 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.069 1.075 jit_kernel_multiply 6 15.9 0.349 1.074 0.349 1.074 dbcsr_complete_redistribute 329 12.2 0.120 0.151 0.767 1.042 transfer_rs2pw 487 10.6 0.004 0.005 0.881 0.979 make_images_data 4572 15.5 0.048 0.052 0.851 0.964 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.752 0.916 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.877 mp_alltoall_d11v 2130 13.8 0.768 0.858 0.768 0.858 acc_transpose_blocks_kernels 27432 16.5 0.275 0.282 0.830 0.847 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.813 cp_fm_cholesky_invert 11 10.9 0.731 0.734 0.731 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.517000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.090909, yerr=3.476010 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 631.906304E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 28.700 28.701 qs_mol_dyn_low 1 2.0 0.003 0.003 28.491 28.499 qs_forces 11 3.9 0.002 0.003 28.251 28.252 qs_energies 11 4.9 0.002 0.002 26.564 26.567 scf_env_do_scf 11 5.9 0.000 0.001 21.654 21.655 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.258 19.260 velocity_verlet 10 3.0 0.002 0.002 14.561 14.576 dbcsr_multiply_generic 2286 12.5 0.093 0.099 12.992 13.120 qs_scf_new_mos 108 7.5 0.001 0.001 11.770 11.791 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.769 11.791 ot_scf_mini 108 9.5 0.002 0.002 11.066 11.092 multiply_cannon 2286 13.5 0.224 0.232 10.497 10.902 multiply_cannon_loop 2286 14.5 0.645 0.665 9.620 9.870 ot_mini 108 10.5 0.001 0.001 6.425 6.454 multiply_cannon_multrec 9144 15.5 1.681 1.916 5.939 6.186 rebuild_ks_matrix 119 8.3 0.000 0.000 5.662 5.681 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.662 5.681 qs_ot_get_derivative 108 11.5 0.001 0.001 5.147 5.172 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.037 5.054 dbcsr_mm_accdrv_process 12550 15.8 3.703 4.158 4.146 4.205 init_scf_run 11 5.9 0.000 0.001 3.481 3.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.480 3.481 sum_up_and_integrate 119 10.3 0.001 0.001 3.372 3.377 integrate_v_rspace 119 11.3 0.003 0.003 3.362 3.367 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.274 3.279 calculate_rho_elec 119 8.7 0.060 0.061 3.273 3.279 qs_ot_get_p 119 10.4 0.001 0.001 2.778 2.816 init_scf_loop 11 6.9 0.000 0.000 2.377 2.378 calculate_first_density_matrix 1 7.0 0.000 0.000 2.369 2.371 mp_waitall_1 115863 16.7 1.417 2.023 1.417 2.023 pw_transfer 1439 11.6 0.066 0.068 1.948 1.958 grid_integrate_task_list 119 12.3 1.849 1.947 1.849 1.947 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.854 1.866 make_m2s 4572 13.5 0.033 0.035 1.726 1.846 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.811 1.814 make_images 4572 14.5 0.269 0.300 1.637 1.756 calculate_dm_sparse 119 9.5 0.000 0.000 1.732 1.752 density_rs2pw 119 9.7 0.003 0.004 1.660 1.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.683 1.706 prepare_preconditioner 11 7.9 0.000 0.000 1.665 1.669 make_preconditioner 11 8.9 0.001 0.001 1.664 1.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.560 1.583 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.569 1.569 fft_wrap_pw1pw2_140 487 13.2 0.321 0.329 1.539 1.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.515 1.516 acc_transpose_blocks 9144 15.5 0.043 0.044 1.441 1.476 grid_collocate_task_list 119 9.7 1.298 1.388 1.298 1.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.367 1.374 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.329 1.341 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.294 1.302 cp_fm_diag_elpa_base 50 14.0 1.267 1.282 1.292 1.300 fft3d_ps 1201 14.6 0.561 0.571 1.269 1.276 ot_diis_step 108 11.5 0.013 0.013 1.265 1.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.215 1.255 apply_single 119 13.6 0.000 0.000 1.215 1.255 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.221 1.222 potential_pw2rs 119 12.3 0.010 0.011 1.194 1.196 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.123 1.128 wfi_extrapolate 11 7.9 0.001 0.001 1.059 1.059 jit_kernel_multiply 6 15.7 0.403 1.021 0.403 1.021 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.925 mp_alltoall_d11v 2130 13.8 0.767 0.918 0.767 0.918 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.738 0.906 make_images_data 4572 15.5 0.042 0.046 0.728 0.860 cp_fm_cholesky_invert 11 10.9 0.831 0.834 0.831 0.834 transfer_rs2pw 487 10.6 0.004 0.005 0.750 0.823 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.811 0.817 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.450 0.770 acc_transpose_blocks_sync 27432 16.5 0.725 0.758 0.725 0.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.746 0.749 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.670 0.721 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 0.659 0.664 transfer_pw2rs 487 13.2 0.003 0.004 0.630 0.632 mp_alltoall_z22v 1201 16.6 0.581 0.617 0.581 0.617 parallel_gemm_fm 81 9.0 0.000 0.000 0.608 0.609 parallel_gemm_fm_cosma 81 10.0 0.607 0.608 0.607 0.608 mp_allgather_i34 2286 14.5 0.195 0.586 0.195 0.586 make_basis_sm 11 9.8 0.000 0.000 0.575 0.576 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.701000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=596.090909, yerr=9.708989 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 781.066240E+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 0.066 0.085 42.308 42.313 qs_mol_dyn_low 1 2.0 0.003 0.005 41.936 41.996 qs_forces 11 3.9 0.003 0.004 41.861 41.861 qs_energies 11 4.9 0.002 0.002 39.903 39.905 scf_env_do_scf 11 5.9 0.001 0.001 34.187 34.187 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.528 26.529 velocity_verlet 10 3.0 0.002 0.003 23.565 23.571 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.096 19.251 qs_scf_new_mos 108 7.5 0.001 0.001 17.323 17.425 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.322 17.424 ot_scf_mini 108 9.5 0.002 0.002 16.163 16.264 multiply_cannon 2286 13.5 0.299 0.303 15.195 16.131 multiply_cannon_loop 2286 14.5 0.878 0.917 13.955 14.918 ot_mini 108 10.5 0.001 0.001 9.936 10.048 multiply_cannon_multrec 9144 15.5 3.356 4.750 8.879 9.105 qs_ot_get_derivative 108 11.5 0.001 0.001 7.914 8.015 init_scf_loop 11 6.9 0.000 0.000 7.634 7.636 rebuild_ks_matrix 119 8.3 0.000 0.000 6.973 7.110 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.972 7.109 dbcsr_mm_accdrv_process 12550 15.8 4.547 6.055 5.381 6.780 prepare_preconditioner 11 7.9 0.000 0.000 6.702 6.715 make_preconditioner 11 8.9 0.001 0.001 6.702 6.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.299 6.598 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.289 6.413 cp_fm_upper_to_full 72 14.2 3.184 4.531 3.184 4.531 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.929 3.935 calculate_rho_elec 119 8.7 0.118 0.120 3.928 3.935 sum_up_and_integrate 119 10.3 0.001 0.001 3.694 3.703 integrate_v_rspace 119 11.3 0.003 0.004 3.684 3.693 init_scf_run 11 5.9 0.000 0.001 3.672 3.672 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.672 3.672 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.966 3.453 qs_ot_get_p 119 10.4 0.001 0.001 3.310 3.444 mp_waitall_1 94719 16.7 2.238 3.376 2.238 3.376 dbcsr_complete_redistribute 329 12.2 0.283 0.291 1.909 2.726 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.270 2.714 apply_single 119 13.6 0.000 0.000 2.270 2.714 pw_transfer 1439 11.6 0.069 0.069 2.525 2.532 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.388 2.453 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.426 2.434 make_m2s 4572 13.5 0.037 0.037 2.273 2.419 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.617 2.414 make_images 4572 14.5 0.356 0.390 2.152 2.298 mp_alltoall_i22 627 13.8 1.443 2.273 1.443 2.273 calculate_first_density_matrix 1 7.0 0.000 0.000 2.246 2.248 calculate_dm_sparse 119 9.5 0.000 0.000 2.231 2.245 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.397 2.186 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.058 2.108 grid_integrate_task_list 119 12.3 2.069 2.106 2.069 2.106 fft_wrap_pw1pw2_140 487 13.2 0.575 0.580 2.061 2.070 ot_diis_step 108 11.5 0.014 0.014 2.000 2.000 density_rs2pw 119 9.7 0.003 0.003 1.973 1.992 qs_ot_p2m_diag 50 11.0 0.044 0.045 1.940 1.943 mp_sum_l 7287 12.8 1.102 1.919 1.102 1.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.910 1.910 acc_transpose_blocks 9144 15.5 0.045 0.047 1.840 1.905 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.771 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.604 1.633 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.581 1.582 grid_collocate_task_list 119 9.7 1.513 1.533 1.513 1.533 fft3d_ps 1201 14.6 0.598 0.609 1.522 1.530 cp_fm_cholesky_invert 11 10.9 1.368 1.372 1.368 1.372 wfi_extrapolate 11 7.9 0.001 0.001 1.361 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.328 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.305 1.305 cp_fm_diag_elpa_base 50 14.0 1.160 1.212 1.303 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.229 1.235 potential_pw2rs 119 12.3 0.014 0.014 1.222 1.223 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.018 1.219 acc_transpose_blocks_sync 27432 16.5 1.129 1.190 1.129 1.190 mp_alltoall_d11v 2130 13.8 1.110 1.135 1.110 1.135 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.102 1.119 make_images_data 4572 15.5 0.047 0.051 0.940 1.115 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.993 1.014 jit_kernel_multiply 6 15.5 0.805 1.010 0.805 1.010 qs_create_task_list 11 7.9 0.001 0.002 0.929 0.941 generate_qs_task_list 11 8.9 0.368 0.388 0.928 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.313000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.181818, yerr=17.245912 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.489088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.084 80.302 80.323 qs_mol_dyn_low 1 2.0 0.003 0.004 79.970 80.029 qs_forces 11 3.9 0.003 0.003 79.901 79.902 qs_energies 11 4.9 0.002 0.002 77.065 77.087 scf_env_do_scf 11 5.9 0.000 0.001 68.359 68.362 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 63.027 63.027 dbcsr_multiply_generic 2055 12.4 0.106 0.109 50.322 50.660 qs_scf_new_mos 99 7.5 0.000 0.001 46.277 46.388 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.276 