=== 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: ad61d0fb47758fedc409ed1d609b4bb070cc6269 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/01 job id: 50582537 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/02 job id: 50582538 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/03 job id: 50582540 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/04 job id: 50582542 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/05 job id: 50582544 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/06 job id: 50582545 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/07 job id: 50582546 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/08 job id: 50582547 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/09 job id: 50582548 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/10 job id: 50582549 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/11 job id: 50582550 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/12 job id: 50582552 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/13 job id: 50582553 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/14 job id: 50582554 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/15 job id: 50582557 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/16 job id: 50582558 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/17 job id: 50582559 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/18 job id: 50582561 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/19 job id: 50582562 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/20 job id: 50582563 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/21 job id: 50582564 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/22 job id: 50582565 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/23 job id: 50582566 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/24 job id: 50582567 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/25 job id: 50582568 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/26 job id: 50582569 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/27 job id: 50582570 --- 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/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.031 135.404 135.405 farming_run 1 2.0 134.947 134.949 135.377 135.380 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.500213E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 115.196 115.196 qs_energies 1 2.0 0.000 0.000 115.001 115.003 mp2_main 1 3.0 0.000 0.000 112.892 112.894 mp2_gpw_main 1 4.0 0.028 0.035 111.888 111.890 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.903 93.222 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.628 54.949 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 41.033 46.067 get_2c_integrals 1 6.0 0.008 0.009 37.019 38.104 integrate_v_rspace 273 8.0 0.436 0.453 24.611 29.529 pw_transfer 6555 10.6 0.378 0.387 26.741 27.235 grid_integrate_task_list 273 9.0 20.524 25.908 20.524 25.908 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.382 25.901 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 22.925 23.424 cp_fm_cholesky_decompose 12 8.2 17.959 19.069 17.959 19.069 rpa_ri_compute_en 1 5.0 0.019 0.021 18.878 19.061 compute_2c_integrals 1 7.0 0.002 0.002 19.012 19.014 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.532 18.708 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.421 18.531 18.707 cholesky_decomp 1 7.0 0.000 0.000 16.868 17.983 fft3d_s 5443 13.4 16.060 16.324 16.081 16.345 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.317 10.771 15.317 calculate_wavefunction 272 8.0 5.431 5.573 12.333 13.046 rpa_num_int 1 6.0 0.000 0.001 10.645 10.645 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.609 10.642 calc_mat_Q 8 8.0 0.000 0.000 9.472 9.579 contract_S_to_Q 8 9.0 0.000 0.000 8.892 9.002 calc_potential_gpw 544 9.5 0.004 0.005 8.206 8.643 parallel_gemm_fm 14 9.1 0.000 0.000 8.490 8.599 parallel_gemm_fm_cosma 14 10.1 8.490 8.599 8.490 8.599 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.038 8.288 potential_pw2rs 545 10.0 0.107 0.109 7.467 8.099 create_integ_mat 1 6.0 0.027 0.033 7.821 7.821 collocate_single_gaussian 272 10.0 0.039 0.042 7.306 7.604 array2fm 1 7.0 0.000 0.000 6.778 7.242 pw_scatter_s 2720 13.7 4.353 4.498 4.353 4.498 pw_gather_s 2722 13.2 3.448 3.834 3.448 3.834 array2fm_buffer_send 1 8.0 2.960 3.211 2.960 3.211 pw_poisson_solve 545 10.5 1.124 1.187 2.190 2.351 mp_sync 25 8.8 1.120 2.332 1.120 2.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.890141, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.046 422.131 422.133 farming_run 1 2.0 421.134 421.141 422.087 422.093 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.236218E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 217.350 217.350 qs_energies 1 2.0 0.008 0.033 217.098 217.114 scf_env_do_scf 1 3.0 0.000 0.000 114.834 114.834 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.852 113.859 rebuild_ks_matrix 4 6.0 0.000 0.000 113.850 113.857 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 113.850 113.857 hfx_ks_matrix 4 8.0 0.001 0.001 113.479 113.483 integrate_four_center 4 9.0 0.152 0.463 113.478 113.483 integrate_four_center_main 4 10.0 0.098 0.492 102.018 105.050 integrate_four_center_bin 265 11.0 101.920 105.048 101.920 105.048 mp2_main 1 3.0 0.014 0.064 101.948 101.959 mp2_gpw_main 1 4.0 0.064 0.154 100.819 100.834 init_scf_loop 1 4.0 0.000 0.000 96.936 96.936 mp2_ri_gpw_compute_in 1 5.0 0.064 0.072 73.971 74.998 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.813 54.834 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 41.595 46.543 integrate_v_rspace 95 8.0 0.394 0.564 28.071 32.887 pw_transfer 2240 10.6 0.145 0.171 29.530 29.982 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.540 28.998 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.608 28.660 grid_integrate_task_list 95 9.0 23.477 28.480 23.477 28.480 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.287 26.759 ao_to_mo_and_store_B_mult_1 91 7.0 10.544 26.062 10.544 26.062 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.897 24.937 24.946 get_2c_integrals 1 6.0 0.000 0.000 20.064 20.100 compute_2c_integrals 1 7.0 0.002 0.003 19.047 19.057 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.647 18.888 mp2_eri_2c_integrate_gpw 1 9.0 1.732 1.866 18.646 18.887 fft3d_s 1823 13.4 18.477 18.868 18.491 18.881 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.896 17.896 calculate_wavefunction 91 8.0 2.025 2.055 9.633 9.831 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.579 8.782 9.261 potential_pw2rs 186 10.0 0.033 0.035 8.461 9.047 local_gemm 172 8.0 8.226 8.691 8.226 8.691 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.172 8.570 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.515 7.926 8.419 calc_potential_gpw 182 9.5 0.002 0.002 7.894 8.156 collocate_single_gaussian 91 10.0 0.017 0.021 7.760 7.985 mp_sync 37 10.5 3.427 6.970 3.427 6.970 integrate_four_center_load 4 10.0 0.000 0.000 6.762 6.765 hfx_load_balance 1 11.0 0.000 0.000 6.761 6.765 mp_sendrecv_dm3 2068 8.0 5.964 6.451 5.964 6.451 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.385 6.334 6.385 pw_gather_s 912 13.2 4.504 4.996 4.504 4.996 pw_scatter_s 910 13.7 3.864 4.422 3.864 4.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.839977, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1496.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.472256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.065 54.283 54.288 qs_mol_dyn_low 1 2.0 0.003 0.006 53.568 53.576 qs_forces 11 3.9 0.006 0.008 53.504 53.505 qs_energies 11 4.9 0.002 0.002 51.962 51.988 scf_env_do_scf 11 5.9 0.001 0.001 45.103 45.103 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.927 42.928 dbcsr_multiply_generic 2286 12.5 0.093 0.100 33.372 33.786 qs_scf_new_mos 108 7.5 0.000 0.001 32.782 33.081 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.781 33.081 ot_scf_mini 108 9.5 0.002 0.002 31.084 31.284 multiply_cannon 2286 13.5 0.187 0.195 26.509 28.109 multiply_cannon_loop 2286 14.5 1.849 1.950 25.789 27.400 velocity_verlet 10 3.0 0.001 0.002 26.566 26.567 ot_mini 108 10.5 0.001 0.001 18.340 18.593 qs_ot_get_derivative 108 11.5 0.001 0.001 15.383 15.547 mp_waitall_1 245248 16.5 8.507 14.519 8.507 14.519 multiply_cannon_metrocomm3 54864 15.5 0.074 0.079 5.884 12.935 multiply_cannon_multrec 54864 15.5 3.673 5.901 7.610 10.966 qs_ot_get_p 119 10.4 0.001 0.001 8.042 8.310 rebuild_ks_matrix 119 8.3 0.000 0.000 7.926 8.093 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.926 8.092 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.964 7.108 mp_sum_l 7287 12.8 4.839 6.459 4.839 6.459 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.540 6.014 multiply_cannon_sync_h2d 54864 15.5 5.196 5.887 5.196 5.887 init_scf_run 11 5.9 0.000 0.001 5.547 5.547 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.547 5.547 dbcsr_mm_accdrv_process 76910 16.1 1.849 2.907 3.850 5.505 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.193 5.225 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.660 4.760 sum_up_and_integrate 119 10.3 0.001 0.002 4.515 4.522 integrate_v_rspace 119 11.3 0.002 0.003 4.504 4.513 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.473 4.473 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.231 4.372 calculate_rho_elec 119 8.7 0.011 0.016 4.231 4.372 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.247 4.247 cp_fm_redistribute_end 50 14.0 2.167 4.218 2.176 4.222 cp_fm_diag_elpa_base 50 14.0 2.038 4.097 2.044 4.106 acc_transpose_blocks 54864 15.5 0.225 0.248 2.420 3.558 calculate_first_density_matrix 1 7.0 0.001 0.003 3.244 3.250 multiply_cannon_metrocomm1 54864 15.5 0.056 0.064 1.904 3.158 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.919 3.127 apply_single 119 13.6 0.000 0.000 2.919 3.127 calculate_dm_sparse 119 9.5 0.000 0.000 2.919 3.109 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.069 3.071 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.737 2.793 ot_diis_step 108 11.5 0.006 0.006 2.689 2.690 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.500 2.545 jit_kernel_multiply 13 15.8 1.936 2.512 1.936 2.512 density_rs2pw 119 9.7 0.004 0.004 2.237 2.343 wfi_extrapolate 11 7.9 0.001 0.001 2.230 2.230 init_scf_loop 11 6.9 0.000 0.000 2.156 2.156 grid_integrate_task_list 119 12.3 2.008 2.109 2.008 2.109 mp_sum_d 4137 12.0 1.320 1.954 1.320 1.954 potential_pw2rs 119 12.3 0.004 0.004 1.874 1.887 acc_transpose_blocks_kernels 54864 16.5 0.248 0.384 0.941 1.817 pw_transfer 1439 11.6 0.051 0.055 1.732 1.802 make_m2s 4572 13.5 0.053 0.055 1.743 1.787 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.657 1.726 make_images 4572 14.5 0.133 0.139 1.661 1.704 mp_alltoall_d11v 2130 13.8 1.367 1.534 1.367 1.534 transfer_rs2pw 487 10.6 0.005 0.006 1.421 1.516 mp_waitany 12084 13.8 1.272 1.473 1.272 1.473 acc_transpose_blocks_sync 164592 16.5 1.210 1.447 1.210 1.447 fft3d_ps 1201 14.6 0.368 0.471 1.368 1.436 jit_kernel_transpose 5 15.5 0.692 1.436 0.692 1.436 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.429 grid_collocate_task_list 119 9.7 1.349 1.412 1.349 1.412 transfer_pw2rs 487 13.2 0.006 0.006 1.366 1.380 fft_wrap_pw1pw2_140 487 13.2 0.109 0.116 1.271 1.342 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.715 1.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.288000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 488.267776E+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 490450. 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 0.006 0.027 40.044 40.044 qs_mol_dyn_low 1 2.0 0.003 0.003 39.837 39.845 qs_forces 11 3.9 0.002 0.003 39.774 39.775 qs_energies 11 4.9 0.002 0.003 38.099 38.103 scf_env_do_scf 11 5.9 0.000 0.001 32.742 32.742 scf_env_do_scf_inner_loop 108 6.5 0.005 0.016 30.267 30.267 dbcsr_multiply_generic 2286 12.5 0.100 0.104 22.484 22.816 qs_scf_new_mos 108 7.5 0.001 0.001 21.007 21.238 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.007 21.237 ot_scf_mini 108 9.5 0.002 0.003 20.075 20.243 multiply_cannon 2286 13.5 0.209 0.215 17.215 18.759 velocity_verlet 10 3.0 0.001 0.002 18.631 18.632 multiply_cannon_loop 2286 14.5 1.195 1.256 16.013 17.467 ot_mini 108 10.5 0.001 0.001 12.329 12.563 mp_waitall_1 200699 16.5 5.866 10.978 5.866 10.978 qs_ot_get_derivative 108 11.5 0.001 0.001 9.959 10.125 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.168 9.646 multiply_cannon_multrec 27432 15.5 1.823 4.100 6.140 9.045 rebuild_ks_matrix 119 8.3 0.000 0.000 6.896 7.030 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.896 7.030 dbcsr_mm_accdrv_process 47894 16.0 3.529 5.858 4.236 6.