=== 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: e39d2370f112907676398aed806e11d93bfbe80f ################# 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, SPGLIB 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # 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: 28.03.2024 # # \ 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; \ rm -rf build; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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 ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(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_f90cstub.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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/01 job id: 53281363 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/02 job id: 53281373 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/03 job id: 53281374 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/04 job id: 53281375 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/05 job id: 53281377 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/06 job id: 53281378 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/07 job id: 53281379 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/08 job id: 53281380 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/09 job id: 53281381 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/10 job id: 53281382 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/11 job id: 53281383 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/12 job id: 53281384 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/13 job id: 53281386 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/14 job id: 53281387 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/15 job id: 53281388 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/16 job id: 53281389 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/17 job id: 53281390 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/18 job id: 53281391 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/19 job id: 53281392 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/20 job id: 53281393 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/21 job id: 53281394 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/22 job id: 53281396 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/23 job id: 53281397 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/24 job id: 53281398 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/25 job id: 53281399 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/26 job id: 53281401 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/27 job id: 53281403 --- 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/e39d2370f112907676398aed806e11d93bfbe80f_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 0.034 131.246 131.246 farming_run 1 2.0 130.605 130.607 131.215 131.219 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.536971E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.026 111.232 111.232 qs_energies 1 2.0 0.001 0.001 111.046 111.047 mp2_main 1 3.0 0.000 0.000 109.241 109.242 mp2_gpw_main 1 4.0 0.034 0.079 108.341 108.341 mp2_ri_gpw_compute_in 1 5.0 0.174 0.175 89.512 89.887 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.242 57.623 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.167 43.673 48.827 get_2c_integrals 1 6.0 0.008 0.009 31.724 32.096 integrate_v_rspace 273 8.0 0.443 0.457 25.542 30.080 fft_wrap_pw1pw2 5465 10.4 0.067 0.071 27.991 28.739 grid_integrate_task_list 273 9.0 20.767 26.264 20.767 26.264 fft_wrap_pw1pw2_100 2178 11.4 1.322 1.455 25.438 26.177 compute_2c_integrals 1 7.0 0.002 0.002 22.440 22.442 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 21.097 22.137 mp2_eri_2c_integrate_gpw 1 9.0 2.361 2.411 21.095 22.136 rpa_ri_compute_en 1 5.0 0.019 0.027 18.701 18.904 fft3d_s 5443 12.4 16.840 17.548 16.862 17.569 ao_to_mo_and_store_B_mult_1 272 7.0 10.777 15.333 10.777 15.333 calculate_wavefunction 272 8.0 5.284 5.432 12.782 13.739 calc_potential_gpw 544 9.5 0.004 0.005 11.302 12.475 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 10.129 11.232 rpa_num_int 1 6.0 0.001 0.010 