=== 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: 668feff676deb4b57bbee5867e9b22be8df45b0d ################# 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/01 job id: 50421474 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/02 job id: 50421475 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/03 job id: 50421476 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/04 job id: 50421477 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/05 job id: 50421478 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/06 job id: 50421479 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/07 job id: 50421480 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/08 job id: 50421483 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/09 job id: 50421490 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/10 job id: 50421497 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/11 job id: 50421505 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/12 job id: 50421509 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/13 job id: 50421516 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/14 job id: 50421517 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/15 job id: 50421518 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/16 job id: 50421519 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/17 job id: 50421521 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/18 job id: 50421522 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/19 job id: 50421523 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/20 job id: 50421524 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/21 job id: 50421525 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/22 job id: 50421526 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/23 job id: 50421527 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/24 job id: 50421528 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/25 job id: 50421530 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/26 job id: 50421531 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/27 job id: 50421532 --- 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/668feff676deb4b57bbee5867e9b22be8df45b0d_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 0.017 0.031 133.661 133.662 farming_run 1 2.0 133.206 133.212 133.633 133.636 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.510040E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 114.832 114.832 qs_energies 1 2.0 0.000 0.000 114.630 114.633 mp2_main 1 3.0 0.000 0.000 112.520 112.523 mp2_gpw_main 1 4.0 0.028 0.033 111.478 111.481 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 92.679 92.788 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.552 54.659 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 40.966 45.738 get_2c_integrals 1 6.0 0.008 0.009 36.970 37.953 integrate_v_rspace 273 8.0 0.439 0.455 24.558 29.003 pw_transfer 6555 10.6 0.379 0.399 26.836 27.282 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.411 25.689 grid_integrate_task_list 273 9.0 20.470 25.345 20.470 25.345 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 22.951 23.235 compute_2c_integrals 1 7.0 0.002 0.002 19.378 19.379 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.684 19.102 mp2_eri_2c_integrate_gpw 1 9.0 2.349 2.416 18.682 19.101 rpa_ri_compute_en 1 5.0 0.019 0.022 18.692 18.784 cp_fm_cholesky_decompose 12 8.2 17.526 18.557 17.526 18.557 cholesky_decomp 1 7.0 0.000 0.000 16.430 17.457 fft3d_s 5443 13.4 16.117 16.301 16.138 16.322 ao_to_mo_and_store_B_mult_1 272 7.0 10.765 15.322 10.765 15.322 calculate_wavefunction 272 8.0 5.412 5.501 12.309 13.012 rpa_num_int 1 6.0 0.000 0.001 10.634 10.643 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.610 10.622 calc_mat_Q 8 8.0 0.000 0.000 9.471 9.555 contract_S_to_Q 8 9.0 0.000 0.000 8.892 8.976 calc_potential_gpw 544 9.5 0.004 0.005 8.255 8.739 parallel_gemm_fm 14 9.1 0.000 0.000 8.472 8.573 parallel_gemm_fm_cosma 14 10.1 8.472 8.573 8.472 8.573 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.139 8.404 potential_pw2rs 545 10.0 0.107 0.108 7.527 8.152 create_integ_mat 1 6.0 0.015 0.029 7.858 7.868 collocate_single_gaussian 272 10.0 0.039 0.041 7.389 7.626 array2fm 1 7.0 0.000 0.000 6.752 7.306 pw_scatter_s 2720 13.7 4.324 4.513 4.324 4.513 pw_gather_s 2722 13.2 3.419 3.793 3.419 3.793 array2fm_buffer_send 1 8.0 2.964 3.146 2.964 3.146 pw_poisson_solve 545 10.5 1.124 1.185 2.254 2.467 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.480828, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2826.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 0.025 0.035 420.615 420.616 farming_run 1 2.0 419.894 419.901 420.573 420.575 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.244647E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 216.679 216.679 qs_energies 1 2.0 0.000 0.000 216.453 216.460 scf_env_do_scf 1 3.0 0.000 0.000 114.591 114.591 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.692 113.698 rebuild_ks_matrix 4 6.0 0.000 0.000 113.690 113.697 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.690 113.697 hfx_ks_matrix 4 8.0 0.001 0.001 113.345 113.350 integrate_four_center 4 9.0 0.153 0.464 113.344 113.350 integrate_four_center_main 4 10.0 0.086 0.501 102.223 105.106 integrate_four_center_bin 266 11.0 102.137 105.092 102.137 105.092 mp2_main 1 3.0 0.000 0.001 101.569 101.576 mp2_gpw_main 1 4.0 0.048 0.071 100.728 100.739 init_scf_loop 1 4.0 0.000 0.000 96.870 96.870 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.913 74.955 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.829 54.871 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.155 41.560 46.388 integrate_v_rspace 95 8.0 0.395 0.559 28.031 32.737 pw_transfer 2240 10.6 0.144 0.167 29.486 29.878 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.494 28.909 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.669 28.770 ao_to_mo_and_store_B_mult_1 91 7.0 10.595 28.610 10.595 28.610 grid_integrate_task_list 95 9.0 23.447 28.386 23.447 28.386 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.247 26.683 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.830 1.890 24.976 24.987 get_2c_integrals 1 6.0 0.000 0.000 19.998 20.019 compute_2c_integrals 1 7.0 0.002 0.003 18.983 18.984 fft3d_s 1823 13.4 18.489 18.908 18.502 18.922 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.617 18.841 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.873 18.616 18.840 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.719 17.719 calculate_wavefunction 91 8.0 2.012 2.043 9.618 9.865 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.585 8.809 9.226 potential_pw2rs 186 10.0 0.033 0.034 8.421 8.971 local_gemm 172 8.0 8.251 8.672 8.251 8.672 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.152 8.530 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.513 7.954 8.429 calc_potential_gpw 182 9.5 0.002 0.002 7.908 8.092 collocate_single_gaussian 91 10.0 0.017 0.021 7.758 7.970 mp_sync 37 10.5 3.097 7.757 3.097 7.757 integrate_four_center_load 4 10.0 0.000 0.000 6.760 6.763 hfx_load_balance 1 11.0 0.000 0.000 6.759 6.763 mp_sendrecv_dm3 2068 8.0 5.987 6.442 5.987 6.442 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.399 6.337 6.399 pw_gather_s 912 13.2 4.451 4.934 4.451 4.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.723585, 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/668feff676deb4b57bbee5867e9b22be8df45b0d_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 452.988928E+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 0.016 0.029 55.890 55.891 qs_mol_dyn_low 1 2.0 0.004 0.008 55.606 55.615 qs_forces 11 3.9 0.003 0.012 55.531 55.532 qs_energies 11 4.9 0.026 0.197 53.913 53.950 scf_env_do_scf 11 5.9 0.000 0.001 47.566 47.567 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.262 45.263 qs_scf_new_mos 108 7.5 0.000 0.001 35.068 35.371 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.068 35.370 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.650 35.208 ot_scf_mini 108 9.5 0.002 0.002 33.287 33.474 multiply_cannon 2286 13.5 0.192 0.203 26.618 28.407 velocity_verlet 10 3.0 0.001 0.002 27.810 27.811 multiply_cannon_loop 2286 14.5 1.822 1.936 25.934 27.775 ot_mini 108 10.5 0.001 0.001 20.241 20.494 qs_ot_get_derivative 108 11.5 0.001 0.001 17.234 17.413 mp_waitall_1 245248 16.5 9.021 15.536 9.021 15.536 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.363 13.405 multiply_cannon_multrec 54864 15.5 3.606 5.711 7.540 10.898 qs_ot_get_p 119 10.4 0.001 0.001 8.195 8.472 rebuild_ks_matrix 119 8.3 0.000 0.000 8.045 8.196 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.048 8.045 8.196 mp_sum_l 7287 12.8 6.023 7.859 6.023 7.859 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.004 7.151 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.267 6.711 multiply_cannon_sync_h2d 54864 15.5 5.123 6.598 5.123 6.598 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.597 5.699 dbcsr_mm_accdrv_process 76910 16.1 1.843 2.909 3.848 5.517 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.245 5.283 init_scf_run 11 5.9 0.000 0.001 4.730 4.730 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.729 4.730 sum_up_and_integrate 119 10.3 0.001 0.003 4.581 4.602 integrate_v_rspace 119 11.3 0.002 0.003 4.570 4.592 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.375 4.540 calculate_rho_elec 119 8.7 0.011 0.016 4.375 4.539 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.415 4.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.208 4.209 cp_fm_redistribute_end 50 14.0 2.147 4.183 2.153 4.186 cp_fm_diag_elpa_base 50 14.0 2.027 4.081 2.031 4.089 multiply_cannon_metrocomm1 54864 15.5 0.056 0.063 1.961 3.243 calculate_dm_sparse 119 9.5 0.000 0.000 3.007 3.175 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.923 3.146 apply_single 119 13.6 0.000 0.000 2.923 3.146 acc_transpose_blocks 54864 15.5 0.229 0.251 2.205 2.728 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.638 2.700 ot_diis_step 108 11.5 0.006 0.006 2.697 2.698 jit_kernel_multiply 13 15.8 1.940 2.517 1.940 2.517 calculate_first_density_matrix 1 7.0 0.000 0.002 2.392 2.404 density_rs2pw 119 9.7 0.004 0.004 2.241 2.396 init_scf_loop 11 6.9 0.000 0.000 2.286 2.286 wfi_extrapolate 11 7.9 0.001 0.001 2.253 2.253 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.200 2.203 grid_integrate_task_list 119 12.3 2.016 2.120 2.016 2.120 potential_pw2rs 119 12.3 0.004 0.004 1.935 1.970 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.893 1.970 mp_sum_d 4137 12.0 1.361 1.945 1.361 1.945 pw_transfer 1439 11.6 0.051 0.058 1.786 1.867 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.711 1.793 make_m2s 4572 13.5 0.053 0.056 1.696 1.749 mp_alltoall_d11v 2130 13.8 1.516 1.727 1.516 1.727 make_images 4572 14.5 0.134 0.140 1.614 1.667 transfer_rs2pw 487 10.6 0.005 0.006 1.439 1.543 fft3d_ps 1201 14.6 0.368 0.473 1.423 1.500 acc_transpose_blocks_sync 164592 16.5 1.201 1.446 1.201 1.446 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.397 1.443 fft_wrap_pw1pw2_140 487 13.2 0.108 0.115 1.346 1.429 grid_collocate_task_list 119 9.7 1.353 1.428 1.353 1.428 transfer_pw2rs 487 13.2 0.006 0.007 1.383 1.407 mp_waitany 12084 13.8 1.289 1.393 1.289 1.393 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.755 1.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.891000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 489.480192E+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 1654462. 