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.079 6.200 qs_ot_get_p 119 10.4 0.001 0.001 4.807 5.025 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.776 4.610 init_scf_run 11 5.9 0.000 0.001 4.155 4.156 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.155 4.155 mp_sum_l 7287 12.8 2.061 4.123 2.061 4.123 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.007 4.050 apply_single 119 13.6 0.000 0.000 3.007 4.050 sum_up_and_integrate 119 10.3 0.001 0.001 3.812 3.818 integrate_v_rspace 119 11.3 0.002 0.003 3.798 3.803 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.695 3.719 calculate_rho_elec 119 8.7 0.021 0.024 3.695 3.718 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.103 3.122 make_m2s 4572 13.5 0.052 0.054 2.755 3.088 make_images 4572 14.5 0.206 0.243 2.665 3.001 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.694 2.694 calculate_first_density_matrix 1 7.0 0.000 0.000 2.610 2.612 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.394 2.478 init_scf_loop 11 6.9 0.000 0.000 2.454 2.454 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.348 2.349 cp_fm_redistribute_end 50 14.0 1.190 2.321 1.194 2.324 ot_diis_step 108 11.5 0.011 0.011 2.322 2.322 cp_fm_diag_elpa_base 50 14.0 1.097 2.223 1.126 2.264 acc_transpose_blocks 27432 15.5 0.115 0.121 1.706 2.234 multiply_cannon_sync_h2d 27432 15.5 1.691 2.221 1.691 2.221 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.181 2.183 calculate_dm_sparse 119 9.5 0.000 0.001 2.072 2.162 density_rs2pw 119 9.7 0.004 0.004 1.993 2.093 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.961 2.003 pw_transfer 1439 11.6 0.065 0.068 1.914 1.948 grid_integrate_task_list 119 12.3 1.845 1.925 1.845 1.925 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.822 1.859 make_images_data 4572 15.5 0.048 0.054 1.282 1.650 prepare_preconditioner 11 7.9 0.000 0.000 1.563 1.588 make_preconditioner 11 8.9 0.000 0.000 1.563 1.588 jit_kernel_multiply 8 16.5 0.649 1.576 0.649 1.576 potential_pw2rs 119 12.3 0.006 0.006 1.549 1.555 hybrid_alltoall_any 4725 16.4 0.053 0.114 1.129 1.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.459 1.519 wfi_extrapolate 11 7.9 0.001 0.001 1.485 1.485 fft3d_ps 1201 14.6 0.521 0.576 1.454 1.485 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.414 1.451 mp_alltoall_d11v 2130 13.8 1.239 1.386 1.239 1.386 grid_collocate_task_list 119 9.7 1.286 1.348 1.286 1.348 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.219 1.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.266 transfer_rs2pw 487 10.6 0.005 0.005 1.141 1.227 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 0.750 1.153 mp_allgather_i34 2286 14.5 0.625 1.032 0.625 1.032 mp_sum_d 4137 12.0 0.539 1.022 0.539 1.022 transfer_pw2rs 487 13.2 0.004 0.005 0.971 0.978 acc_transpose_blocks_sync 82296 16.5 0.816 0.940 0.816 0.940 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.926 0.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.874 0.887 jit_kernel_transpose 5 15.5 0.562 0.880 0.562 0.880 make_images_sizes 4572 15.5 0.005 0.005 0.602 0.871 mp_alltoall_i44 4572 16.5 0.597 0.866 0.597 0.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.044000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 520.929280E+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 0.020 0.035 33.187 33.188 qs_mol_dyn_low 1 2.0 0.003 0.004 32.905 32.912 qs_forces 11 3.9 0.002 0.002 32.833 32.834 qs_energies 11 4.9 0.003 0.007 31.274 31.275 scf_env_do_scf 11 5.9 0.001 0.002 26.074 26.075 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.527 23.527 dbcsr_multiply_generic 2286 12.5 0.097 0.114 17.286 17.355 qs_scf_new_mos 108 7.5 0.001 0.001 15.716 15.730 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.715 15.729 velocity_verlet 10 3.0 0.001 0.002 15.452 15.454 ot_scf_mini 108 9.5 0.002 0.003 14.954 14.966 multiply_cannon 2286 13.5 0.195 0.201 13.887 14.643 multiply_cannon_loop 2286 14.5 0.869 0.908 13.095 13.883 ot_mini 108 10.5 0.001 0.001 9.169 9.182 qs_ot_get_derivative 108 11.5 0.001 0.001 7.666 7.678 multiply_cannon_multrec 18288 15.5 1.893 2.841 6.987 7.211 rebuild_ks_matrix 119 8.3 0.000 0.000 6.026 6.037 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.025 6.036 dbcsr_mm_accdrv_process 38222 16.0 4.888 5.781 4.999 5.846 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.313 5.323 mp_waitall_1 158411 16.6 2.679 3.823 2.679 3.823 init_scf_run 11 5.9 0.000 0.001 3.794 3.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.794 3.794 sum_up_and_integrate 119 10.3 0.001 0.001 3.584 3.589 integrate_v_rspace 119 11.3 0.003 0.003 3.571 3.578 qs_ot_get_p 119 10.4 0.001 0.002 3.529 3.557 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.872 3.468 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.352 3.360 calculate_rho_elec 119 8.7 0.031 0.032 3.352 3.359 calculate_first_density_matrix 1 7.0 0.001 0.004 2.535 2.537 init_scf_loop 11 6.9 0.001 0.003 2.530 2.531 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.434 2.471 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.366 2.371 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.009 2.354 apply_single 119 13.6 0.000 0.000 2.009 2.354 make_m2s 4572 13.5 0.044 0.046 1.919 2.064 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.051 2.051 make_images 4572 14.5 0.193 0.206 1.832 1.975 density_rs2pw 119 9.7 0.004 0.004 1.909 1.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.949 1.954 pw_transfer 1439 11.6 0.065 0.068 1.909 1.920 grid_integrate_task_list 119 12.3 1.798 1.899 1.798 1.899 calculate_dm_sparse 119 9.5 0.000 0.001 1.857 1.865 acc_transpose_blocks 18288 15.5 0.080 0.082 1.813 1.861 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.816 1.829 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.782 1.793 cp_fm_diag_elpa_base 50 14.0 1.758 1.770 1.780 1.791 prepare_preconditioner 11 7.9 0.000 0.000 1.767 1.770 make_preconditioner 11 8.9 0.000 0.000 1.767 1.770 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.755 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.754 1.760 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.618 1.705 ot_diis_step 108 11.5 0.011 0.012 1.487 1.487 mp_sum_l 7287 12.8 1.091 1.463 1.091 1.463 fft_wrap_pw1pw2_140 487 13.2 0.179 0.182 1.442 1.454 potential_pw2rs 119 12.3 0.007 0.008 1.380 1.384 fft3d_ps 1201 14.6 0.549 0.569 1.371 1.382 grid_collocate_task_list 119 9.7 1.235 1.295 1.235 1.295 transfer_rs2pw 487 10.6 0.005 0.005 1.144 1.253 wfi_extrapolate 11 7.9 0.001 0.001 1.203 1.203 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.197 1.202 multiply_cannon_sync_h2d 18288 15.5 1.037 1.168 1.037 1.168 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.165 1.166 make_images_data 4572 15.5 0.047 0.051 0.826 0.992 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.970 0.988 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.955 0.972 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.722 0.913 mp_alltoall_d11v 2130 13.8 0.712 0.839 0.712 0.839 transfer_pw2rs 487 13.2 0.004 0.004 0.834 0.838 acc_transpose_blocks_sync 54864 16.5 0.745 0.797 0.745 0.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.787 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.732 0.784 mp_waitany 9880 13.7 0.653 0.782 0.653 0.782 jit_kernel_transpose 5 15.6 0.753 0.771 0.753 0.771 cp_fm_cholesky_invert 11 10.9 0.748 0.752 0.748 0.752 transfer_rs2pw_140 130 11.5 0.115 0.121 0.640 0.748 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.665 0.742 mp_alltoall_z22v 1201 16.6 0.669 0.737 0.669 0.737 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.188000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.363636, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 559.919104E+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 0.011 0.029 37.090 37.091 qs_mol_dyn_low 1 2.0 0.003 0.003 36.847 36.853 qs_forces 11 3.9 0.002 0.002 36.787 36.787 qs_energies 11 4.9 0.002 0.005 35.052 35.056 scf_env_do_scf 11 5.9 0.001 0.002 29.659 29.660 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.194 26.196 dbcsr_multiply_generic 2286 12.5 0.100 0.102 20.127 20.272 velocity_verlet 10 3.0 0.001 0.002 18.635 18.637 qs_scf_new_mos 108 7.5 0.001 0.001 18.024 18.082 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.023 18.081 ot_scf_mini 108 9.5 0.002 0.003 16.981 17.032 multiply_cannon 2286 13.5 0.219 0.227 16.282 16.747 multiply_cannon_loop 2286 14.5 1.529 1.629 15.315 15.674 ot_mini 108 10.5 0.001 0.001 10.380 10.444 multiply_cannon_multrec 27432 15.5 2.458 3.145 9.008 9.282 qs_ot_get_derivative 108 11.5 0.001 0.001 8.519 8.572 dbcsr_mm_accdrv_process 47916 15.9 6.158 7.712 6.447 7.842 rebuild_ks_matrix 119 8.3 0.000 0.000 6.408 6.458 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 6.407 6.458 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.658 5.702 init_scf_run 11 5.9 0.000 0.001 3.872 3.872 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.872 3.872 qs_ot_get_p 119 10.4 0.001 0.002 3.743 3.807 sum_up_and_integrate 119 10.3 0.001 0.002 3.567 3.579 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.105 3.574 integrate_v_rspace 119 11.3 0.003 0.003 3.555 3.568 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.415 3.448 calculate_rho_elec 119 8.7 0.040 0.046 3.415 3.448 init_scf_loop 11 6.9 0.001 0.003 3.445 3.447 acc_transpose_blocks 27432 15.5 0.120 0.123 2.592 2.768 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.171 2.643 apply_single 119 13.6 0.000 0.000 2.171 2.642 prepare_preconditioner 11 7.9 0.000 0.000 2.601 2.609 make_preconditioner 11 8.9 0.000 0.001 2.601 2.609 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.204 2.536 calculate_first_density_matrix 1 7.0 0.000 0.002 2.468 2.471 make_m2s 4572 13.5 0.054 0.056 2.334 2.436 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.361 2.387 make_images 4572 14.5 0.274 0.336 2.226 2.328 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.281 2.294 mp_waitall_1 137007 16.6 1.790 2.287 1.790 2.287 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.197 pw_transfer 1439 11.6 0.065 0.070 2.061 2.099 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.054 2.056 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.969 2.010 density_rs2pw 119 9.7 0.004 0.004 1.870 1.970 grid_integrate_task_list 119 12.3 1.831 1.902 1.831 1.902 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.886 1.899 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.863 1.864 ot_diis_step 108 11.5 0.012 0.012 1.818 1.818 mp_sum_l 7287 12.8 1.086 1.754 1.086 1.754 fft_wrap_pw1pw2_140 487 13.2 0.226 0.234 1.614 1.662 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.607 1.616 cp_fm_diag_elpa_base 50 14.0 1.573 1.590 1.604 1.613 acc_transpose_blocks_sync 82296 16.5 1.424 1.530 1.424 1.530 fft3d_ps 1201 14.6 0.592 0.651 1.459 1.488 potential_pw2rs 119 12.3 0.008 0.009 1.370 1.378 wfi_extrapolate 11 7.9 0.001 0.001 1.349 1.349 grid_collocate_task_list 119 9.7 1.248 1.330 1.248 1.330 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.308 1.326 qs_energies_init_hamiltonians 11 5.9 0.005 0.007 1.251 1.251 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.228 1.239 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.713 1.206 cp_fm_upper_to_full 72 14.2 0.834 1.186 0.834 1.186 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 1.021 1.089 dbcsr_complete_redistribute 329 12.2 0.116 0.144 0.762 1.034 transfer_rs2pw 487 10.6 0.004 0.005 0.925 1.026 make_images_data 4572 15.5 0.048 0.053 0.873 0.995 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.752 0.934 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.800 0.877 mp_alltoall_d11v 2130 13.8 0.763 0.862 0.763 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.845 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.564 0.832 jit_kernel_transpose 5 15.6 0.751 0.816 0.751 0.816 mp_alltoall_z22v 1201 16.6 0.749 0.784 0.749 0.784 transfer_pw2rs 487 13.2 0.004 0.004 0.768 0.773 cp_fm_cholesky_invert 11 10.9 0.757 0.761 0.757 0.761 jit_kernel_multiply 4 16.3 0.223 0.752 0.223 0.752 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.683 0.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.091000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.090909, yerr=4.521756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 628.023296E+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 0.008 0.029 30.688 30.689 qs_mol_dyn_low 1 2.0 0.003 0.003 30.477 30.490 qs_forces 11 3.9 0.002 0.002 30.239 30.240 qs_energies 11 4.9 0.002 0.002 28.525 28.527 scf_env_do_scf 11 5.9 0.000 0.001 22.305 22.305 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.821 19.822 velocity_verlet 10 3.0 0.002 0.002 15.021 15.035 dbcsr_multiply_generic 2286 12.5 0.093 0.095 14.438 14.502 multiply_cannon 2286 13.5 0.224 0.230 11.836 12.301 qs_scf_new_mos 108 7.5 0.001 0.001 12.074 12.096 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.073 12.095 ot_scf_mini 108 9.5 0.002 0.002 11.361 11.386 multiply_cannon_loop 2286 14.5 0.648 0.671 10.936 11.122 ot_mini 108 10.5 0.001 0.001 6.583 6.616 multiply_cannon_multrec 9144 15.5 1.852 2.041 6.277 6.479 rebuild_ks_matrix 119 8.3 0.000 0.000 5.840 5.862 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.839 5.862 qs_ot_get_derivative 108 11.5 0.001 0.001 5.282 5.308 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.194 5.214 init_scf_run 11 5.9 0.000 0.001 4.776 4.776 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.775 4.775 dbcsr_mm_accdrv_process 12550 15.8 3.678 4.320 4.313 4.392 calculate_first_density_matrix 1 7.0 0.000 0.000 3.630 3.631 sum_up_and_integrate 119 10.3 0.001 0.001 3.476 3.481 integrate_v_rspace 119 11.3 0.003 0.007 3.466 3.470 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.395 3.404 calculate_rho_elec 119 8.7 0.060 0.061 3.394 3.403 qs_ot_get_p 119 10.4 0.002 0.002 2.880 2.915 acc_transpose_blocks 9144 15.5 0.043 0.044 2.364 2.