10.458 10.458 rpa_num_int_RPA_matrix_operati 8 7.0 0.010 0.012 10.450 10.454 cp_fm_cholesky_decompose 12 8.2 9.437 9.808 9.437 9.808 potential_pw2rs 545 10.0 0.106 0.108 8.403 9.590 calc_mat_Q 8 8.0 0.000 0.000 9.433 9.510 contract_S_to_Q 8 9.0 0.000 0.000 8.854 8.932 cholesky_decomp 1 7.0 0.000 0.000 8.474 8.845 parallel_gemm_fm 14 9.1 0.000 0.000 8.469 8.567 parallel_gemm_fm_cosma 14 10.1 8.469 8.567 8.469 8.567 collocate_single_gaussian 272 10.0 0.040 0.042 7.791 8.267 create_integ_mat 1 6.0 0.007 0.009 7.773 7.773 array2fm 1 7.0 0.000 0.000 6.680 7.222 pw_poisson_solve 545 10.5 0.010 0.012 5.013 6.226 pw_scatter_s 2720 12.7 4.496 4.893 4.496 4.893 pw_gather_s 2722 12.2 3.811 4.524 3.811 4.524 pw_poisson_set 548 11.5 0.019 0.020 3.618 4.480 pw_copy 4911 11.6 2.813 3.507 2.813 3.507 array2fm_buffer_send 1 8.0 2.978 3.143 2.978 3.143 pw_derive 1641 12.5 2.145 2.694 2.145 2.694 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.081 1.245 2.027 2.232 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=108.340723, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 0.074 0.406 428.583 428.586 farming_run 1 2.0 427.174 427.495 428.475 428.532 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.277264E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171443. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 222.707 222.708 qs_energies 1 2.0 0.045 0.120 222.478 222.484 scf_env_do_scf 1 3.0 0.000 0.000 114.237 114.237 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.348 113.355 rebuild_ks_matrix 4 6.0 0.000 0.000 113.347 113.354 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.347 113.354 hfx_ks_matrix 4 8.0 0.001 0.001 113.000 113.003 integrate_four_center 4 9.0 0.153 0.458 112.999 113.002 mp2_main 1 3.0 0.000 0.000 107.905 107.956 mp2_gpw_main 1 4.0 0.067 0.140 106.988 107.001 integrate_four_center_main 4 10.0 0.139 0.543 101.841 104.798 integrate_four_center_bin 269 11.0 101.702 104.444 101.702 104.444 init_scf_loop 1 4.0 0.000 0.000 96.549 96.549 mp2_ri_gpw_compute_in 1 5.0 0.082 0.125 80.015 81.216 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.799 57.981 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.162 44.625 49.411 integrate_v_rspace 95 8.0 0.397 0.564 28.294 32.989 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.500 31.115 ao_to_mo_and_store_B_mult_1 91 7.0 10.492 29.132 10.492 29.132 mp2_ri_gpw_compute_en 1 5.0 0.055 0.067 26.803 29.025 fft_wrap_pw1pw2_100 730 11.4 0.629 0.755 28.213 28.752 grid_integrate_task_list 95 9.0 23.595 28.486 23.595 28.486 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.889 24.965 24.975 get_2c_integrals 1 6.0 0.000 0.000 23.116 23.170 compute_2c_integrals 1 7.0 0.002 0.003 22.284 22.322 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.722 22.159 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.813 21.721 22.159 fft3d_s 1823 12.4 19.238 19.701 19.252 19.714 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.687 17.687 calc_potential_gpw 182 9.5 0.002 0.002 12.445 13.462 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.721 11.480 calculate_wavefunction 91 8.0 1.980 2.023 10.003 10.800 potential_pw2rs 186 10.0 0.033 0.034 8.972 10.383 collocate_single_gaussian 91 10.0 0.017 0.024 8.278 9.186 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.830 9.175 local_gemm 172 8.0 8.274 8.615 8.274 8.615 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.519 7.927 8.317 mp_sync 37 10.5 3.124 7.418 3.124 7.418 integrate_four_center_load 4 10.0 0.000 0.000 6.760 6.763 hfx_load_balance 1 11.0 0.000 0.000 6.760 6.763 mp2_ri_gpw_compute_en_ener 172 7.0 6.323 6.438 6.323 6.438 mp_sendrecv_dm3 2068 8.0 5.956 6.334 5.956 6.334 pw_poisson_solve 186 10.4 0.005 0.006 5.568 5.997 pw_gather_s 912 12.2 4.773 5.482 4.773 5.482 pw_scatter_s 910 12.7 4.149 4.798 4.149 4.798 pw_poisson_set 189 11.4 0.009 0.010 4.127 4.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.989029, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1530.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 446.185472E+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 3690 72628. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.124 58.166 58.170 qs_mol_dyn_low 1 2.0 0.011 0.055 57.683 57.718 qs_forces 11 3.9 0.017 0.107 57.494 57.498 qs_energies 11 4.9 0.002 0.004 55.864 55.888 scf_env_do_scf 11 5.9 0.000 0.001 49.227 49.228 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 47.082 47.083 dbcsr_multiply_generic 2286 12.5 0.094 0.098 36.939 37.393 qs_scf_new_mos 108 7.5 0.000 0.000 36.471 36.789 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.470 36.788 ot_scf_mini 108 9.5 0.002 0.002 34.678 34.883 multiply_cannon 2286 13.5 0.186 0.191 28.530 29.820 multiply_cannon_loop 2286 14.5 1.827 1.932 27.808 29.140 velocity_verlet 10 3.0 0.012 0.044 27.952 27.963 ot_mini 108 10.5 0.001 0.001 21.426 21.640 qs_ot_get_derivative 108 11.5 0.001 0.001 18.391 18.538 mp_waitall_1 245248 16.5 9.965 16.573 9.965 16.573 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.914 14.308 multiply_cannon_multrec 54864 15.5 3.583 5.550 8.205 11.813 qs_ot_get_p 119 10.4 0.001 0.001 8.372 8.654 rebuild_ks_matrix 119 8.3 0.000 0.000 8.178 8.331 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.016 8.177 8.330 mp_sum_l 7287 12.8 6.122 7.641 6.122 7.641 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.232 7.370 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.672 7.023 dbcsr_mm_accdrv_process 76910 16.1 1.856 2.916 4.536 6.425 multiply_cannon_sync_h2d 54864 15.5 5.057 6.203 5.057 6.203 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.679 5.778 init_scf_run 11 5.9 0.000 0.001 5.344 5.344 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.344 5.344 qs_ot_p2m_diag 50 11.0 0.007 0.028 5.266 5.314 sum_up_and_integrate 119 10.3 0.001 0.004 4.665 4.671 integrate_v_rspace 119 11.3 0.002 0.003 4.654 4.661 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.455 4.586 calculate_rho_elec 119 8.7 0.011 0.016 4.455 4.585 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.451 4.454 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.180 4.180 cp_fm_redistribute_end 50 14.0 2.134 4.153 2.140 4.155 cp_fm_diag_elpa_base 50 14.0 2.010 4.041 2.014 4.044 calculate_dm_sparse 119 9.5 0.000 0.001 3.452 3.592 multiply_cannon_metrocomm1 54864 15.5 0.055 0.062 2.165 3.541 jit_kernel_multiply 13 15.8 2.618 3.518 2.618 3.518 acc_transpose_blocks 54864 15.5 0.219 0.239 2.710 3.476 calculate_first_density_matrix 1 7.0 0.002 0.013 3.284 3.293 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.950 3.272 apply_single 119 13.6 0.000 0.000 2.950 3.272 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.761 2.775 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.684 2.755 ot_diis_step 108 11.5 0.006 0.006 2.745 2.746 density_rs2pw 119 9.7 0.003 0.004 2.375 2.471 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.351 2.419 grid_integrate_task_list 119 12.3 2.007 2.128 2.007 2.128 init_scf_loop 11 6.9 0.000 0.000 2.126 2.126 make_m2s 4572 13.5 0.054 0.056 2.002 2.066 mp_sum_d 4139 12.0 1.403 2.058 1.403 2.058 potential_pw2rs 119 12.3 0.004 0.004 2.004 2.041 make_images 4572 14.5 0.133 0.139 1.919 1.983 wfi_extrapolate 11 7.9 0.001 0.004 1.981 1.981 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.833 1.862 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.773 1.858 mp_alltoall_d11v 2130 13.8 1.556 1.771 1.556 1.771 acc_transpose_blocks_kernels 54864 16.5 0.250 0.384 1.233 1.756 transfer_rs2pw 487 10.6 0.006 0.007 1.572 1.702 fft3d_ps 1201 13.6 0.367 0.471 1.548 1.637 mp_waitany 12084 13.8 1.313 1.522 1.313 1.522 transfer_pw2rs 487 13.2 0.005 0.006 1.474 1.506 fft_wrap_pw1pw2_140 487 12.2 0.048 0.055 1.372 1.455 acc_transpose_blocks_sync 164592 16.5 1.212 1.452 1.212 1.452 grid_collocate_task_list 119 9.7 1.351 1.425 1.351 1.425 jit_kernel_transpose 5 15.5 0.983 1.382 0.983 1.382 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.758 1.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.170000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=425.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 484.810752E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 493917. 