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 0.022 0.043 40.046 40.047 qs_mol_dyn_low 1 2.0 0.004 0.010 39.760 39.771 qs_forces 11 3.9 0.003 0.008 39.694 39.696 qs_energies 11 4.9 0.003 0.008 38.012 38.018 scf_env_do_scf 11 5.9 0.001 0.002 32.665 32.666 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.080 30.080 dbcsr_multiply_generic 2286 12.5 0.126 0.131 22.344 22.699 qs_scf_new_mos 108 7.5 0.001 0.001 21.051 21.280 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.050 21.280 ot_scf_mini 108 9.5 0.003 0.003 20.109 20.271 velocity_verlet 10 3.0 0.001 0.004 19.240 19.241 multiply_cannon 2286 13.5 0.210 0.218 16.992 18.469 multiply_cannon_loop 2286 14.5 1.191 1.252 15.741 17.136 ot_mini 108 10.5 0.001 0.001 12.348 12.584 mp_waitall_1 200699 16.5 5.753 10.718 5.753 10.718 qs_ot_get_derivative 108 11.5 0.001 0.001 9.971 10.133 multiply_cannon_metrocomm3 27432 15.5 0.073 0.075 4.044 9.302 multiply_cannon_multrec 27432 15.5 1.823 4.219 6.129 8.932 rebuild_ks_matrix 119 8.3 0.000 0.000 6.993 7.131 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.992 7.130 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.172 6.298 dbcsr_mm_accdrv_process 47894 16.0 3.536 5.794 4.224 6.282 qs_ot_get_p 119 10.4 0.001 0.002 4.767 4.989 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.826 4.675 mp_sum_l 7287 12.8 2.115 4.132 2.115 4.132 init_scf_run 11 5.9 0.000 0.001 4.049 4.049 scf_env_initial_rho_setup 11 6.9 0.001 0.004 4.049 4.049 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.007 4.044 apply_single 119 13.6 0.000 0.000 3.006 4.043 sum_up_and_integrate 119 10.3 0.001 0.002 3.894 3.900 integrate_v_rspace 119 11.3 0.002 0.003 3.880 3.885 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.850 3.883 calculate_rho_elec 119 8.7 0.021 0.024 3.850 3.883 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.073 3.091 make_m2s 4572 13.5 0.052 0.054 2.772 3.091 make_images 4572 14.5 0.207 0.248 2.682 3.003 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.673 2.674 init_scf_loop 11 6.9 0.001 0.003 2.562 2.562 calculate_first_density_matrix 1 7.0 0.000 0.003 2.473 2.476 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.334 2.420 ot_diis_step 108 11.5 0.014 0.034 2.328 2.346 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.340 2.341 cp_fm_redistribute_end 50 14.0 1.185 2.310 1.190 2.314 cp_fm_diag_elpa_base 50 14.0 1.091 2.216 1.120 2.253 density_rs2pw 119 9.7 0.004 0.004 2.130 2.188 multiply_cannon_sync_h2d 27432 15.5 1.689 2.174 1.689 2.174 calculate_dm_sparse 119 9.5 0.000 0.001 2.084 2.166 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.055 2.058 pw_transfer 1439 11.6 0.065 0.068 1.971 2.009 acc_transpose_blocks 27432 15.5 0.116 0.122 1.616 1.979 grid_integrate_task_list 119 12.3 1.838 1.920 1.838 1.920 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.879 1.919 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.772 1.811 jit_kernel_multiply 9 16.3 0.630 1.705 0.630 1.705 prepare_preconditioner 11 7.9 0.000 0.000 1.639 1.664 make_preconditioner 11 8.9 0.000 0.001 1.639 1.664 make_images_data 4572 15.5 0.049 0.055 1.279 1.652 potential_pw2rs 119 12.3 0.006 0.006 1.609 1.617 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.538 1.595 fft3d_ps 1201 14.6 0.522 0.578 1.511 1.542 wfi_extrapolate 11 7.9 0.001 0.004 1.517 1.519 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.133 1.512 fft_wrap_pw1pw2_140 487 13.2 0.127 0.133 1.458 1.498 mp_alltoall_d11v 2130 13.8 1.308 1.435 1.308 1.435 grid_collocate_task_list 119 9.7 1.289 1.337 1.289 1.337 transfer_rs2pw 487 10.6 0.005 0.005 1.245 1.310 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.243 1.289 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.251 1.259 mp_sum_d 4137 12.0 0.591 1.076 0.591 1.076 transfer_pw2rs 487 13.2 0.004 0.005 1.034 1.046 mp_allgather_i34 2286 14.5 0.624 1.029 0.624 1.029 qs_energies_init_hamiltonians 11 5.9 0.006 0.016 1.005 1.008 acc_transpose_blocks_sync 82296 16.5 0.812 0.942 0.812 0.942 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.663 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.880 0.893 make_images_sizes 4572 15.5 0.005 0.005 0.621 0.891 mp_alltoall_i44 4572 16.5 0.616 0.886 0.616 0.886 mp_alltoall_z22v 1201 16.6 0.755 0.831 0.755 0.831 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.047000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=2.092884 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 522.014720E+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 0.011 0.028 32.747 32.749 qs_mol_dyn_low 1 2.0 0.003 0.004 32.493 32.502 qs_forces 11 3.9 0.002 0.002 32.431 32.432 qs_energies 11 4.9 0.005 0.025 30.868 30.870 scf_env_do_scf 11 5.9 0.001 0.002 25.572 25.573 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.114 23.114 dbcsr_multiply_generic 2286 12.5 0.102 0.106 16.871 16.939 qs_scf_new_mos 108 7.5 0.001 0.001 15.372 15.387 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.372 15.386 velocity_verlet 10 3.0 0.001 0.002 15.379 15.381 ot_scf_mini 108 9.5 0.002 0.003 14.614 14.629 multiply_cannon 2286 13.5 0.194 0.198 13.671 14.365 multiply_cannon_loop 2286 14.5 0.861 0.885 12.898 13.611 ot_mini 108 10.5 0.001 0.001 9.069 9.084 qs_ot_get_derivative 108 11.5 0.001 0.001 7.585 7.601 multiply_cannon_multrec 18288 15.5 1.883 2.886 6.938 7.257 rebuild_ks_matrix 119 8.3 0.000 0.000 6.028 6.054 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 6.028 6.054 dbcsr_mm_accdrv_process 38222 16.0 4.902 5.689 4.961 5.754 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.304 5.327 init_scf_run 11 5.9 0.000 0.001 3.963 3.963 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.963 3.963 mp_waitall_1 158411 16.6 2.687 3.761 2.687 3.761 sum_up_and_integrate 119 10.3 0.001 0.002 3.587 3.593 integrate_v_rspace 119 11.3 0.003 0.003 3.574 3.581 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.923 3.467 qs_ot_get_p 119 10.4 0.001 0.002 3.306 3.321 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.308 3.313 calculate_rho_elec 119 8.7 0.031 0.031 3.308 3.313 calculate_first_density_matrix 1 7.0 0.000 0.003 2.700 2.701 init_scf_loop 11 6.9 0.001 0.003 2.441 2.442 multiply_cannon_metrocomm3 18288 15.5 0.049 0.051 1.503 2.413 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.944 2.343 apply_single 119 13.6 0.000 0.000 1.944 2.343 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.200 2.205 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.071 2.073 make_m2s 4572 13.5 0.044 0.045 1.831 1.978 density_rs2pw 119 9.7 0.004 0.004 1.872 1.932 grid_integrate_task_list 119 12.3 1.796 1.909 1.796 1.909 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.904 1.905 pw_transfer 1439 11.6 0.065 0.068 1.883 1.892 make_images 4572 14.5 0.193 0.207 1.746 1.891 calculate_dm_sparse 119 9.5 0.000 0.001 1.840 1.846 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.791 1.801 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.690 1.697 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.665 1.676 cp_fm_diag_elpa_base 50 14.0 1.642 1.658 1.663 1.674 prepare_preconditioner 11 7.9 0.000 0.000 1.668 1.671 make_preconditioner 11 8.9 0.000 0.001 1.668 1.671 acc_transpose_blocks 18288 15.5 0.080 0.082 1.616 1.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.524 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.506 1.511 ot_diis_step 108 11.5 0.011 0.012 1.470 1.470 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.440 1.452 mp_sum_l 7287 12.8 0.975 1.392 0.975 1.392 potential_pw2rs 119 12.3 0.007 0.008 1.380 1.385 fft3d_ps 1201 14.6 0.547 0.566 1.345 1.355 grid_collocate_task_list 119 9.7 1.233 1.311 1.233 1.311 wfi_extrapolate 11 7.9 0.001 0.001 1.209 1.209 multiply_cannon_sync_h2d 18288 15.5 1.051 1.185 1.051 1.185 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.182 transfer_rs2pw 487 10.6 0.005 0.005 1.048 1.145 qs_energies_init_hamiltonians 11 5.9 0.023 0.029 1.033 1.034 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.942 0.959 make_images_data 4572 15.5 0.049 0.053 0.780 0.951 dbcsr_complete_redistribute 329 12.2 0.424 0.495 0.845 0.950 hybrid_alltoall_any 4725 16.4 0.059 0.116 0.679 0.882 transfer_pw2rs 487 13.2 0.004 0.004 0.846 0.851 mp_alltoall_d11v 2130 13.8 0.691 0.809 0.691 0.809 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.700 0.805 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.796 0.798 acc_transpose_blocks_kernels 18288 16.5 0.218 0.224 0.786 0.794 acc_transpose_blocks_sync 54864 16.5 0.731 0.781 0.731 0.781 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.673 0.731 mp_alltoall_z22v 1201 16.6 0.649 0.719 0.649 0.719 cp_fm_cholesky_invert 11 10.9 0.711 0.715 0.711 0.715 mp_waitany 9880 13.7 0.563 0.677 0.563 0.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.749000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 558.718976E+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 0.010 0.029 36.680 36.680 qs_mol_dyn_low 1 2.0 0.003 0.003 36.494 36.503 qs_forces 11 3.9 0.002 0.002 36.433 36.433 qs_energies 11 4.9 0.002 0.002 34.715 34.721 scf_env_do_scf 11 5.9 0.000 0.001 29.567 29.569 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.095 26.096 dbcsr_multiply_generic 2286 12.5 0.109 0.111 20.009 20.112 velocity_verlet 10 3.0 0.001 0.001 18.734 18.736 qs_scf_new_mos 108 7.5 0.001 0.001 17.952 18.007 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.951 18.006 ot_scf_mini 108 9.5 0.002 0.003 16.915 16.971 multiply_cannon 2286 13.5 0.218 0.225 16.210 16.567 multiply_cannon_loop 2286 14.5 1.521 1.629 15.225 15.553 ot_mini 108 10.5 0.001 0.001 10.428 10.490 multiply_cannon_multrec 27432 15.5 2.454 3.085 9.090 9.338 qs_ot_get_derivative 108 11.5 0.001 0.001 8.557 8.615 dbcsr_mm_accdrv_process 47916 15.9 6.026 7.669 6.534 7.863 rebuild_ks_matrix 119 8.3 0.000 0.001 6.350 6.405 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.350 6.404 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.623 5.674 init_scf_run 11 5.9 0.000 0.001 3.803 3.803 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.803 3.803 qs_ot_get_p 119 10.4 0.001 0.001 3.647 3.731 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.061 3.537 sum_up_and_integrate 119 10.3 0.001 0.001 3.527 3.536 integrate_v_rspace 119 11.3 0.003 0.003 3.516 3.525 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.427 3.466 calculate_rho_elec 119 8.7 0.040 0.046 3.426 3.465 init_scf_loop 11 6.9 0.000 0.000 3.449 3.449 prepare_preconditioner 11 7.9 0.000 0.000 2.609 2.617 make_preconditioner 11 8.9 0.000 0.001 2.609 2.617 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.189 2.599 apply_single 119 13.6 0.000 0.000 2.189 2.599 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.213 2.543 acc_transpose_blocks 27432 15.5 0.123 0.126 2.385 2.537 make_m2s 4572 13.5 0.054 0.056 2.312 2.422 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.327 2.356 calculate_first_density_matrix 1 7.0 0.000 0.000 2.353 2.355 make_images 4572 14.5 0.275 0.336 2.204 2.314 calculate_dm_sparse 119 9.5 0.000 0.000 2.186 2.241 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.222 2.230 mp_waitall_1 137007 16.6 1.783 2.079 1.783 2.079 pw_transfer 1439 11.6 0.065 0.070 2.021 2.056 density_rs2pw 119 9.7 0.004 0.004 1.865 1.985 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.929 1.968 grid_integrate_task_list 119 12.3 1.838 1.902 1.838 1.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.901 1.902 ot_diis_step 108 11.5 0.012 0.012 1.834 1.834 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.816 1.817 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.709 1.721 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.587 1.628 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.572 1.584 cp_fm_diag_elpa_base 50 14.0 1.538 1.559 1.569 1.582 acc_transpose_blocks_sync 82296 16.5 1.412 1.559 1.412 1.559 mp_sum_l 7287 12.8 1.049 1.551 1.049 1.551 fft3d_ps 1201 14.6 0.592 0.651 1.418 1.445 wfi_extrapolate 11 7.9 0.001 0.001 1.396 1.396 potential_pw2rs 119 12.3 0.009 0.009 1.325 1.328 grid_collocate_task_list 119 9.7 1.249 1.322 1.249 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.293 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.289 cp_fm_upper_to_full 72 14.2 0.820 1.168 0.820 1.168 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.742 1.156 jit_kernel_multiply 6 16.0 0.443 1.123 0.443 1.123 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.076 1.077 dbcsr_complete_redistribute 329 12.2 0.118 0.144 0.784 1.057 transfer_rs2pw 487 10.6 0.004 0.005 0.922 1.041 make_images_data 4572 15.5 0.050 0.054 0.847 0.974 hybrid_alltoall_any 4725 16.4 0.067 0.157 0.729 0.893 mp_alltoall_d11v 2130 13.8 0.770 0.889 0.770 0.889 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.878 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.583 0.854 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.822 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.821 cp_fm_cholesky_invert 11 10.9 0.791 0.794 0.791 0.794 transfer_pw2rs 487 13.2 0.004 0.004 0.748 0.750 mp_alltoall_z22v 1201 16.6 0.708 0.734 0.708 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.680000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.454545, yerr=3.394210 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 624.054272E+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 0.021 0.125 29.387 29.474 qs_mol_dyn_low 1 2.0 0.003 0.003 29.189 29.202 qs_forces 11 3.9 0.002 0.002 29.058 29.059 qs_energies 11 4.9 0.001 0.002 27.331 27.334 scf_env_do_scf 11 5.9 0.000 0.001 22.364 22.364 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.749 19.750 velocity_verlet 10 3.0 0.002 0.002 15.345 15.352 dbcsr_multiply_generic 2286 12.5 0.096 0.100 13.412 13.482 qs_scf_new_mos 108 7.5 0.001 0.001 11.934 11.958 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.934 11.957 ot_scf_mini 108 9.5 0.002 0.002 11.220 11.241 multiply_cannon 2286 13.5 0.224 0.231 10.735 11.214 multiply_cannon_loop 2286 14.5 0.644 0.665 9.773 9.916 ot_mini 108 10.5 0.001 0.001 6.429 6.456 multiply_cannon_multrec 9144 15.5 1.710 1.924 6.160 6.411 rebuild_ks_matrix 119 8.3 0.000 0.000 5.877 5.897 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.876 5.896 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.219 5.237 qs_ot_get_derivative 108 11.5 0.001 0.001 5.111 5.132 dbcsr_mm_accdrv_process 12550 15.8 3.793 4.355 4.340 4.415 init_scf_run 11 5.9 0.000 0.001 3.523 3.523 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.523 3.523 sum_up_and_integrate 119 10.3 0.001 0.001 3.494 3.499 integrate_v_rspace 119 11.3 0.003 0.003 3.484 3.489 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.462 3.468 calculate_rho_elec 119 8.7 0.060 0.061 3.461 3.468 qs_ot_get_p 119 10.4 0.001 0.001 2.868 2.904 init_scf_loop 11 6.9 0.000 0.000 2.592 2.592 calculate_first_density_matrix 1 7.0 0.000 0.001 2.324 2.325 pw_transfer 1439 11.6 0.065 0.068 2.176 2.185 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.083 2.092 make_m2s 4572 13.5 0.033 0.034 1.869 2.021 mp_waitall_1 115863 16.7 1.455 1.985 1.455 1.985 grid_integrate_task_list 119 12.3 1.859 1.951 1.859 1.951 density_rs2pw 119 9.7 0.003 0.004 1.842 1.937 make_images 4572 14.5 0.269 0.301 1.779 1.930 calculate_dm_sparse 119 9.5 0.000 0.000 1.847 1.868 prepare_preconditioner 11 7.9 0.000 0.000 1.843 1.846 make_preconditioner 11 8.9 0.000 0.001 1.843 1.846 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.839 1.841 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.811 1.825 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.727 1.752 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.719 1.731 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.684 1.685 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.596 1.596 acc_transpose_blocks 9144 15.5 0.044 0.045 1.496 1.