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.487 2.488 init_scf_loop 11 6.9 0.000 0.000 2.463 2.464 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.325 2.333 pw_transfer 1439 11.6 0.065 0.068 2.166 2.174 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.072 2.083 mp_waitall_1 115863 16.7 1.575 2.061 1.575 2.061 calculate_dm_sparse 119 9.5 0.000 0.000 1.955 1.973 make_m2s 4572 13.5 0.034 0.035 1.812 1.952 grid_integrate_task_list 119 12.3 1.867 1.920 1.867 1.920 density_rs2pw 119 9.7 0.003 0.003 1.813 1.900 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.877 1.880 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.849 1.868 make_images 4572 14.5 0.270 0.299 1.723 1.860 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.581 1.770 prepare_preconditioner 11 7.9 0.000 0.000 1.734 1.739 make_preconditioner 11 8.9 0.000 0.000 1.734 1.739 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.717 1.728 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.628 1.655 jit_kernel_transpose 5 15.6 1.464 1.651 1.464 1.651 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.631 1.632 fft3d_ps 1201 14.6 0.652 0.667 1.424 1.435 grid_collocate_task_list 119 9.7 1.296 1.405 1.296 1.405 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.366 1.379 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.362 1.370 cp_fm_diag_elpa_base 50 14.0 1.334 1.352 1.360 1.368 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.340 1.344 potential_pw2rs 119 12.3 0.010 0.011 1.288 1.293 ot_diis_step 108 11.5 0.013 0.013 1.285 1.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.222 1.264 apply_single 119 13.6 0.000 0.000 1.222 1.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.229 1.230 jit_kernel_multiply 6 15.7 0.595 1.105 0.595 1.105 wfi_extrapolate 11 7.9 0.001 0.001 1.093 1.093 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.773 0.969 make_images_data 4572 15.5 0.042 0.046 0.781 0.947 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.423 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.923 transfer_rs2pw 487 10.6 0.004 0.004 0.796 0.881 cp_fm_cholesky_invert 11 10.9 0.873 0.876 0.873 0.876 mp_alltoall_d11v 2130 13.8 0.735 0.845 0.735 0.845 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.836 0.843 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.766 0.768 acc_transpose_blocks_sync 27432 16.5 0.724 0.760 0.724 0.760 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.678 0.726 parallel_gemm_fm 81 9.0 0.000 0.000 0.705 0.707 parallel_gemm_fm_cosma 81 10.0 0.705 0.706 0.705 0.706 transfer_pw2rs 487 13.2 0.003 0.004 0.679 0.683 mp_alltoall_z22v 1201 16.6 0.645 0.683 0.645 0.683 make_basis_sm 11 9.8 0.000 0.000 0.673 0.674 mp_allgather_i34 2286 14.5 0.216 0.633 0.216 0.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.689000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.454545, yerr=6.932973 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 787.193856E+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 0.011 0.031 44.516 44.517 qs_mol_dyn_low 1 2.0 0.003 0.003 44.283 44.290 qs_forces 11 3.9 0.002 0.002 44.221 44.222 qs_energies 11 4.9 0.002 0.002 42.206 42.209 scf_env_do_scf 11 5.9 0.001 0.001 35.029 35.029 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.116 27.118 velocity_verlet 10 3.0 0.002 0.002 24.267 24.273 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.087 20.252 qs_scf_new_mos 108 7.5 0.001 0.001 17.393 17.493 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.393 17.492 multiply_cannon 2286 13.5 0.297 0.306 16.071 16.917 ot_scf_mini 108 9.5 0.002 0.002 16.222 16.327 multiply_cannon_loop 2286 14.5 0.863 0.880 14.808 15.658 ot_mini 108 10.5 0.001 0.001 9.896 10.018 multiply_cannon_multrec 9144 15.5 3.458 4.850 9.012 9.204 qs_ot_get_derivative 108 11.5 0.001 0.001 7.827 7.932 init_scf_loop 11 6.9 0.000 0.000 7.886 7.889 rebuild_ks_matrix 119 8.3 0.000 0.000 7.298 7.450 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.297 7.450 prepare_preconditioner 11 7.9 0.000 0.000 6.918 6.932 make_preconditioner 11 8.9 0.000 0.000 6.918 6.932 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.454 6.811 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.579 6.717 dbcsr_mm_accdrv_process 12550 15.8 4.644 6.652 5.411 6.684 init_scf_run 11 5.9 0.000 0.001 5.117 5.118 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.117 5.117 cp_fm_upper_to_full 72 14.2 3.228 4.638 3.228 4.638 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.206 4.212 calculate_rho_elec 119 8.7 0.117 0.120 4.205 4.211 sum_up_and_integrate 119 10.3 0.001 0.001 3.852 3.859 integrate_v_rspace 119 11.3 0.003 0.004 3.842 3.849 calculate_first_density_matrix 1 7.0 0.000 0.000 3.645 3.647 qs_ot_get_p 119 10.4 0.002 0.002 3.394 3.533 mp_waitall_1 94719 16.7 2.571 3.506 2.571 3.506 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.828 3.227 pw_transfer 1439 11.6 0.068 0.069 3.049 3.054 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.951 2.955 dbcsr_complete_redistribute 329 12.2 0.281 0.286 1.997 2.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.785 2.785 acc_transpose_blocks 9144 15.5 0.045 0.046 2.411 2.766 make_m2s 4572 13.5 0.037 0.038 2.422 2.571 mp_alltoall_i22 627 13.8 1.561 2.569 1.561 2.569 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.215 2.521 apply_single 119 13.6 0.000 0.000 2.214 2.521 fft_wrap_pw1pw2_140 487 13.2 0.648 0.651 2.512 2.519 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.692 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.417 2.502 calculate_dm_sparse 119 9.5 0.000 0.000 2.445 2.460 make_images 4572 14.5 0.354 0.388 2.300 2.450 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.458 2.269 density_rs2pw 119 9.7 0.003 0.003 2.225 2.245 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.476 2.152 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.074 2.129 grid_integrate_task_list 119 12.3 2.076 2.119 2.076 2.119 ot_diis_step 108 11.5 0.014 0.015 2.042 2.042 qs_ot_p2m_diag 50 11.0 0.044 0.045 1.997 1.999 fft3d_ps 1201 14.6 0.869 0.890 1.873 1.877 mp_sum_l 7287 12.8 1.048 1.819 1.048 1.819 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.768 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.639 1.639 acc_transpose_blocks_kernels 9144 16.5 0.120 0.121 1.251 1.635 grid_collocate_task_list 119 9.7 1.519 1.539 1.519 1.539 jit_kernel_transpose 5 15.6 1.131 1.514 1.131 1.514 cp_fm_cholesky_invert 11 10.9 1.452 1.456 1.452 1.456 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.438 1.446 wfi_extrapolate 11 7.9 0.001 0.001 1.406 1.406 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.146 1.382 potential_pw2rs 119 12.3 0.014 0.014 1.369 1.370 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.349 1.349 cp_fm_diag_elpa_base 50 14.0 1.202 1.254 1.348 1.348 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.307 1.327 make_images_data 4572 15.5 0.046 0.049 1.084 1.277 mp_alltoall_d11v 2130 13.8 1.173 1.207 1.173 1.207 acc_transpose_blocks_sync 27432 16.5 1.098 1.126 1.098 1.126 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.094 1.114 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.993 1.040 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.813 1.029 jit_kernel_multiply 6 15.5 0.738 1.009 0.738 1.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.929 0.943 qs_create_task_list 11 7.9 0.001 0.001 0.926 0.938 generate_qs_task_list 11 8.9 0.367 0.388 0.926 0.938 mp_alltoall_z22v 1201 16.6 0.866 0.895 0.866 0.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.517000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=735.727273, yerr=17.199846 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 501.587968E+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 2395560. 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 0.045 0.070 83.567 83.568 qs_mol_dyn_low 1 2.0 0.003 0.004 83.176 83.186 qs_forces 11 3.9 0.004 0.008 83.097 83.098 qs_energies 11 4.9 0.002 0.007 80.200 80.214 scf_env_do_scf 11 5.9 0.001 0.002 70.875 70.877 scf_env_do_scf_inner_loop 99 6.5 0.002 0.009 65.422 65.423 dbcsr_multiply_generic 2055 12.4 0.104 0.107 52.533 52.721 qs_scf_new_mos 99 7.5 0.000 0.001 48.419 48.514 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.419 48.513 ot_scf_mini 99 9.5 0.002 0.003 46.011 46.125 multiply_cannon 2055 13.4 0.182 0.188 43.386 44.275 velocity_verlet 10 3.0 0.001 0.001 43.922 43.922 multiply_cannon_loop 2055 14.4 1.806 1.853 42.455 43.350 ot_mini 99 10.5 0.001 0.001 27.643 27.743 qs_ot_get_derivative 99 11.5 0.001 0.001 20.779 20.892 multiply_cannon_multrec 49320 15.4 11.333 11.986 17.348 18.058 rebuild_ks_matrix 110 8.3 0.000 0.001 14.427 14.523 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 14.427 14.522 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.642 12.735 mp_waitall_1 220248 16.4 11.360 12.416 11.360 12.416 multiply_cannon_sync_h2d 49320 15.4 9.579 10.111 9.579 10.111 qs_ot_get_p 110 10.4 0.001 0.002 9.827 9.940 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.045 9.027 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.718 8.213 apply_single 110 13.6 0.000 0.001 7.718 8.212 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.850 7.915 init_scf_run 11 5.9 0.000 0.001 7.216 7.216 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.216 7.216 sum_up_and_integrate 110 10.3 0.002 0.004 6.937 6.950 integrate_v_rspace 110 11.3 0.003 0.004 6.911 6.930 ot_diis_step 99 11.5 0.005 0.006 6.602 6.602 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.473 6.488 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.332 6.481 calculate_rho_elec 110 8.6 0.021 0.025 6.331 6.481 dbcsr_mm_accdrv_process 87628 16.1 3.054 3.176 5.885 6.188 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.650 5.650 mp_sum_l 6594 12.7 4.784 5.620 4.784 5.620 init_scf_loop 11 6.9 0.001 0.003 5.427 5.429 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.352 5.404 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.080 5.105 cp_fm_diag_elpa_base 48 14.0 5.067 5.094 5.078 5.103 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.036 4.041 make_m2s 4110 13.4 0.060 0.065 3.910 4.022 wfi_extrapolate 11 7.9 0.001 0.001 4.014 4.015 make_images 4110 14.4 0.178 0.190 3.815 3.933 calculate_dm_sparse 110 9.5 0.001 0.001 3.782 3.899 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.711 3.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.647 3.675 density_rs2pw 110 9.6 0.004 0.004 3.309 3.474 grid_integrate_task_list 110 12.3 3.255 3.394 3.255 3.394 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.276 3.340 acc_transpose_blocks 49320 15.4 0.211 0.219 2.845 3.319 prepare_preconditioner 11 7.9 0.000 0.000 3.250 3.276 make_preconditioner 11 8.9 0.000 0.001 3.250 3.276 pw_transfer 1331 11.6 0.055 0.065 3.031 3.109 calculate_first_density_matrix 1 7.0 0.001 0.013 3.097 3.106 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.055 3.094 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.943 3.020 jit_kernel_multiply 13 15.9 2.550 2.876 2.550 2.876 fft_wrap_pw1pw2_140 451 13.1 0.311 0.335 2.519 2.606 mp_alltoall_d11v 2046 13.8 2.097 2.592 2.097 2.592 potential_pw2rs 110 12.3 0.005 0.006 2.534 2.548 grid_collocate_task_list 110 9.6 2.158 2.292 2.158 2.292 fft3d_ps 1111 14.6 0.791 0.885 2.233 2.290 transfer_rs2pw 451 10.6 0.005 0.006 1.906 2.056 mp_waitany 14300 13.8 1.854 2.037 1.854 2.037 mp_sum_d 3891 11.9 1.495 1.970 1.495 1.970 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.931 make_images_data 4110 15.4 0.044 0.048 1.779 1.914 hybrid_alltoall_any 4261 16.3 0.085 0.488 1.550 1.867 acc_transpose_blocks_kernels 49320 16.4 0.423 0.437 1.320 1.778 cp_fm_cholesky_invert 11 10.9 1.767 1.771 1.767 1.771 transfer_pw2rs 451 13.1 0.006 0.007 1.710 1.718 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.648 1.686 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.568000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 594.751488E+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 1636728. 