46.387 ot_scf_mini 99 9.5 0.002 0.002 43.911 44.006 multiply_cannon 2055 13.4 0.187 0.193 42.140 43.181 multiply_cannon_loop 2055 14.4 1.784 1.819 41.224 42.289 velocity_verlet 10 3.0 0.003 0.006 42.205 42.206 ot_mini 99 10.5 0.001 0.001 26.117 26.237 qs_ot_get_derivative 99 11.5 0.001 0.001 19.325 19.428 multiply_cannon_multrec 49320 15.4 11.381 12.057 17.400 18.104 rebuild_ks_matrix 110 8.3 0.000 0.001 14.366 14.553 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.366 14.553 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.599 12.771 mp_waitall_1 220248 16.4 10.302 11.181 10.302 11.181 multiply_cannon_sync_h2d 49320 15.4 9.610 10.153 9.610 10.153 qs_ot_get_p 110 10.4 0.001 0.001 9.387 9.515 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.346 7.914 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.246 7.699 apply_single 110 13.6 0.000 0.001 7.246 7.698 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 6.314 7.433 sum_up_and_integrate 110 10.3 0.002 0.003 6.882 6.895 integrate_v_rspace 110 11.3 0.003 0.003 6.857 6.876 init_scf_run 11 5.9 0.000 0.001 6.695 6.696 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.695 6.696 ot_diis_step 99 11.5 0.006 0.006 6.586 6.586 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.369 6.398 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.138 6.275 calculate_rho_elec 110 8.6 0.020 0.024 6.137 6.275 dbcsr_mm_accdrv_process 87628 16.1 3.072 3.140 5.887 6.136 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.519 5.519 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.268 5.318 init_scf_loop 11 6.9 0.000 0.000 5.307 5.307 cp_fm_diag_elpa 48 13.0 0.000 0.001 4.965 4.986 cp_fm_diag_elpa_base 48 14.0 4.951 4.972 4.962 4.984 mp_sum_l 6594 12.7 3.897 4.812 3.897 4.812 wfi_extrapolate 11 7.9 0.001 0.001 3.995 3.995 make_m2s 4110 13.4 0.060 0.065 3.849 3.948 make_images 4110 14.4 0.177 0.190 3.754 3.857 calculate_dm_sparse 110 9.5 0.001 0.001 3.730 3.824 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.520 3.524 grid_integrate_task_list 110 12.3 3.259 3.403 3.259 3.403 density_rs2pw 110 9.6 0.004 0.005 3.213 3.362 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.228 3.349 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.176 3.237 prepare_preconditioner 11 7.9 0.000 0.000 3.154 3.181 make_preconditioner 11 8.9 0.000 0.001 3.154 3.181 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.108 3.154 pw_transfer 1331 11.6 0.055 0.065 2.977 3.052 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.945 2.981 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.888 2.967 acc_transpose_blocks 49320 15.4 0.216 0.222 2.563 2.615 calculate_first_density_matrix 1 7.0 0.000 0.001 2.603 2.607 fft_wrap_pw1pw2_140 451 13.1 0.390 0.433 2.490 2.577 jit_kernel_multiply 13 15.9 2.534 2.558 2.534 2.558 potential_pw2rs 110 12.3 0.005 0.006 2.488 2.501 mp_alltoall_d11v 2046 13.8 1.969 2.449 1.969 2.449 grid_collocate_task_list 110 9.6 2.155 2.271 2.155 2.271 fft3d_ps 1111 14.6 0.797 0.884 2.184 2.235 transfer_rs2pw 451 10.6 0.005 0.006 1.805 1.963 mp_waitany 14300 13.8 1.796 1.949 1.796 1.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.895 1.910 make_images_data 4110 15.4 0.043 0.047 1.758 1.885 mp_sum_d 3889 11.9 1.303 1.858 1.303 1.858 hybrid_alltoall_any 4261 16.3 0.085 0.480 1.517 1.776 cp_fm_cholesky_invert 11 10.9 1.722 1.726 1.722 1.726 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.649 1.690 transfer_pw2rs 451 13.1 0.006 0.007 1.682 1.689 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.323000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=3.117532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 591.912960E+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 3075086. 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 0.048 0.156 68.970 68.977 qs_mol_dyn_low 1 2.0 0.003 0.006 68.300 68.514 qs_forces 11 3.9 0.003 0.005 68.181 68.182 qs_energies 11 4.9 0.002 0.004 64.877 64.880 scf_env_do_scf 11 5.9 0.000 0.001 55.993 55.997 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 48.492 48.494 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.712 37.889 velocity_verlet 10 3.0 0.001 0.002 35.661 35.663 qs_scf_new_mos 99 7.5 0.001 0.001 32.817 32.985 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.816 32.984 multiply_cannon 2055 13.4 0.226 0.244 31.226 32.244 ot_scf_mini 99 9.5 0.003 0.004 31.150 31.299 multiply_cannon_loop 2055 14.4 1.167 1.189 30.123 31.214 ot_mini 99 10.5 0.004 0.025 18.229 18.395 multiply_cannon_multrec 24660 15.4 7.052 8.475 14.125 15.460 rebuild_ks_matrix 110 8.3 0.000 0.000 13.500 13.623 qs_ks_build_kohn_sham_matrix 110 9.3 0.022 0.093 13.499 13.622 qs_ot_get_derivative 99 11.5 0.001 0.002 12.471 12.625 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.884 11.993 mp_waitall_1 176588 16.5 7.315 9.734 7.315 9.734 multiply_cannon_metrocomm3 24660 15.4 0.072 0.075 5.020 7.746 multiply_cannon_sync_h2d 24660 15.4 6.492 7.541 6.492 7.541 init_scf_loop 11 6.9 0.000 0.000 7.468 7.482 dbcsr_mm_accdrv_process 52282 16.1 5.514 6.453 6.899 7.211 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.443 7.097 apply_single 110 13.6 0.000 0.001 6.442 7.097 init_scf_run 11 5.9 0.000 0.001 6.510 6.510 scf_env_initial_rho_setup 11 6.9 0.009 0.067 6.509 6.510 qs_ot_get_p 110 10.4 0.001 0.003 6.297 6.485 sum_up_and_integrate 110 10.3 0.002 0.003 6.275 6.286 integrate_v_rspace 110 11.3 0.002 0.003 6.248 6.261 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.851 5.868 calculate_rho_elec 110 8.6 0.039 0.047 5.850 5.867 ot_diis_step 99 11.5 0.010 0.013 5.709 5.709 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.709 5.429 prepare_preconditioner 11 7.9 0.000 0.000 5.379 5.400 make_preconditioner 11 8.9 0.000 0.001 5.379 5.400 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.976 5.133 qs_ot_p2m_diag 48 11.0 0.030 0.044 4.441 4.462 make_m2s 4110 13.4 0.056 0.059 4.136 4.453 make_images 4110 14.4 0.405 0.467 4.024 4.338 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.981 3.983 pw_transfer 1331 11.6 0.066 0.074 3.476 3.611 wfi_extrapolate 11 7.9 0.009 0.069 3.529 3.529 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.369 3.506 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.421 3.432 cp_fm_diag_elpa_base 48 14.0 3.375 3.388 3.418 3.429 grid_integrate_task_list 110 12.3 3.162 3.335 3.162 3.335 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.239 3.327 density_rs2pw 110 9.6 0.004 0.004 3.050 3.304 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.227 3.230 fft_wrap_pw1pw2_140 451 13.1 0.458 0.473 2.909 3.050 calculate_dm_sparse 110 9.5 0.001 0.001 2.970 2.995 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.900 2.938 calculate_first_density_matrix 1 7.0 0.000 0.001 2.823 2.827 fft3d_ps 1111 14.6 1.112 1.340 2.467 2.617 make_images_data 4110 15.4 0.049 0.054 2.255 2.591 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.962 2.529 cp_fm_cholesky_invert 11 10.9 2.499 2.506 2.499 2.506 mp_sum_l 6594 12.7 1.733 2.504 1.733 2.504 grid_collocate_task_list 110 9.6 2.169 2.307 2.169 2.307 potential_pw2rs 110 12.3 0.008 0.009 2.170 2.178 mp_alltoall_d11v 2046 13.8 1.738 2.050 1.738 2.050 acc_transpose_blocks 24660 15.4 0.118 0.122 1.985 2.010 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.954 1.980 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.868 1.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.772 1.786 jit_kernel_multiply 9 16.3 1.016 1.772 1.016 1.772 multiply_cannon_metrocomm4 22605 15.4 0.078 0.084 0.757 1.763 transfer_rs2pw 451 10.6 0.006 0.007 1.405 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.545 mp_irecv_dv 57340 16.2 0.626 1.534 0.626 1.534 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.491 mp_allgather_i34 2055 14.4 0.413 1.420 0.413 1.420 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.977000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.000000, yerr=6.888726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.165248E+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 0.017 0.065 59.684 59.706 qs_mol_dyn_low 1 2.0 0.003 0.004 59.412 59.424 qs_forces 11 3.9 0.003 0.004 59.187 59.190 qs_energies 11 4.9 0.002 0.002 56.027 56.032 scf_env_do_scf 11 5.9 0.000 0.001 47.835 47.836 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.416 39.417 velocity_verlet 10 3.0 0.001 0.003 32.350 32.356 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.860 29.153 qs_scf_new_mos 99 7.5 0.001 0.001 24.959 25.087 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.958 25.086 ot_scf_mini 99 9.5 0.002 0.002 23.710 23.859 multiply_cannon 2055 13.4 0.215 0.222 22.729 23.844 multiply_cannon_loop 2055 14.4 0.812 0.849 21.546 22.732 ot_mini 99 10.5 0.001 0.001 13.663 13.813 rebuild_ks_matrix 110 8.3 0.000 0.000 12.048 12.232 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.047 12.231 multiply_cannon_multrec 16440 15.4 3.589 4.982 9.765 11.103 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.602 10.771 mp_waitall_1 139946 16.5 6.994 10.212 6.994 10.212 qs_ot_get_derivative 99 11.5 0.001 0.001 9.161 9.309 init_scf_loop 11 6.9 0.000 0.000 8.387 8.387 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.492 7.417 prepare_preconditioner 11 7.9 0.000 0.000 6.643 6.658 make_preconditioner 11 8.9 0.000 0.001 6.643 6.658 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.957 6.320 sum_up_and_integrate 110 10.3 0.001 0.002 6.197 6.222 integrate_v_rspace 110 11.3 0.003 0.003 6.171 6.197 dbcsr_mm_accdrv_process 34862 16.1 5.247 5.556 6.020 6.174 init_scf_run 11 5.9 0.000 0.001 5.817 5.