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 0.115 0.675 41.892 41.894 qs_mol_dyn_low 1 2.0 0.004 0.009 40.810 40.855 qs_forces 11 3.9 0.005 0.028 40.744 40.746 qs_energies 11 4.9 0.008 0.057 39.049 39.057 scf_env_do_scf 11 5.9 0.001 0.001 32.894 32.897 scf_env_do_scf_inner_loop 108 6.5 0.005 0.022 30.466 30.467 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.226 23.587 qs_scf_new_mos 108 7.5 0.001 0.001 21.609 21.838 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.609 21.838 ot_scf_mini 108 9.5 0.002 0.003 20.679 20.841 multiply_cannon 2286 13.5 0.209 0.216 17.857 19.393 velocity_verlet 10 3.0 0.001 0.004 18.443 18.445 multiply_cannon_loop 2286 14.5 1.188 1.240 16.663 18.251 ot_mini 108 10.5 0.001 0.001 12.864 13.092 mp_waitall_1 200699 16.5 6.118 11.462 6.118 11.462 qs_ot_get_derivative 108 11.5 0.001 0.001 10.505 10.668 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 4.470 10.088 multiply_cannon_multrec 27432 15.5 1.854 4.401 6.567 9.555 rebuild_ks_matrix 119 8.3 0.000 0.000 6.982 7.118 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.061 6.982 7.118 dbcsr_mm_accdrv_process 47894 16.0 3.821 6.225 4.632 7.072 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.175 6.298 qs_ot_get_p 119 10.4 0.001 0.001 4.886 5.113 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.970 4.845 init_scf_run 11 5.9 0.000 0.001 4.647 4.647 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.647 4.647 mp_sum_l 7287 12.8 2.165 4.264 2.165 4.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.026 4.184 apply_single 119 13.6 0.000 0.000 3.025 4.184 sum_up_and_integrate 119 10.3 0.001 0.002 3.797 3.803 integrate_v_rspace 119 11.3 0.002 0.003 3.784 3.789 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.664 3.695 calculate_rho_elec 119 8.7 0.021 0.024 3.664 3.695 qs_ot_p2m_diag 50 11.0 0.009 0.014 3.152 3.171 calculate_first_density_matrix 1 7.0 0.001 0.009 3.119 3.123 make_m2s 4572 13.5 0.052 0.054 2.770 3.114 make_images 4572 14.5 0.206 0.242 2.681 3.027 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.748 2.748 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.392 2.478 init_scf_loop 11 6.9 0.008 0.063 2.408 2.411 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.403 2.403 cp_fm_redistribute_end 50 14.0 1.215 2.367 1.221 2.370 ot_diis_step 108 11.5 0.011 0.011 2.310 2.311 calculate_dm_sparse 119 9.5 0.000 0.000 2.234 2.307 cp_fm_diag_elpa_base 50 14.0 1.114 2.261 1.144 2.302 multiply_cannon_sync_h2d 27432 15.5 1.700 2.278 1.700 2.278 acc_transpose_blocks 27432 15.5 0.114 0.119 1.684 2.270 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.161 2.164 jit_kernel_multiply 9 16.4 0.753 2.131 0.753 2.131 density_rs2pw 119 9.7 0.004 0.004 1.945 2.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.951 1.989 grid_integrate_task_list 119 12.3 1.841 1.944 1.841 1.944 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.750 1.782 make_images_data 4572 15.5 0.047 0.054 1.276 1.623 potential_pw2rs 119 12.3 0.006 0.006 1.537 1.544 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.143 1.535 fft3d_ps 1201 13.6 0.513 0.566 1.469 1.498 wfi_extrapolate 11 7.9 0.001 0.003 1.447 1.447 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.416 1.425 prepare_preconditioner 11 7.9 0.000 0.000 1.388 1.418 make_preconditioner 11 8.9 0.000 0.003 1.388 1.418 transfer_rs2pw 487 10.6 0.005 0.006 1.294 1.405 mp_alltoall_d11v 2130 13.8 1.280 1.398 1.280 1.398 fft_wrap_pw1pw2_140 487 12.2 0.047 0.050 1.335 1.366 grid_collocate_task_list 119 9.7 1.292 1.343 1.292 1.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.292 1.341 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.213 1.259 qs_energies_init_hamiltonians 11 5.9 0.008 0.034 1.199 1.203 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.727 1.192 mp_sum_d 4139 12.0 0.610 1.122 0.610 1.122 mp_allgather_i34 2286 14.5 0.616 1.036 0.616 1.036 transfer_pw2rs 487 13.2 0.004 0.005 0.997 1.002 acc_transpose_blocks_sync 82296 16.5 0.817 0.940 0.817 0.940 jit_kernel_transpose 5 15.5 0.538 0.920 0.538 0.920 make_images_sizes 4572 15.5 0.005 0.005 0.631 0.893 mp_alltoall_i44 4572 16.5 0.626 0.889 0.626 0.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.894000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.090909, yerr=1.621141 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 518.410240E+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 3629 73909. 