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.514 1.521 fft3d_ps 1201 14.6 0.651 0.663 1.435 1.447 grid_collocate_task_list 119 9.7 1.296 1.383 1.296 1.383 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.346 1.359 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.314 1.323 cp_fm_diag_elpa_base 50 14.0 1.288 1.305 1.313 1.321 ot_diis_step 108 11.5 0.013 0.013 1.305 1.305 potential_pw2rs 119 12.3 0.010 0.010 1.295 1.300 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.244 1.272 apply_single 119 13.6 0.000 0.000 1.244 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.234 1.239 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.220 1.221 jit_kernel_multiply 6 16.3 0.507 1.156 0.507 1.156 wfi_extrapolate 11 7.9 0.001 0.001 1.147 1.147 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.814 1.035 make_images_data 4572 15.5 0.043 0.047 0.831 0.999 mp_alltoall_d11v 2130 13.8 0.831 0.965 0.831 0.965 cp_fm_cholesky_invert 11 10.9 0.933 0.935 0.933 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.924 transfer_rs2pw 487 10.6 0.004 0.004 0.807 0.901 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.840 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.779 0.781 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.719 0.741 acc_transpose_blocks_sync 27432 16.5 0.718 0.740 0.718 0.740 mp_allgather_i34 2286 14.5 0.258 0.722 0.258 0.722 qs_env_update_s_mstruct 11 6.9 0.001 0.004 0.668 0.718 mp_alltoall_z22v 1201 16.6 0.658 0.695 0.658 0.695 transfer_pw2rs 487 13.2 0.003 0.004 0.682 0.685 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.336 0.681 jit_kernel_transpose 5 15.6 0.601 0.623 0.601 0.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.474000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.363636, yerr=4.118092 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 771.915776E+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 0.016 0.032 43.060 43.061 qs_mol_dyn_low 1 2.0 0.003 0.003 42.863 42.871 qs_forces 11 3.9 0.002 0.003 42.796 42.797 qs_energies 11 4.9 0.002 0.003 40.786 40.789 scf_env_do_scf 11 5.9 0.001 0.001 34.861 34.861 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.062 27.064 velocity_verlet 10 3.0 0.002 0.002 24.299 24.305 dbcsr_multiply_generic 2286 12.5 0.103 0.106 19.051 19.273 qs_scf_new_mos 108 7.5 0.001 0.001 17.227 17.329 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.226 17.328 ot_scf_mini 108 9.5 0.002 0.002 16.068 16.167 multiply_cannon 2286 13.5 0.295 0.304 15.055 15.985 multiply_cannon_loop 2286 14.5 0.864 0.884 13.792 14.756 ot_mini 108 10.5 0.001 0.001 9.750 9.866 multiply_cannon_multrec 9144 15.5 3.331 4.607 8.794 8.855 qs_ot_get_derivative 108 11.5 0.001 0.001 7.713 7.813 init_scf_loop 11 6.9 0.000 0.000 7.772 7.776 rebuild_ks_matrix 119 8.3 0.000 0.000 7.359 7.500 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.358 7.500 prepare_preconditioner 11 7.9 0.000 0.000 6.798 6.812 make_preconditioner 11 8.9 0.000 0.001 6.798 6.812 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.633 6.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.373 6.693 dbcsr_mm_accdrv_process 12550 15.8 4.699 6.506 5.327 6.538 cp_fm_upper_to_full 72 14.2 3.151 4.525 3.151 4.525 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.272 4.277 calculate_rho_elec 119 8.7 0.117 0.120 4.271 4.276 sum_up_and_integrate 119 10.3 0.001 0.001 3.939 3.946 integrate_v_rspace 119 11.3 0.003 0.004 3.929 3.935 init_scf_run 11 5.9 0.000 0.001 3.867 3.867 scf_env_initial_rho_setup 11 6.9 0.003 0.003 3.867 3.867 qs_ot_get_p 119 10.4 0.001 0.002 3.387 3.522 mp_waitall_1 94719 16.7 2.263 3.275 2.263 3.275 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.854 3.246 pw_transfer 1439 11.6 0.068 0.069 3.094 3.102 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.995 3.003 dbcsr_complete_redistribute 329 12.2 0.283 0.288 1.963 2.780 make_m2s 4572 13.5 0.037 0.038 2.450 2.594 fft_wrap_pw1pw2_140 487 13.2 0.651 0.655 2.544 2.554 make_images 4572 14.5 0.356 0.389 2.330 2.474 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.178 2.471 apply_single 119 13.6 0.000 0.000 2.178 2.471 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.652 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.394 2.396 density_rs2pw 119 9.7 0.003 0.004 2.295 2.316 calculate_dm_sparse 119 9.5 0.000 0.000 2.257 2.269 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.420 2.228 mp_alltoall_i22 627 13.8 1.418 2.207 1.418 2.207 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.293 2.195 grid_integrate_task_list 119 12.3 2.078 2.148 2.078 2.148 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.057 2.108 ot_diis_step 108 11.5 0.014 0.015 2.014 2.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.987 1.988 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.986 1.987 fft3d_ps 1201 14.6 0.869 0.887 1.915 1.920 acc_transpose_blocks 9144 15.5 0.045 0.046 1.853 1.901 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.769 1.770 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.720 1.746 mp_sum_l 7287 12.8 0.992 1.745 0.992 1.745 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.629 1.629 grid_collocate_task_list 119 9.7 1.520 1.564 1.520 1.564 cp_fm_cholesky_invert 11 10.9 1.442 1.446 1.442 1.446 potential_pw2rs 119 12.3 0.014 0.014 1.416 1.418 wfi_extrapolate 11 7.9 0.001 0.001 1.408 1.408 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.345 1.345 cp_fm_diag_elpa_base 50 14.0 1.201 1.252 1.343 1.343 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.319 1.337 hybrid_alltoall_any 4725 16.4 0.092 0.152 1.083 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.269 mp_alltoall_d11v 2130 13.8 1.201 1.237 1.201 1.237 make_images_data 4572 15.5 0.048 0.051 1.015 1.221 acc_transpose_blocks_sync 27432 16.5 1.103 1.135 1.103 1.135 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.101 1.115 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.997 1.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.934 0.947 mp_alltoall_z22v 1201 16.6 0.909 0.938 0.909 0.938 qs_create_task_list 11 7.9 0.000 0.000 0.927 0.937 generate_qs_task_list 11 8.9 0.367 0.387 0.926 0.936 jit_kernel_multiply 5 15.4 0.599 0.886 0.599 0.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.061000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=726.000000, yerr=14.314646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 501.084160E+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 2370112. 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 0.009 0.034 81.680 81.682 qs_mol_dyn_low 1 2.0 0.003 0.003 81.431 81.442 qs_forces 11 3.9 0.003 0.004 81.363 81.364 qs_energies 11 4.9 0.002 0.003 78.513 78.532 scf_env_do_scf 11 5.9 0.000 0.001 69.652 69.654 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.288 64.288 dbcsr_multiply_generic 2055 12.4 0.105 0.108 51.197 51.450 qs_scf_new_mos 99 7.5 0.000 0.001 47.382 47.516 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.381 47.515 ot_scf_mini 99 9.5 0.002 0.002 44.957 45.091 multiply_cannon 2055 13.4 0.184 0.188 42.705 43.498 velocity_verlet 10 3.0 0.001 0.002 43.213 43.214 multiply_cannon_loop 2055 14.4 1.782 1.828 41.767 42.600 ot_mini 99 10.5 0.001 0.001 26.782 26.891 qs_ot_get_derivative 99 11.5 0.001 0.001 19.927 20.047 multiply_cannon_multrec 49320 15.4 11.362 12.011 17.386 17.937 rebuild_ks_matrix 110 8.3 0.000 0.001 14.426 14.581 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.426 14.580 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.646 12.775 mp_waitall_1 220248 16.4 10.918 11.754 10.918 11.754 multiply_cannon_sync_h2d 49320 15.4 9.612 10.143 9.612 10.143 qs_ot_get_p 110 10.4 0.001 0.001 9.614 9.784 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.874 8.399 multiply_cannon_metrocomm3 49320 15.4 0.082 0.088 6.704 7.985 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.203 7.679 apply_single 110 13.6 0.000 0.001 7.203 7.679 sum_up_and_integrate 110 10.3 0.001 0.002 6.960 6.975 integrate_v_rspace 110 11.3 0.002 0.003 6.935 6.956 init_scf_run 11 5.9 0.000 0.001 6.847 6.847 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.847 6.847 ot_diis_step 99 11.5 0.006 0.006 6.597 6.597 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.428 6.465 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.240 6.345 calculate_rho_elec 110 8.6 0.021 0.024 6.239 6.345 dbcsr_mm_accdrv_process 87628 16.1 3.060 3.154 5.893 6.196 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.559 5.560 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.383 5.446 init_scf_loop 11 6.9 0.000 0.000 5.337 5.338 mp_sum_l 6594 12.7 4.218 5.033 4.218 5.033 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.997 5.018 cp_fm_diag_elpa_base 48 14.0 4.985 5.006 4.996 5.016 wfi_extrapolate 11 7.9 0.001 0.001 3.990 3.990 make_m2s 4110 13.4 0.060 0.066 3.816 3.911 calculate_dm_sparse 110 9.5 0.001 0.001 3.813 3.910 make_images 4110 14.4 0.178 0.190 3.721 3.819 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.423 3.685 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.491 3.496 density_rs2pw 110 9.6 0.004 0.004 3.258 3.460 grid_integrate_task_list 110 12.3 3.269 3.449 3.269 3.449 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.265 3.326 prepare_preconditioner 11 7.9 0.000 0.000 3.188 3.206 make_preconditioner 11 8.9 0.000 0.000 3.188 3.206 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.170 3.206 pw_transfer 1331 11.6 0.055 0.067 3.004 3.083 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.993 3.048 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.916 2.997 calculate_first_density_matrix 1 7.0 0.000 0.001 2.751 2.763 acc_transpose_blocks 49320 15.4 0.213 0.223 2.549 2.625 fft_wrap_pw1pw2_140 451 13.1 0.310 0.333 2.511 2.598 jit_kernel_multiply 13 15.9 2.551 2.573 2.551 2.573 potential_pw2rs 110 12.3 0.005 0.006 2.543 2.558 mp_alltoall_d11v 2046 13.8 2.040 2.404 2.040 2.404 fft3d_ps 1111 14.6 0.798 0.890 2.209 2.267 grid_collocate_task_list 110 9.6 2.159 2.260 2.159 2.260 mp_waitany 14300 13.8 1.815 2.097 1.815 2.097 transfer_rs2pw 451 10.6 0.005 0.006 1.843 2.036 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.909 1.934 mp_sum_d 3891 11.9 1.395 1.894 1.395 1.894 make_images_data 4110 15.4 0.045 0.048 1.734 1.851 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.509 1.762 cp_fm_cholesky_invert 11 10.9 1.744 1.748 1.744 1.748 transfer_pw2rs 451 13.1 0.006 0.007 1.725 1.733 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.663 1.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.682000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.454545, yerr=2.230517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.576512E+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 1365208. 