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.817 5.817 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.642 5.653 calculate_rho_elec 110 8.6 0.059 0.059 5.641 5.653 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.995 5.508 apply_single 110 13.6 0.000 0.000 4.995 5.508 qs_ot_get_p 110 10.4 0.001 0.001 5.222 5.408 make_m2s 4110 13.4 0.048 0.051 4.105 4.518 ot_diis_step 99 11.5 0.011 0.011 4.473 4.473 make_images 4110 14.4 0.397 0.518 3.989 4.401 multiply_cannon_sync_h2d 16440 15.4 3.232 3.892 3.232 3.892 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.143 3.761 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.630 3.633 pw_transfer 1331 11.6 0.066 0.073 3.516 3.531 grid_integrate_task_list 110 12.3 3.196 3.443 3.196 3.443 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.409 3.425 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.265 3.265 density_rs2pw 110 9.6 0.004 0.004 2.893 3.071 fft_wrap_pw1pw2_140 451 13.1 0.577 0.582 2.965 2.985 wfi_extrapolate 11 7.9 0.001 0.001 2.938 2.938 make_images_data 4110 15.4 0.046 0.050 2.324 2.817 calculate_first_density_matrix 1 7.0 0.000 0.000 2.790 2.792 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.772 2.774 calculate_dm_sparse 110 9.5 0.001 0.001 2.732 2.765 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.747 2.756 cp_fm_diag_elpa_base 48 14.0 2.682 2.713 2.745 2.754 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.050 2.733 cp_fm_cholesky_invert 11 10.9 2.574 2.579 2.574 2.579 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.918 2.522 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.446 2.501 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.406 2.476 mp_irecv_dv 48980 15.7 0.840 2.385 0.840 2.385 grid_collocate_task_list 110 9.6 2.223 2.376 2.223 2.376 fft3d_ps 1111 14.6 1.091 1.102 2.353 2.372 potential_pw2rs 110 12.3 0.011 0.011 2.052 2.066 mp_alltoall_d11v 2046 13.8 1.719 2.004 1.719 2.004 mp_sum_l 6594 12.7 1.410 1.995 1.410 1.995 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.947 1.950 dbcsr_complete_redistribute 325 12.2 0.323 0.368 1.441 1.891 acc_transpose_blocks 16440 15.4 0.078 0.080 1.723 1.837 cp_fm_upper_to_full 70 14.2 1.399 1.784 1.399 1.784 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.668 1.682 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.651 mp_allgather_i34 2055 14.4 0.495 1.558 0.495 1.558 cp_fm_cholesky_decompose 22 10.9 1.520 1.540 1.520 1.540 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.476 transfer_rs2pw 451 10.6 0.005 0.006 1.217 1.423 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.967 1.415 mp_waitany 17072 13.8 1.095 1.343 1.095 1.343 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.270 1.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.706000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.454545, yerr=8.228152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 739.020800E+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 0.033 0.067 65.610 65.623 qs_mol_dyn_low 1 2.0 0.003 0.003 65.360 65.370 qs_forces 11 3.9 0.003 0.003 65.297 65.298 qs_energies 11 4.9 0.003 0.003 61.927 61.930 scf_env_do_scf 11 5.9 0.000 0.001 53.570 53.572 scf_env_do_scf_inner_loop 99 6.5 0.026 0.036 42.075 42.075 velocity_verlet 10 3.0 0.001 0.002 36.883 36.886 dbcsr_multiply_generic 2055 12.4 0.114 0.120 30.890 31.114 qs_scf_new_mos 99 7.5 0.001 0.001 27.583 27.680 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.582 27.680 ot_scf_mini 99 9.5 0.002 0.003 25.896 25.992 multiply_cannon 2055 13.4 0.240 0.256 23.822 25.170 multiply_cannon_loop 2055 14.4 1.442 1.480 22.475 23.074 ot_mini 99 10.5 0.001 0.001 14.945 15.069 multiply_cannon_multrec 24660 15.4 4.129 6.855 13.061 14.242 rebuild_ks_matrix 110 8.3 0.000 0.000 11.899 12.003 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.899 12.002 init_scf_loop 11 6.9 0.000 0.000 11.457 11.458 qs_ot_get_derivative 99 11.5 0.001 0.001 10.799 10.899 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.512 10.604 dbcsr_mm_accdrv_process 52304 16.0 7.734 8.806 8.772 9.765 prepare_preconditioner 11 7.9 0.000 0.000 9.722 9.735 make_preconditioner 11 8.9 0.000 0.001 9.722 9.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.922 9.408 sum_up_and_integrate 110 10.3 0.002 0.003 6.178 6.198 mp_waitall_1 121746 16.5 4.365 6.181 4.365 6.181 integrate_v_rspace 110 11.3 0.003 0.003 6.153 6.171 qs_ot_get_p 110 10.4 0.001 0.001 5.808 5.951 make_m2s 4110 13.4 0.059 0.061 5.472 5.768 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.753 5.759 calculate_rho_elec 110 8.6 0.078 0.081 5.752 5.759 init_scf_run 11 5.9 0.000 0.001 5.650 5.650 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.649 5.650 make_images 4110 14.4 0.574 0.699 5.331 5.620 cp_fm_upper_to_full 70 14.2 3.362 4.891 3.362 4.891 ot_diis_step 99 11.5 0.011 0.011 4.108 4.108 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.037 4.089 apply_single 110 13.6 0.000 0.000 4.037 4.089 qs_ot_p2m_diag 48 11.0 0.055 0.065 3.936 3.951 dbcsr_complete_redistribute 325 12.2 0.418 0.462 2.641 3.793 pw_transfer 1331 11.6 0.065 0.074 3.595 3.638 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.487 3.542 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.489 3.535 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.361 3.416 grid_integrate_task_list 110 12.3 3.280 3.402 3.280 3.402 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.373 3.374 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.139 3.280 calculate_dm_sparse 110 9.5 0.001 0.001 3.132 3.159 make_images_data 4110 15.4 0.049 0.053 2.745 3.068 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.513 3.059 fft_wrap_pw1pw2_140 451 13.1 0.608 0.629 3.006 3.057 hybrid_alltoall_any 4261 16.3 0.122 0.456 2.278 2.997 density_rs2pw 110 9.6 0.004 0.004 2.869 2.975 wfi_extrapolate 11 7.9 0.001 0.001 2.959 2.959 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.790 2.916 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.844 2.854 cp_fm_diag_elpa_base 48 14.0 2.695 2.748 2.841 2.851 mp_alltoall_i22 605 13.7 1.671 2.839 1.671 2.839 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.789 2.791 acc_transpose_blocks 24660 15.4 0.118 0.119 2.545 2.656 calculate_first_density_matrix 1 7.0 0.000 0.001 2.596 2.599 cp_fm_cholesky_invert 11 10.9 2.569 2.579 2.569 2.579 multiply_cannon_sync_h2d 24660 15.4 2.379 2.531 2.379 2.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.476 fft3d_ps 1111 14.6 1.083 1.114 2.373 2.396 grid_collocate_task_list 110 9.6 2.267 2.368 2.267 2.368 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.220 2.220 potential_pw2rs 110 12.3 0.012 0.013 1.985 2.002 mp_alltoall_d11v 2046 13.8 1.760 1.968 1.760 1.968 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.803 1.838 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.710 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.685 1.700 cp_fm_cholesky_decompose 22 10.9 1.601 1.644 1.601 1.644 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.636 mp_allgather_i34 2055 14.4 0.432 1.612 0.432 1.612 jit_kernel_multiply 9 15.7 0.687 1.588 0.687 1.588 acc_transpose_blocks_sync 73980 16.4 1.454 1.578 1.454 1.578 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.836 1.529 mp_sum_l 6594 12.7 0.909 1.507 0.909 1.507 mp_irecv_dv 62702 16.1 0.729 1.447 0.729 1.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.623000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.818182, yerr=10.381961 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 849.920000E+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 0.010 0.028 55.329 55.330 qs_mol_dyn_low 1 2.0 0.003 0.003 55.096 55.106 qs_forces 11 3.9 0.003 0.005 54.987 54.989 qs_energies 11 4.9 0.002 0.002 51.349 51.353 scf_env_do_scf 11 5.9 0.000 0.001 43.003 43.004 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.156 35.158 velocity_verlet 10 3.0 0.002 0.002 31.215 31.221 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.348 23.513 qs_scf_new_mos 99 7.5 0.001 0.001 20.698 20.738 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.697 20.738 ot_scf_mini 99 9.5 0.002 0.002 19.432 19.452 multiply_cannon 2055 13.4 0.238 0.252 17.674 18.899 multiply_cannon_loop 2055 14.4 0.605 0.626 16.394 16.657 rebuild_ks_matrix 110 8.3 0.000 0.000 11.578 11.604 qs_ks_build_kohn_sham_matrix 110 9.3 0.026 0.033 11.578 11.604 ot_mini 99 10.5 0.001 0.001 10.780 10.793 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.302 10.324 multiply_cannon_multrec 8220 15.4 3.150 4.381 7.600 8.546 init_scf_loop 11 6.9 0.000 0.000 7.800 7.802 mp_waitall_1 103326 16.6 5.840 7.546 5.840 7.546 qs_ot_get_derivative 99 11.5 0.001 0.001 7.097 7.117 sum_up_and_integrate 110 10.3 0.001 0.002 6.190 6.203 integrate_v_rspace 110 11.3 0.003 0.003 6.163 6.177 prepare_preconditioner 11 7.9 0.000 0.000 6.149 6.156 make_preconditioner 11 8.9 0.006 0.019 6.149 6.156 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.914 5.930 calculate_rho_elec 110 8.6 0.113 0.114 5.914 5.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.730 5.811 dbcsr_mm_accdrv_process 17442 15.9 3.097 3.834 4.310 5.226 init_scf_run 11 5.9 0.000 0.001 5.214 5.214 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.213 5.214 qs_ot_get_p 110 10.4 0.001 0.001 4.665 4.684 make_m2s 4110 13.4 0.037 0.039 4.331 4.617 make_images 4110 14.4 0.643 0.704 4.202 4.487 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.943 4.298 pw_transfer 1331 11.6 0.066 0.072 3.870 3.891 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.762 3.786 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.718 3.766 apply_single 110 13.6 0.000 0.000 3.717 3.766 ot_diis_step 99 11.5 0.012 0.012 3.660 3.660 grid_integrate_task_list 110 12.3 3.385 3.511 3.385 3.511 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.302 3.306 fft_wrap_pw1pw2_140 451 13.1 0.773 0.787 3.256 3.292 density_rs2pw 110 9.6 0.004 0.004 2.914 3.083 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.960 2.961 cp_fm_cholesky_invert 11 10.9 2.943 2.947 2.943 2.947 hybrid_alltoall_any 4261 16.3 0.199 0.849 2.282 2.801 wfi_extrapolate 11 7.9 0.001 0.001 2.720 2.720 make_images_data 4110 15.4 0.041 0.046 2.307 2.704 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.665 2.665 calculate_dm_sparse 110 9.5 0.001 0.001 2.530 2.570 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.562 2.564 grid_collocate_task_list 110 9.6 2.366 2.561 2.366 2.561 multiply_cannon_sync_h2d 8220 15.4 2.369 2.470 2.369 2.470 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.444 2.451 cp_fm_diag_elpa_base 48 14.0 2.386 2.414 2.441 2.449 fft3d_ps 1111 14.6 1.137 1.179 2.433 2.441 calculate_first_density_matrix 1 7.0 0.000 0.000 2.398 2.