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 0.045 35.701 35.702 qs_mol_dyn_low 1 2.0 0.021 0.150 34.608 34.625 qs_forces 11 3.9 0.002 0.003 34.476 34.494 qs_energies 11 4.9 0.023 0.031 32.881 32.901 scf_env_do_scf 11 5.9 0.000 0.001 27.074 27.074 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.238 24.239 dbcsr_multiply_generic 2286 12.5 0.096 0.098 18.915 19.026 qs_scf_new_mos 108 7.5 0.001 0.001 16.628 16.647 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.627 16.646 multiply_cannon 2286 13.5 0.193 0.202 15.272 15.942 ot_scf_mini 108 9.5 0.002 0.002 15.852 15.871 multiply_cannon_loop 2286 14.5 0.862 0.907 14.516 15.183 velocity_verlet 10 3.0 0.005 0.027 15.160 15.165 ot_mini 108 10.5 0.001 0.001 10.211 10.231 qs_ot_get_derivative 108 11.5 0.001 0.001 8.683 8.703 multiply_cannon_multrec 18288 15.5 2.115 3.358 8.117 8.478 dbcsr_mm_accdrv_process 38222 16.0 5.850 7.062 5.908 7.126 rebuild_ks_matrix 119 8.3 0.000 0.000 6.489 6.509 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.489 6.509 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.767 5.785 init_scf_run 11 5.9 0.000 0.001 4.326 4.327 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.326 4.326 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.111 4.014 mp_waitall_1 158411 16.6 2.766 3.529 2.766 3.529 sum_up_and_integrate 119 10.3 0.001 0.001 3.501 3.506 integrate_v_rspace 119 11.3 0.003 0.003 3.489 3.496 qs_ot_get_p 119 10.4 0.010 0.040 3.440 3.469 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.202 3.207 calculate_rho_elec 119 8.7 0.032 0.042 3.202 3.207 calculate_first_density_matrix 1 7.0 0.000 0.000 3.110 3.111 init_scf_loop 11 6.9 0.000 0.000 2.813 2.815 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.323 2.791 apply_single 119 13.6 0.000 0.000 2.323 2.791 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.641 2.420 calculate_dm_sparse 119 9.5 0.000 0.001 2.340 2.348 acc_transpose_blocks 18288 15.5 0.078 0.081 1.966 2.291 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.233 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.036 2.037 make_m2s 4572 13.5 0.044 0.045 1.840 1.972 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.937 1.938 grid_integrate_task_list 119 12.3 1.807 1.897 1.807 1.897 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.879 1.887 make_images 4572 14.5 0.192 0.204 1.755 1.886 density_rs2pw 119 9.7 0.003 0.004 1.765 1.876 mp_sum_l 7287 12.8 1.422 1.861 1.422 1.861 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.737 1.744 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.698 1.708 cp_fm_diag_elpa_base 50 14.0 1.675 1.684 1.697 1.707 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.667 1.671 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.645 1.654 ot_diis_step 108 11.5 0.011 0.011 1.510 1.511 prepare_preconditioner 11 7.9 0.000 0.000 1.465 1.466 make_preconditioner 11 8.9 0.000 0.000 1.465 1.466 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.348 1.403 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 1.120 1.362 fft3d_ps 1201 13.6 0.542 0.564 1.329 1.338 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.316 1.325 grid_collocate_task_list 119 9.7 1.231 1.304 1.231 1.304 transfer_rs2pw 487 10.6 0.005 0.005 1.157 1.301 potential_pw2rs 119 12.3 0.007 0.007 1.293 1.296 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.228 1.246 multiply_cannon_sync_h2d 18288 15.5 1.038 1.240 1.038 1.240 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 jit_kernel_transpose 5 15.6 0.903 1.142 0.903 1.142 mp_sum_d 4139 12.0 0.605 1.026 0.605 1.026 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.960 0.976 make_images_data 4572 15.5 0.047 0.051 0.783 0.933 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.678 0.872 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.782 0.866 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.367 0.850 acc_transpose_blocks_sync 54864 16.5 0.747 0.830 0.747 0.830 transfer_pw2rs 487 13.2 0.004 0.004 0.828 0.830 transfer_rs2pw_140 130 11.5 0.120 0.134 