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 0.011 0.028 68.439 68.440 qs_mol_dyn_low 1 2.0 0.003 0.004 68.199 68.210 qs_forces 11 3.9 0.004 0.015 68.123 68.132 qs_energies 11 4.9 0.002 0.002 64.802 64.806 scf_env_do_scf 11 5.9 0.000 0.001 56.411 56.415 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.950 48.950 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.822 38.002 velocity_verlet 10 3.0 0.002 0.003 35.729 35.730 qs_scf_new_mos 99 7.5 0.001 0.001 33.141 33.284 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.140 33.283 multiply_cannon 2055 13.4 0.225 0.243 31.272 32.363 ot_scf_mini 99 9.5 0.003 0.004 31.473 31.603 multiply_cannon_loop 2055 14.4 1.163 1.193 30.033 30.848 ot_mini 99 10.5 0.001 0.001 18.316 18.453 multiply_cannon_multrec 24660 15.4 6.994 8.461 14.067 15.785 rebuild_ks_matrix 110 8.3 0.000 0.000 13.486 13.579 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 13.486 13.579 qs_ot_get_derivative 99 11.5 0.001 0.002 12.545 12.680 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.862 11.939 mp_waitall_1 176588 16.5 7.447 9.976 7.447 9.976 multiply_cannon_metrocomm3 24660 15.4 0.075 0.077 5.073 7.875 multiply_cannon_sync_h2d 24660 15.4 6.413 7.725 6.413 7.725 init_scf_loop 11 6.9 0.000 0.000 7.428 7.428 dbcsr_mm_accdrv_process 52282 16.1 5.513 6.086 6.902 7.198 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.467 7.033 apply_single 110 13.6 0.000 0.001 6.467 7.033 qs_ot_get_p 110 10.4 0.001 0.001 6.509 6.682 sum_up_and_integrate 110 10.3 0.002 0.003 6.337 6.349 integrate_v_rspace 110 11.3 0.002 0.003 6.310 6.324 init_scf_run 11 5.9 0.000 0.001 6.042 6.043 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.042 6.043 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.889 5.898 calculate_rho_elec 110 8.6 0.039 0.047 5.888 5.897 ot_diis_step 99 11.5 0.011 0.013 5.725 5.726 prepare_preconditioner 11 7.9 0.000 0.000 5.426 5.444 make_preconditioner 11 8.9 0.000 0.001 5.426 5.444 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.685 5.388 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.997 5.152 make_m2s 4110 13.4 0.057 0.059 4.201 4.647 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.618 4.637 make_images 4110 14.4 0.408 0.465 4.089 4.532 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.155 4.155 pw_transfer 1331 11.6 0.066 0.073 3.528 3.665 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.583 3.597 cp_fm_diag_elpa_base 48 14.0 3.536 3.550 3.580 3.593 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.422 3.563 wfi_extrapolate 11 7.9 0.001 0.001 3.480 3.480 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.284 3.359 density_rs2pw 110 9.6 0.004 0.004 3.134 3.320 grid_integrate_task_list 110 12.3 3.159 3.297 3.159 3.297 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.206 3.207 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.935 3.074 calculate_dm_sparse 110 9.5 0.001 0.001 2.968 2.995 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.884 2.940 hybrid_alltoall_any 4261 16.3 0.106 0.449 1.978 2.726 make_images_data 4110 15.4 0.050 0.054 2.289 2.721 fft3d_ps 1111 14.6 1.115 1.348 2.514 2.665 cp_fm_cholesky_invert 11 10.9 2.521 2.528 2.521 2.528 calculate_first_density_matrix 1 7.0 0.000 0.000 2.465 2.469 grid_collocate_task_list 110 9.6 2.172 2.378 2.172 2.378 mp_sum_l 6594 12.7 1.718 2.292 1.718 2.292 potential_pw2rs 110 12.3 0.008 0.009 2.223 2.237 acc_transpose_blocks 24660 15.4 0.118 0.122 1.964 2.002 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.961 1.974 jit_kernel_multiply 10 16.3 1.020 1.961 1.020 1.961 mp_alltoall_d11v 2046 13.8 1.727 1.916 1.727 1.916 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.848 1.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.780 1.799 multiply_cannon_metrocomm4 22605 15.4 0.075 0.080 0.767 1.682 transfer_rs2pw 451 10.6 0.006 0.007 1.436 1.607 mp_irecv_dv 57340 16.2 0.641 1.566 0.641 1.566 mp_allgather_i34 2055 14.4 0.548 1.559 0.548 1.559 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.545 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.492 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.440000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.545455, yerr=7.278407 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 669.278208E+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 0.015 0.028 60.219 60.221 qs_mol_dyn_low 1 2.0 0.003 0.003 59.906 59.915 qs_forces 11 3.9 0.007 0.034 59.835 59.836 qs_energies 11 4.9 0.003 0.009 56.583 56.587 scf_env_do_scf 11 5.9 0.001 0.001 48.726 48.727 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.174 40.174 velocity_verlet 10 3.0 0.001 0.001 32.578 32.580 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.708 28.973 qs_scf_new_mos 99 7.5 0.001 0.001 25.495 25.595 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.494 25.595 ot_scf_mini 99 9.5 0.003 0.003 24.244 24.368 multiply_cannon 2055 13.4 0.211 0.220 22.500 23.504 multiply_cannon_loop 2055 14.4 0.819 0.843 21.438 22.546 ot_mini 99 10.5 0.001 0.001 13.776 13.896 rebuild_ks_matrix 110 8.3 0.000 0.000 12.236 12.374 qs_ks_build_kohn_sham_matrix 110 9.3 0.030 0.086 12.236 12.373 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.729 10.855 multiply_cannon_multrec 16440 15.4 3.592 4.586 9.775 10.681 mp_waitall_1 139946 16.5 7.099 9.923 7.099 9.923 qs_ot_get_derivative 99 11.5 0.001 0.001 9.313 9.433 init_scf_loop 11 6.9 0.001 0.004 8.520 8.521 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.553 7.364 prepare_preconditioner 11 7.9 0.000 0.000 6.769 6.787 make_preconditioner 11 8.9 0.000 0.001 6.769 6.787 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.113 6.478 sum_up_and_integrate 110 10.3 0.002 0.003 6.306 6.319 integrate_v_rspace 110 11.3 0.003 0.004 6.280 6.293 dbcsr_mm_accdrv_process 34862 16.1 5.381 5.757 6.027 6.176 qs_ot_get_p 110 10.4 0.001 0.003 5.632 5.801 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.779 5.788 calculate_rho_elec 110 8.6 0.059 0.059 5.778 5.788 init_scf_run 11 5.9 0.000 0.001 5.441 5.441 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.441 5.441 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.870 5.377 apply_single 110 13.6 0.000 0.000 4.869 5.377 make_m2s 4110 13.4 0.049 0.051 4.155 4.443 ot_diis_step 99 11.5 0.011 0.011 4.430 4.430 make_images 4110 14.4 0.399 0.523 4.038 4.326 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.963 3.967 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.262 3.850 multiply_cannon_sync_h2d 16440 15.4 3.218 3.832 3.218 3.832 pw_transfer 1331 11.6 0.065 0.070 3.657 3.667 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.591 3.592 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.551 3.563 grid_integrate_task_list 110 12.3 3.205 3.454 3.205 3.454 density_rs2pw 110 9.6 0.004 0.004 3.000 3.223 fft_wrap_pw1pw2_140 451 13.1 0.461 0.471 3.073 3.084 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.041 3.056 cp_fm_diag_elpa_base 48 14.0 2.974 3.003 3.039 3.053 wfi_extrapolate 11 7.9 0.001 0.001 2.982 2.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.753 2.755 make_images_data 4110 15.4 0.048 0.052 2.318 2.689 cp_fm_cholesky_invert 11 10.9 2.619 2.625 2.619 2.625 hybrid_alltoall_any 4261 16.3 0.109 0.376 2.047 2.605 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.557 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.457 2.519 fft3d_ps 1111 14.6 1.120 1.132 2.467 2.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.418 2.480 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.894 2.437 grid_collocate_task_list 110 9.6 2.223 2.420 2.223 2.420 calculate_first_density_matrix 1 7.0 0.001 0.003 2.368 2.370 mp_irecv_dv 48980 15.7 0.818 2.296 0.818 2.296 potential_pw2rs 110 12.3 0.011 0.011 2.157 2.163 mp_sum_l 6594 12.7 1.430 2.149 1.430 2.149 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.979 1.980 mp_alltoall_d11v 2046 13.8 1.763 1.957 1.763 1.957 dbcsr_complete_redistribute 325 12.2 0.324 0.353 1.434 1.888 cp_fm_upper_to_full 70 14.2 1.401 1.774 1.401 1.774 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.720 acc_transpose_blocks 16440 15.4 0.080 0.082 1.578 1.624 cp_fm_cholesky_decompose 22 10.9 1.602 1.623 1.602 1.623 transfer_rs2pw 451 10.6 0.005 0.005 1.265 1.499 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.474 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.473 mp_allgather_i34 2055 14.4 0.377 1.420 0.377 1.420 mp_waitany 17072 13.8 1.170 1.402 1.170 1.402 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.937 1.384 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.288 1.297 transfer_pw2rs 451 13.1 0.005 0.005 1.236 1.256 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.221000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=634.090909, yerr=9.069976 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 744.845312E+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 0.009 0.026 66.288 66.289 qs_mol_dyn_low 1 2.0 0.003 0.003 66.062 66.071 qs_forces 11 3.9 0.003 0.003 65.994 65.995 qs_energies 11 4.9 0.014 0.017 62.599 62.603 scf_env_do_scf 11 5.9 0.000 0.001 54.075 54.078 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.524 42.525 velocity_verlet 10 3.0 0.001 0.002 37.252 37.254 dbcsr_multiply_generic 2055 12.4 0.307 0.313 30.997 31.254 qs_scf_new_mos 99 7.5 0.001 0.001 27.790 27.913 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.790 27.913 ot_scf_mini 99 9.5 0.002 0.003 26.096 26.193 multiply_cannon 2055 13.4 0.238 0.258 23.668 24.756 multiply_cannon_loop 2055 14.4 1.421 1.478 22.277 22.989 ot_mini 99 10.5 0.001 0.001 15.019 15.136 multiply_cannon_multrec 24660 15.4 4.069 6.831 12.991 14.032 rebuild_ks_matrix 110 8.3 0.000 0.000 12.062 12.163 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.062 12.163 init_scf_loop 11 6.9 0.000 0.000 11.513 11.514 qs_ot_get_derivative 99 11.5 0.001 0.001 10.839 10.936 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.668 10.758 dbcsr_mm_accdrv_process 52304 16.0 7.745 9.124 8.764 9.778 prepare_preconditioner 11 7.9 0.000 0.000 9.755 9.768 make_preconditioner 11 8.9 0.000 0.001 9.754 9.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.984 9.440 sum_up_and_integrate 110 10.3 0.001 0.002 6.302 6.313 integrate_v_rspace 110 11.3 0.003 0.003 6.276 6.286 mp_waitall_1 121746 16.5 4.363 6.185 4.363 6.185 qs_ot_get_p 110 10.4 0.001 0.002 5.963 6.103 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.868 5.877 calculate_rho_elec 110 8.6 0.078 0.081 5.867 5.876 make_m2s 4110 13.4 0.059 0.061 5.479 5.774 init_scf_run 11 5.9 0.000 0.001 5.731 5.732 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.731 5.731 make_images 4110 14.4 0.577 0.701 5.338 5.628 cp_fm_upper_to_full 70 14.2 3.318 4.819 3.318 4.819 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.054 4.148 apply_single 110 13.6 0.000 0.000 4.054 4.148 ot_diis_step 99 11.5 0.011 0.011 4.143 4.144 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.048 4.064 pw_transfer 1331 11.6 0.065 0.073 3.813 3.848 dbcsr_complete_redistribute 325 12.2 0.492 0.554 2.707 3.830 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.706 3.744 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.518 3.569 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.480 3.481 grid_integrate_task_list 110 12.3 3.272 3.436 3.272 3.436 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.358 3.412 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.197 3.312 fft_wrap_pw1pw2_140 451 13.1 0.531 0.543 3.172 3.213 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.479 3.173 density_rs2pw 110 9.6 0.004 0.004 2.982 3.151 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.083 3.084 make_images_data 4110 15.4 0.050 0.055 2.711 3.038 calculate_dm_sparse 110 9.5 0.001 0.001 2.992 3.030 hybrid_alltoall_any 4261 16.3 0.123 0.461 2.259 3.000 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.941 2.950 cp_fm_diag_elpa_base 48 14.0 2.793 2.854 2.938 2.947 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.762 2.861 mp_alltoall_i22 605 13.7 1.645 2.801 1.645 2.801 calculate_first_density_matrix 1 7.0 0.000 0.001 2.685 2.689 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.643 2.681 cp_fm_cholesky_invert 11 10.9 2.632 2.641 2.632 2.641 fft3d_ps 1111 14.6 1.153 1.196 2.532 2.548 multiply_cannon_sync_h2d 24660 15.4 2.367 2.534 2.367 2.534 acc_transpose_blocks 24660 15.4 0.114 0.119 2.425 2.514 grid_collocate_task_list 110 9.6 2.261 2.444 2.261 2.444 qs_energies_init_hamiltonians 11 5.9 0.012 0.051 2.292 2.293 potential_pw2rs 110 12.3 0.012 0.013 2.094 2.101 mp_alltoall_d11v 2046 13.8 1.803 2.040 1.803 2.040 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.804 1.844 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.720 cp_fm_cholesky_decompose 22 10.9 1.626 1.666 1.626 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.646 mp_sum_l 6594 12.7 0.965 1.627 0.965 1.627 mp_allgather_i34 2055 14.4 0.466 1.574 0.466 1.574 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.558 acc_transpose_blocks_sync 73980 16.4 1.408 1.491 1.408 1.491 jit_kernel_multiply 8 15.6 0.670 1.484 0.670 1.484 multiply_cannon_metrocomm4 20550 15.4 0.061 0.064 0.870 1.472 mp_irecv_dv 62702 16.1 0.764 1.385 0.764 1.385 transfer_rs2pw 451 10.6 0.005 0.006 1.168 1.