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.206 2.216 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.060 2.085 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.782 1.999 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.971 1.981 potential_pw2rs 110 12.3 0.015 0.015 1.976 1.980 mp_alltoall_d11v 2046 13.8 1.675 1.827 1.675 1.827 cp_fm_cholesky_decompose 22 10.9 1.734 1.758 1.734 1.758 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.496 1.613 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.606 1.609 mp_allgather_i34 2055 14.4 0.453 1.570 0.453 1.570 dbcsr_complete_redistribute 325 12.2 0.546 0.560 1.416 1.509 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.480 acc_transpose_blocks 8220 15.4 0.040 0.042 1.407 1.428 qs_create_task_list 11 7.9 0.000 0.001 1.211 1.310 generate_qs_task_list 11 8.9 0.376 0.444 1.210 1.309 transfer_rs2pw 451 10.6 0.005 0.005 1.101 1.287 mp_waitany 9240 13.8 1.056 1.251 1.056 1.251 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.161 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.330000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=802.636364, yerr=10.002479 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 1.365385E+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 0.036 0.084 85.889 85.906 qs_mol_dyn_low 1 2.0 0.003 0.004 85.551 85.566 qs_forces 11 3.9 0.005 0.012 85.472 85.477 qs_energies 11 4.9 0.002 0.002 81.398 81.406 scf_env_do_scf 11 5.9 0.000 0.001 71.462 71.462 velocity_verlet 10 3.0 0.009 0.017 54.994 55.000 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.452 43.453 dbcsr_multiply_generic 2055 12.4 0.120 0.125 29.265 29.361 init_scf_loop 11 6.9 0.000 0.000 27.941 27.944 qs_scf_new_mos 99 7.5 0.001 0.001 26.417 26.453 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.416 26.452 prepare_preconditioner 11 7.9 0.000 0.000 25.969 25.981 make_preconditioner 11 8.9 0.005 0.015 25.969 25.981 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.198 25.448 ot_scf_mini 99 9.5 0.002 0.002 24.629 24.657 multiply_cannon 2055 13.4 0.335 0.366 22.341 23.129 multiply_cannon_loop 2055 14.4 0.819 0.837 20.570 20.843 cp_fm_upper_to_full 70 14.2 12.693 18.188 12.693 18.188 ot_mini 99 10.5 0.001 0.001 13.958 13.984 rebuild_ks_matrix 110 8.3 0.000 0.001 13.415 13.452 qs_ks_build_kohn_sham_matrix 110 9.3 0.023 0.032 13.415 13.451 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.174 12.208 dbcsr_complete_redistribute 325 12.2 1.002 1.015 7.281 10.509 multiply_cannon_multrec 8220 15.4 4.044 4.204 9.761 9.855 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.298 9.528 qs_ot_get_derivative 99 11.5 0.001 0.001 9.401 9.428 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.751 8.957 mp_alltoall_i22 605 13.7 5.390 8.598 5.390 8.598 mp_waitall_1 84994 16.7 7.331 8.116 7.331 8.116 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.971 7.012 calculate_rho_elec 110 8.6 0.222 0.223 6.971 7.011 sum_up_and_integrate 110 10.3 0.002 0.002 6.625 6.640 integrate_v_rspace 110 11.3 0.004 0.004 6.597 6.613 dbcsr_mm_accdrv_process 11614 15.7 3.949 4.547 5.568 5.769 make_m2s 4110 13.4 0.043 0.043 5.245 5.735 init_scf_run 11 5.9 0.000 0.001 5.671 5.671 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.670 5.670 make_images 4110 14.4 0.880 0.934 5.055 5.544 qs_ot_get_p 110 10.4 0.001 0.001 5.406 5.439 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.708 5.171 apply_single 110 13.6 0.000 0.000 4.708 5.170 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.779 5.119 cp_fm_cholesky_invert 11 10.9 5.073 5.077 5.073 5.077 pw_transfer 1331 11.6 0.074 0.075 4.812 4.817 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.696 4.700 ot_diis_step 99 11.5 0.015 0.016 4.541 4.541 fft_wrap_pw1pw2_140 451 13.1 1.280 1.285 4.118 4.122 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.880 3.887 grid_integrate_task_list 110 12.3 3.672 3.731 3.672 3.731 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.665 3.671 hybrid_alltoall_any 4261 16.3 0.263 0.560 2.770 3.480 density_rs2pw 110 9.6 0.004 0.004 3.451 3.469 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.996 3.453 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.368 3.369 make_images_data 4110 15.4 0.046 0.049 2.740 3.334 wfi_extrapolate 11 7.9 0.001 0.001 3.261 3.261 calculate_dm_sparse 110 9.5 0.001 0.001 3.185 3.210 multiply_cannon_sync_h2d 8220 15.4 3.137 3.157 3.137 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.840 2.842 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.796 2.796 cp_fm_diag_elpa_base 48 14.0 2.259 2.453 2.793 2.793 fft3d_ps 1111 14.6 1.297 1.308 2.748 2.756 grid_collocate_task_list 110 9.6 2.663 2.678 2.663 2.678 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.611 2.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.361 2.375 calculate_first_density_matrix 1 7.0 0.000 0.000 2.296 2.297 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.182 2.245 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.194 potential_pw2rs 110 12.3 0.021 0.021 2.177 2.180 mp_alltoall_d11v 2046 13.8 1.966 2.028 1.966 2.028 cp_fm_cholesky_decompose 22 10.9 1.914 1.935 1.914 1.935 qs_create_task_list 11 7.9 0.001 0.001 1.879 1.924 generate_qs_task_list 11 8.9 0.734 0.790 1.878 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.842 jit_kernel_multiply 10 15.2 1.421 1.769 1.421 1.769 acc_transpose_blocks 8220 15.4 0.042 0.042 1.708 1.730 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.702 1.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.906000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1241.727273, yerr=49.414924 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 631.853056E+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 1700233. 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 0.037 209.305 209.306 qs_mol_dyn_low 1 2.0 0.003 0.003 208.788 208.819 qs_forces 11 3.9 0.007 0.033 208.698 208.699 qs_energies 11 4.9 0.003 0.013 203.082 203.099 scf_env_do_scf 11 5.9 0.001 0.001 186.214 186.218 scf_env_do_scf_inner_loop 117 6.6 0.008 0.050 165.648 165.651 dbcsr_multiply_generic 2507 12.6 0.182 0.185 126.676 127.181 qs_scf_new_mos 117 7.6 0.001 0.001 126.188 126.407 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.187 126.406 velocity_verlet 10 3.0 0.002 0.010 123.707 123.708 ot_scf_mini 117 9.6 0.003 0.003 119.441 119.677 multiply_cannon 2507 13.6 0.247 0.266 101.978 103.271 multiply_cannon_loop 2507 14.6 2.416 2.473 99.751 101.129 ot_mini 117 10.6 0.001 0.002 67.230 67.448 multiply_cannon_multrec 60168 15.6 31.620 33.761 41.502 43.277 qs_ot_get_derivative 117 11.6 0.001 0.001 42.418 42.623 rebuild_ks_matrix 128 8.3 0.001 0.001 33.647 34.025 qs_ks_build_kohn_sham_matrix 128 9.3 0.024 0.094 33.646 34.024 mp_waitall_1 267128 16.5 29.433 32.492 29.433 32.492 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.218 30.569 qs_ot_get_p 128 10.4 0.001 0.001 29.891 30.202 multiply_cannon_sync_h2d 60168 15.6 26.055 27.591 26.055 27.591 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.254 24.834 apply_single 128 13.6 0.001 0.001 24.254 24.834 ot_diis_step 117 11.6 0.008 0.008 24.516 24.517 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.189 23.288 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.629 20.781 init_scf_loop 11 6.9 0.000 0.000 20.493 20.496 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.415 20.416 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 16.472 18.848 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.423 17.454 cp_fm_diag_elpa_base 83 14.4 17.352 17.402 17.419 17.451 prepare_preconditioner 11 7.9 0.000 0.000 15.779 15.827 make_preconditioner 11 8.9 0.000 0.001 15.779 15.827 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.984 15.188 sum_up_and_integrate 128 10.3 0.005 0.027 14.114 14.147 make_m2s 5014 13.6 0.105 0.115 13.791 14.128 integrate_v_rspace 128 11.3 0.003 0.004 14.053 14.091 make_images 5014 14.6 0.395 0.414 13.604 13.949 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.188 13.382 calculate_rho_elec 128 8.7 0.045 0.064 13.188 13.381 init_scf_run 11 5.9 0.000 0.001 12.661 12.661 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.661 12.661 mp_sum_l 7950 12.9 9.649 10.901 9.649 10.901 dbcsr_mm_accdrv_process 124484 16.2 4.794 4.946 9.460 9.980 wfi_extrapolate 11 7.9 0.001 0.001 9.095 9.095 cp_fm_cholesky_invert 11 10.9 8.959 8.966 8.959 8.966 calculate_dm_sparse 128 9.5 0.001 0.001 8.563 8.639 multiply_cannon_metrocomm1 60168 15.6 0.098 0.102 6.513 8.398 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.268 8.384 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.171 8.321 pw_transfer 1547 11.6 0.074 0.088 7.684 7.885 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.482 7.675 make_images_data 5014 15.6 0.065 0.072 6.718 7.619 grid_integrate_task_list 128 12.3 7.068 7.408 7.068 7.408 density_rs2pw 128 9.7 0.006 0.007 6.751 7.278 hybrid_alltoall_any 5200 16.5 0.296 2.275 5.870 7.139 fft_wrap_pw1pw2_140 523 13.2 1.124 1.163 6.560 6.745 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.727 6.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.846 5.980 fft3d_ps 1291 14.7 2.201 2.845 5.491 5.849 mp_alltoall_d11v 2415 14.1 4.427 5.377 4.427 5.377 grid_collocate_task_list 128 9.7 4.844 5.289 4.844 5.289 potential_pw2rs 128 12.3 0.009 0.010 4.610 4.648 cp_fm_cholesky_decompose 22 10.9 4.612 4.626 4.612 4.626 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.306000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.363636, yerr=6.414086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/16/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 2.199914E+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 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 839.847936E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5014659. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 189.591 189.592 qs_mol_dyn_low 1 2.0 0.003 0.004 189.251 189.266 qs_forces 11 3.9 0.004 0.005 189.162 189.164 qs_energies 11 4.9 0.017 0.053 182.402 182.411 scf_env_do_scf 11 5.9 0.001 0.001 165.229 165.238 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.758 132.761 velocity_verlet 10 3.0 0.006 0.014 118.996 118.998 dbcsr_multiply_generic 2527 12.6 0.193 0.202 97.846 99.139 qs_scf_new_mos 118 7.6 0.001 0.001 94.204 94.870 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.203 94.869 ot_scf_mini 118 9.6 0.004 0.004 89.357 90.105 multiply_cannon 2527 13.6 0.506 0.557 77.780 82.430 multiply_cannon_loop 2527 14.6 1.589 1.658 74.498 77.319 ot_mini 118 10.6 0.001 0.001 50.567 51.219 mp_waitall_1 216438 16.6 24.176 38.650 24.176 38.650 multiply_cannon_multrec 30324 15.6 21.060 26.357 31.760 37.279 rebuild_ks_matrix 129 8.3 0.001 0.001 32.366 32.972 qs_ks_build_kohn_sham_matrix 129 9.3 0.056 0.103 32.365 32.971 init_scf_loop 11 6.9 0.000 0.000 32.385 32.386 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.101 29.678 qs_ot_get_derivative 118 11.6 0.001 0.002 28.265 29.009 multiply_cannon_metrocomm3 30324 15.6 0.098 0.104 15.831 28.726 prepare_preconditioner 11 7.9 0.000 0.000 28.072 28.189 make_preconditioner 11 8.9 0.000 0.001 28.072 28.189 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.781 27.365 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.415 23.618 apply_single 129 13.6 0.001 0.001 22.414 23.618 ot_diis_step 118 11.6 0.014 0.015 22.130 22.132 qs_ot_get_p 129 10.4 0.001 0.001 20.673 21.461 multiply_cannon_sync_h2d 30324 15.6 18.168 21.091 18.168 21.091 cp_fm_cholesky_invert 11 10.9 16.360 16.372 16.360 16.372 qs_ot_p2m_diag 83 11.4 0.189 0.217 15.821 15.855 make_m2s 5054 13.6 0.087 0.093 14.224 15.798 make_images 5054 14.6 1.169 1.375 14.005 15.577 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.619 14.620 sum_up_and_integrate 129 10.3 0.002 0.004 13.965 14.002 integrate_v_rspace 129 11.3 0.003 0.004 13.904 13.945 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.332 13.368 calculate_rho_elec 129 8.7 0.087 0.104 13.332 13.367 init_scf_run 11 5.9 0.000 0.001 12.195 12.196 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.195 12.196 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.110 11.652 cp_fm_diag_elpa 83 13.4 0.001 0.002 11.504 11.540 cp_fm_diag_elpa_base 83 14.4 11.253 11.341 11.497 11.529 multiply_cannon_metrocomm4 27797 15.6 0.105 0.121 3.891 11.312 mp_irecv_dv 70031 16.3 3.681 10.895 3.681 10.895 dbcsr_mm_accdrv_process 62734 16.2 5.501 6.416 10.140 10.797 make_images_data 5054 15.6 0.067 0.078 8.389 10.207 hybrid_alltoall_any 5240 16.5 0.351 1.494 7.186 9.895 pw_transfer 1559 11.6 0.085 0.095 8.571 8.640 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.346 8.416 wfi_extrapolate 11 7.9 0.001 0.001 8.275 8.275 grid_integrate_task_list 129 12.3 7.224 7.554 7.224 7.554 density_rs2pw 129 9.7 0.006 0.007 7.000 7.536 fft_wrap_pw1pw2_140 527 13.2 1.220 1.247 7.427 7.525 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.399 7.203 cp_fm_cholesky_decompose 22 10.9 6.816 6.902 6.816 6.902 calculate_dm_sparse 129 9.5 0.001 0.001 6.507 6.655 mp_sum_l 8010 12.9 4.028 6.172 4.028 6.172 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.073 6.082 fft3d_ps 1301 14.7 2.860 3.054 5.889 5.942 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.442 5.507 grid_collocate_task_list 129 9.7 5.050 5.435 5.050 5.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.159 5.339 mp_allgather_i34 2527 14.6 1.804 4.952 1.804 4.952 mp_alltoall_d11v 2423 14.1 4.072 4.696 4.072 4.696 potential_pw2rs 129 12.3 0.015 0.017 4.430 4.447 dbcsr_complete_redistribute 395 12.7 0.788 0.883 3.104 3.958 mp_sum_d 4499 12.1 2.634 3.906 2.634 3.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.592000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.909091, yerr=0.287480 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 933.183488E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.059 173.749 173.750 qs_mol_dyn_low 1 2.0 0.003 0.004 173.207 173.235 qs_forces 11 3.9 0.004 0.006 173.111 173.113 qs_energies 11 4.9 0.004 0.023 166.416 166.428 scf_env_do_scf 11 5.9 0.007 0.051 150.898 150.900 scf_env_do_scf_inner_loop 116 6.6 0.020 0.040 115.974 115.977 velocity_verlet 10 3.0 0.005 0.009 110.487 110.489 dbcsr_multiply_generic 2485 12.5 0.181 0.185 81.305 82.503 qs_scf_new_mos 116 7.6 0.001 0.001 79.481 79.890 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.480 79.890 ot_scf_mini 116 9.6 0.003 0.004 75.287 75.709 multiply_cannon 2485 13.5 0.504 0.526 61.500 66.231 multiply_cannon_loop 2485 14.5 1.120 1.192 58.825 61.867 ot_mini 116 10.6 0.001 0.001 42.108 42.530 init_scf_loop 11 6.9 0.001 0.004 34.785 34.786 mp_waitall_1 169034 16.6 24.713 34.703 24.713 34.703 prepare_preconditioner 11 7.9 0.000 0.000 30.709 30.762 make_preconditioner 11 8.9 0.001 0.003 30.709 30.762 rebuild_ks_matrix 127 8.3 0.001 0.001 29.947 30.543 qs_ks_build_kohn_sham_matrix 127 9.3 0.027 0.092 29.947 30.542 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.379 29.775 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.917 27.464 multiply_cannon_multrec 19880 15.5 12.957 16.100 22.271 25.452 multiply_cannon_metrocomm3 19880 15.5 0.062 0.065 15.314 25.096 qs_ot_get_derivative 116 11.6 0.002 0.002 22.662 23.089 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.512 20.429 apply_single 127 13.6 0.001 0.001 19.512 20.429 ot_diis_step 116 11.6 0.018 0.019 19.334 19.334 qs_ot_get_p 127 10.4 0.002 0.006 18.628 19.160 multiply_cannon_sync_h2d 19880 15.5 13.509 15.361 13.509 15.361 make_m2s 4970 13.5 0.076 0.081 14.734 15.361 make_images 4970 14.5 1.134 1.220 14.499 15.122 qs_ot_p2m_diag 82 11.4 0.272 0.345 14.367 14.379 cp_fm_cholesky_invert 11 10.9 14.252 14.261 14.252 14.261 sum_up_and_integrate 127 10.3 0.002 0.003 13.787 13.816 integrate_v_rspace 127 11.3 0.003 0.004 13.727 13.756 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.305 13.343 calculate_rho_elec 127 8.7 0.127 0.142 13.305 13.342 cp_dbcsr_syevd 82 12.4 0.006 0.008 13.262 13.264 init_scf_run 11 5.9 0.000 0.001 10.416 10.416 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.416 10.416 cp_fm_diag_elpa 82 13.4 0.001 0.001 10.104 10.123 cp_fm_diag_elpa_base 82 14.4 9.696 9.849 10.100 10.119 make_images_data 4970 15.5 0.061 0.072 8.913 9.971 hybrid_alltoall_any 5155 16.4 0.444 2.018 7.666 9.628 multiply_cannon_metrocomm4 17395 15.5 0.066 0.078 3.462 9.612 mp_irecv_dv 49801 16.2 3.331 9.344 3.331 9.344 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.922 9.240 dbcsr_mm_accdrv_process 41158 16.2 5.585 5.915 8.775 8.939 pw_transfer 1535 11.6 0.085 0.102 8.569 8.673 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.347 8.458 fft_wrap_pw1pw2_140 519 13.2 1.290 1.310 7.430 7.548 grid_integrate_task_list 127 12.3 7.256 7.531 7.256 7.531 wfi_extrapolate 11 7.9 0.001 0.001 7.385 7.385 density_rs2pw 127 9.7 0.005 0.006 6.754 7.333 cp_fm_cholesky_decompose 22 10.9 7.258 7.288 7.258 7.288 cp_fm_upper_to_full 104 14.8 5.779 7.267 5.779 7.267 dbcsr_complete_redistribute 393 12.7 1.155 1.181 4.525 6.282 fft3d_ps 1281 14.7 2.741 2.962 5.748 5.853 calculate_dm_sparse 127 9.5 0.001 0.001 5.779 5.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.431 5.436 grid_collocate_task_list 127 9.7 5.123 5.427 5.123 5.427 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.595 5.275 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.369 5.124 mp_sum_l 7884 12.9 3.265 4.733 3.265 4.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.545 4.721 mp_alltoall_d11v 2401 14.1 4.198 4.451 4.198 4.451 mp_allgather_i34 2485 14.5 1.177 4.399 1.177 4.399 potential_pw2rs 127 12.3 0.020 0.021 4.312 4.326 transfer_fm_to_dbcsr 11 9.9 0.020 0.028 2.309 4.018 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.956 3.990 mp_alltoall_i22 712 14.1 1.954 3.816 1.954 3.816 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.749 3.751 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.526 3.576 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.750000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=884.090909, yerr=8.039160 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.148621E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 187.480 187.482 qs_mol_dyn_low 1 2.0 0.003 0.004 187.056 187.069 qs_forces 11 3.9 0.018 0.122 186.725 186.734 qs_energies 11 4.9 0.013 0.095 179.445 179.455 scf_env_do_scf 11 5.9 0.001 0.005 162.223 162.240 velocity_verlet 10 3.0 0.002 0.002 124.264 124.277 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.065 116.066 dbcsr_multiply_generic 2507 12.6 0.188 0.195 80.598 81.542 qs_scf_new_mos 117 7.6 0.001 0.001 80.143 80.436 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.142 80.435 ot_scf_mini 117 9.6 0.005 0.018 75.583 75.918 multiply_cannon 2507 13.6 0.569 0.627 56.099 59.311 multiply_cannon_loop 2507 14.6 1.871 1.950 52.255 54.431 init_scf_loop 11 6.9 0.001 0.005 46.038 46.041 ot_mini 117 10.6 0.001 0.001 43.064 43.378 prepare_preconditioner 11 7.9 0.000 0.000 41.909 41.952 make_preconditioner 11 8.9 0.001 0.003 41.909 41.952 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.544 40.586 multiply_cannon_multrec 30084 15.6 13.571 18.843 26.531 31.501 rebuild_ks_matrix 128 8.3 0.001 0.001 29.112 29.492 qs_ks_build_kohn_sham_matrix 128 9.3 0.026 0.093 29.111 29.491 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.228 26.566 mp_waitall_1 147882 16.7 16.797 25.150 16.797 25.150 qs_ot_get_derivative 117 11.6 0.002 0.002 23.382 23.721 make_m2s 5014 13.6 0.093 0.098 20.021 21.180 make_images 5014 14.6 1.948 2.309 19.712 20.869 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.047 19.739 apply_single 128 13.6 0.001 0.001 19.047 19.739 ot_diis_step 117 11.6 0.018 0.019 19.558 19.559 qs_ot_get_p 128 10.4 0.001 0.001 18.670 19.013 cp_fm_upper_to_full 105 14.8 11.478 16.927 11.478 16.927 cp_fm_cholesky_invert 11 10.9 15.869 15.878 15.869 15.878 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.211 15.059 qs_ot_p2m_diag 83 11.4 0.344 0.390 14.486 14.534 sum_up_and_integrate 128 10.3 0.002 0.005 13.872 13.903 integrate_v_rspace 128 11.3 0.003 0.004 13.812 13.844 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.613 13.641 calculate_rho_elec 128 8.7 0.170 0.185 13.613 13.641 dbcsr_mm_accdrv_process 62264 16.2 8.471 9.316 12.529 13.103 cp_dbcsr_syevd 83 12.4 0.007 0.022 13.079 13.080 dbcsr_complete_redistribute 395 12.7 1.485 1.629 9.085 12.856 make_images_data 5014 15.6 0.066 0.072 10.684 12.513 multiply_cannon_sync_h2d 30084 15.6 10.511 11.557 10.511 11.557 copy_fm_to_dbcsr 209 11.7 0.001 0.001 7.756 11.515 init_scf_run 11 5.9 0.000 0.001 11.069 11.070 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.069 11.070 hybrid_alltoall_any 5200 16.5 0.527 2.180 9.620 11.063 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.821 10.055 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.342 10.043 cp_fm_diag_elpa 83 13.4 0.007 0.031 9.902 9.920 cp_fm_diag_elpa_base 83 14.4 8.918 9.244 9.867 9.883 mp_alltoall_i22 716 14.1 5.638 9.548 5.638 9.548 pw_transfer 1547 11.6 0.085 0.100 8.871 8.944 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.646 8.724 grid_integrate_task_list 128 12.3 7.551 7.921 7.551 7.921 fft_wrap_pw1pw2_140 523 13.2 1.445 1.480 7.731 7.812 wfi_extrapolate 11 7.9 0.001 0.001 7.721 7.721 cp_fm_cholesky_decompose 22 10.9 7.470 7.583 7.470 7.583 multiply_cannon_metrocomm4 25070 15.6 0.083 0.094 2.831 7.034 density_rs2pw 128 9.7 0.005 0.006 6.660 6.980 mp_irecv_dv 76098 16.2 2.671 6.742 2.671 6.742 calculate_dm_sparse 128 9.5 0.001 0.001 6.291 6.389 fft3d_ps 1291 14.7 2.848 2.907 5.821 5.