0.668 0.813 mp_waitany 9880 13.7 0.678 0.811 0.678 0.811 mp_alltoall_d11v 2130 13.8 0.694 0.798 0.694 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.794 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.669 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.702000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=493.000000, yerr=2.662876 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 555.692032E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.036 0.054 36.930 36.931 qs_mol_dyn_low 1 2.0 0.003 0.004 36.417 36.425 qs_forces 11 3.9 0.030 0.089 36.343 36.344 qs_energies 11 4.9 0.002 0.003 34.572 34.602 scf_env_do_scf 11 5.9 0.000 0.001 29.119 29.119 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.268 26.269 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.441 20.569 qs_scf_new_mos 108 7.5 0.001 0.001 18.233 18.284 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.232 18.283 velocity_verlet 10 3.0 0.004 0.023 17.934 17.935 ot_scf_mini 108 9.5 0.002 0.003 17.199 17.256 multiply_cannon 2286 13.5 0.218 0.228 16.531 17.000 multiply_cannon_loop 2286 14.5 1.497 1.636 15.570 15.903 ot_mini 108 10.5 0.001 0.001 10.741 10.810 multiply_cannon_multrec 27432 15.5 2.556 3.270 9.329 9.605 qs_ot_get_derivative 108 11.5 0.001 0.001 8.870 8.927 dbcsr_mm_accdrv_process 47916 15.9 6.240 7.978 6.671 8.073 rebuild_ks_matrix 119 8.3 0.000 0.000 6.325 6.380 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.324 6.379 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.606 5.656 init_scf_run 11 5.9 0.000 0.001 3.987 3.988 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.987 3.987 qs_ot_get_p 119 10.4 0.001 0.001 3.611 3.707 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.211 3.704 sum_up_and_integrate 119 10.3 0.001 0.001 3.422 3.429 integrate_v_rspace 119 11.3 0.003 0.003 3.411 3.417 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.330 3.370 calculate_rho_elec 119 8.7 0.040 0.046 3.329 3.369 init_scf_loop 11 6.9 0.000 0.000 2.827 2.827 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.203 2.638 apply_single 119 13.6 0.000 0.000 2.203 2.638 acc_transpose_blocks 27432 15.5 0.116 0.122 2.393 2.624 calculate_first_density_matrix 1 7.0 0.000 0.000 2.621 2.622 mp_waitall_1 137007 16.6 1.930 2.476 1.930 2.476 make_m2s 4572 13.5 0.054 0.056 2.335 2.457 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.344 2.371 make_images 4572 14.5 0.272 0.332 2.227 2.346 calculate_dm_sparse 119 9.5 0.000 0.000 2.177 2.236 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.155 2.167 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.953 1.954 prepare_preconditioner 11 7.9 0.000 0.000 1.937 1.944 make_preconditioner 11 8.9 0.000 0.000 1.937 1.944 mp_sum_l 7287 12.8 1.176 1.929 1.176 1.929 grid_integrate_task_list 119 12.3 1.819 1.906 1.819 1.906 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.676 1.877 density_rs2pw 119 9.7 0.003 0.004 1.748 1.839 ot_diis_step 108 11.5 0.012 0.012 1.830 1.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.779 1.794 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.735 1.765 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.757 1.757 acc_transpose_blocks_sync 82296 16.5 1.390 1.598 1.390 1.598 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.514 1.515 jit_kernel_multiply 7 16.5 0.366 1.501 0.366 1.501 cp_fm_redistribute_end 50 14.0 0.759 1.492 0.763 1.494 cp_fm_diag_elpa_base 50 14.0 0.704 1.426 0.729 1.467 fft_wrap_pw1pw2_140 487 12.2 0.070 0.072 1.429 1.461 fft3d_ps 1201 13.6 0.578 0.640 1.397 1.421 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.822 1.421 grid_collocate_task_list 119 9.7 1.249 1.320 1.249 1.320 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.295 1.316 wfi_extrapolate 11 7.9 0.001 0.001 1.315 1.315 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.284 potential_pw2rs 119 12.3 0.008 0.009 1.232 1.234 qs_energies_init_hamiltonians 11 5.9 0.031 0.047 1.195 1.222 transfer_rs2pw 487 10.6 0.005 0.005 0.985 1.082 make_images_data 4572 15.5 0.047 0.051 0.877 1.006 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.735 0.948 mp_alltoall_d11v 2130 13.8 0.807 0.941 0.807 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.879 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.857 0.877 dbcsr_complete_redistribute 329 12.2 0.115 0.131 0.660 0.