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.289000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.090909, yerr=11.508531 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 865.406976E+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 0.027 57.324 57.325 qs_mol_dyn_low 1 2.0 0.003 0.003 57.093 57.103 qs_forces 11 3.9 0.003 0.003 57.019 57.020 qs_energies 11 4.9 0.002 0.002 53.268 53.275 scf_env_do_scf 11 5.9 0.000 0.001 45.017 45.017 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 37.270 37.270 velocity_verlet 10 3.0 0.002 0.002 32.014 32.018 dbcsr_multiply_generic 2055 12.4 0.108 0.110 23.620 23.791 qs_scf_new_mos 99 7.5 0.001 0.001 22.047 22.108 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.047 22.108 ot_scf_mini 99 9.5 0.002 0.002 20.773 20.813 multiply_cannon 2055 13.4 0.238 0.247 17.743 19.258 multiply_cannon_loop 2055 14.4 0.605 0.631 16.421 16.782 rebuild_ks_matrix 110 8.3 0.000 0.000 12.004 12.030 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.004 12.030 ot_mini 99 10.5 0.001 0.001 10.991 11.025 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.643 10.667 multiply_cannon_multrec 8220 15.4 3.182 4.442 7.576 8.518 mp_waitall_1 103326 16.6 6.090 7.765 6.090 7.765 init_scf_loop 11 6.9 0.000 0.000 7.695 7.696 qs_ot_get_derivative 99 11.5 0.001 0.001 7.287 7.327 sum_up_and_integrate 110 10.3 0.001 0.002 6.467 6.479 integrate_v_rspace 110 11.3 0.003 0.003 6.440 6.452 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.355 6.369 calculate_rho_elec 110 8.6 0.113 0.113 6.355 6.368 prepare_preconditioner 11 7.9 0.000 0.000 6.049 6.053 make_preconditioner 11 8.9 0.001 0.001 6.049 6.053 qs_ot_get_p 110 10.4 0.001 0.001 5.739 5.779 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.631 5.705 dbcsr_mm_accdrv_process 17442 15.9 3.008 4.087 4.254 5.172 init_scf_run 11 5.9 0.000 0.001 5.084 5.084 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.083 5.084 make_m2s 4110 13.4 0.037 0.039 4.429 4.717 make_images 4110 14.4 0.645 0.708 4.299 4.588 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.931 4.450 pw_transfer 1331 11.6 0.066 0.070 4.400 4.408 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.292 4.304 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.215 4.219 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.858 3.859 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.731 3.783 apply_single 110 13.6 0.000 0.000 3.731 3.783 fft_wrap_pw1pw2_140 451 13.1 0.720 0.736 3.689 3.702 ot_diis_step 99 11.5 0.012 0.012 3.679 3.680 grid_integrate_task_list 110 12.3 3.381 3.540 3.381 3.540 density_rs2pw 110 9.6 0.004 0.004 3.255 3.419 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.240 3.248 cp_fm_diag_elpa_base 48 14.0 3.182 3.210 3.238 3.246 cp_fm_cholesky_invert 11 10.9 2.908 2.911 2.908 2.911 fft3d_ps 1111 14.6 1.292 1.300 2.837 2.846 hybrid_alltoall_any 4261 16.3 0.200 0.848 2.352 2.830 make_images_data 4110 15.4 0.042 0.047 2.388 2.767 wfi_extrapolate 11 7.9 0.001 0.001 2.713 2.713 qs_energies_init_hamiltonians 11 5.9 0.013 0.017 2.679 2.679 grid_collocate_task_list 110 9.6 2.366 2.559 2.366 2.559 calculate_dm_sparse 110 9.5 0.001 0.001 2.505 2.547 multiply_cannon_sync_h2d 8220 15.4 2.370 2.503 2.370 2.503 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.501 2.502 calculate_first_density_matrix 1 7.0 0.000 0.000 2.273 2.276 potential_pw2rs 110 12.3 0.015 0.015 2.180 2.183 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.136 2.151 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.115 2.149 mp_alltoall_d11v 2046 13.8 1.902 2.091 1.902 2.091 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.055 2.083 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.013 cp_fm_cholesky_decompose 22 10.9 1.692 1.709 1.692 1.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.695 mp_allgather_i34 2055 14.4 0.494 1.662 0.494 1.662 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.497 1.614 dbcsr_complete_redistribute 325 12.2 0.561 0.583 1.466 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.447 1.462 acc_transpose_blocks 8220 15.4 0.042 0.043 1.368 1.403 jit_kernel_multiply 8 15.7 0.925 1.376 0.925 1.376 transfer_rs2pw 451 10.6 0.005 0.005 1.170 1.337 mp_alltoall_z22v 1111 16.6 1.313 1.330 1.313 1.330 multiply_cannon_metrocomm1 8220 15.4 0.023 0.023 0.887 1.325 qs_create_task_list 11 7.9 0.001 0.002 1.210 1.308 generate_qs_task_list 11 8.9 0.373 0.443 1.208 1.307 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.223 1.258 mp_waitany 9240 13.8 1.057 1.234 1.057 1.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.325000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=817.636364, yerr=12.085098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 1.395032E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.040 90.171 90.173 qs_mol_dyn_low 1 2.0 0.003 0.003 89.858 89.872 qs_forces 11 3.9 0.004 0.005 89.787 89.788 qs_energies 11 4.9 0.002 0.002 85.607 85.609 scf_env_do_scf 11 5.9 0.000 0.001 75.473 75.474 velocity_verlet 10 3.0 0.002 0.004 56.230 56.236 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 47.079 47.081 dbcsr_multiply_generic 2055 12.4 0.123 0.127 30.224 30.288 qs_scf_new_mos 99 7.5 0.001 0.001 28.752 28.781 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.751 28.780 init_scf_loop 11 6.9 0.000 0.000 28.322 28.326 ot_scf_mini 99 9.5 0.002 0.003 26.914 26.938 prepare_preconditioner 11 7.9 0.000 0.000 26.302 26.309 make_preconditioner 11 8.9 0.000 0.000 26.302 26.309 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.612 25.785 multiply_cannon 2055 13.4 0.337 0.369 22.809 23.612 multiply_cannon_loop 2055 14.4 0.825 0.845 20.973 21.361 cp_fm_upper_to_full 70 14.2 12.640 18.059 12.640 18.059 ot_mini 99 10.5 0.001 0.001 14.629 14.653 rebuild_ks_matrix 110 8.3 0.000 0.001 14.194 14.224 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.194 14.223 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.886 12.911 dbcsr_complete_redistribute 325 12.2 1.004 1.022 7.275 10.428 qs_ot_get_derivative 99 11.5 0.001 0.001 9.898 9.923 multiply_cannon_multrec 8220 15.4 4.058 4.223 9.745 9.849 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.240 9.388 mp_waitall_1 84994 16.7 7.967 8.803 7.967 8.803 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.674 8.793 mp_alltoall_i22 605 13.7 5.304 8.473 5.304 8.473 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.658 7.694 calculate_rho_elec 110 8.6 0.222 0.223 7.657 7.694 sum_up_and_integrate 110 10.3 0.002 0.002 7.035 7.050 integrate_v_rspace 110 11.3 0.004 0.004 7.007 7.022 qs_ot_get_p 110 10.4 0.003 0.003 6.839 6.867 make_m2s 4110 13.4 0.043 0.044 5.521 6.033 pw_transfer 1331 11.6 0.074 0.075 5.985 5.997 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.869 5.880 init_scf_run 11 5.9 0.000 0.001 5.841 5.841 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.841 5.841 make_images 4110 14.4 0.882 0.941 5.331 5.840 dbcsr_mm_accdrv_process 11614 15.7 3.855 4.122 5.538 5.735 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.199 5.568 cp_fm_cholesky_invert 11 10.9 5.390 5.395 5.390 5.395 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.846 5.287 apply_single 110 13.6 0.000 0.000 4.846 5.287 fft_wrap_pw1pw2_140 451 13.1 1.368 1.375 5.115 5.129 qs_ot_p2m_diag 48 11.0 0.151 0.155 5.054 5.061 ot_diis_step 99 11.5 0.015 0.015 4.707 4.707 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.501 4.502 density_rs2pw 110 9.6 0.004 0.004 4.094 4.113 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.843 3.843 cp_fm_diag_elpa_base 48 14.0 3.296 3.494 3.841 3.841 grid_integrate_task_list 110 12.3 3.685 3.743 3.685 3.743 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.679 3.680 hybrid_alltoall_any 4261 16.3 0.264 0.563 2.939 3.621 fft3d_ps 1111 14.6 1.880 1.887 3.572 3.580 make_images_data 4110 15.4 0.047 0.050 2.893 3.578 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.031 3.484 wfi_extrapolate 11 7.9 0.001 0.001 3.347 3.347 calculate_dm_sparse 110 9.5 0.001 0.001 3.202 3.234 multiply_cannon_sync_h2d 8220 15.4 3.129 3.146 3.129 3.146 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.924 2.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.862 grid_collocate_task_list 110 9.6 2.673 2.690 2.673 2.690 potential_pw2rs 110 12.3 0.021 0.021 2.535 2.538 calculate_first_density_matrix 1 7.0 0.000 0.000 2.377 2.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.364 2.375 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.199 2.248 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.221 mp_alltoall_d11v 2046 13.8 2.117 2.175 2.117 2.175 cp_fm_cholesky_decompose 22 10.9 2.056 2.069 2.056 2.069 qs_create_task_list 11 7.9 0.002 0.003 1.882 1.926 generate_qs_task_list 11 8.9 0.730 0.784 1.879 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.908 1.912 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.788 1.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.173000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1245.181818, yerr=68.688782 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.416512E+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 3386795. 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 0.009 0.028 206.424 206.425 qs_mol_dyn_low 1 2.0 0.003 0.003 206.002 206.016 qs_forces 11 3.9 0.019 0.034 205.915 205.916 qs_energies 11 4.9 0.002 0.003 200.317 200.347 scf_env_do_scf 11 5.9 0.001 0.001 183.560 183.564 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.222 163.224 dbcsr_multiply_generic 2507 12.6 0.177 0.182 125.224 125.921 qs_scf_new_mos 117 7.6 0.001 0.001 123.882 124.007 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.881 124.006 velocity_verlet 10 3.0 0.001 0.001 123.064 123.065 ot_scf_mini 117 9.6 0.003 0.003 117.186 117.323 multiply_cannon 2507 13.6 0.236 0.243 101.441 103.102 multiply_cannon_loop 2507 14.6 2.391 2.443 99.293 101.058 ot_mini 117 10.6 0.001 0.001 65.767 65.925 multiply_cannon_multrec 60168 15.6 31.825 34.003 41.611 43.554 qs_ot_get_derivative 117 11.6 0.001 0.001 40.924 41.050 rebuild_ks_matrix 128 8.3 0.001 0.001 33.503 33.996 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.502 33.996 mp_waitall_1 267128 16.5 28.746 31.846 28.746 31.846 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.077 30.475 qs_ot_get_p 128 10.4 0.001 0.001 29.077 29.391 multiply_cannon_sync_h2d 60168 15.6 26.327 28.080 26.327 28.080 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.331 24.914 apply_single 128 13.6 0.001 0.001 24.331 24.914 ot_diis_step 117 11.6 0.008 0.008 24.559 24.560 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.537 22.582 init_scf_loop 11 6.9 0.000 0.000 20.265 20.266 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.668 19.668 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.466 19.619 multiply_cannon_metrocomm3 60168 15.6 0.118 0.121 15.976 18.401 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.696 16.735 cp_fm_diag_elpa_base 83 14.4 16.629 16.669 16.692 16.732 prepare_preconditioner 11 7.9 0.000 0.000 15.714 15.751 make_preconditioner 11 8.9 0.000 0.000 15.714 15.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.953 15.154 make_m2s 5014 13.6 0.105 0.115 13.741 14.102 sum_up_and_integrate 128 10.3 0.002 0.004 13.962 13.978 make_images 5014 14.6 0.398 0.418 13.557 13.922 integrate_v_rspace 128 11.3 0.003 0.004 13.904 13.921 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.117 13.235 calculate_rho_elec 128 8.7 0.045 0.064 13.116 13.234 init_scf_run 11 5.9 0.000 0.001 12.530 12.531 scf_env_initial_rho_setup 11 6.9 0.061 0.136 12.530 12.531 dbcsr_mm_accdrv_process 124484 16.2 4.727 4.863 9.352 9.877 mp_sum_l 7950 12.9 8.778 9.824 8.778 9.824 wfi_extrapolate 11 7.9 0.001 0.001 9.054 9.054 cp_fm_cholesky_invert 11 10.9 9.010 9.017 9.010 9.017 calculate_dm_sparse 128 9.5 0.001 0.001 8.497 8.608 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.209 8.305 multiply_cannon_metrocomm1 60168 15.6 0.094 0.100 6.306 8.285 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.097 8.247 pw_transfer 1547 11.6 0.073 0.093 7.606 7.790 make_images_data 5014 15.6 0.068 0.073 6.767 7.616 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.405 7.584 grid_integrate_task_list 128 12.3 7.064 7.434 7.064 7.434 density_rs2pw 128 9.7 0.006 0.006 6.747 7.342 hybrid_alltoall_any 5200 16.5 0.293 2.271 5.929 7.187 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.783 6.795 fft_wrap_pw1pw2_140 523 13.2 0.856 0.896 6.529 6.697 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.744 5.828 fft3d_ps 1291 14.7 2.193 2.846 5.415 5.751 mp_alltoall_d11v 2415 14.1 4.308 5.594 4.308 5.594 grid_collocate_task_list 128 9.7 4.843 5.205 4.843 5.205 cp_fm_cholesky_decompose 22 10.9 4.538 4.553 4.538 4.553 potential_pw2rs 128 12.3 0.009 0.010 4.459 4.479 mp_sum_d 4467 12.1 3.495 4.211 3.495 4.211 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.425000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.636364, yerr=4.995866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 841.076736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 4842752. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 189.408 189.410 qs_mol_dyn_low 1 2.0 0.003 0.003 189.007 189.022 qs_forces 11 3.9 0.005 0.011 188.921 188.922 qs_energies 11 4.9 0.002 0.003 182.170 182.181 scf_env_do_scf 11 5.9 0.001 0.001 165.950 165.960 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 133.294 133.296 velocity_verlet 10 3.0 0.001 0.002 118.716 118.718 dbcsr_multiply_generic 2507 12.6 0.189 0.196 97.435 98.741 qs_scf_new_mos 117 7.6 0.001 0.001 94.885 95.509 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.884 95.508 ot_scf_mini 117 9.6 0.004 0.004 90.053 90.672 multiply_cannon 2507 13.6 0.502 0.561 77.226 81.763 multiply_cannon_loop 2507 14.6 1.570 1.638 73.865 76.427 ot_mini 117 10.6 0.001 0.001 50.227 50.841 mp_waitall_1 214728 16.6 24.186 39.457 24.186 39.457 multiply_cannon_multrec 30084 15.6 20.887 25.727 31.526 36.634 rebuild_ks_matrix 128 8.3 0.001 0.001 32.202 32.971 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.025 32.201 32.971 init_scf_loop 11 6.9 0.000 0.000 32.566 32.567 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.958 29.653 multiply_cannon_metrocomm3 30084 15.6 0.100 0.106 15.706 29.225 qs_ot_get_derivative 117 11.6 0.001 0.002 28.073 28.664 prepare_preconditioner 11 7.9 0.000 0.000 28.260 28.336 make_preconditioner 11 8.9 0.000 0.002 28.260 28.336 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.944 27.456 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.227 23.341 apply_single 128 13.6 0.001 0.001 22.227 23.340 qs_ot_get_p 128 10.4 0.001 0.002 21.804 22.545 ot_diis_step 117 11.6 0.015 0.015 21.980 21.982 multiply_cannon_sync_h2d 30084 15.6 17.954 20.658 17.954 20.658 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.863 16.896 cp_fm_cholesky_invert 11 10.9 16.475 16.488 16.475 16.488 make_m2s 5014 13.6 0.086 0.093 14.270 15.942 make_images 5014 14.6 1.168 1.379 14.054 15.721 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.639 15.640 sum_up_and_integrate 128 10.3 0.003 0.005 13.914 13.946 integrate_v_rspace 128 11.3 0.003 0.005 13.853 13.889 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.308 13.346 calculate_rho_elec 128 8.7 0.087 0.103 13.307 13.346 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.440 12.473 cp_fm_diag_elpa_base 83 14.4 12.186 12.274 12.435 12.462 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.135 11.575 init_scf_run 11 5.9 0.000 0.001 11.423 11.424 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.423 11.424 multiply_cannon_metrocomm4 27577 15.6 0.101 0.115 3.897 11.304 mp_irecv_dv 69486 16.3 3.694 10.898 3.694 10.898 dbcsr_mm_accdrv_process 62242 16.2 5.472 6.344 10.083 10.659 make_images_data 5014 15.6 0.068 0.078 8.496 10.441 hybrid_alltoall_any 5200 16.5 0.349 1.535 7.290 9.915 pw_transfer 1547 11.6 0.084 0.098 8.531 8.593 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.308 8.370 wfi_extrapolate 11 7.9 0.001 0.001 8.327 8.327 density_rs2pw 128 9.7 0.006 0.006 7.020 7.638 grid_integrate_task_list 128 12.3 7.178 7.467 7.178 7.467 fft_wrap_pw1pw2_140 523 13.2 0.934 0.960 7.372 7.433 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.250 7.045 cp_fm_cholesky_decompose 22 10.9 6.852 6.935 6.852 6.935 calculate_dm_sparse 128 9.5 0.001 0.001 6.489 6.644 mp_sum_l 7950 12.9 4.137 6.286 4.137 6.286 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.077 6.086 fft3d_ps 1291 14.7 2.826 3.020 5.866 5.933 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.422 5.491 grid_collocate_task_list 128 9.7 5.006 5.413 5.006 5.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.201 5.397 mp_allgather_i34 2507 14.6 1.889 5.146 1.889 5.146 mp_alltoall_d11v 2415 14.1 4.112 4.987 4.112 4.987 potential_pw2rs 128 12.3 0.015 0.018 4.432 4.451 mp_sum_d 4472 12.1 2.669 4.112 2.669 4.112 dbcsr_complete_redistribute 395 12.7 0.762 0.888 3.148 3.999 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.410000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.363636, yerr=2.346089 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 941.379584E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.107 174.041 174.091 qs_mol_dyn_low 1 2.0 0.003 0.004 173.546 173.611 qs_forces 11 3.9 0.006 0.016 173.436 173.442 qs_energies 11 4.9 0.003 0.010 166.924 166.939 scf_env_do_scf 11 5.9 0.001 0.006 151.475 151.476 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 116.926 116.927 velocity_verlet 10 3.0 0.001 0.002 111.423 111.425 dbcsr_multiply_generic 2507 12.6 0.185 0.188 81.827 82.926 qs_scf_new_mos 117 7.6 0.001 0.001 79.989 80.277 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.988 80.277 ot_scf_mini 117 9.6 0.004 0.005 75.736 76.123 multiply_cannon 2507 13.6 0.508 0.527 62.130 66.729 multiply_cannon_loop 2507 14.6 1.135 1.203 59.030 61.656 ot_mini 117 10.6 0.001 0.001 42.478 42.864 init_scf_loop 11 6.9 0.001 0.010 34.445 34.446 mp_waitall_1 170520 16.6 24.344 33.526 24.344 33.526 rebuild_ks_matrix 128 8.3 0.001 0.001 30.085 30.567 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.023 30.084 30.567 prepare_preconditioner 11 7.9 0.000 0.000 30.454 30.510 make_preconditioner 11 8.9 0.000 0.002 30.454 30.510 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.152 29.512 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.120 27.555 multiply_cannon_multrec 20056 15.6 13.127 16.343 22.488 25.720 multiply_cannon_metrocomm3 20056 15.6 0.062 0.066 15.038 24.432 qs_ot_get_derivative 117 11.6 0.001 0.002 22.951 23.330 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.598 20.567 apply_single 128 13.6 0.001 0.001 19.597 20.567 ot_diis_step 117 11.6 0.018 0.018 19.417 19.417 qs_ot_get_p 128 10.4 0.002 0.003 18.624 19.126 make_m2s 5014 13.6 0.078 0.082 14.654 15.987 make_images 5014 14.6 1.154 1.243 14.418 15.749 multiply_cannon_sync_h2d 20056 15.6 13.666 15.576 13.666 15.576 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.319 14.327 cp_fm_cholesky_invert 11 10.9 14.143 14.152 14.143 14.152 sum_up_and_integrate 128 10.3 0.002 0.004 13.850 13.878 integrate_v_rspace 128 11.3 0.003 0.005 13.791 13.820 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.507 13.548 calculate_rho_elec 128 8.7 0.129 0.143 13.507 13.548 cp_dbcsr_syevd 83 12.4 0.005 0.007 13.271 13.272 make_images_data 5014 15.6 0.066 0.075 8.762 10.564 init_scf_run 11 5.9 0.000 0.001 10.400 10.400 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.400 10.400 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.125 10.146 cp_fm_diag_elpa_base 83 14.4 9.719 9.884 10.121 10.142 hybrid_alltoall_any 5200 16.5 0.448 2.018 7.653 9.403 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.071 9.344 multiply_cannon_metrocomm4 17549 15.6 0.066 0.075 3.449 9.220 dbcsr_mm_accdrv_process 41502 16.2 5.596 5.939 8.821 8.975 mp_irecv_dv 50230 16.2 3.318 8.962 3.318 8.962 pw_transfer 1547 11.6 0.084 0.101 8.763 8.864 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.540 8.646 grid_integrate_task_list 128 12.3 7.341 7.759 7.341 7.759 fft_wrap_pw1pw2_140 523 13.2 1.066 1.101 7.603 7.718 wfi_extrapolate 11 7.9 0.001 0.001 7.410 7.410 density_rs2pw 128 9.7 0.005 0.006 6.868 7.238 cp_fm_upper_to_full 105 14.8 5.744 7.233 5.744 7.233 cp_fm_cholesky_decompose 22 10.9 7.200 7.230 7.200 7.230 dbcsr_complete_redistribute 395 12.7 1.161 1.187 4.455 6.213 calculate_dm_sparse 128 9.5 0.001 0.001 5.845 5.968 fft3d_ps 1291 14.7 2.820 3.051 5.850 5.953 grid_collocate_task_list 128 9.7 5.161 5.528 5.161 5.528 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.425 5.430 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.630 5.307 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.324 5.076 mp_sum_l 7950 12.9 3.211 4.862 3.211 4.862 mp_allgather_i34 2507 14.6 1.604 4.826 1.604 4.826 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.549 4.684 mp_alltoall_d11v 2415 14.1 4.161 4.573 4.161 4.573 potential_pw2rs 128 12.3 0.020 0.022 4.313 4.322 transfer_fm_to_dbcsr 11 9.9 0.020 0.024 2.282 3.985 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.915 3.937 mp_alltoall_i22 716 14.1 1.914 3.805 1.914 3.805 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.743 3.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.464 3.511 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.091000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.090909, yerr=9.346410 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.154060E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509575. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 188.896 188.897 qs_mol_dyn_low 1 2.0 0.003 0.003 188.544 188.559 qs_forces 11 3.9 0.004 0.004 188.433 188.444 qs_energies 11 4.9 0.002 0.003 181.359 181.368 scf_env_do_scf 11 5.9 0.001 0.001 164.723 164.732 velocity_verlet 10 3.0 0.001 0.002 125.280 125.283 scf_env_do_scf_inner_loop 118 6.6 0.005 0.008 118.445 118.448 qs_scf_new_mos 118 7.6 0.001 0.001 81.750 82.041 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.750 82.040 dbcsr_multiply_generic 2529 12.6 0.195 0.199 81.106 81.981 ot_scf_mini 118 9.6 0.003 0.003 77.135 77.450 multiply_cannon 2529 13.6 0.557 0.604 56.082 60.260 multiply_cannon_loop 2529 14.6 1.851 1.937 52.451 54.326 init_scf_loop 11 6.9 0.000 0.000 46.157 46.164 ot_mini 118 10.6 0.001 0.001 43.356 43.654 prepare_preconditioner 11 7.9 0.000 0.000 41.906 41.939 make_preconditioner 11 8.9 0.000 0.001 41.906 41.939 make_full_inverse_cholesky 11 9.9 0.011 0.020 35.435 40.491 multiply_cannon_multrec 30348 15.6 13.846 19.058 26.844 31.719 rebuild_ks_matrix 129 8.3 0.001 0.001 29.512 29.821 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.023 29.512 29.821 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.656 26.929 mp_waitall_1 149172 16.7 17.108 25.944 17.108 25.944 qs_ot_get_derivative 118 11.6 0.001 0.002 23.622 23.933 make_m2s 5058 13.6 0.093 0.097 20.499 21.488 make_images 5058 14.6 1.962 2.293 20.186 21.173 qs_ot_get_p 129 10.4 0.068 0.140 19.972 20.320 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.070 19.668 apply_single 129 13.6 0.001 0.001 19.069 19.668 ot_diis_step 118 11.6 0.018 0.019 19.599 19.601 cp_fm_upper_to_full 106 14.8 11.517 16.986 11.517 16.986 cp_fm_cholesky_invert 11 10.9 15.828 15.836 15.828 15.836 qs_ot_p2m_diag 84 11.4 0.346 0.394 15.519 15.574 multiply_cannon_metrocomm3 30348 15.6 0.049 0.052 6.220 14.910 sum_up_and_integrate 129 10.3 0.002 0.003 14.093 14.120 cp_dbcsr_syevd 84 12.4 0.005 0.006 14.095 14.096 integrate_v_rspace 129 11.3 0.003 0.004 14.033 14.062 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.926 13.973 calculate_rho_elec 129 8.7 0.171 0.186 13.925 13.972 dbcsr_mm_accdrv_process 62780 16.2 8.504 9.225 12.568 13.124 dbcsr_complete_redistribute 397 12.7 1.483 1.600 9.190 13.053 make_images_data 5058 15.6 0.068 0.074 10.972 12.659 multiply_cannon_sync_h2d 30348 15.6 10.619 11.603 10.619 11.603 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.830 11.601 hybrid_alltoall_any 5245 16.5 0.525 2.093 9.765 11.358 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.837 10.856 cp_fm_diag_elpa_base 84 14.4 9.868 10.199 10.830 10.848 init_scf_run 11 5.9 0.000 0.001 10.722 10.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.722 10.723 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.995 10.219 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.448 10.116 mp_alltoall_i22 720 14.1 5.650 9.563 5.650 9.563 pw_transfer 1559 11.6 0.086 0.102 9.286 9.353 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.058 9.131 fft_wrap_pw1pw2_140 527 13.2 1.261 1.296 8.049 8.135 grid_integrate_task_list 129 12.3 7.606 8.003 7.606 8.003 wfi_extrapolate 11 7.9 0.001 0.001 7.661 7.661 cp_fm_cholesky_decompose 22 10.9 7.399 7.514 7.399 7.514 density_rs2pw 129 9.7 0.005 0.006 6.911 7.197 multiply_cannon_metrocomm4 25290 15.6 0.082 0.090 2.752 6.608 calculate_dm_sparse 129 9.5 0.001 0.001 6.298 6.404 mp_irecv_dv 76751 16.2 2.597 6.338 2.597 6.338 fft3d_ps 1301 14.7 3.016 3.078 6.079 6.118 grid_collocate_task_list 129 9.7 5.317 5.692 5.317 5.692 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.436 5.494 mp_alltoall_d11v 2429 14.1 5.014 5.360 5.014 5.360 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.500 4.593 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.341 4.430 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.415 4.416 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.287 4.346 potential_pw2rs 129 12.3 0.023 0.023 4.293 4.304 mp_allgather_i34 2529 14.6 1.704 3.964 1.704 3.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.897000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.363636, yerr=19.868992 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.543221E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57807. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 170.616 170.617 qs_mol_dyn_low 1 2.0 0.003 0.003 170.233 170.247 qs_forces 11 3.9 0.004 0.004 170.107 170.114 qs_energies 11 4.9 0.002 0.002 162.719 162.727 scf_env_do_scf 11 5.9 0.001 0.001 145.616 145.625 velocity_verlet 10 3.0 0.002 0.002 111.783 111.789 scf_env_do_scf_inner_loop 118 6.6 0.005 0.010 110.204 110.205 dbcsr_multiply_generic 2527 12.6 0.186 0.194 72.697 73.185 qs_scf_new_mos 118 7.6 0.001 0.001 72.783 72.884 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.783 72.883 ot_scf_mini 118 9.6 0.004 0.004 68.322 68.410 multiply_cannon 2527 13.6 0.563 0.580 54.031 57.114 multiply_cannon_loop 2527 14.6 0.817 0.859 50.861 51.938 ot_mini 118 10.6 0.001 0.001 38.128 38.223 init_scf_loop 11 6.9 0.000 0.000 35.265 35.268 mp_waitall_1 126780 16.7 25.002 31.621 25.002 31.621 prepare_preconditioner 11 7.9 0.000 0.000 31.432 31.466 make_preconditioner 11 8.9 0.000 0.001 31.432 31.466 rebuild_ks_matrix 129 8.3 0.001 0.001 29.401 29.536 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.400 29.535 make_full_inverse_cholesky 11 9.9 0.014 0.028 29.278 29.532 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.840 26.965 multiply_cannon_multrec 10108 15.6 10.394 16.239 18.050 22.427 qs_ot_get_derivative 118 11.6 0.002 0.002 20.703 20.787 multiply_cannon_metrocomm3 10108 15.6 0.025 0.026 12.997 20.444 cp_fm_cholesky_invert 11 10.9 17.954 17.961 17.954 17.961 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.479 17.646 apply_single 129 13.6 0.001 0.001 17.479 17.646 qs_ot_get_p 129 10.4 0.003 0.003 17.388 17.504 ot_diis_step 118 11.6 0.020 0.021 17.351 17.351 make_m2s 5054 13.6 0.065 0.070 14.750 15.753 make_images 5054 14.6 2.182 2.595 14.439 15.437 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.320 14.349 calculate_rho_elec 129 8.7 0.255 0.267 14.320 14.348 sum_up_and_integrate 129 10.3 0.002 0.002 14.293 14.336 integrate_v_rspace 129 11.3 0.004 0.004 14.232 14.276 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.563 13.578 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.397 12.398 multiply_cannon_sync_h2d 10108 15.6 10.805 11.300 10.805 11.300 init_scf_run 11 5.9 0.000 0.001 10.306 10.307 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.306 10.306 pw_transfer 1559 11.6 0.085 0.094 9.962 10.000 make_images_data 5054 15.6 0.057 0.067 8.453 9.887 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.736 9.782 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.498 9.509 cp_fm_diag_elpa_base 83 14.4 9.260 9.343 9.492 9.503 hybrid_alltoall_any 5240 16.5 0.847 3.803 8.306 9.478 fft_wrap_pw1pw2_140 527 13.2 1.684 1.726 8.632 8.678 grid_integrate_task_list 129 12.3 7.842 8.166 7.842 8.166 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.072 8.137 dbcsr_mm_accdrv_process 20926 16.1 3.351 4.419 7.303 8.022 cp_fm_cholesky_decompose 22 10.9 7.764 7.869 7.764 7.869 multiply_cannon_metrocomm1 10108 15.6 0.031 0.032 4.331 7.556 wfi_extrapolate 11 7.9 0.001 0.001 7.313 7.313 density_rs2pw 129 9.7 0.005 0.005 6.994 7.183 calculate_dm_sparse 129 9.5 0.001 0.001 6.124 6.205 fft3d_ps 1301 14.7 3.164 3.265 6.160 6.202 multiply_cannon_metrocomm4 7581 15.6 0.027 0.030 1.885 5.890 grid_collocate_task_list 129 9.7 5.608 5.830 5.608 5.830 mp_irecv_dv 29086 15.9 1.843 5.792 1.843 5.792 dbcsr_complete_redistribute 395 12.7 2.091 2.168 5.253 5.617 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 5.227 5.