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 0.012 0.029 69.320 69.323 qs_mol_dyn_low 1 2.0 0.003 0.007 69.033 69.043 qs_forces 11 3.9 0.004 0.014 68.932 68.933 qs_energies 11 4.9 0.001 0.003 65.615 65.619 scf_env_do_scf 11 5.9 0.001 0.002 56.764 56.767 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.253 49.253 dbcsr_multiply_generic 2055 12.4 0.112 0.116 38.668 38.856 velocity_verlet 10 3.0 0.001 0.002 35.895 35.897 qs_scf_new_mos 99 7.5 0.001 0.001 33.477 33.652 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.477 33.652 multiply_cannon 2055 13.4 0.223 0.243 31.850 33.179 ot_scf_mini 99 9.5 0.003 0.003 31.808 31.954 multiply_cannon_loop 2055 14.4 1.166 1.204 30.591 31.635 ot_mini 99 10.5 0.001 0.002 18.748 18.910 multiply_cannon_multrec 24660 15.4 6.990 8.523 14.095 15.438 rebuild_ks_matrix 110 8.3 0.000 0.000 13.496 13.596 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 13.495 13.595 qs_ot_get_derivative 99 11.5 0.001 0.001 12.970 13.121 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.875 11.965 mp_waitall_1 176588 16.5 7.685 10.479 7.685 10.479 multiply_cannon_metrocomm3 24660 15.4 0.075 0.077 5.275 8.733 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.713 7.606 apply_single 110 13.6 0.000 0.001 6.713 7.605 init_scf_loop 11 6.9 0.000 0.000 7.476 7.477 dbcsr_mm_accdrv_process 52282 16.1 5.572 6.368 6.934 7.472 multiply_cannon_sync_h2d 24660 15.4 6.419 7.402 6.419 7.402 qs_ot_get_p 110 10.4 0.001 0.001 6.415 6.628 init_scf_run 11 5.9 0.000 0.001 6.439 6.439 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.438 6.439 sum_up_and_integrate 110 10.3 0.002 0.003 6.331 6.346 integrate_v_rspace 110 11.3 0.003 0.004 6.304 6.320 qs_rho_update_rho_low 110 7.6 0.001 0.002 5.843 5.851 calculate_rho_elec 110 8.6 0.039 0.047 5.843 5.851 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.891 5.818 ot_diis_step 99 11.5 0.010 0.013 5.731 5.732 prepare_preconditioner 11 7.9 0.000 0.000 5.466 5.495 make_preconditioner 11 8.9 0.000 0.000 5.466 5.495 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.033 5.190 make_m2s 4110 13.4 0.057 0.060 4.207 4.643 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.512 4.532 make_images 4110 14.4 0.408 0.474 4.096 4.529 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.050 4.051 pw_transfer 1331 11.6 0.065 0.070 3.528 3.666 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.615 3.617 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.422 3.558 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.487 3.501 cp_fm_diag_elpa_base 48 14.0 3.440 3.453 3.485 3.497 wfi_extrapolate 11 7.9 0.001 0.004 3.495 3.495 grid_integrate_task_list 110 12.3 3.171 3.394 3.171 3.394 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.285 3.369 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.284 3.338 density_rs2pw 110 9.6 0.004 0.004 3.093 3.286 fft_wrap_pw1pw2_140 451 13.1 0.358 0.376 2.934 3.071 calculate_dm_sparse 110 9.5 0.001 0.001 2.964 2.995 calculate_first_density_matrix 1 7.0 0.000 0.001 2.848 2.850 make_images_data 4110 15.4 0.049 0.053 2.290 2.753 mp_sum_l 6594 12.7 1.994 2.751 1.994 2.751 hybrid_alltoall_any 4261 16.3 0.105 0.458 2.014 2.702 fft3d_ps 1111 14.6 1.114 1.350 2.513 2.677 cp_fm_cholesky_invert 11 10.9 2.546 2.553 2.546 2.553 acc_transpose_blocks 24660 15.4 0.117 0.120 2.230 2.366 grid_collocate_task_list 110 9.6 2.167 2.311 2.167 2.311 potential_pw2rs 110 12.3 0.008 0.009 2.208 2.216 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.958 1.981 mp_alltoall_d11v 2046 13.8 1.746 1.957 1.746 1.957 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.898 1.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.776 1.787 jit_kernel_multiply 9 16.3 0.993 1.767 0.993 1.767 transfer_rs2pw 451 10.6 0.006 0.007 1.442 1.617 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.759 1.601 mp_allgather_i34 2055 14.4 0.551 1.562 0.551 1.562 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.529 1.540 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.494 mp_irecv_dv 57340 16.2 0.635 1.484 0.635 1.484 mp_waitany 10164 13.8 1.198 1.398 1.198 1.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.323000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=562.000000, yerr=6.522688 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.161152E+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 0.008 0.028 61.081 61.082 qs_mol_dyn_low 1 2.0 0.003 0.003 60.818 60.827 qs_forces 11 3.9 0.003 0.003 60.753 60.754 qs_energies 11 4.9 0.001 0.002 57.602 57.605 scf_env_do_scf 11 5.9 0.000 0.001 49.614 49.614 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.817 40.818 velocity_verlet 10 3.0 0.001 0.002 33.051 33.052 dbcsr_multiply_generic 2055 12.4 0.108 0.110 29.289 29.574 qs_scf_new_mos 99 7.5 0.001 0.001 25.892 26.022 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.892 26.021 ot_scf_mini 99 9.5 0.002 0.003 24.622 24.765 multiply_cannon 2055 13.4 0.213 0.223 22.766 24.065 multiply_cannon_loop 2055 14.4 0.821 0.842 21.531 22.590 ot_mini 99 10.5 0.001 0.001 14.061 14.204 rebuild_ks_matrix 110 8.3 0.000 0.000 12.303 12.453 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.303 12.453 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.844 10.983 multiply_cannon_multrec 16440 15.4 3.535 4.700 9.731 10.819 mp_waitall_1 139946 16.5 7.307 10.286 7.307 10.286 qs_ot_get_derivative 99 11.5 0.001 0.001 9.529 9.671 init_scf_loop 11 6.9 0.000 0.000 8.764 8.766 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.614 7.342 prepare_preconditioner 11 7.9 0.000 0.000 7.000 7.018 make_preconditioner 11 8.9 0.000 0.000 7.000 7.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.318 6.716 sum_up_and_integrate 110 10.3 0.001 0.002 6.364 6.379 integrate_v_rspace 110 11.3 0.003 0.003 6.338 6.353 dbcsr_mm_accdrv_process 34862 16.1 5.341 5.695 6.041 6.189 qs_ot_get_p 110 10.4 0.001 0.001 5.658 5.832 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.797 5.809 calculate_rho_elec 110 8.6 0.059 0.060 5.796 5.808 init_scf_run 11 5.9 0.000 0.001 5.598 5.598 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.598 5.598 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.056 5.536 apply_single 110 13.6 0.000 0.000 5.056 5.536 make_m2s 4110 13.4 0.049 0.051 4.294 4.700 make_images 4110 14.4 0.401 0.525 4.178 4.585 ot_diis_step 99 11.5 0.011 0.011 4.495 4.495 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.938 3.945 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.243 3.918 multiply_cannon_sync_h2d 16440 15.4 3.203 3.883 3.203 3.883 pw_transfer 1331 11.6 0.065 0.072 3.660 3.674 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.553 3.570 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.555 3.555 grid_integrate_task_list 110 12.3 3.200 3.396 3.200 3.396 density_rs2pw 110 9.6 0.004 0.004 3.028 3.130 fft_wrap_pw1pw2_140 451 13.1 0.459 0.474 3.051 3.075 wfi_extrapolate 11 7.9 0.001 0.001 3.014 3.014 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.002 3.013 cp_fm_diag_elpa_base 48 14.0 2.931 2.963 2.999 3.010 make_images_data 4110 15.4 0.047 0.051 2.430 2.944 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.173 2.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.839 2.840 cp_fm_cholesky_invert 11 10.9 2.756 2.762 2.756 2.762 calculate_dm_sparse 110 9.5 0.001 0.001 2.580 2.617 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.527 2.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.498 2.554 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.920 2.514 calculate_first_density_matrix 1 7.0 0.000 0.000 2.491 2.493 fft3d_ps 1111 14.6 1.122 1.163 2.471 2.490 grid_collocate_task_list 110 9.6 2.219 2.400 2.219 2.400 mp_irecv_dv 48980 15.7 0.844 2.380 0.844 2.380 mp_sum_l 6594 12.7 1.601 2.268 1.601 2.268 potential_pw2rs 110 12.3 0.011 0.011 2.199 2.204 mp_alltoall_d11v 2046 13.8 1.820 2.102 1.820 2.102 dbcsr_complete_redistribute 325 12.2 0.325 0.347 1.481 1.972 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.941 1.941 cp_fm_upper_to_full 70 14.2 1.462 1.856 1.462 1.856 mp_allgather_i34 2055 14.4 0.543 1.667 0.543 1.667 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.652 1.667 acc_transpose_blocks 16440 15.4 0.079 0.081 1.634 1.665 cp_fm_cholesky_decompose 22 10.9 1.621 1.644 1.621 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.499 1.513 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.003 1.467 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.427 transfer_rs2pw 451 10.6 0.005 0.006 1.286 1.410 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.308 1.318 mp_waitany 17072 13.8 1.141 1.291 1.141 1.291 transfer_pw2rs 451 13.1 0.005 0.005 1.218 1.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.082000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.181818, yerr=8.788610 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 740.872192E+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 0.024 0.045 66.999 67.001 qs_mol_dyn_low 1 2.0 0.003 0.003 66.699 66.709 qs_forces 11 3.9 0.003 0.003 66.634 66.634 qs_energies 11 4.9 0.001 0.002 63.217 63.220 scf_env_do_scf 11 5.9 0.001 0.001 54.782 54.784 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.791 42.792 velocity_verlet 10 3.0 0.001 0.002 37.923 37.925 dbcsr_multiply_generic 2055 12.4 0.115 0.120 30.955 31.230 qs_scf_new_mos 99 7.5 0.001 0.001 27.990 28.099 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.989 28.099 ot_scf_mini 99 9.5 0.002 0.003 26.294 26.400 multiply_cannon 2055 13.4 0.239 0.258 23.650 24.672 multiply_cannon_loop 2055 14.4 1.407 1.489 22.184 23.016 ot_mini 99 10.5 0.001 0.001 14.940 15.075 multiply_cannon_multrec 24660 15.4 4.103 6.910 13.023 14.109 rebuild_ks_matrix 110 8.3 0.000 0.000 12.102 12.218 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.101 12.218 init_scf_loop 11 6.9 0.000 0.000 11.949 11.950 qs_ot_get_derivative 99 11.5 0.001 0.001 10.755 10.867 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.694 10.797 prepare_preconditioner 11 7.9 0.000 0.000 10.202 10.220 make_preconditioner 11 8.9 0.000 0.000 10.202 10.220 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.377 9.886 dbcsr_mm_accdrv_process 52304 16.0 7.747 9.099 8.762 9.716 sum_up_and_integrate 110 10.3 0.002 0.002 6.325 6.341 qs_ot_get_p 110 10.4 0.001 0.001 6.174 6.335 integrate_v_rspace 110 11.3 0.003 0.003 6.299 6.314 mp_waitall_1 121746 16.5 4.308 6.104 4.308 6.104 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.935 5.944 calculate_rho_elec 110 8.6 0.078 0.081 5.935 5.943 make_m2s 4110 13.4 0.059 0.062 5.558 5.915 make_images 4110 14.4 0.576 0.696 5.417 5.769 init_scf_run 11 5.9 0.000 0.001 5.693 5.693 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.692 5.693 cp_fm_upper_to_full 70 14.2 3.436 4.996 3.436 4.996 qs_ot_p2m_diag 48 11.0 0.055 0.065 4.220 4.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.042 4.149 apply_single 110 13.6 0.000 0.000 4.042 4.148 ot_diis_step 99 11.5 0.011 0.011 4.145 4.145 pw_transfer 1331 11.6 0.065 0.074 3.846 3.881 dbcsr_complete_redistribute 325 12.2 0.409 0.451 2.697 3.879 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.740 3.777 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.642 3.644 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.560 3.614 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.441 3.501 grid_integrate_task_list 110 12.3 3.268 3.445 3.268 3.445 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.183 3.347 fft_wrap_pw1pw2_140 451 13.1 0.532 0.545 3.189 3.229 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.374 3.209 density_rs2pw 110 9.6 0.004 0.004 3.046 3.162 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.318 3.148 make_images_data 4110 15.4 0.049 0.053 2.759 3.124 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.068 3.079 cp_fm_diag_elpa_base 48 14.0 2.912 2.975 3.065 3.078 wfi_extrapolate 11 7.9 0.001 0.001 3.071 3.071 calculate_dm_sparse 110 9.5 0.001 0.001 3.000 3.034 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.816 2.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.932 2.934 mp_alltoall_i22 605 13.7 1.723 2.919 1.723 2.919 cp_fm_cholesky_invert 11 10.9 2.896 2.904 2.896 2.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.549 2.595 acc_transpose_blocks 24660 15.4 0.112 0.114 2.474 2.589 fft3d_ps 1111 14.6 1.153 1.193 2.564 2.584 multiply_cannon_sync_h2d 24660 15.4 2.365 2.564 2.365 2.564 calculate_first_density_matrix 1 7.0 0.001 0.006 2.526 2.529 grid_collocate_task_list 110 9.6 2.266 2.465 2.266 2.465 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 2.237 2.237 mp_alltoall_d11v 2046 13.8 1.846 2.124 1.846 2.124 potential_pw2rs 110 12.3 0.012 0.013 2.111 2.117 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.827 1.871 cp_fm_cholesky_decompose 22 10.9 1.669 1.719 1.669 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.716 mp_sum_l 6594 12.7 1.057 1.662 1.057 1.662 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.648 1.660 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.551 1.567 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.852 1.537 mp_allgather_i34 2055 14.4 0.551 1.522 0.551 1.522 acc_transpose_blocks_sync 73980 16.4 1.387 1.517 1.387 1.517 mp_irecv_dv 62702 16.1 0.749 1.457 0.749 1.457 transfer_rs2pw 451 10.6 0.005 0.005 1.195 1.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.001000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.818182, yerr=10.675530 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 856.887296E+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 0.029 0.034 56.902 56.903 qs_mol_dyn_low 1 2.0 0.003 0.003 56.358 56.367 qs_forces 11 3.9 0.003 0.003 56.293 56.294 qs_energies 11 4.9 0.001 0.001 52.621 52.627 scf_env_do_scf 11 5.9 0.000 0.001 44.019 44.019 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.024 36.024 velocity_verlet 10 3.0 0.002 0.002 31.739 31.742 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.935 24.116 qs_scf_new_mos 99 7.5 0.001 0.001 21.254 21.340 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.254 21.339 ot_scf_mini 99 9.5 0.002 0.002 19.984 20.048 multiply_cannon 2055 13.4 0.238 0.251 18.234 19.453 multiply_cannon_loop 2055 14.4 0.604 0.620 16.962 17.317 rebuild_ks_matrix 110 8.3 0.000 0.000 11.770 11.843 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.769 11.842 ot_mini 99 10.5 0.001 0.001 11.070 11.128 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.466 10.529 multiply_cannon_multrec 8220 15.4 3.180 4.716 7.584 8.760 init_scf_loop 11 6.9 0.000 0.000 7.947 7.948 mp_waitall_1 103326 16.6 6.238 7.837 6.238 7.837 qs_ot_get_derivative 99 11.5 0.001 0.001 7.265 7.327 sum_up_and_integrate 110 10.3 0.001 0.002 6.304 6.317 prepare_preconditioner 11 7.9 0.000 0.000 6.287 6.293 make_preconditioner 11 8.9 0.000 0.000 6.287 6.293 integrate_v_rspace 110 11.3 0.003 0.005 6.277 6.289 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.104 6.123 calculate_rho_elec 110 8.6 0.113 0.113 6.104 6.