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.811 multiply_cannon_metrocomm1 27432 15.5 0.035 0.035 0.258 0.789 transfer_pw2rs 487 13.2 0.004 0.004 0.737 0.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.931000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.818182, yerr=2.328409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 619.896832E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.048 0.118 29.855 29.856 qs_mol_dyn_low 1 2.0 0.003 0.004 29.102 29.110 qs_forces 11 3.9 0.002 0.003 29.043 29.044 qs_energies 11 4.9 0.002 0.003 27.355 27.358 scf_env_do_scf 11 5.9 0.000 0.001 22.029 22.029 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.625 19.626 velocity_verlet 10 3.0 0.002 0.003 14.251 14.254 dbcsr_multiply_generic 2286 12.5 0.095 0.098 14.046 14.161 qs_scf_new_mos 108 7.5 0.001 0.001 12.213 12.238 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.212 12.237 multiply_cannon 2286 13.5 0.224 0.229 11.434 11.897 ot_scf_mini 108 9.5 0.002 0.002 11.510 11.542 multiply_cannon_loop 2286 14.5 0.643 0.661 10.550 10.798 ot_mini 108 10.5 0.001 0.001 6.851 6.889 multiply_cannon_multrec 9144 15.5 1.740 1.938 6.448 6.711 rebuild_ks_matrix 119 8.3 0.000 0.000 5.706 5.727 qs_ks_build_kohn_sham_matrix 119 9.3 0.020 0.027 5.705 5.726 qs_ot_get_derivative 108 11.5 0.001 0.001 5.569 5.602 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.083 5.102 dbcsr_mm_accdrv_process 12550 15.8 3.789 4.661 4.599 4.730 init_scf_run 11 5.9 0.000 0.001 3.863 3.863 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.862 3.863 sum_up_and_integrate 119 10.3 0.001 0.001 3.344 3.348 integrate_v_rspace 119 11.3 0.003 0.003 3.334 3.338 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.176 3.186 calculate_rho_elec 119 8.7 0.060 0.061 3.176 3.185 qs_ot_get_p 119 10.4 0.021 0.060 2.884 2.932 calculate_first_density_matrix 1 7.0 0.000 0.000 2.758 2.758 init_scf_loop 11 6.9 0.000 0.000 2.382 2.384 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.072 2.073 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.006 2.072 mp_waitall_1 115863 16.7 1.515 1.921 1.515 1.921 grid_integrate_task_list 119 12.3 1.867 1.919 1.867 1.919 make_m2s 4572 13.5 0.034 0.035 1.776 1.897 calculate_dm_sparse 119 9.5 0.000 0.000 1.837 1.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.843 1.850 acc_transpose_blocks 9144 15.5 0.041 0.042 1.801 1.843 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.817 1.820 make_images 4572 14.5 0.269 0.300 1.687 1.806 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.743 1.751 density_rs2pw 119 9.7 0.003 0.004 1.580 1.634 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.576 1.577 prepare_preconditioner 11 7.9 0.000 0.000 1.517 1.525 make_preconditioner 11 8.9 0.000 0.000 1.517 1.525 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.441 1.449 jit_kernel_multiply 6 15.5 0.771 1.442 0.771 1.442 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.412 1.442 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.260 1.393 apply_single 119 13.6 0.000 0.000 1.259 1.393 grid_collocate_task_list 119 9.7 1.294 1.386 1.294 1.386 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.352 1.370 fft3d_ps 1201 13.6 0.638 0.653 1.355 1.366 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.310 1.317 cp_fm_diag_elpa_base 50 14.0 1.283 1.299 1.308 1.316 ot_diis_step 108 11.5 0.012 0.013 1.268 1.268 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.243 1.244 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.233 1.238 potential_pw2rs 119 12.3 0.010 0.010 1.155 1.157 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 1.025 1.060 wfi_extrapolate 11 7.9 0.001 0.001 1.055 1.055 jit_kernel_transpose 5 15.6 0.907 0.941 0.907 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.928 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.717 0.891 make_images_data 4572 15.5 0.042 0.047 0.744 0.882 mp_alltoall_d11v 2130 13.8 0.749 0.837 0.749 0.837 transfer_rs2pw 487 10.6 0.004 0.005 