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.095 5.123 mp_alltoall_d11v 2423 14.1 4.657 5.046 4.657 5.046 mp_allgather_i34 2527 14.6 1.353 4.670 1.353 4.670 potential_pw2rs 129 12.3 0.026 0.027 4.398 4.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.018 4.071 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.586 3.874 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.689 3.774 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.402 3.716 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.665 3.712 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.444 3.522 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.466 3.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.617000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1446.636364, yerr=25.909888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.177107E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3989 59322. MP_Allreduce 11050 1505. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.038 286.851 286.852 qs_mol_dyn_low 1 2.0 0.003 0.003 286.332 286.346 qs_forces 11 3.9 0.005 0.005 286.230 286.235 qs_energies 11 4.9 0.002 0.003 277.361 277.368 scf_env_do_scf 11 5.9 0.001 0.001 254.792 254.803 velocity_verlet 10 3.0 0.002 0.002 207.458 207.466 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 131.931 131.933 init_scf_loop 11 6.9 0.000 0.000 122.586 122.588 prepare_preconditioner 11 7.9 0.000 0.000 117.724 117.747 make_preconditioner 11 8.9 0.028 0.054 117.724 117.747 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.380 114.845 qs_scf_new_mos 116 7.6 0.001 0.001 87.649 87.690 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.649 87.689 ot_scf_mini 116 9.6 0.004 0.004 82.843 82.851 dbcsr_multiply_generic 2485 12.5 0.215 0.225 80.782 81.198 cp_fm_upper_to_full 104 14.8 52.327 74.587 52.327 74.587 multiply_cannon 2485 13.5 0.675 0.727 58.139 59.727 multiply_cannon_loop 2485 14.5 1.019 1.037 54.172 55.367 ot_mini 116 10.6 0.001 0.001 43.269 43.279 dbcsr_complete_redistribute 393 12.7 3.977 4.052 29.279 42.144 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.790 38.625 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.274 35.904 rebuild_ks_matrix 127 8.3 0.001 0.001 34.397 34.411 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.034 34.396 34.410 mp_alltoall_i22 712 14.1 21.035 33.985 21.035 33.985 cp_fm_cholesky_invert 11 10.9 32.996 33.002 32.996 33.002 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.875 31.900 mp_waitall_1 102768 16.8 27.398 31.864 27.398 31.864 qs_ot_get_p 127 10.4 0.002 0.002 24.596 24.638 qs_ot_get_derivative 116 11.6 0.002 0.002 24.168 24.176 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.630 20.658 make_m2s 4970 13.5 0.073 0.078 18.207 19.681 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.496 19.619 make_images 4970 14.5 3.035 3.231 17.728 19.202 ot_diis_step 116 11.6 0.022 0.022 19.077 19.077 cp_dbcsr_syevd 82 12.4 0.006 0.006 18.874 18.875 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.427 18.508 apply_single 127 13.6 0.001 0.001 18.427 18.507 multiply_cannon_multrec 9940 15.5 10.203 12.056 17.931 18.003 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.116 17.124 calculate_rho_elec 127 8.7 0.475 0.475 17.115 17.124 sum_up_and_integrate 127 10.3 0.002 0.002 16.262 16.350 integrate_v_rspace 127 11.3 0.005 0.005 16.200 16.288 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.780 15.780 cp_fm_diag_elpa_base 82 14.4 11.450 13.036 15.776 15.776 multiply_cannon_sync_h2d 9940 15.5 14.219 14.242 14.219 14.242 pw_transfer 1535 11.6 0.094 0.096 12.975 12.984 hybrid_alltoall_any 5155 16.4 1.303 3.021 10.511 12.797 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.739 12.747 make_images_data 4970 15.5 0.065 0.072 10.232 12.368 init_scf_run 11 5.9 0.000 0.001 12.208 12.209 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.208 12.209 fft_wrap_pw1pw2_140 519 13.2 3.194 3.233 11.365 11.373 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.490 9.495 dbcsr_mm_accdrv_process 20590 16.0 4.183 5.944 7.480 9.289 wfi_extrapolate 11 7.9 0.001 0.001 9.108 9.108 cp_fm_cholesky_decompose 22 10.9 8.820 8.833 8.820 8.833 grid_integrate_task_list 127 12.3 8.492 8.689 8.492 8.689 density_rs2pw 127 9.7 0.005 0.005 8.217 8.271 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 8.191 8.192 fft3d_ps 1281 14.7 3.951 3.961 7.242 7.288 mp_alltoall_d11v 2401 14.1 6.868 6.982 6.868 6.982 calculate_dm_sparse 127 9.5 0.001 0.001 6.678 6.735 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.418 6.470 grid_collocate_task_list 127 9.7 6.364 6.452 6.364 6.452 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.105 6.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.852000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2891.181818, yerr=151.994508 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 1.260962E+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 0.013 0.030 84.611 84.612 qs_energies 1 2.0 0.000 0.000 84.166 84.173 ls_scf 1 3.0 0.000 0.000 83.273 83.279 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.356 72.545 multiply_cannon 111 7.7 0.017 0.020 55.776 57.182 multiply_cannon_loop 111 8.7 0.229 0.242 52.406 53.957 ls_scf_main 1 4.0 0.000 0.000 52.135 52.147 density_matrix_trs4 2 5.0 0.002 0.003 46.642 46.731 ls_scf_init_scf 1 4.0 0.000 0.000 28.103 28.104 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.982 27.042 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.908 24.925 mp_waitall_1 11031 10.9 22.448 24.834 22.448 24.834 multiply_cannon_multrec 2664 9.7 8.130 8.875 15.565 17.258 multiply_cannon_sync_h2d 2664 9.7 13.407 15.307 13.407 15.307 make_m2s 222 7.7 0.009 0.012 12.994 13.524 make_images 222 8.7 0.098 0.108 12.971 13.504 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.737 12.120 make_images_data 222 9.7 0.004 0.005 7.570 8.160 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.477 8.136 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.690 7.054 8.037 hybrid_alltoall_any 227 10.6 0.216 1.811 6.495 7.641 dbcsr_mm_accdrv_process_sort 4760 11.4 6.271 7.156 6.271 7.156 calculate_norms 4752 9.8 5.504 6.145 5.504 6.145 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.013 5.115 mp_sum_l 887 5.1 3.045 4.169 3.045 4.169 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.062 3.695 mp_irecv_dv 6231 10.9 2.046 3.666 2.046 3.666 make_images_sizes 222 9.7 0.000 0.000 0.707 3.395 mp_alltoall_i44 222 10.7 0.706 3.395 0.706 3.395 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.290 3.294 arnoldi_extremal 4 6.8 0.000 0.000 3.232 3.262 arnoldi_normal_ev 4 7.8 0.001 0.003 3.232 3.262 build_subspace 16 8.4 0.009 0.012 3.143 3.145 ls_scf_post 1 4.0 0.000 0.000 3.034 3.041 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.363 2.770 dbcsr_merge_single_wm 555 10.7 0.453 0.583 2.355 2.763 make_images_pack 222 9.7 2.208 2.628 2.210 2.629 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.310 2.554 dbcsr_sort_data 658 11.4 2.158 2.509 2.158 2.509 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.465 2.073 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.274 2.380 buffer_matrices_ensure_size 222 8.7 1.739 2.035 1.739 2.035 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.755 1.756 rebuild_ks_matrix 3 7.3 0.000 0.000 1.745 1.747 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.745 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.612000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 2.146943E+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 0.025 0.042 89.326 89.328 qs_energies 1 2.0 0.000 0.000 88.877 88.882 ls_scf 1 3.0 0.000 0.002 87.544 87.548 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.773 74.043 multiply_cannon 111 7.7 0.027 0.041 52.633 56.652 ls_scf_main 1 4.0 0.000 0.002 54.229 54.233 multiply_cannon_loop 111 8.7 0.135 0.147 49.804 52.536 density_matrix_trs4 2 5.0 0.002 0.003 48.626 48.810 ls_scf_init_scf 1 4.0 0.000 0.001 29.665 29.667 mp_waitall_1 9105 10.9 20.540 28.717 20.540 28.717 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.463 28.555 multiply_cannon_multrec 1332 9.7 13.316 17.276 22.669 27.752 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.005 26.124 26.135 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.384 20.004 make_m2s 222 7.7 0.006 0.007 14.897 15.484 make_images 222 8.7 1.370 1.694 14.866 15.455 dbcsr_mm_accdrv_process 4041 10.4 0.345 0.522 8.947 10.515 dbcsr_mm_accdrv_process_sort 4041 11.4 8.466 9.988 8.466 9.988 make_images_data 222 9.7 0.004 0.005 8.647 9.574 hybrid_alltoall_any 227 10.6 0.541 2.543 8.027 9.387 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.171 7.798 mp_irecv_dv 3311 11.0 3.150 7.733 3.150 7.733 mp_sum_l 887 5.1 4.767 7.351 4.767 7.351 calculate_norms 2376 9.8 6.052 6.813 6.052 6.813 multiply_cannon_sync_h2d 1332 9.7 4.884 6.106 4.884 6.106 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.617 5.975 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.821 5.014 arnoldi_extremal 4 6.8 0.000 0.000 4.616 4.634 arnoldi_normal_ev 4 7.8 0.001 0.005 4.616 4.634 build_subspace 16 8.4 0.014 0.021 4.364 4.368 ls_scf_post 1 4.0 0.000 0.001 3.649 3.655 ls_scf_store_result 1 5.0 0.000 0.000 3.362 3.476 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.120 3.355 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.222 2.741 3.224 mp_allgather_i34 111 8.7 0.815 3.003 0.815 3.003 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.155 2.904 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.595 2.679 dbcsr_data_new 4174 10.1 2.119 2.405 2.119 2.405 make_images_pack 222 9.7 1.817 2.127 1.820 2.130 dbcsr_sort_data 436 11.2 1.825 2.034 1.825 2.034 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.857 1.859 rebuild_ks_matrix 3 7.3 0.000 0.000 1.844 1.846 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.007 1.844 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.328000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1761.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 2.920686E+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 0.024 0.044 91.923 91.925 qs_energies 1 2.0 0.000 0.000 91.399 91.403 ls_scf 1 3.0 0.000 0.000 89.980 89.984 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.609 74.957 ls_scf_main 1 4.0 0.000 0.000 56.009 56.013 multiply_cannon 111 7.7 0.035 0.073 51.834 55.708 multiply_cannon_loop 111 8.7 0.116 0.129 49.100 53.292 density_matrix_trs4 2 5.0 0.002 0.003 50.098 50.262 mp_waitall_1 7281 11.0 23.247 33.152 23.247 33.152 ls_scf_init_scf 1 4.0 0.000 0.000 30.252 30.255 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.023 29.136 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.681 26.694 multiply_cannon_multrec 888 9.7 12.623 15.113 21.275 24.511 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.820 22.665 make_m2s 222 7.7 0.006 0.007 16.187 16.939 make_images 222 8.7 1.583 1.890 16.149 16.899 make_images_data 222 9.7 0.004 0.005 9.447 10.506 hybrid_alltoall_any 227 10.6 0.640 2.945 9.030 10.240 dbcsr_mm_accdrv_process 3754 10.4 0.299 0.491 8.170 9.382 dbcsr_mm_accdrv_process_sort 3754 11.4 7.734 8.892 7.734 8.892 mp_sum_l 887 5.1 4.891 8.189 4.891 8.189 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.439 mp_irecv_dv 2335 11.1 2.456 7.381 2.456 7.381 multiply_cannon_sync_h2d 888 9.7 6.100 7.006 6.100 7.006 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.622 6.839 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.769 6.465 arnoldi_extremal 4 6.8 0.000 0.000 5.087 5.103 arnoldi_normal_ev 4 7.8 0.001 0.004 5.087 5.103 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.754 4.994 calculate_norms 1584 9.8 4.338 4.804 4.338 4.804 build_subspace 16 8.4 0.014 0.020 4.786 4.795 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.435 3.759 ls_scf_post 1 4.0 0.000 0.000 3.719 3.724 mp_allgather_i34 111 8.7 0.846 3.666 0.846 3.666 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.593 3.023 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.442 3.543 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.870 2.983 dbcsr_data_new 4116 9.9 2.105 2.463 2.105 2.463 dbcsr_sort_data 325 11.1 1.869 2.142 1.869 2.142 make_images_sizes 222 9.7 0.000 0.000 0.917 2.119 mp_alltoall_i44 222 10.7 0.917 2.119 0.917 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 make_images_pack 222 9.7 1.623 1.885 1.626 1.888 rebuild_ks_matrix 3 7.3 0.000 0.000 1.869 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.869 1.871 dbcsr_finalize 304 7.8 0.026 0.032 1.607 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.925000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2254.