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.846 5.921 init_scf_run 11 5.9 0.000 0.001 5.459 5.459 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.459 5.459 dbcsr_mm_accdrv_process 17442 15.9 3.168 4.126 4.264 5.171 qs_ot_get_p 110 10.4 0.001 0.002 4.935 4.991 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.126 4.558 make_m2s 4110 13.4 0.038 0.039 4.264 4.468 make_images 4110 14.4 0.649 0.708 4.134 4.334 pw_transfer 1331 11.6 0.066 0.070 4.191 4.205 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.083 4.099 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.813 3.898 apply_single 110 13.6 0.000 0.000 3.813 3.898 ot_diis_step 99 11.5 0.012 0.015 3.780 3.781 fft_wrap_pw1pw2_140 451 13.1 0.719 0.735 3.511 3.534 grid_integrate_task_list 110 12.3 3.386 3.529 3.386 3.529 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.499 3.504 density_rs2pw 110 9.6 0.004 0.004 3.112 3.211 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.144 3.144 cp_fm_cholesky_invert 11 10.9 3.052 3.055 3.052 3.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.850 2.852 wfi_extrapolate 11 7.9 0.001 0.001 2.764 2.764 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 2.682 2.683 hybrid_alltoall_any 4261 16.3 0.199 0.851 2.227 2.647 fft3d_ps 1111 14.6 1.293 1.311 2.626 2.641 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.627 2.635 cp_fm_diag_elpa_base 48 14.0 2.571 2.594 2.625 2.633 calculate_first_density_matrix 1 7.0 0.000 0.000 2.597 2.598 make_images_data 4110 15.4 0.041 0.047 2.269 2.595 multiply_cannon_sync_h2d 8220 15.4 2.355 2.574 2.355 2.574 calculate_dm_sparse 110 9.5 0.001 0.001 2.493 2.539 grid_collocate_task_list 110 9.6 2.366 2.504 2.366 2.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.472 2.499 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.074 2.121 potential_pw2rs 110 12.3 0.015 0.015 2.080 2.084 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.985 2.017 mp_alltoall_d11v 2046 13.8 1.713 2.005 1.713 2.005 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.999 cp_fm_cholesky_decompose 22 10.9 1.760 1.779 1.760 1.779 acc_transpose_blocks 8220 15.4 0.041 0.042 1.603 1.727 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.999 1.721 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.645 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.507 1.617 dbcsr_complete_redistribute 325 12.2 0.552 0.587 1.475 1.570 mp_allgather_i34 2055 14.4 0.443 1.509 0.443 1.509 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.431 1.446 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.477 1.378 mp_irecv_dv 24056 15.7 0.451 1.335 0.451 1.335 qs_create_task_list 11 7.9 0.000 0.001 1.209 1.306 generate_qs_task_list 11 8.9 0.374 0.442 1.209 1.306 transfer_rs2pw 451 10.6 0.005 0.005 1.154 1.266 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.167 1.210 mp_waitany 9240 13.8 1.081 1.185 1.081 1.185 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.903000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.727273, yerr=11.670405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 1.406464E+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 0.012 0.034 89.011 89.012 qs_mol_dyn_low 1 2.0 0.003 0.003 88.667 88.676 qs_forces 11 3.9 0.003 0.004 88.507 88.507 qs_energies 11 4.9 0.001 0.001 84.322 84.325 scf_env_do_scf 11 5.9 0.000 0.001 73.987 73.988 velocity_verlet 10 3.0 0.002 0.002 56.337 56.355 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.537 45.539 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.380 30.417 init_scf_loop 11 6.9 0.000 0.000 28.379 28.383 qs_scf_new_mos 99 7.5 0.001 0.001 27.354 27.384 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.353 27.383 prepare_preconditioner 11 7.9 0.000 0.000 26.341 26.349 make_preconditioner 11 8.9 0.000 0.000 26.341 26.348 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.487 25.806 ot_scf_mini 99 9.5 0.002 0.002 25.540 25.569 multiply_cannon 2055 13.4 0.338 0.365 23.107 23.852 multiply_cannon_loop 2055 14.4 0.837 0.850 21.313 21.803 cp_fm_upper_to_full 70 14.2 12.805 18.369 12.805 18.369 ot_mini 99 10.5 0.001 0.001 14.569 14.592 rebuild_ks_matrix 110 8.3 0.000 0.001 14.136 14.178 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.135 14.177 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.828 12.866 dbcsr_complete_redistribute 325 12.2 1.020 1.046 7.408 10.595 multiply_cannon_multrec 8220 15.4 4.088 4.353 9.780 9.929 qs_ot_get_derivative 99 11.5 0.001 0.001 9.886 9.915 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.409 9.602 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.838 8.998 mp_waitall_1 84994 16.7 8.021 8.904 8.021 8.904 mp_alltoall_i22 605 13.7 5.451 8.654 5.451 8.654 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.585 7.624 calculate_rho_elec 110 8.6 0.222 0.223 7.584 7.623 sum_up_and_integrate 110 10.3 0.002 0.002 7.032 7.049 integrate_v_rspace 110 11.3 0.004 0.004 7.004 7.021 init_scf_run 11 5.9 0.000 0.001 6.054 6.054 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.054 6.054 make_m2s 4110 13.4 0.043 0.044 5.506 5.988 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.194 5.897 pw_transfer 1331 11.6 0.074 0.075 5.840 5.847 make_images 4110 14.4 0.882 0.922 5.317 5.800 dbcsr_mm_accdrv_process 11614 15.7 3.942 4.148 5.543 5.766 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.723 5.730 qs_ot_get_p 110 10.4 0.002 0.002 5.609 5.644 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.855 5.490 apply_single 110 13.6 0.000 0.000 4.855 5.489 cp_fm_cholesky_invert 11 10.9 5.197 5.201 5.197 5.201 fft_wrap_pw1pw2_140 451 13.1 1.368 1.371 5.030 5.038 ot_diis_step 99 11.5 0.015 0.016 4.660 4.660 density_rs2pw 110 9.6 0.004 0.004 3.982 4.060 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.026 4.034 grid_integrate_task_list 110 12.3 3.742 3.818 3.742 3.818 hybrid_alltoall_any 4261 16.3 0.265 0.570 2.998 3.723 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.225 3.722 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.672 3.672 make_images_data 4110 15.4 0.045 0.048 2.986 3.606 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.505 3.505 fft3d_ps 1111 14.6 1.884 1.889 3.428 3.434 wfi_extrapolate 11 7.9 0.001 0.001 3.337 3.337 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.238 3.242 calculate_dm_sparse 110 9.5 0.001 0.001 3.171 3.200 multiply_cannon_sync_h2d 8220 15.4 3.104 3.128 3.104 3.128 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.912 2.912 cp_fm_diag_elpa_base 48 14.0 2.374 2.568 2.910 2.910 grid_collocate_task_list 110 9.6 2.722 2.776 2.722 2.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.720 2.728 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.658 2.672 calculate_first_density_matrix 1 7.0 0.000 0.000 2.596 2.596 potential_pw2rs 110 12.3 0.021 0.021 2.456 2.462 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.183 2.244 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.116 2.228 mp_alltoall_d11v 2046 13.8 2.113 2.214 2.113 2.214 acc_transpose_blocks 8220 15.4 0.042 0.043 2.033 2.171 cp_fm_cholesky_decompose 22 10.9 1.969 1.980 1.969 1.980 qs_create_task_list 11 7.9 0.000 0.001 1.873 1.919 generate_qs_task_list 11 8.9 0.731 0.783 1.872 1.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.906 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.012000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1263.818182, yerr=54.229844 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.914496E+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 2171993. 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 0.019 0.036 208.982 208.983 qs_mol_dyn_low 1 2.0 0.003 0.003 208.530 208.544 qs_forces 11 3.9 0.005 0.005 208.119 208.119 qs_energies 11 4.9 0.002 0.002 202.507 202.526 scf_env_do_scf 11 5.9 0.001 0.001 185.884 185.888 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 165.196 165.197 dbcsr_multiply_generic 2507 12.6 0.175 0.179 126.339 126.888 qs_scf_new_mos 117 7.6 0.001 0.001 125.436 125.644 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.435 125.643 velocity_verlet 10 3.0 0.001 0.002 125.143 125.147 ot_scf_mini 117 9.6 0.003 0.003 118.697 118.925 multiply_cannon 2507 13.6 0.237 0.247 102.043 103.549 multiply_cannon_loop 2507 14.6 2.411 2.463 99.825 101.290 ot_mini 117 10.6 0.001 0.001 66.364 66.601 multiply_cannon_multrec 60168 15.6 31.933 33.905 41.694 43.453 qs_ot_get_derivative 117 11.6 0.001 0.001 41.558 41.774 rebuild_ks_matrix 128 8.3 0.001 0.001 33.869 34.049 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.869 34.048 mp_waitall_1 267128 16.5 29.084 32.361 29.084 32.361 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.437 30.618 qs_ot_get_p 128 10.4 0.001 0.001 30.015 30.425 multiply_cannon_sync_h2d 60168 15.6 26.515 28.617 26.515 28.617 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.438 25.164 apply_single 128 13.6 0.001 0.001 24.437 25.163 ot_diis_step 117 11.6 0.007 0.008 24.573 24.574 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.291 23.352 init_scf_loop 11 6.9 0.000 0.000 20.611 20.612 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.426 20.427 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.985 20.186 multiply_cannon_metrocomm3 60168 15.6 0.123 0.127 16.258 18.132 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.439 17.474 cp_fm_diag_elpa_base 83 14.4 17.369 17.407 17.435 17.471 prepare_preconditioner 11 7.9 0.000 0.000 15.937 15.981 make_preconditioner 11 8.9 0.000 0.000 15.937 15.981 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.150 15.336 make_m2s 5014 13.6 0.104 0.112 13.957 14.315 sum_up_and_integrate 128 10.3 0.004 0.013 14.148 14.167 make_images 5014 14.6 0.402 0.421 13.778 14.151 integrate_v_rspace 128 11.3 0.003 0.004 14.087 14.106 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.259 13.426 calculate_rho_elec 128 8.7 0.045 0.064 13.258 13.425 init_scf_run 11 5.9 0.000 0.001 12.444 12.444 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.444 12.444 mp_sum_l 7950 12.9 9.074 10.401 9.074 10.401 dbcsr_mm_accdrv_process 124484 16.2 4.694 4.846 9.327 9.857 wfi_extrapolate 11 7.9 0.001 0.001 9.141 9.141 cp_fm_cholesky_invert 11 10.9 9.045 9.052 9.045 9.052 calculate_dm_sparse 128 9.5 0.001 0.001 8.546 8.612 multiply_cannon_metrocomm1 60168 15.6 0.095 0.100 6.196 8.410 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.158 8.278 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.089 8.173 pw_transfer 1547 11.6 0.074 0.094 7.738 7.918 make_images_data 5014 15.6 0.068 0.074 6.840 7.821 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.536 7.717 density_rs2pw 128 9.7 0.006 0.006 6.897 7.558 grid_integrate_task_list 128 12.3 7.064 7.505 7.064 7.505 hybrid_alltoall_any 5200 16.5 0.298 2.281 5.982 7.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.752 6.765 fft_wrap_pw1pw2_140 523 13.2 0.862 0.904 6.551 6.725 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.901 6.015 fft3d_ps 1291 14.7 2.191 2.841 5.534 5.862 mp_alltoall_d11v 2415 14.1 4.340 5.503 4.340 5.503 grid_collocate_task_list 128 9.7 4.834 5.177 4.834 5.177 cp_fm_cholesky_decompose 22 10.9 4.679 4.693 4.679 4.693 potential_pw2rs 128 12.3 0.009 0.010 4.641 4.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.983000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=6.501112 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 840.060928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 6155390. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.040 187.711 187.712 qs_mol_dyn_low 1 2.0 0.003 0.004 187.297 187.309 qs_forces 11 3.9 0.021 0.030 187.215 187.216 qs_energies 11 4.9 0.001 0.002 180.470 180.484 scf_env_do_scf 11 5.9 0.001 0.001 164.195 164.205 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.495 131.498 velocity_verlet 10 3.0 0.001 0.002 117.492 117.494 dbcsr_multiply_generic 2485 12.5 0.184 0.190 96.511 97.813 qs_scf_new_mos 116 7.6 0.001 0.001 93.310 93.864 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.310 93.863 ot_scf_mini 116 9.6 0.004 0.004 88.553 89.179 multiply_cannon 2485 13.5 0.498 0.552 76.595 80.970 multiply_cannon_loop 2485 14.5 1.555 1.623 73.280 76.009 ot_mini 116 10.6 0.001 0.001 49.879 50.428 mp_waitall_1 212858 16.6 23.814 38.489 23.814 38.489 multiply_cannon_multrec 29820 15.5 20.599 24.151 31.148 35.144 rebuild_ks_matrix 127 8.3 0.001 0.001 32.040 32.761 qs_ks_build_kohn_sham_matrix 127 9.3 0.023 0.072 32.040 32.761 init_scf_loop 11 6.9 0.000 0.000 32.610 32.611 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.808 29.464 multiply_cannon_metrocomm3 29820 15.5 0.100 0.106 15.647 28.964 qs_ot_get_derivative 116 11.6 0.001 0.002 28.087 28.700 prepare_preconditioner 11 7.9 0.000 0.000 28.209 28.273 make_preconditioner 11 8.9 0.000 0.000 28.209 28.273 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.914 27.447 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.904 23.095 apply_single 127 13.6 0.001 0.001 21.903 23.094 ot_diis_step 116 11.6 0.014 0.015 21.619 21.621 qs_ot_get_p 127 10.4 0.001 0.002 20.926 21.593 multiply_cannon_sync_h2d 29820 15.5 17.781 20.385 17.781 20.385 cp_fm_cholesky_invert 11 10.9 16.522 16.534 16.522 16.534 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.996 16.029 make_m2s 4970 13.5 0.086 0.091 14.006 15.640 make_images 4970 14.5 1.164 1.369 13.793 15.432 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.799 14.800 sum_up_and_integrate 127 10.3 0.004 0.008 13.872 13.905 integrate_v_rspace 127 11.3 0.003 0.004 13.811 13.849 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.319 13.346 calculate_rho_elec 127 8.7 0.087 0.104 13.319 13.345 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.579 11.610 cp_fm_diag_elpa_base 82 14.4 11.323 11.422 11.573 11.600 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.085 11.564 init_scf_run 11 5.9 0.000 0.001 11.450 11.