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.544 5.609 grid_collocate_task_list 128 9.7 5.286 5.558 5.286 5.558 mp_alltoall_d11v 2415 14.1 4.897 5.271 4.897 5.271 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.507 4.619 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.431 4.493 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.420 4.421 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.238 4.300 potential_pw2rs 128 12.3 0.024 0.024 4.145 4.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.482000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.272727, yerr=18.906905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 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 1.559417E+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 5.532238E+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 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58089. MP_Allreduce 11102 1166. MP_Sync 87 MP_Alltoall 1712 18838222. 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 0.103 0.114 169.594 169.600 qs_mol_dyn_low 1 2.0 0.003 0.004 169.052 169.075 qs_forces 11 3.9 0.007 0.008 168.934 168.938 qs_energies 11 4.9 0.002 0.002 161.507 161.512 scf_env_do_scf 11 5.9 0.001 0.001 144.274 144.282 velocity_verlet 10 3.0 0.030 0.035 111.115 111.119 scf_env_do_scf_inner_loop 117 6.6 0.037 0.047 108.641 108.642 dbcsr_multiply_generic 2507 12.6 0.183 0.197 71.882 72.268 qs_scf_new_mos 117 7.6 0.001 0.001 71.959 72.073 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.959 72.073 ot_scf_mini 117 9.6 0.003 0.004 67.536 67.608 multiply_cannon 2507 13.6 0.566 0.600 53.182 56.594 multiply_cannon_loop 2507 14.6 0.806 0.846 50.188 50.856 ot_mini 117 10.6 0.001 0.001 37.312 37.372 init_scf_loop 11 6.9 0.000 0.000 35.483 35.484 prepare_preconditioner 11 7.9 0.000 0.000 31.572 31.624 make_preconditioner 11 8.9 0.001 0.001 31.572 31.624 mp_waitall_1 125778 16.7 24.426 30.728 24.426 30.728 make_full_inverse_cholesky 11 9.9 0.016 0.026 29.421 29.682 rebuild_ks_matrix 128 8.3 0.001 0.001 29.063 29.145 qs_ks_build_kohn_sham_matrix 128 9.3 0.052 0.059 29.062 29.144 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.509 26.587 multiply_cannon_multrec 10028 15.6 10.397 14.562 18.105 21.028 qs_ot_get_derivative 117 11.6 0.002 0.002 20.397 20.463 multiply_cannon_metrocomm3 10028 15.6 0.026 0.028 12.406 19.461 cp_fm_cholesky_invert 11 10.9 17.971 17.976 17.971 17.976 qs_ot_get_p 128 10.4 0.001 0.001 17.552 17.631 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.928 17.040 apply_single 128 13.6 0.001 0.001 16.927 17.039 ot_diis_step 117 11.6 0.020 0.024 16.843 16.844 make_m2s 5014 13.6 0.064 0.069 14.779 15.679 make_images 5014 14.6 2.155 2.580 14.469 15.363 sum_up_and_integrate 128 10.3 0.002 0.002 14.071 14.124 integrate_v_rspace 128 11.3 0.004 0.004 14.011 14.065 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.923 13.952 calculate_rho_elec 128 8.7 0.253 0.265 13.922 13.951 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.761 13.777 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.584 12.586 multiply_cannon_sync_h2d 10028 15.6 10.908 11.210 10.908 11.210 init_scf_run 11 5.9 0.000 0.001 10.445 10.446 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.445 10.446 hybrid_alltoall_any 5200 16.5 0.840 3.773 8.343 9.892 make_images_data 5014 15.6 0.055 0.065 8.521 9.846 cp_fm_diag_elpa 83 13.4 0.001 0.001 9.697 9.708 cp_fm_diag_elpa_base 83 14.4 9.455 9.532 9.690 9.701 pw_transfer 1547 11.6 0.085 0.093 9.233 9.261 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.011 9.045 grid_integrate_task_list 128 12.3 7.786 8.096 7.786 8.096 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.025 8.085 dbcsr_mm_accdrv_process 20762 16.1 3.323 4.560 7.356 8.045 fft_wrap_pw1pw2_140 523 13.2 1.800 1.844 7.983 8.018 cp_fm_cholesky_decompose 22 10.9 7.888 7.994 7.888 7.994 wfi_extrapolate 11 7.9 0.001 0.001 7.376 7.376 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.264 7.189 density_rs2pw 128 9.7 0.005 0.006 6.650 6.998 calculate_dm_sparse 128 9.5 0.001 0.001 6.124 6.183 grid_collocate_task_list 128 9.7 5.576 5.858 5.576 5.858 fft3d_ps 1291 14.7 2.764 2.836 5.745 5.784 dbcsr_complete_redistribute 395 12.7 2.088 2.143 5.257 5.633 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.220 5.221 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.161 5.182 mp_alltoall_d11v 2415 14.1 4.616 4.941 4.616 4.941 mp_allgather_i34 2507 14.6 1.192 4.661 1.192 4.661 potential_pw2rs 128 12.3 0.026 0.026 4.249 4.258 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.071 4.113 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.585 3.923 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.421 3.737 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.581 3.668 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.562 3.590 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.427 3.511 multiply_cannon_metrocomm4 7521 15.6 0.027 0.030 1.705 3.507 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.452 3.467 mp_irecv_dv 28860 15.9 1.663 3.433 1.663 3.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.600000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1460.727273, yerr=32.924406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 11.786060E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 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 3.223179E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144969E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318398E+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 726592338480 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - 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 0.040 0.141 289.062 289.072 qs_mol_dyn_low 1 2.0 0.003 0.003 288.441 288.458 qs_forces 11 3.9 0.005 0.005 288.344 288.346 qs_energies 11 4.9 0.026 0.026 279.577 279.582 scf_env_do_scf 11 5.9 0.001 0.001 257.039 257.044 velocity_verlet 10 3.0 0.015 0.018 209.244 209.253 scf_env_do_scf_inner_loop 118 6.6 0.016 0.020 133.489 133.491 init_scf_loop 11 6.9 0.000 0.000 123.277 123.280 prepare_preconditioner 11 7.9 0.000 0.000 118.499 118.517 make_preconditioner 11 8.9 0.001 0.001 118.499 118.517 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.844 115.655 qs_scf_new_mos 118 7.6 0.001 0.001 90.016 90.116 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.015 90.115 ot_scf_mini 118 9.6 0.004 0.004 85.143 85.196 dbcsr_multiply_generic 2529 12.6 0.215 0.228 82.423 82.909 cp_fm_upper_to_full 106 14.8 52.557 75.190 52.557 75.190 multiply_cannon 2529 13.6 0.699 0.743 59.203 60.543 multiply_cannon_loop 2529 14.6 1.060 1.088 55.160 56.519 ot_mini 118 10.6 0.001 0.001 44.982 45.054 dbcsr_complete_redistribute 397 12.7 3.996 4.047 29.563 42.343 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.095 38.826 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.612 36.211 rebuild_ks_matrix 129 8.3 0.001 0.001 33.969 34.008 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.019 33.968 34.007 mp_alltoall_i22 720 14.1 21.373 33.993 21.373 33.993 cp_fm_cholesky_invert 11 10.9 33.299 33.306 33.299 33.306 mp_waitall_1 104580 16.8 28.057 32.084 28.057 32.084 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.591 31.636 qs_ot_get_p 129 10.4 0.001 0.001 25.284 25.374 qs_ot_get_derivative 118 11.6 0.002 0.002 24.866 24.917 qs_ot_p2m_diag 84 11.4 0.890 0.896 21.228 21.260 make_m2s 5058 13.6 0.075 0.078 18.670 20.157 multiply_cannon_metrocomm3 10116 15.6 0.025 0.028 18.820 20.115 ot_diis_step 118 11.6 0.022 0.023 20.082 20.082 make_images 5058 14.6 3.073 3.265 18.183 19.671 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.350 19.501 apply_single 129 13.6 0.001 0.001 19.350 19.501 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.413 19.415 multiply_cannon_multrec 10116 15.6 10.389 12.234 18.298 18.432 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.433 16.446 calculate_rho_elec 129 8.7 0.482 0.482 16.432 16.446 cp_fm_diag_elpa 84 13.4 0.002 0.003 16.247 16.248 cp_fm_diag_elpa_base 84 14.4 11.802 13.377 16.241 16.241 sum_up_and_integrate 129 10.3 0.002 0.002 16.049 16.141 integrate_v_rspace 129 11.3 0.004 0.005 15.986 16.078 multiply_cannon_sync_h2d 10116 15.6 14.413 14.445 14.413 14.445 hybrid_alltoall_any 5245 16.5 1.321 3.091 10.770 12.893 make_images_data 5058 15.6 0.065 0.070 10.557 12.714 init_scf_run 11 5.9 0.000 0.001 12.275 12.276 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.274 12.275 pw_transfer 1559 11.6 0.096 0.097 11.243 11.248 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.004 11.010 fft_wrap_pw1pw2_140 527 13.2 3.057 3.099 9.830 9.841 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.786 9.827 dbcsr_mm_accdrv_process 20934 16.1 4.320 6.254 7.657 9.536 cp_fm_cholesky_decompose 22 10.9 8.834 8.855 8.834 8.855 grid_integrate_task_list 129 12.3 8.619 8.807 8.619 8.807 wfi_extrapolate 11 7.9 0.001 0.001 8.771 8.771 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 8.098 8.099 density_rs2pw 129 9.7 0.005 0.005 7.418 7.463 mp_alltoall_d11v 2429 14.1 6.816 6.971 6.816 6.971 calculate_dm_sparse 129 9.5 0.001 0.001 6.723 6.761 grid_collocate_task_list 129 9.7 6.464 6.537 6.464 6.