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 3.385483E+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 0.027 0.045 97.165 97.166 qs_energies 1 2.0 0.000 0.000 96.596 96.600 ls_scf 1 3.0 0.000 0.000 94.932 94.937 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.584 78.833 ls_scf_main 1 4.0 0.000 0.000 58.878 58.879 multiply_cannon 111 7.7 0.044 0.083 51.841 56.293 density_matrix_trs4 2 5.0 0.002 0.003 52.760 52.854 multiply_cannon_loop 111 8.7 0.152 0.168 46.733 49.557 ls_scf_init_scf 1 4.0 0.000 0.000 32.822 32.824 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.619 31.680 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.151 29.166 mp_waitall_1 6369 11.0 22.835 28.977 22.835 28.977 multiply_cannon_multrec 1332 9.7 14.104 17.247 22.079 24.833 make_m2s 222 7.7 0.006 0.008 21.064 22.466 make_images 222 8.7 3.132 3.593 21.014 22.419 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.263 17.295 make_images_data 222 9.7 0.004 0.004 11.786 13.414 hybrid_alltoall_any 227 10.6 0.796 3.746 11.100 12.942 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.481 7.607 9.150 dbcsr_mm_accdrv_process_sort 3641 11.4 7.166 8.646 7.166 8.646 mp_sum_l 887 5.1 4.091 7.613 4.091 7.613 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.081 6.121 multiply_cannon_sync_h2d 1332 9.7 5.440 6.074 5.440 6.074 mp_irecv_dv 3229 10.9 2.057 6.055 2.057 6.055 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.206 6.053 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.614 5.851 arnoldi_extremal 4 6.8 0.000 0.000 5.322 5.336 arnoldi_normal_ev 4 7.8 0.001 0.004 5.322 5.336 build_subspace 16 8.4 0.014 0.021 4.985 4.992 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.525 4.706 mp_allgather_i34 111 8.7 2.234 4.597 2.234 4.597 calculate_norms 2376 9.8 4.157 4.518 4.157 4.518 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.598 3.926 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.682 3.189 3.684 dbcsr_sort_data 658 11.4 3.064 3.338 3.064 3.338 ls_scf_post 1 4.0 0.000 0.000 3.232 3.236 dbcsr_special_finalize 555 9.7 0.006 0.007 2.807 3.139 dbcsr_merge_single_wm 555 10.7 0.535 0.652 2.798 3.131 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.025 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.976 3.036 dbcsr_data_release 10477 10.7 1.585 2.455 1.585 2.455 dbcsr_finalize 304 7.8 0.049 0.061 1.804 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.166000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 4.755042E+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 0.038 0.058 92.225 92.228 qs_energies 1 2.0 0.000 0.000 91.456 91.461 ls_scf 1 3.0 0.000 0.000 89.506 89.511 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.005 71.197 ls_scf_main 1 4.0 0.000 0.000 56.600 56.601 multiply_cannon 111 7.7 0.096 0.129 52.819 55.957 multiply_cannon_loop 111 8.7 0.087 0.092 50.238 51.359 density_matrix_trs4 2 5.0 0.002 0.003 49.584 49.650 ls_scf_init_scf 1 4.0 0.000 0.000 29.463 29.464 mp_waitall_1 5436 11.0 24.546 28.954 24.546 28.954 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.210 28.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.117 26.124 multiply_cannon_multrec 444 9.7 13.681 16.389 20.832 22.303 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.078 15.524 make_m2s 222 7.7 0.005 0.005 13.422 14.361 make_images 222 8.7 2.038 2.474 13.354 14.293 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.079 13.476 make_images_data 222 9.7 0.003 0.004 8.201 9.714 hybrid_alltoall_any 227 10.6 0.800 3.797 8.159 9.426 multiply_cannon_sync_h2d 444 9.7 6.749 8.138 6.749 8.138 dbcsr_mm_accdrv_process 3003 10.4 0.363 0.409 6.845 7.966 dbcsr_mm_accdrv_process_sort 3003 11.4 6.482 7.570 6.482 7.570 arnoldi_extremal 4 6.8 0.000 0.000 5.860 5.866 arnoldi_normal_ev 4 7.8 0.001 0.005 5.859 5.866 build_subspace 16 8.4 0.015 0.020 5.470 5.479 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.487 4.672 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.206 4.394 dbcsr_matrix_vector_mult_local 304 10.0 3.693 4.178 3.695 4.180 mp_sum_l 887 5.1 2.775 4.176 2.775 4.176 calculate_norms 792 9.8 3.607 3.682 3.607 3.682 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.544 3.611 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.537 3.606 mp_allgather_i34 111 8.7 1.159 3.601 1.159 3.601 mp_irecv_dv 1241 11.2 1.523 3.573 1.523 3.573 ls_scf_post 1 4.0 0.000 0.000 3.442 3.448 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.913 3.420 ls_scf_store_result 1 5.0 0.000 0.000 3.235 3.266 make_images_sizes 222 9.7 0.000 0.000 0.838 3.257 mp_alltoall_i44 222 10.7 0.838 3.257 0.838 3.257 dbcsr_finalize 304 7.8 0.062 0.077 2.200 2.306 dbcsr_data_new 4608 9.7 1.785 2.249 1.785 2.249 dbcsr_merge_all 275 8.9 0.477 0.524 2.058 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.037 2.037 rebuild_ks_matrix 3 7.3 0.000 0.000 2.005 2.006 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.005 2.005 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 1.934 1.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.228000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 8.791343E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.090 105.716 105.717 qs_energies 1 2.0 0.000 0.000 104.379 104.387 ls_scf 1 3.0 0.000 0.000 101.387 101.395 dbcsr_multiply_generic 111 6.7 0.023 0.027 74.824 74.975 ls_scf_main 1 4.0 0.000 0.000 63.881 63.881 density_matrix_trs4 2 5.0 0.002 0.003 54.811 54.866 multiply_cannon 111 7.7 0.130 0.177 48.354 50.411 multiply_cannon_loop 111 8.7 0.097 0.098 45.417 46.145 ls_scf_init_scf 1 4.0 0.000 0.000 33.725 33.726 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.158 32.187 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.427 29.437 mp_waitall_1 4527 11.1 21.973 25.409 21.973 25.409 make_m2s 222 7.7 0.005 0.005 22.772 23.816 make_images 222 8.7 3.561 3.866 22.665 23.707 multiply_cannon_multrec 444 9.7 17.805 18.467 22.498 23.174 hybrid_alltoall_any 227 10.6 1.650 3.604 12.923 15.890 make_images_data 222 9.7 0.003 0.004 13.167 15.269 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.195 10.710 multiply_cannon_sync_h2d 444 9.7 8.787 8.847 8.787 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.544 7.557 arnoldi_normal_ev 4 7.8 0.002 0.008 7.544 7.557 build_subspace 16 8.4 0.026 0.036 6.983 6.994 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.536 5.684 dbcsr_matrix_vector_mult_local 304 10.0 5.049 5.373 5.052 5.376 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.182 5.267 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.902 5.131 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.362 4.503 4.641 dbcsr_mm_accdrv_process_sort 1814 11.4 4.166 4.298 4.166 4.298 ls_scf_post 1 4.0 0.000 0.000 3.781 3.790 make_images_sizes 222 9.7 0.000 0.000 1.487 3.651 mp_alltoall_i44 222 10.7 1.486 3.651 1.486 3.651 ls_scf_store_result 1 5.0 0.000 0.000 3.501 3.537 mp_allgather_i34 111 8.7 1.053 3.359 1.053 3.359 calculate_norms 792 9.8 3.235 3.292 3.235 3.292 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.164 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 2.961 2.961 dbcsr_merge_all 275 8.9 0.883 0.913 2.856 2.942 dbcsr_complete_redistribute 5 7.6 1.431 1.471 2.744 2.855 matrix_ls_to_qs 2 6.0 0.000 0.000 2.385 2.503 dbcsr_sort_data 325 11.1 2.436 2.498 2.436 2.498 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.466 2.468 rebuild_ks_matrix 3 7.3 0.000 0.000 2.401 2.403 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.401 2.403 dbcsr_data_new 6591 9.6 1.828 2.286 1.828 2.286 dbcsr_new_transposed 4 7.5 0.241 0.251 2.271 2.281 dbcsr_frobenius_norm 74 6.6 2.056 2.129 2.205 2.240 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.201 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.200 dbcsr_data_release 12724 10.6 1.977 2.195 1.977 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.717000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6958.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/668feff676deb4b57bbee5867e9b22be8df45b0d_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 594.296832E+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 4808143. 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 0.046 0.155 245.248 245.253 qs_mol_dyn_low 1 2.0 0.004 0.014 244.022 244.110 qs_forces 5 3.8 0.008 0.055 243.920 243.935 qs_energies 5 4.8 0.004 0.047 240.744 240.775 scf_env_do_scf 5 5.8 0.000 0.002 226.135 226.137 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 197.750 197.751 qs_scf_new_mos 105 7.6 0.000 0.001 155.205 155.356 qs_scf_loop_do_ot 105 8.6 0.001 0.001 155.204 155.356 ot_scf_mini 105 9.6 0.003 0.008 145.371 145.518 dbcsr_multiply_generic 1445 12.2 0.129 0.137 133.871 134.462 multiply_cannon 1445 13.2 0.274 0.285 114.370 116.511 multiply_cannon_loop 1445 14.2 2.820 2.985 112.503 114.014 velocity_verlet 4 3.0 0.001 0.017 108.916 108.917 ot_mini 105 10.6 0.001 0.007 60.855 60.986 qs_ot_get_p 112 10.4 0.001 0.002 54.532 54.832 multiply_cannon_multrec 69360 15.2 29.648 34.642 39.472 44.495 qs_ot_p2m_diag 40 11.0 0.020 0.030 43.187 43.313 mp_waitall_1 488190 16.1 35.329 42.064 35.329 42.064 cp_dbcsr_syevd 40 12.0 0.002 0.003 39.941 39.942 qs_ot_get_derivative 55 11.6 0.001 0.002 38.980 39.103 cp_fm_syevd 40 13.0 0.000 0.001 34.516 34.665 multiply_cannon_sync_h2d 69360 15.2 29.229 33.326 29.229 33.326 multiply_cannon_metrocomm3 69360 15.2 0.203 0.213 25.751 32.225 rebuild_ks_matrix 110 8.4 0.000 0.000 30.117 30.304 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.043 30.116 30.303 cp_fm_redistribute_end 40 14.0 14.787 29.515 14.793 29.518 cp_fm_syevd_base 40 14.0 14.715 29.447 14.715 29.447 init_scf_loop 7 6.6 0.001 0.010 28.341 28.342 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.586 27.752 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.180 23.436 apply_single 62 13.6 0.000 0.000 23.180 23.436 prepare_preconditioner 7 7.6 0.000 0.000 23.306 23.338 make_preconditioner 7 8.6 0.000 0.002 23.306 23.338 ot_new_cg_direction 55 11.6 0.001 0.003 21.170 21.171 qs_rho_update_rho_low 110 7.6 0.001 0.002 18.494 18.865 calculate_rho_elec 110 8.6 0.030 0.032 18.493 18.865 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.895 15.959 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.736 14.898 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.000 14.130 density_rs2pw 110 9.6 0.005 0.006 12.620 13.081 mp_sum_l 4764 12.2 12.247 12.974 12.247 12.974 pw_transfer 1645 12.4 0.079 0.102 12.730 12.967 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.592 12.835 init_scf_run 5 5.8 0.000 0.001 11.966 11.967 scf_env_initial_rho_setup 5 6.8 0.002 0.005 11.965 11.966 calculate_dm_sparse 110 9.5 0.000 0.001 11.384 11.542 fft_wrap_pw1pw2_240 915 15.0 0.816 0.902 11.146 11.388 qs_vxc_create 110 10.4 0.002 0.004 10.952 11.000 cp_fm_cholesky_invert 7 10.6 10.850 10.859 10.850 10.859 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.573 10.685 dbcsr_mm_accdrv_process 154766 15.8 6.224 6.432 9.691 10.533 check_diag 80 13.5 8.686 8.941 9.919 10.056 fft3d_pb 915 16.0 2.375 2.646 9.379 9.630 transfer_rs2pw 445 10.6 0.007 0.008 8.975 9.451 sum_up_and_integrate 60 10.3 0.001 0.004 9.037 9.064 integrate_v_rspace 60 11.3 0.002 0.003 9.019 9.048 acc_transpose_blocks 69360 15.2 0.364 0.389 7.570 8.117 xc_rho_set_and_dset_create 110 12.4 0.075 0.095 7.670 7.938 multiply_cannon_metrocomm1 69360 15.2 0.096 0.104 4.573 7.675 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.541 7.558 make_m2s 2890 13.2 0.078 0.087 6.842 7.448 make_images 2890 14.2 0.238 0.259 6.734 7.340 xc_vxc_pw_create 60 11.3 0.039 0.049 7.283 7.330 make_full_single_inverse 7 9.6 0.001 0.004 7.091 7.123 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.047 7.091 calculate_first_density_matrix 1 7.0 0.000 0.004 7.065 7.082 mp_alltoall_z22v 2340 17.7 6.472 6.748 6.472 6.748 xc_pw_derive 510 13.4 0.005 0.007 6.664 6.740 mp_waitany 7680 13.5 4.496 5.276 4.496 5.276 acc_transpose_blocks_kernels 69360 16.2 0.849 0.901 4.817 5.255 potential_pw2rs 60 12.3 0.002 0.003 5.157 5.190 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=245.253000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=564.400000, yerr=2.939388 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 668feff676deb4b57bbee5867e9b22be8df45b0d Summary: empty Status: OK