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.450 11.452 multiply_cannon_metrocomm4 27335 15.5 0.099 0.113 3.901 10.966 dbcsr_mm_accdrv_process 61726 16.2 5.449 6.185 10.000 10.587 mp_irecv_dv 68888 16.3 3.701 10.558 3.701 10.558 make_images_data 4970 15.5 0.067 0.076 8.176 10.132 hybrid_alltoall_any 5155 16.4 0.346 1.509 6.991 9.761 pw_transfer 1535 11.6 0.084 0.098 8.490 8.552 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.269 8.335 wfi_extrapolate 11 7.9 0.001 0.001 8.314 8.314 fft_wrap_pw1pw2_140 519 13.2 0.929 0.950 7.339 7.423 grid_integrate_task_list 127 12.3 7.132 7.389 7.132 7.389 density_rs2pw 127 9.7 0.006 0.006 6.981 7.324 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.306 7.130 cp_fm_cholesky_decompose 22 10.9 6.853 6.928 6.853 6.928 calculate_dm_sparse 127 9.5 0.001 0.001 6.411 6.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.102 6.109 mp_sum_l 7884 12.9 4.123 5.965 4.123 5.965 fft3d_ps 1281 14.7 2.806 2.990 5.839 5.896 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.325 5.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.212 5.393 grid_collocate_task_list 127 9.7 4.966 5.370 4.966 5.370 mp_allgather_i34 2485 14.5 1.846 4.773 1.846 4.773 mp_alltoall_d11v 2401 14.1 4.310 4.709 4.310 4.709 potential_pw2rs 127 12.3 0.015 0.017 4.411 4.427 mp_sum_d 4454 12.1 2.702 4.125 2.702 4.125 dbcsr_complete_redistribute 393 12.7 0.768 0.842 3.068 3.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.712000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.181818, yerr=2.166614 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 971.255808E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.111 194.779 194.781 qs_mol_dyn_low 1 2.0 0.004 0.008 194.233 194.247 qs_forces 11 3.9 0.004 0.006 194.134 194.136 qs_energies 11 4.9 0.002 0.008 187.637 187.647 scf_env_do_scf 11 5.9 0.001 0.004 151.397 151.398 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.542 116.543 velocity_verlet 10 3.0 0.002 0.002 111.653 111.655 dbcsr_multiply_generic 2507 12.6 0.183 0.190 101.976 103.051 multiply_cannon 2507 13.6 0.505 0.523 82.646 86.500 multiply_cannon_loop 2507 14.6 1.142 1.213 79.896 82.759 qs_scf_new_mos 117 7.6 0.001 0.001 79.669 80.113 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.668 80.112 ot_scf_mini 117 9.6 0.004 0.006 75.445 75.873 ot_mini 117 10.6 0.001 0.003 42.364 42.784 mp_waitall_1 170520 16.6 25.926 36.627 25.926 36.627 init_scf_loop 11 6.9 0.001 0.006 34.758 34.758 init_scf_run 11 5.9 0.000 0.001 31.134 31.134 scf_env_initial_rho_setup 11 6.9 0.001 0.003 31.133 31.134 prepare_preconditioner 11 7.9 0.000 0.000 30.770 30.820 make_preconditioner 11 8.9 0.000 0.002 30.770 30.820 rebuild_ks_matrix 128 8.3 0.001 0.001 30.067 30.634 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.025 30.066 30.633 multiply_cannon_metrocomm3 20056 15.6 0.063 0.068 16.419 30.079 make_full_inverse_cholesky 11 9.9 0.000 0.002 28.324 29.833 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.101 27.615 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 25.922 25.930 multiply_cannon_multrec 20056 15.6 12.999 16.176 22.364 25.519 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 25.025 25.179 calculate_first_density_matrix 1 7.0 0.001 0.004 23.448 23.450 qs_ot_get_derivative 117 11.6 0.002 0.003 22.724 23.143 acc_transpose_blocks 20056 15.6 0.103 0.108 21.062 22.390 acc_transpose_blocks_kernels 20056 16.6 0.282 0.287 20.028 21.366 jit_kernel_transpose 5 15.6 19.746 21.086 19.746 21.086 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.709 20.755 apply_single 128 13.6 0.001 0.001 19.708 20.755 ot_diis_step 117 11.6 0.019 0.024 19.531 19.531 qs_ot_get_p 128 10.4 0.002 0.003 18.533 19.034 make_m2s 5014 13.6 0.079 0.083 14.377 15.263 multiply_cannon_sync_h2d 20056 15.6 13.543 15.161 13.543 15.161 make_images 5014 14.6 1.156 1.235 14.143 15.024 qs_ot_p2m_diag 83 11.4 0.266 0.273 14.333 14.343 cp_fm_cholesky_invert 11 10.9 14.129 14.138 14.129 14.138 sum_up_and_integrate 128 10.3 0.002 0.004 13.863 13.892 integrate_v_rspace 128 11.3 0.003 0.004 13.803 13.828 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.587 13.625 calculate_rho_elec 128 8.7 0.130 0.144 13.586 13.624 cp_dbcsr_syevd 83 12.4 0.006 0.008 13.278 13.279 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.106 10.124 cp_fm_diag_elpa_base 83 14.4 9.685 9.835 10.101 10.120 make_images_data 5014 15.6 0.065 0.075 8.500 9.800 multiply_cannon_metrocomm4 17549 15.6 0.065 0.077 3.523 9.540 mp_irecv_dv 50230 16.2 3.393 9.271 3.393 9.271 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.948 9.259 hybrid_alltoall_any 5200 16.5 0.452 2.063 7.480 9.123 dbcsr_mm_accdrv_process 41502 16.2 5.673 5.952 8.825 8.944 pw_transfer 1547 11.6 0.084 0.103 8.769 8.871 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.546 8.653 fft_wrap_pw1pw2_140 523 13.2 1.066 1.100 7.571 7.681 cp_fm_upper_to_full 105 14.8 6.006 7.587 6.006 7.587 grid_integrate_task_list 128 12.3 7.294 7.582 7.294 7.582 wfi_extrapolate 11 7.9 0.001 0.004 7.499 7.499 multiply_cannon_metrocomm1 20056 15.6 0.039 0.040 1.380 7.343 density_rs2pw 128 9.7 0.006 0.006 6.998 7.314 cp_fm_cholesky_decompose 22 10.9 7.129 7.196 7.129 7.196 dbcsr_complete_redistribute 395 12.7 1.162 1.204 4.645 6.493 fft3d_ps 1291 14.7 2.823 3.051 5.853 5.950 calculate_dm_sparse 128 9.5 0.001 0.001 5.832 5.917 grid_collocate_task_list 128 9.7 5.150 5.459 5.150 5.459 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.499 5.345 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.623 5.239 mp_alltoall_d11v 2415 14.1 4.258 4.699 4.258 4.699 mp_sum_l 7950 12.9 3.176 4.530 3.176 4.530 potential_pw2rs 128 12.3 0.020 0.021 4.318 4.331 transfer_fm_to_dbcsr 11 9.9 0.018 0.024 2.427 4.258 mp_allgather_i34 2507 14.6 1.245 4.041 1.245 4.041 mp_alltoall_i22 716 14.1 2.014 4.017 2.014 4.017 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.932 3.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=194.781000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=906.272727, yerr=21.925871 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/18/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 4.353788E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.146241E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.116 193.454 193.455 qs_mol_dyn_low 1 2.0 0.003 0.003 192.937 192.950 qs_forces 11 3.9 0.004 0.004 192.823 192.834 qs_energies 11 4.9 0.002 0.002 185.701 185.711 scf_env_do_scf 11 5.9 0.001 0.001 162.549 162.563 velocity_verlet 10 3.0 0.002 0.003 124.152 124.155 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.165 116.166 dbcsr_multiply_generic 2507 12.6 0.192 0.197 86.629 87.175 qs_scf_new_mos 117 7.6 0.001 0.001 79.985 80.331 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.984 80.330 ot_scf_mini 117 9.6 0.003 0.004 75.418 75.784 multiply_cannon 2507 13.6 0.553 0.591 62.286 64.950 multiply_cannon_loop 2507 14.6 1.820 1.926 58.507 59.911 init_scf_loop 11 6.9 0.000 0.000 46.262 46.263 ot_mini 117 10.6 0.001 0.001 42.747 43.078 prepare_preconditioner 11 7.9 0.000 0.000 42.175 42.202 make_preconditioner 11 8.9 0.000 0.000 42.175 42.202 make_full_inverse_cholesky 11 9.9 0.012 0.024 35.767 40.846 multiply_cannon_multrec 30084 15.6 13.721 19.036 26.638 31.608 rebuild_ks_matrix 128 8.3 0.001 0.001 29.234 29.522 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.233 29.521 mp_waitall_1 147882 16.7 17.297 26.669 17.297 26.669 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.362 26.615 qs_ot_get_derivative 117 11.6 0.001 0.002 23.077 23.434 make_m2s 5014 13.6 0.093 0.097 20.027 21.151 make_images 5014 14.6 1.954 2.267 19.717 20.842 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.044 19.536 apply_single 128 13.6 0.001 0.001 19.044 19.536 ot_diis_step 117 11.6 0.018 0.030 19.534 19.535 qs_ot_get_p 128 10.4 0.002 0.002 18.760 19.144 init_scf_run 11 5.9 0.000 0.001 17.245 17.246 scf_env_initial_rho_setup 11 6.9 0.001 0.001 17.245 17.246 cp_fm_upper_to_full 105 14.8 11.576 17.054 11.576 17.054 cp_fm_cholesky_invert 11 10.9 15.993 16.002 15.993 16.002 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.450 14.823 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.709 14.761 sum_up_and_integrate 128 10.3 0.002 0.003 13.958 13.983 integrate_v_rspace 128 11.3 0.003 0.004 13.898 13.925 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.742 13.778 calculate_rho_elec 128 8.7 0.171 0.186 13.742 13.777 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.313 13.314 dbcsr_mm_accdrv_process 62264 16.2 8.440 9.279 12.485 12.988 dbcsr_complete_redistribute 395 12.7 1.499 1.632 9.119 12.969 make_images_data 5014 15.6 0.066 0.072 10.694 12.410 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 11.940 11.984 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.766 11.613 hybrid_alltoall_any 5200 16.5 0.531 2.221 9.682 11.285 multiply_cannon_sync_h2d 30084 15.6 10.499 11.166 10.499 11.166 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 10.846 10.932 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.386 10.160 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.120 10.141 cp_fm_diag_elpa_base 83 14.4 9.151 9.484 10.113 10.132 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.678 9.936 mp_alltoall_i22 716 14.1 5.664 9.607 5.664 9.607 acc_transpose_blocks 30084 15.6 0.148 0.152 8.832 9.439 calculate_first_density_matrix 1 7.0 0.000 0.000 9.369 9.374 pw_transfer 1547 11.6 0.085 0.101 9.178 9.251 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.953 9.030 fft_wrap_pw1pw2_140 523 13.2 1.248 1.287 7.988 8.087 grid_integrate_task_list 128 12.3 7.532 7.867 7.532 7.867 wfi_extrapolate 11 7.9 0.001 0.001 7.674 7.674 cp_fm_cholesky_decompose 22 10.9 7.465 7.554 7.465 7.554 acc_transpose_blocks_kernels 30084 16.6 0.398 0.404 6.867 7.437 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.814 7.271 density_rs2pw 128 9.7 0.005 0.006 6.799 7.105 jit_kernel_transpose 5 15.6 6.469 7.036 6.469 7.036 mp_irecv_dv 76098 16.2 2.662 6.995 2.662 6.995 calculate_dm_sparse 128 9.5 0.001 0.001 6.254 6.358 fft3d_ps 1291 14.7 2.995 3.047 5.998 6.048 grid_collocate_task_list 128 9.7 5.284 5.591 5.284 5.591 mp_alltoall_d11v 2415 14.1 4.861 5.263 4.861 5.263 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.439 4.541 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 4.419 4.419 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.250 4.317 potential_pw2rs 128 12.3 0.023 0.024 4.247 4.256 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=193.455000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.363636, yerr=14.027718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.524965E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58159. MP_Allreduce 11090 1166. MP_Sync 86 MP_Alltoall 1700 18828189. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 172.657 172.658 qs_mol_dyn_low 1 2.0 0.003 0.003 172.214 172.227 qs_forces 11 3.9 0.004 0.005 172.113 172.118 qs_energies 11 4.9 0.002 0.002 164.730 164.737 scf_env_do_scf 11 5.9 0.001 0.001 145.084 145.092 velocity_verlet 10 3.0 0.002 0.002 110.905 110.909 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 109.189 109.190 dbcsr_multiply_generic 2485 12.5 0.182 0.187 74.846 75.247 qs_scf_new_mos 116 7.6 0.001 0.001 72.224 72.357 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.223 72.356 ot_scf_mini 116 9.6 0.003 0.004 67.838 67.991 multiply_cannon 2485 13.5 0.557 0.575 55.965 59.205 multiply_cannon_loop 2485 14.5 0.807 0.838 52.979 53.827 ot_mini 116 10.6 0.001 0.001 37.947 38.087 init_scf_loop 11 6.9 0.000 0.000 35.747 35.749 prepare_preconditioner 11 7.9 0.000 0.000 31.806 31.838 make_preconditioner 11 8.9 0.000 0.000 31.806 31.838 mp_waitall_1 124680 16.7 25.326 31.817 25.326 31.817 make_full_inverse_cholesky 11 9.9 0.012 0.028 29.765 30.030 rebuild_ks_matrix 127 8.3 0.001 0.001 29.119 29.244 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.118 29.244 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.553 26.666 multiply_cannon_multrec 9940 15.5 10.281 14.933 18.074 22.768 qs_ot_get_derivative 116 11.6 0.002 0.002 20.866 21.007 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.990 20.372 cp_fm_cholesky_invert 11 10.9 18.256 18.264 18.256 18.264 qs_ot_get_p 127 10.4 0.026 0.036 17.258 17.429 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.126 17.295 apply_single 127 13.6 0.001 0.001 17.126 17.294 ot_diis_step 116 11.6 0.020 0.021 17.010 17.010 make_m2s 4970 13.5 0.063 0.069 14.989 15.846 make_images 4970 14.5 2.155 2.560 14.685 15.549 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.200 14.229 calculate_rho_elec 127 8.7 0.252 0.263 14.199 14.229 sum_up_and_integrate 127 10.3 0.002 0.002 14.108 14.156 integrate_v_rspace 127 11.3 0.004 0.004 14.048 14.099 qs_ot_p2m_diag 82 11.4 0.490 0.495 13.397 13.413 init_scf_run 11 5.9 0.000 0.001 12.871 12.872 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.871 12.871 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.219 12.220 multiply_cannon_sync_h2d 9940 15.5 10.603 11.232 10.603 11.232 make_images_data 4970 15.5 0.055 0.067 8.693 9.918 hybrid_alltoall_any 5155 16.4 0.837 3.763 8.478 9.898 pw_transfer 1535 11.6 0.084 0.094 9.852 9.885 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.631 9.671 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.306 9.317 cp_fm_diag_elpa_base 82 14.4 9.068 9.147 9.302 9.314 fft_wrap_pw1pw2_140 519 13.2 1.657 1.699 8.525 8.569 dbcsr_mm_accdrv_process 20590 16.1 3.536 4.662 7.444 8.199 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.031 8.129 cp_fm_cholesky_decompose 22 10.9 8.017 8.112 8.017 8.112 grid_integrate_task_list 127 12.3 7.728 8.066 7.728 8.066 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.482 7.491 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.401 7.488 wfi_extrapolate 11 7.9 0.001 0.001 7.381 7.381 density_rs2pw 127 9.7 0.005 0.005 6.930 7.159 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.413 6.470 calculate_dm_sparse 127 9.5 0.001 0.001 6.135 6.228 fft3d_ps 1281 14.7 3.118 3.228 6.113 6.138 grid_collocate_task_list 127 9.7 5.524 5.844 5.524 5.844 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.853 5.603 mp_irecv_dv 28618 15.9 1.812 5.504 1.812 5.504 dbcsr_complete_redistribute 393 12.7 2.076 2.138 5.099 5.427 calculate_first_density_matrix 1 7.0 0.000 0.000 5.253 5.255 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.217 5.220 mp_alltoall_d11v 2401 14.1 4.519 4.978 4.519 4.978 mp_allgather_i34 2485 14.5 1.186 4.737 1.186 4.737 potential_pw2rs 127 12.3 0.026 0.026 4.347 4.351 acc_transpose_blocks 9940 15.5 0.054 0.055 3.932 3.965 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.578 3.866 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.651 3.731 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.617 3.655 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.287 3.585 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.391 3.501 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.463 3.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.658000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1444.454545, yerr=13.950325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 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 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.150615E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58586. MP_Allreduce 11179 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.063 298.987 298.989 qs_mol_dyn_low 1 2.0 0.003 0.003 298.407 298.424 qs_forces 11 3.9 0.005 0.005 298.312 298.314 qs_energies 11 4.9 0.002 0.003 289.373 289.378 scf_env_do_scf 11 5.9 0.001 0.001 263.162 263.170 velocity_verlet 10 3.0 0.002 0.002 214.047 214.055 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 136.664 136.666 init_scf_loop 11 6.9 0.000 0.000 126.212 126.214 prepare_preconditioner 11 7.9 0.000 0.000 121.362 121.388 make_preconditioner 11 8.9 0.000 0.000 121.362 121.388 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.780 118.453 qs_scf_new_mos 118 7.6 0.001 0.001 91.242 91.317 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.241 91.316 dbcsr_multiply_generic 2535 12.6 0.218 0.227 86.579 87.066 ot_scf_mini 118 9.6 0.004 0.004 86.294 86.342 cp_fm_upper_to_full 106 14.8 54.041 77.572 54.041 77.572 multiply_cannon 2535 13.6 0.693 0.725 63.713 64.705 multiply_cannon_loop 2535 14.6 1.060 1.091 59.677 61.083 ot_mini 118 10.6 0.001 0.001 44.862 44.914 dbcsr_complete_redistribute 397 12.7 3.966 3.997 30.488 44.444 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.995 40.931 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.539 38.359 mp_alltoall_i22 720 14.1 22.293 36.080 22.293 36.080 rebuild_ks_matrix 129 8.3 0.001 0.001 35.090 35.131 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 35.089 35.130 cp_fm_cholesky_invert 11 10.9 33.682 33.688 33.682 33.688 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.586 32.620 mp_waitall_1 104820 16.8 28.139 32.422 28.139 32.422 qs_ot_get_p 129 10.4 0.002 0.002 26.226 26.281 qs_ot_get_derivative 118 11.6 0.002 0.002 25.047 25.093 qs_ot_p2m_diag 84 11.4 0.890 0.897 22.136 22.168 multiply_cannon_metrocomm3 10140 15.6 0.025 0.026 19.293 20.698 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.256 20.303 ot_diis_step 118 11.6 0.023 0.025 19.786 19.786 make_m2s 5070 13.6 0.075 0.079 18.254 19.721 make_images 5070 14.6 3.088 3.282 17.769 19.236 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.067 19.177 apply_single 129 13.6 0.001 0.001 19.067 19.177 multiply_cannon_multrec 10140 15.6 10.505 12.483 18.536 18.611 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.567 17.579 calculate_rho_elec 129 8.7 0.482 0.484 17.566 17.578 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.011 17.012 cp_fm_diag_elpa_base 84 14.4 12.471 14.078 17.007 17.007 sum_up_and_integrate 129 10.3 0.002 0.002 16.676 16.764 integrate_v_rspace 129 11.3 0.005 0.005 16.612 16.701 init_scf_run 11 5.9 0.000 0.001 15.910 15.912 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.910 15.911 multiply_cannon_sync_h2d 10140 15.6 14.421 14.464 14.421 14.464 pw_transfer 1559 11.6 0.096 0.097 13.382 13.389 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 13.142 13.149 hybrid_alltoall_any 5257 16.5 1.323 3.074 10.267 12.370 make_images_data 5070 15.6 0.066 0.071 10.120 12.268 fft_wrap_pw1pw2_140 527 13.2 3.242 3.282 11.679 11.690 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.031 10.071 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.809 9.847 dbcsr_mm_accdrv_process 20958 16.1 4.438 6.440 7.777 9.771 cp_fm_cholesky_decompose 22 10.9 9.050 9.068 9.050 9.068 wfi_extrapolate 11 7.9 0.001 0.001 8.977 8.978 grid_integrate_task_list 129 12.3 8.620 8.813 8.620 8.813 density_rs2pw 129 9.7 0.005 0.005 8.508 8.557 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.159 8.160 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.096 8.121 fft3d_ps 1301 14.7 4.005 4.015 7.566 7.600 mp_alltoall_d11v 2429 14.1 6.907 7.019 6.907 7.019 calculate_dm_sparse 129 9.5 0.001 0.001 6.890 6.964 calculate_first_density_matrix 1 7.0 0.000 0.000 6.662 6.662 grid_collocate_task_list 129 9.7 6.469 6.546 6.469 6.546 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.133 6.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.989000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2845.363636, yerr=154.498175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 1.260667E+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 0.012 0.102 85.517 85.549 qs_energies 1 2.0 0.000 0.000 85.027 85.053 ls_scf 1 3.0 0.000 0.000 84.133 84.159 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.197 73.389 multiply_cannon 111 7.7 0.017 0.020 56.450 57.546 multiply_cannon_loop 111 8.7 0.231 0.246 53.045 54.490 ls_scf_main 1 4.0 0.000 0.000 52.256 52.256 density_matrix_trs4 2 5.0 0.002 0.003 46.726 46.811 ls_scf_init_scf 1 4.0 0.000 0.000 28.826 28.827 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.706 27.755 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.557 25.575 mp_waitall_1 11031 10.9 22.357 24.841 22.357 24.841 multiply_cannon_multrec 2664 9.7 8.148 8.955 15.667 17.491 multiply_cannon_sync_h2d 2664 9.7 13.649 15.433 13.649 15.433 make_m2s 222 7.7 0.009 0.012 13.090 13.623 make_images 222 8.7 0.100 0.110 13.068 13.603 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 9.616 11.897 make_images_data 222 9.7 0.005 0.006 7.660 8.196 dbcsr_mm_accdrv_process 4760 10.4 0.597 0.722 7.135 8.133 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.422 8.001 hybrid_alltoall_any 227 10.6 0.219 1.858 6.544 7.671 dbcsr_mm_accdrv_process_sort 4760 11.4 6.335 7.272 6.335 7.272 calculate_norms 4752 9.8 5.533 6.242 5.533 6.242 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.049 5.200 mp_sum_l 887 5.1 3.147 4.738 3.147 4.738 make_images_sizes 222 9.7 0.000 0.000 0.784 3.487 mp_alltoall_i44 222 10.7 0.784 3.486 0.784 3.486 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.321 3.476 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.065 3.213 arnoldi_extremal 4 6.8 0.000 0.000 3.189 3.208 arnoldi_normal_ev 4 7.8 0.001 0.003 3.189 3.208 mp_irecv_dv 6231 10.9 2.049 3.186 2.049 3.186 build_subspace 16 8.4 0.009 0.012 3.095 3.097 ls_scf_post 1 4.0 0.000 0.000 3.052 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.908 dbcsr_special_finalize 555 9.7 0.005 0.006 2.284 2.702 dbcsr_merge_single_wm 555 10.7 0.461 0.588 2.276 2.693 make_images_pack 222 9.7 2.213 2.670 2.215 2.672 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.308 2.549 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.455 2.064 2.457 dbcsr_sort_data 658 11.4 2.073 2.440 2.073 2.440 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.308 2.414 buffer_matrices_ensure_size 222 8.7 1.770 2.104 1.770 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.766 1.767 rebuild_ks_matrix 3 7.3 0.000 0.000 1.757 1.758 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.756 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.549000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1129.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 2.153992E+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 0.021 0.040 90.082 90.083 qs_energies 1 2.0 0.000 0.000 89.654 89.660 ls_scf 1 3.0 0.000 0.000 88.323 88.330 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.507 74.815 multiply_cannon 111 7.7 0.026 0.037 53.140 57.297 ls_scf_main 1 4.0 0.000 0.000 54.386 54.392 multiply_cannon_loop 111 8.7 0.136 0.147 50.341 53.186 density_matrix_trs4 2 5.0 0.002 0.003 48.757 48.948 ls_scf_init_scf 1 4.0 0.000 0.000 30.274 30.276 mp_waitall_1 9105 10.9 20.717 29.279 20.717 29.279 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.078 29.175 multiply_cannon_multrec 1332 9.7 13.210 17.233 22.586 27.768 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.723 26.731 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.483 20.277 make_m2s 222 7.7 0.007 0.007 15.008 15.511 make_images 222 8.7 1.378 1.700 14.978 15.482 dbcsr_mm_accdrv_process 4041 10.4 0.325 0.518 8.974 10.577 dbcsr_mm_accdrv_process_sort 4041 11.4 8.493 10.052 8.493 10.052 make_images_data 222 9.7 0.004 0.005 8.692 9.506 hybrid_alltoall_any 227 10.6 0.543 2.559 8.109 9.273 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.191 7.935 mp_irecv_dv 3311 11.0 3.170 7.879 3.170 7.879 mp_sum_l 887 5.1 4.863 7.624 4.863 7.624 calculate_norms 2376 9.8 6.074 6.785 6.074 6.785 multiply_cannon_sync_h2d 1332 9.7 4.889 6.737 4.889 6.737 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.697 6.095 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.875 5.105 arnoldi_extremal 4 6.8 0.000 0.000 4.649 4.678 arnoldi_normal_ev 4 7.8 0.001 0.005 4.649 4.678 build_subspace 16 8.4 0.014 0.021 4.399 4.401 ls_scf_post 1 4.0 0.000 0.000 3.663 3.669 ls_scf_store_result 1 5.0 0.000 0.000 3.390 3.487 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.169 3.401 dbcsr_matrix_vector_mult_local 304 10.0 2.778 3.265 2.780 3.267 mp_allgather_i34 111 8.7 0.782 3.021 0.782 3.021 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.595 2.694 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.196 2.668 dbcsr_data_new 4174 10.1 2.117 2.401 2.117 2.401 make_images_pack 222 9.7 1.820 2.135 1.822 2.137 dbcsr_sort_data 436 11.2 1.858 2.109 1.858 2.109 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.853 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.841 1.843 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.841 1.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.083000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 2.836546E+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 0.023 0.045 92.031 92.032 qs_energies 1 2.0 0.000 0.000 91.484 91.488 ls_scf 1 3.0 0.000 0.000 90.061 90.066 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.748 75.140 ls_scf_main 1 4.0 0.000 0.000 55.731 55.734 multiply_cannon 111 7.7 0.043 0.111 52.106 55.715 multiply_cannon_loop 111 8.7 0.117 0.129 49.353 53.446 density_matrix_trs4 2 5.0 0.002 0.003 49.777 49.955 mp_waitall_1 7281 11.0 22.865 33.337 22.865 33.337 ls_scf_init_scf 1 4.0 0.000 0.000 30.612 30.615 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.422 29.498 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.060 27.068 multiply_cannon_multrec 888 9.7 12.787 15.381 21.491 24.708 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.556 23.072 make_m2s 222 7.7 0.006 0.007 15.944 16.694 make_images 222 8.7 1.589 1.860 15.906 16.653 make_images_data 222 9.7 0.004 0.005 9.269 10.284 hybrid_alltoall_any 227 10.6 0.643 2.948 8.899 9.960 dbcsr_mm_accdrv_process 3754 10.4 0.307 0.493 8.224 9.489 dbcsr_mm_accdrv_process_sort 3754 11.4 7.787 8.996 7.787 8.996 mp_sum_l 887 5.1 5.009 8.090 5.009 8.090 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.154 mp_irecv_dv 2335 11.1 2.456 7.096 2.456 7.096 multiply_cannon_sync_h2d 888 9.7 6.040 7.058 6.040 7.058 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.685 6.688 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.861 6.247 arnoldi_extremal 4 6.8 0.000 0.000 5.094 5.106 arnoldi_normal_ev 4 7.8 0.001 0.006 5.094 5.106 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.763 4.987 build_subspace 16 8.4 0.014 0.020 4.790 4.796 calculate_norms 1584 9.8 4.359 4.763 4.359 4.763 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.467 3.784 ls_scf_post 1 4.0 0.000 0.000 3.718 3.723 mp_allgather_i34 111 8.7 0.860 3.632 0.860 3.632 dbcsr_matrix_vector_mult_local 304 10.0 3.055 3.625 3.057 3.627 ls_scf_store_result 1 5.0 0.000 0.000 3.447 3.539 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.890 3.017 dbcsr_data_new 4116 9.9 2.103 2.435 2.103 2.435 dbcsr_sort_data 325 11.1 1.864 2.133 1.864 2.133 make_images_sizes 222 9.7 0.000 0.000 0.831 1.960 mp_alltoall_i44 222 10.7 0.831 1.960 0.831 1.960 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.892 1.894 make_images_pack 222 9.7 1.626 1.885 1.629 1.888 rebuild_ks_matrix 3 7.3 0.000 0.000 1.874 1.876 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.874 1.876 dbcsr_finalize 304 7.8 0.026 0.032 1.596 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.032000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2213.