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.254 6.285 fft3d_ps 1301 14.7 2.867 2.882 6.227 6.265 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.119 6.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.072000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2926.909091, yerr=163.850516 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 1.260700E+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 0.041 0.057 85.042 85.043 qs_energies 1 2.0 0.000 0.000 84.529 84.541 ls_scf 1 3.0 0.000 0.000 83.602 83.609 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.525 72.717 multiply_cannon 111 7.7 0.024 0.074 55.759 56.792 multiply_cannon_loop 111 8.7 0.227 0.242 52.369 53.505 ls_scf_main 1 4.0 0.000 0.000 52.292 52.304 density_matrix_trs4 2 5.0 0.002 0.003 46.758 46.847 ls_scf_init_scf 1 4.0 0.000 0.000 28.268 28.275 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.087 27.130 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.987 25.005 mp_waitall_1 11031 10.9 22.274 24.913 22.274 24.913 multiply_cannon_multrec 2664 9.7 8.134 8.916 15.633 17.401 multiply_cannon_sync_h2d 2664 9.7 13.615 15.183 13.615 15.183 make_m2s 222 7.7 0.008 0.011 13.110 13.590 make_images 222 8.7 0.098 0.108 13.088 13.572 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.534 13.023 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.365 8.563 make_images_data 222 9.7 0.004 0.005 7.709 8.362 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.703 7.117 8.107 hybrid_alltoall_any 227 10.6 0.215 1.836 6.601 7.852 dbcsr_mm_accdrv_process_sort 4760 11.4 6.328 7.215 6.328 7.215 calculate_norms 4752 9.8 5.525 6.244 5.525 6.244 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.184 mp_sum_l 887 5.1 3.117 4.396 3.117 4.396 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.040 3.567 mp_irecv_dv 6231 10.9 2.024 3.539 2.024 3.539 make_images_sizes 222 9.7 0.000 0.000 0.742 3.481 mp_alltoall_i44 222 10.7 0.742 3.480 0.742 3.480 arnoldi_extremal 4 6.8 0.000 0.000 3.220 3.243 arnoldi_normal_ev 4 7.8 0.001 0.002 3.220 3.243 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.346 3.224 build_subspace 16 8.4 0.009 0.012 3.121 3.122 ls_scf_post 1 4.0 0.000 0.000 3.042 3.049 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.307 2.787 dbcsr_merge_single_wm 555 10.7 0.455 0.590 2.299 2.778 make_images_pack 222 9.7 2.206 2.621 2.207 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.312 2.558 dbcsr_sort_data 658 11.4 2.102 2.514 2.102 2.514 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.459 2.071 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.300 2.375 buffer_matrices_ensure_size 222 8.7 1.753 2.094 1.753 2.094 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.786 1.792 rebuild_ks_matrix 3 7.3 0.000 0.000 1.777 1.783 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.777 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.043000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 2.192282E+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 0.028 0.044 89.735 89.736 qs_energies 1 2.0 0.000 0.000 89.150 89.160 ls_scf 1 3.0 0.000 0.000 87.807 87.811 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.856 74.136 multiply_cannon 111 7.7 0.043 0.072 52.746 56.816 ls_scf_main 1 4.0 0.000 0.000 54.289 54.294 multiply_cannon_loop 111 8.7 0.135 0.146 49.897 52.563 density_matrix_trs4 2 5.0 0.002 0.003 48.642 48.845 ls_scf_init_scf 1 4.0 0.000 0.000 29.813 29.814 mp_waitall_1 9105 10.9 20.550 29.219 20.550 29.219 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.591 28.687 multiply_cannon_multrec 1332 9.7 13.300 16.555 22.702 27.175 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.148 26.158 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.412 20.284 make_m2s 222 7.7 0.006 0.007 14.927 15.507 make_images 222 8.7 1.368 1.695 14.896 15.477 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.513 8.992 10.574 dbcsr_mm_accdrv_process_sort 4041 11.4 8.511 10.047 8.511 10.047 make_images_data 222 9.7 0.004 0.005 8.647 9.514 hybrid_alltoall_any 227 10.6 0.541 2.530 8.053 9.417 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.189 7.737 mp_irecv_dv 3311 11.0 3.169 7.685 3.169 7.685 mp_sum_l 887 5.1 4.701 7.337 4.701 7.337 calculate_norms 2376 9.8 6.071 6.751 6.071 6.751 multiply_cannon_sync_h2d 1332 9.7 4.893 6.444 4.893 6.444 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.497 5.704 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.884 5.080 arnoldi_extremal 4 6.8 0.000 0.000 4.673 4.688 arnoldi_normal_ev 4 7.8 0.001 0.004 4.673 4.688 build_subspace 16 8.4 0.014 0.021 4.388 4.390 ls_scf_post 1 4.0 0.000 0.000 3.706 3.710 ls_scf_store_result 1 5.0 0.000 0.000 3.384 3.530 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.172 3.406 dbcsr_matrix_vector_mult_local 304 10.0 2.794 3.275 2.796 3.277 mp_allgather_i34 111 8.7 0.822 3.050 0.822 3.050 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.593 2.700 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.141 2.412 dbcsr_data_new 4174 10.1 2.118 2.400 2.118 2.400 make_images_pack 222 9.7 1.820 2.132 1.823 2.135 dbcsr_sort_data 436 11.2 1.854 2.060 1.854 2.060 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.888 1.890 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.877 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.875 1.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.736000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 2.927018E+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 0.064 0.103 94.627 94.628 qs_energies 1 2.0 0.001 0.012 93.785 93.964 ls_scf 1 3.0 0.010 0.072 92.266 92.442 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.320 75.710 ls_scf_main 1 4.0 0.012 0.110 56.825 56.848 multiply_cannon 111 7.7 0.040 0.086 52.098 55.783 multiply_cannon_loop 111 8.7 0.116 0.128 49.326 53.562 density_matrix_trs4 2 5.0 0.005 0.037 50.919 51.110 mp_waitall_1 7281 11.0 23.456 33.540 23.456 33.540 ls_scf_init_scf 1 4.0 0.003 0.026 31.539 31.578 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.019 29.120 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.599 26.610 multiply_cannon_multrec 888 9.7 12.675 15.083 21.349 24.404 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.924 23.345 make_m2s 222 7.7 0.006 0.007 16.243 17.030 make_images 222 8.7 1.583 1.852 16.205 16.990 make_images_data 222 9.7 0.004 0.004 9.432 10.573 hybrid_alltoall_any 227 10.6 0.641 2.929 8.997 10.134 dbcsr_mm_accdrv_process 3754 10.4 0.297 0.490 8.212 9.479 dbcsr_mm_accdrv_process_sort 3754 11.4 7.778 8.989 7.778 8.989 mp_sum_l 887 5.1 5.296 8.789 5.296 8.789 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.747 7.642 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.503 7.138 multiply_cannon_sync_h2d 888 9.7 5.973 7.098 5.973 7.098 mp_irecv_dv 2335 11.1 2.486 7.084 2.486 7.084 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.008 6.917 arnoldi_extremal 4 6.8 0.000 0.000 5.108 5.126 arnoldi_normal_ev 4 7.8 0.001 0.005 5.108 5.126 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.716 4.944 calculate_norms 1584 9.8 4.365 4.834 4.365 4.834 build_subspace 16 8.4 0.014 0.021 4.803 4.809 ls_scf_post 1 4.0 0.006 0.027 3.892 4.059 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.482 3.817 mp_allgather_i34 111 8.7 0.887 3.760 0.887 3.760 dbcsr_matrix_vector_mult_local 304 10.0 3.080 3.658 3.082 3.660 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.531 ls_scf_dm_to_ks 2 5.0 0.001 0.029 2.823 2.917 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.631 2.664 rebuild_ks_matrix 3 7.3 0.000 0.000 2.612 2.645 qs_ks_build_kohn_sham_matrix 3 8.3 0.068 0.262 2.612 2.645 ls_scf_initial_guess 1 5.0 0.000 0.000 2.513 2.615 ls_scf_qs_atomic_guess 1 6.0 0.002 0.039 2.512 2.615 dbcsr_data_new 4116 9.9 2.106 2.455 2.106 2.455 dbcsr_sort_data 325 11.1 1.877 2.208 1.877 2.208 make_images_sizes 222 9.7 0.000 0.000 0.972 2.193 mp_alltoall_i44 222 10.7 0.972 2.192 0.972 2.192 mp_sum_d 1061 5.9 1.483 1.945 1.483 1.945 dbcsr_finalize 304 7.8 0.026 0.032 1.605 1.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.628000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2299.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 3.402600E+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 0.050 96.847 96.848 qs_energies 1 2.0 0.000 0.000 96.217 96.221 ls_scf 1 3.0 0.000 0.000 94.576 94.582 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.376 78.564 ls_scf_main 1 4.0 0.000 0.000 58.675 58.678 multiply_cannon 111 7.7 0.050 0.145 51.758 55.653 density_matrix_trs4 2 5.0 0.002 0.003 52.588 52.694 multiply_cannon_loop 111 8.7 0.153 0.171 46.731 49.473 ls_scf_init_scf 1 4.0 0.000 0.000 32.665 32.669 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.442 31.514 mp_waitall_1 6369 11.0 22.762 30.091 22.762 30.091 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.960 28.971 multiply_cannon_multrec 1332 9.7 14.110 17.068 22.126 24.860 make_m2s 222 7.7 0.007 0.008 21.167 22.524 make_images 222 8.7 3.145 3.599 21.117 22.476 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.282 16.513 make_images_data 222 9.7 0.004 0.004 11.841 13.426 hybrid_alltoall_any 227 10.6 0.795 3.774 11.158 12.969 dbcsr_mm_accdrv_process 3641 10.4 0.277 0.481 7.650 9.209 dbcsr_mm_accdrv_process_sort 3641 11.4 7.212 8.714 7.212 8.714 mp_sum_l 887 5.1 3.879 6.973 3.879 6.973 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.089 6.015 mp_irecv_dv 3229 10.9 2.064 5.966 2.064 5.966 multiply_cannon_sync_h2d 1332 9.7 5.492 5.937 5.492 5.937 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.966 5.367 arnoldi_extremal 4 6.8 0.000 0.000 5.245 5.258 arnoldi_normal_ev 4 7.8 0.001 0.005 5.245 5.258 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.469 5.172 build_subspace 16 8.4 0.014 0.021 4.912 4.918 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.527 4.704 calculate_norms 2376 9.8 4.177 4.548 4.177 4.548 mp_allgather_i34 111 8.7 2.143 4.473 2.143 4.473 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.634 3.937 dbcsr_matrix_vector_mult_local 304 10.0 3.245 3.755 3.247 3.757 dbcsr_sort_data 658 11.4 3.070 3.429 3.070 3.429 dbcsr_special_finalize 555 9.7 0.007 0.030 2.828 3.249 dbcsr_merge_single_wm 555 10.7 0.541 0.664 2.818 3.241 ls_scf_post 1 4.0 0.000 0.000 3.236 3.239 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.999 3.044 ls_scf_store_result 1 5.0 0.000 0.000 2.987 3.041 dbcsr_data_release 10477 10.7 1.586 2.364 1.586 2.364 dbcsr_finalize 304 7.8 0.049 0.061 1.794 1.