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 3.389673E+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 0.026 0.046 96.861 96.862 qs_energies 1 2.0 0.000 0.000 96.295 96.300 ls_scf 1 3.0 0.000 0.000 94.638 94.644 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.273 78.527 ls_scf_main 1 4.0 0.000 0.000 58.611 58.611 multiply_cannon 111 7.7 0.044 0.099 51.589 56.104 density_matrix_trs4 2 5.0 0.002 0.003 52.464 52.598 multiply_cannon_loop 111 8.7 0.153 0.166 46.527 48.840 ls_scf_init_scf 1 4.0 0.000 0.000 32.797 32.800 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.578 31.651 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 29.103 29.116 mp_waitall_1 6369 11.0 22.242 28.566 22.242 28.566 multiply_cannon_multrec 1332 9.7 14.209 17.594 22.226 25.010 make_m2s 222 7.7 0.008 0.008 21.059 22.510 make_images 222 8.7 3.138 3.610 21.008 22.461 multiply_cannon_metrocomm3 1332 9.7 0.004 0.004 8.766 16.641 make_images_data 222 9.7 0.004 0.005 11.736 13.293 hybrid_alltoall_any 227 10.6 0.800 3.771 11.074 12.879 dbcsr_mm_accdrv_process 3641 10.4 0.286 0.478 7.646 9.190 dbcsr_mm_accdrv_process_sort 3641 11.4 7.207 8.694 7.207 8.694 mp_sum_l 887 5.1 4.015 7.610 4.015 7.610 multiply_cannon_sync_h2d 1332 9.7 5.483 6.070 5.483 6.070 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.090 6.012 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.066 5.945 mp_irecv_dv 3229 10.9 2.064 5.934 2.064 5.934 arnoldi_extremal 4 6.8 0.000 0.000 5.308 5.319 arnoldi_normal_ev 4 7.8 0.001 0.004 5.308 5.319 build_subspace 16 8.4 0.014 0.021 4.969 4.977 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.574 4.777 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.550 4.765 mp_allgather_i34 111 8.7 2.169 4.544 2.169 4.544 calculate_norms 2376 9.8 4.183 4.507 4.183 4.507 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.627 3.950 dbcsr_matrix_vector_mult_local 304 10.0 3.236 3.758 3.239 3.760 dbcsr_sort_data 658 11.4 3.106 3.320 3.106 3.320 ls_scf_post 1 4.0 0.000 0.000 3.230 3.235 dbcsr_special_finalize 555 9.7 0.006 0.007 2.851 3.164 dbcsr_merge_single_wm 555 10.7 0.545 0.658 2.843 3.156 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.030 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.031 dbcsr_data_release 10477 10.7 1.600 2.388 1.600 2.388 dbcsr_finalize 304 7.8 0.050 0.062 1.822 1.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.862000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 4.729573E+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 0.034 0.052 92.999 93.000 qs_energies 1 2.0 0.000 0.000 92.275 92.282 ls_scf 1 3.0 0.000 0.000 90.322 90.328 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.884 72.044 ls_scf_main 1 4.0 0.000 0.000 57.178 57.178 multiply_cannon 111 7.7 0.090 0.189 53.256 56.180 multiply_cannon_loop 111 8.7 0.089 0.112 50.631 52.186 density_matrix_trs4 2 5.0 0.002 0.003 50.199 50.275 ls_scf_init_scf 1 4.0 0.000 0.000 29.724 29.726 mp_waitall_1 5436 11.0 24.895 29.229 24.895 29.229 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.465 28.501 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.357 26.368 multiply_cannon_multrec 444 9.7 13.756 16.203 20.916 22.052 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.006 16.110 make_m2s 222 7.7 0.005 0.005 13.730 14.708 make_images 222 8.7 2.046 2.486 13.662 14.640 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.231 14.236 hybrid_alltoall_any 227 10.6 0.805 3.849 8.259 9.987 make_images_data 222 9.7 0.003 0.004 8.478 9.848 multiply_cannon_sync_h2d 444 9.7 6.737 8.463 6.737 8.463 dbcsr_mm_accdrv_process 3003 10.4 0.364 0.403 6.856 7.968 dbcsr_mm_accdrv_process_sort 3003 11.4 6.492 7.569 6.492 7.569 arnoldi_extremal 4 6.8 0.000 0.000 5.825 5.836 arnoldi_normal_ev 4 7.8 0.002 0.005 5.825 5.836 build_subspace 16 8.4 0.015 0.020 5.434 5.447 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.415 4.640 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.197 4.379 mp_sum_l 887 5.1 2.893 4.282 2.893 4.282 dbcsr_matrix_vector_mult_local 304 10.0 3.735 4.181 3.737 4.183 mp_allgather_i34 111 8.7 1.203 3.811 1.203 3.811 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.549 3.787 mp_irecv_dv 1241 11.2 1.536 3.751 1.536 3.751 calculate_norms 792 9.8 3.617 3.717 3.617 3.717 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.536 3.583 ls_scf_post 1 4.0 0.000 0.000 3.420 3.427 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.075 3.381 make_images_sizes 222 9.7 0.000 0.000 0.850 3.322 mp_alltoall_i44 222 10.7 0.850 3.322 0.850 3.322 ls_scf_store_result 1 5.0 0.000 0.000 3.207 3.245 dbcsr_finalize 304 7.8 0.062 0.077 2.207 2.313 dbcsr_data_new 4608 9.7 1.766 2.193 1.766 2.193 dbcsr_merge_all 275 8.9 0.483 0.530 2.064 2.149 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.031 2.031 rebuild_ks_matrix 3 7.3 0.000 0.000 1.998 1.999 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.998 1.999 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.937 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.000000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 8.889647E+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 0.071 0.086 105.782 105.783 qs_energies 1 2.0 0.000 0.000 104.370 104.379 ls_scf 1 3.0 0.000 0.000 101.325 101.334 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.844 75.010 ls_scf_main 1 4.0 0.000 0.000 63.809 63.810 density_matrix_trs4 2 5.0 0.002 0.003 54.745 54.814 multiply_cannon 111 7.7 0.201 0.321 48.544 50.552 multiply_cannon_loop 111 8.7 0.099 0.100 45.534 46.174 ls_scf_init_scf 1 4.0 0.000 0.000 33.723 33.724 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.173 32.191 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.477 29.492 mp_waitall_1 4527 11.1 21.707 25.747 21.707 25.747 make_m2s 222 7.7 0.005 0.006 22.630 23.725 make_images 222 8.7 3.570 3.869 22.520 23.614 multiply_cannon_multrec 444 9.7 17.872 18.549 22.565 23.254 hybrid_alltoall_any 227 10.6 1.653 3.654 12.736 15.675 make_images_data 222 9.7 0.004 0.004 13.012 15.174 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.114 10.958 multiply_cannon_sync_h2d 444 9.7 8.789 8.847 8.789 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.459 7.475 arnoldi_normal_ev 4 7.8 0.003 0.009 7.459 7.475 build_subspace 16 8.4 0.026 0.036 6.902 6.915 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.548 5.688 dbcsr_matrix_vector_mult_local 304 10.0 5.123 5.414 5.125 5.416 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.198 5.300 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.857 5.109 dbcsr_mm_accdrv_process 1814 10.4 0.221 0.357 4.514 4.649 dbcsr_mm_accdrv_process_sort 1814 11.4 4.177 4.312 4.177 4.312 ls_scf_post 1 4.0 0.000 0.000 3.793 3.802 make_images_sizes 222 9.7 0.000 0.000 1.473 3.645 mp_alltoall_i44 222 10.7 1.473 3.644 1.473 3.644 mp_allgather_i34 111 8.7 1.109 3.637 1.109 3.637 ls_scf_store_result 1 5.0 0.000 0.000 3.511 3.548 calculate_norms 792 9.8 3.235 3.277 3.235 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.072 3.165 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 3.014 3.014 dbcsr_merge_all 275 8.9 0.885 0.919 2.856 2.942 dbcsr_complete_redistribute 5 7.6 1.427 1.473 2.734 2.860 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.555 dbcsr_sort_data 325 11.1 2.433 2.494 2.433 2.494 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.465 2.466 rebuild_ks_matrix 3 7.3 0.000 0.000 2.397 2.399 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.397 2.399 dbcsr_new_transposed 4 7.5 0.241 0.252 2.306 2.315 dbcsr_frobenius_norm 74 6.6 2.057 2.132 2.187 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.202 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.202 dbcsr_data_release 12724 10.6 1.980 2.202 1.980 2.202 dbcsr_data_new 6591 9.6 1.799 2.186 1.799 2.186 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.783000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7052.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad61d0fb47758fedc409ed1d609b4bb070cc6269_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 594.227200E+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 1377646758. 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 0.035 0.102 360.329 360.354 qs_mol_dyn_low 1 2.0 0.004 0.045 359.272 359.344 qs_forces 5 3.8 0.006 0.087 359.091 359.116 qs_energies 5 4.8 0.003 0.048 355.982 356.015 dbcsr_multiply_generic 1445 12.2 0.126 0.134 257.104 257.803 multiply_cannon 1445 13.2 0.274 0.285 237.744 240.294 multiply_cannon_loop 1445 14.2 2.845 3.004 235.877 237.468 scf_env_do_scf 5 5.8 0.001 0.052 217.675 217.692 scf_env_do_scf_inner_loop 105 6.6 0.005 0.083 189.318 189.349 qs_scf_new_mos 105 7.6 0.000 0.001 147.529 147.707 qs_scf_loop_do_ot 105 8.6 0.001 0.001 147.528 147.707 ot_scf_mini 105 9.6 0.003 0.013 137.593 137.774 init_scf_run 5 5.8 0.000 0.001 135.525 135.535 scf_env_initial_rho_setup 5 6.8 0.002 0.013 135.524 135.535 acc_transpose_blocks 69360 15.2 0.360 0.377 130.904 131.256 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.011 130.634 130.654 calculate_first_density_matrix 1 7.0 0.000 0.009 130.493 130.509 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 130.213 130.283 acc_transpose_blocks_kernels 69360 16.2 0.846 0.890 128.153 128.411 jit_kernel_transpose 5 15.0 127.308 127.541 127.308 127.541 velocity_verlet 4 3.0 0.006 0.038 108.846 108.854 ot_mini 105 10.6 0.001 0.019 60.419 60.528 qs_ot_get_p 112 10.4 0.002 0.030 47.069 47.376 multiply_cannon_multrec 69360 15.2 29.760 34.245 39.769 44.678 mp_waitall_1 488190 16.1 34.927 42.640 34.927 42.640 qs_ot_get_derivative 55 11.6 0.001 0.014 38.677 38.835 qs_ot_p2m_diag 40 11.0 0.020 0.030 36.136 36.221 multiply_cannon_sync_h2d 69360 15.2 29.136 33.554 29.136 33.554 multiply_cannon_metrocomm3 69360 15.2 0.206 0.216 25.613 33.133 cp_dbcsr_syevd 40 12.0 0.002 0.004 32.811 32.814 rebuild_ks_matrix 110 8.4 0.000 0.000 29.729 29.918 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.066 29.729 29.918 init_scf_loop 7 6.6 0.001 0.021 28.318 28.323 cp_fm_syevd 40 13.0 0.000 0.001 27.536 27.669 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.290 27.461 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.026 23.272 apply_single 62 13.6 0.000 0.000 23.026 23.272 prepare_preconditioner 7 7.6 0.000 0.000 23.096 23.134 make_preconditioner 7 8.6 0.001 0.058 23.096 23.134 cp_fm_redistribute_end 40 14.0 11.370 22.682 11.376 22.684 cp_fm_syevd_base 40 14.0 11.301 22.626 11.301 22.626 ot_new_cg_direction 55 11.6 0.001 0.004 21.049 21.051 qs_rho_update_rho_low 110 7.6 0.001 0.010 18.267 18.557 calculate_rho_elec 110 8.6 0.030 0.065 18.266 18.556 make_full_inverse_cholesky 7 9.6 0.001 0.031 15.579 15.653 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.900 15.115 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.124 14.202 density_rs2pw 110 9.6 0.005 0.006 12.569 13.047 mp_sum_l 4764 12.2 12.192 12.847 12.192 12.847 pw_transfer 1645 12.4 0.079 0.103 12.013 12.239 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.875 12.101 calculate_dm_sparse 110 9.5 0.000 0.001 11.597 11.798 cp_fm_cholesky_invert 7 10.6 11.009 11.027 11.009 11.027 dbcsr_mm_accdrv_process 154766 15.8 6.206 6.393 9.876 10.723 qs_vxc_create 110 10.4 0.003 0.032 10.564 10.601 fft_wrap_pw1pw2_240 915 15.0 0.834 0.913 10.337 10.583 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.209 10.275 check_diag 80 13.5 8.684 8.936 9.711 9.843 transfer_rs2pw 445 10.6 0.007 0.008 9.149 9.702 sum_up_and_integrate 60 10.3 0.001 0.012 9.131 9.153 integrate_v_rspace 60 11.3 0.003 0.070 9.113 9.136 fft3d_pb 915 16.0 2.390 2.583 8.523 8.756 xc_rho_set_and_dset_create 110 12.4 0.077 0.120 7.388 7.657 multiply_cannon_metrocomm1 69360 15.2 0.095 0.101 4.506 7.539 make_m2s 2890 13.2 0.078 0.087 6.689 7.261 make_full_single_inverse 7 9.6 0.001 0.002 7.219 7.257 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=360.354000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=564.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ad61d0fb47758fedc409ed1d609b4bb070cc6269 Summary: empty Status: OK