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.848000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2761.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 4.745548E+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 0.098 0.183 92.223 92.277 qs_energies 1 2.0 0.000 0.000 91.380 91.398 ls_scf 1 3.0 0.000 0.000 89.432 89.435 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.886 71.099 ls_scf_main 1 4.0 0.000 0.000 56.742 56.743 multiply_cannon 111 7.7 0.082 0.150 52.930 56.391 multiply_cannon_loop 111 8.7 0.089 0.095 50.353 51.672 density_matrix_trs4 2 5.0 0.002 0.003 49.813 49.870 ls_scf_init_scf 1 4.0 0.000 0.000 29.293 29.296 mp_waitall_1 5436 11.0 24.577 28.257 24.577 28.257 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.994 28.024 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.815 25.831 multiply_cannon_multrec 444 9.7 13.625 16.726 20.807 22.364 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.952 15.730 make_m2s 222 7.7 0.005 0.005 13.448 14.373 make_images 222 8.7 2.039 2.477 13.380 14.306 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.269 13.014 hybrid_alltoall_any 227 10.6 0.802 3.797 8.133 9.634 make_images_data 222 9.7 0.003 0.004 8.174 9.628 dbcsr_mm_accdrv_process 3003 10.4 0.362 0.404 6.870 8.021 multiply_cannon_sync_h2d 444 9.7 6.781 7.898 6.781 7.898 dbcsr_mm_accdrv_process_sort 3003 11.4 6.507 7.617 6.507 7.617 arnoldi_extremal 4 6.8 0.000 0.000 5.861 5.878 arnoldi_normal_ev 4 7.8 0.001 0.005 5.861 5.878 build_subspace 16 8.4 0.015 0.020 5.458 5.470 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.403 4.590 dbcsr_matrix_vector_mult 304 9.0 0.012 0.022 4.223 4.425 dbcsr_matrix_vector_mult_local 304 10.0 3.769 4.238 3.771 4.240 mp_sum_l 887 5.1 2.524 4.113 2.524 4.113 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.568 3.969 mp_irecv_dv 1241 11.2 1.556 3.935 1.556 3.935 mp_allgather_i34 111 8.7 1.155 3.746 1.155 3.746 calculate_norms 792 9.8 3.635 3.740 3.635 3.740 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.518 3.599 ls_scf_post 1 4.0 0.000 0.000 3.397 3.400 ls_scf_store_result 1 5.0 0.000 0.000 3.176 3.217 make_images_sizes 222 9.7 0.000 0.000 0.871 3.131 mp_alltoall_i44 222 10.7 0.870 3.131 0.870 3.131 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.764 3.047 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.310 dbcsr_data_new 4608 9.7 1.794 2.246 1.794 2.246 dbcsr_merge_all 275 8.9 0.477 0.523 2.054 2.154 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.069 2.070 rebuild_ks_matrix 3 7.3 0.000 0.000 2.038 2.039 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.038 2.039 qs_energies_init_hamiltonians 1 3.0 0.012 0.018 1.931 1.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.277000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 8.917398E+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 0.089 0.161 104.964 104.983 qs_energies 1 2.0 0.000 0.000 103.467 103.485 ls_scf 1 3.0 0.000 0.000 100.534 100.549 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.119 74.277 ls_scf_main 1 4.0 0.000 0.000 63.267 63.279 density_matrix_trs4 2 5.0 0.002 0.003 54.338 54.388 multiply_cannon 111 7.7 0.156 0.258 48.228 50.318 multiply_cannon_loop 111 8.7 0.100 0.108 45.227 45.774 ls_scf_init_scf 1 4.0 0.000 0.000 33.481 33.487 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.959 31.981 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.263 29.272 mp_waitall_1 4527 11.1 21.375 25.042 21.375 25.042 make_m2s 222 7.7 0.005 0.005 22.305 23.412 make_images 222 8.7 3.568 3.873 22.198 23.304 multiply_cannon_multrec 444 9.7 17.728 18.377 22.438 23.115 hybrid_alltoall_any 227 10.6 1.655 3.627 12.549 15.298 make_images_data 222 9.7 0.004 0.004 12.715 14.846 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.087 10.367 multiply_cannon_sync_h2d 444 9.7 8.788 8.839 8.788 8.839 arnoldi_extremal 4 6.8 0.000 0.000 7.466 7.480 arnoldi_normal_ev 4 7.8 0.002 0.008 7.466 7.480 build_subspace 16 8.4 0.026 0.036 6.921 6.935 dbcsr_matrix_vector_mult 304 9.0 0.018 0.034 5.552 5.713 dbcsr_matrix_vector_mult_local 304 10.0 5.133 5.459 5.136 5.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.066 5.157 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.854 5.093 dbcsr_mm_accdrv_process 1814 10.4 0.291 0.357 4.526 4.657 dbcsr_mm_accdrv_process_sort 1814 11.4 4.189 4.326 4.189 4.326 ls_scf_post 1 4.0 0.000 0.000 3.787 3.790 make_images_sizes 222 9.7 0.000 0.000 1.462 3.573 mp_alltoall_i44 222 10.7 1.462 3.572 1.462 3.572 mp_allgather_i34 111 8.7 1.118 3.540 1.118 3.540 ls_scf_store_result 1 5.0 0.000 0.000 3.499 3.534 calculate_norms 792 9.8 3.233 3.268 3.233 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.163 dbcsr_merge_all 275 8.9 0.883 0.908 2.864 2.942 qs_energies_init_hamiltonians 1 3.0 0.012 0.030 2.902 2.904 dbcsr_complete_redistribute 5 7.6 1.434 1.478 2.751 2.865 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.525 dbcsr_sort_data 325 11.1 2.437 2.500 2.437 2.500 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.389 2.392 dbcsr_new_transposed 4 7.5 0.277 0.324 2.362 2.382 dbcsr_data_new 6591 9.6 1.838 2.353 1.838 2.353 rebuild_ks_matrix 3 7.3 0.000 0.000 2.324 2.326 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.324 2.326 dbcsr_frobenius_norm 74 6.6 2.063 2.142 2.209 2.243 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.202 dbcsr_add_anytype 103 7.2 0.859 0.893 2.126 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=104.983000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6995.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac71c107aafb4ece195b773ce8dc175d9d1cdc7c_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 593.768448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.114 238.247 238.269 qs_mol_dyn_low 1 2.0 0.005 0.017 237.002 237.070 qs_forces 5 3.8 0.005 0.022 236.880 236.903 qs_energies 5 4.8 0.005 0.052 233.808 233.829 scf_env_do_scf 5 5.8 0.002 0.059 217.805 217.824 scf_env_do_scf_inner_loop 105 6.6 0.013 0.102 188.674 188.697 qs_scf_new_mos 105 7.6 0.000 0.001 147.963 148.133 qs_scf_loop_do_ot 105 8.6 0.002 0.066 147.963 148.132 ot_scf_mini 105 9.6 0.004 0.057 137.987 138.165 dbcsr_multiply_generic 1445 12.2 0.126 0.133 135.019 135.511 multiply_cannon 1445 13.2 0.275 0.292 115.178 117.441 multiply_cannon_loop 1445 14.2 2.857 3.006 113.393 114.512 velocity_verlet 4 3.0 0.002 0.011 106.332 106.344 ot_mini 105 10.6 0.003 0.043 60.612 60.767 qs_ot_get_p 112 10.4 0.002 0.015 47.234 47.633 multiply_cannon_multrec 69360 15.2 29.614 34.282 39.738 44.858 mp_waitall_1 488190 16.1 35.505 43.877 35.505 43.877 qs_ot_get_derivative 55 11.6 0.003 0.079 38.735 38.873 qs_ot_p2m_diag 40 11.0 0.021 0.045 36.185 36.270 multiply_cannon_metrocomm3 69360 15.2 0.204 0.213 26.128 34.141 multiply_cannon_sync_h2d 69360 15.2 28.880 33.259 28.880 33.259 cp_dbcsr_syevd 40 12.0 0.005 0.110 32.722 32.727 rebuild_ks_matrix 110 8.4 0.000 0.000 29.953 30.154 qs_ks_build_kohn_sham_matrix 110 9.4 0.021 0.153 29.952 30.153 init_scf_loop 7 6.6 0.009 0.172 29.099 29.196 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.525 27.684 cp_fm_syevd 40 13.0 0.019 0.104 27.163 27.313 apply_preconditioner_dbcsr 62 12.6 0.001 0.017 23.143 23.401 apply_single 62 13.6 0.000 0.000 23.142 23.401 prepare_preconditioner 7 7.6 0.000 0.000 23.266 23.300 make_preconditioner 7 8.6 0.011 0.194 23.266 23.300 cp_fm_redistribute_end 40 14.0 11.122 22.199 11.128 22.200 cp_fm_syevd_base 40 14.0 11.062 22.138 11.062 22.138 ot_new_cg_direction 55 11.6 0.003 0.086 21.187 21.189 qs_rho_update_rho_low 110 7.6 0.001 0.041 17.319 17.744 calculate_rho_elec 110 8.6 0.031 0.083 17.318 17.744 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.903 15.163 make_full_inverse_cholesky 7 9.6 0.001 0.070 14.599 14.712 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.894 14.026 mp_sum_l 4764 12.2 12.925 13.686 12.925 13.686 init_scf_run 5 5.8 0.000 0.003 13.243 13.245 scf_env_initial_rho_setup 5 6.8 0.006 0.100 13.243 13.245 pw_transfer 1645 12.4 0.079 0.104 12.840 13.139 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.701 12.999 density_rs2pw 110 9.6 0.005 0.006 11.403 11.796 calculate_dm_sparse 110 9.5 0.000 0.001 11.585 11.739 fft_wrap_pw1pw2_240 915 15.0 1.134 1.234 11.268 11.553 qs_vxc_create 110 10.4 0.012 0.191 11.170 11.223 dbcsr_mm_accdrv_process 154766 15.8 6.297 6.472 9.991 10.942 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.493 10.578 cp_fm_cholesky_invert 7 10.6 10.100 10.108 10.100 10.108 check_diag 80 13.5 8.599 8.887 9.711 9.873 fft3d_pb 915 16.0 2.385 2.671 9.492 9.784 sum_up_and_integrate 60 10.3 0.002 0.028 8.703 8.724 integrate_v_rspace 60 11.3 0.002 0.035 8.685 8.708 transfer_rs2pw 445 10.6 0.007 0.008 7.964 8.414 make_full_single_inverse 7 9.6 0.010 0.181 8.308 8.352 calculate_first_density_matrix 1 7.0 0.005 0.091 8.258 8.274 acc_transpose_blocks 69360 15.2 0.363 0.375 7.729 8.235 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.018 8.133 8.152 xc_rho_set_and_dset_create 110 12.4 0.079 0.139 7.787 8.150 multiply_cannon_metrocomm1 69360 15.2 0.100 0.107 4.922 7.759 xc_vxc_pw_create 60 11.3 0.040 0.075 7.561 7.624 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.293 7.349 make_m2s 2890 13.2 0.078 0.087 6.487 7.105 make_images 2890 14.2 0.238 0.259 6.379 6.999 xc_pw_derive 510 13.4 0.005 0.007 6.882 6.989 mp_alltoall_z22v 2340 17.7 6.542 6.856 6.542 6.856 acc_transpose_blocks_kernels 69360 16.2 0.862 0.905 4.936 5.347 mp_waitany 7680 13.5 4.478 4.993 4.478 4.993 multiply_cannon_metrocomm4 67915 15.2 0.184 0.203 2.031 4.931 potential_pw2rs 60 12.3 0.003 0.003 4.852 4.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=238.269000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ac71c107aafb4ece195b773ce8dc175d9d1cdc7c Summary: empty Status: OK