0.770 0.821 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.821 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.467 0.785 cp_fm_cholesky_invert 11 10.9 0.743 0.748 0.743 0.748 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.742 0.745 acc_transpose_blocks_sync 27432 16.5 0.718 0.740 0.718 0.740 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.685 0.737 transfer_pw2rs 487 13.2 0.003 0.004 0.652 0.654 mp_alltoall_z22v 1201 15.6 0.592 0.630 0.592 0.630 dbcsr_complete_redistribute 329 12.2 0.190 0.217 0.586 0.613 mp_sum_l 7287 12.8 0.451 0.611 0.451 0.611 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.856000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.636364, yerr=4.695699 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 774.627328E+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 3629 74142. 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 0.057 0.073 43.919 43.919 qs_mol_dyn_low 1 2.0 0.004 0.015 43.191 43.200 qs_forces 11 3.9 0.007 0.014 42.858 42.860 qs_energies 11 4.9 0.002 0.002 40.785 40.793 scf_env_do_scf 11 5.9 0.001 0.001 33.566 33.566 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.178 27.180 velocity_verlet 10 3.0 0.244 0.251 22.377 22.387 dbcsr_multiply_generic 2286 12.5 0.103 0.104 21.082 21.321 multiply_cannon 2286 13.5 0.299 0.312 17.076 18.099 qs_scf_new_mos 108 7.5 0.001 0.001 17.918 18.013 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.917 18.013 multiply_cannon_loop 2286 14.5 0.871 0.887 15.824 16.881 ot_scf_mini 108 9.5 0.002 0.002 16.758 16.858 ot_mini 108 10.5 0.001 0.001 10.446 10.561 multiply_cannon_multrec 9144 15.5 3.839 5.165 10.126 10.555 qs_ot_get_derivative 108 11.5 0.001 0.001 8.397 8.497 dbcsr_mm_accdrv_process 12550 15.8 5.350 7.599 6.152 7.799 rebuild_ks_matrix 119 8.3 0.000 0.000 7.080 7.240 qs_ks_build_kohn_sham_matrix 119 9.3 0.034 0.054 7.080 7.239 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.371 6.518 init_scf_loop 11 6.9 0.000 0.000 6.361 6.364 init_scf_run 11 5.9 0.000 0.001 5.183 5.183 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.183 5.183 prepare_preconditioner 11 7.9 0.000 0.000 4.957 4.991 make_preconditioner 11 8.9 0.000 0.000 4.957 4.990 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.013 4.875 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.966 3.971 calculate_rho_elec 119 8.7 0.118 0.122 3.966 3.971 sum_up_and_integrate 119 10.3 0.001 0.001 3.732 3.739 calculate_first_density_matrix 1 7.0 0.000 0.000 3.734 3.735 integrate_v_rspace 119 11.3 0.004 0.004 3.722 3.729 qs_ot_get_p 119 10.4 0.004 0.004 3.375 3.512 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.934 3.420 mp_waitall_1 94719 16.7 2.525 3.400 2.525 3.400 calculate_dm_sparse 119 9.5 0.000 0.000 2.985 2.999 cp_fm_upper_to_full 72 14.2 2.102 2.954 2.102 2.954 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.280 2.627 apply_single 119 13.6 0.000 0.000 2.280 2.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.583 2.583 make_m2s 4572 13.5 0.037 0.038 2.363 2.510 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.399 2.406 make_images 4572 14.5 0.354 0.388 2.242 2.388 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.370 2.380 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.316 2.368 acc_transpose_blocks 9144 15.5 0.044 0.045 2.271 2.319 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.103 2.154 grid_integrate_task_list 119 12.3 2.085 2.123 2.085 2.123 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.041 2.046 ot_diis_step 108 11.5 0.014 0.014 2.025 2.026 density_rs2pw 119 9.7 0.003 0.003 1.986 2.020 dbcsr_complete_redistribute 329 12.2 0.260 0.264 1.495 2.012 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.980 1.988 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.971 1.974 mp_sum_l 7287 12.8 1.101 1.926 1.101 1.926 fft3d_ps 1201 13.6 0.860 0.882 1.851 1.857 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.765 1.772 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.182 1.685 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.615 1.616 grid_collocate_task_list 119 9.7 1.525 1.551 1.525 1.551 mp_alltoall_i22 627 13.8 0.930 1.500 0.930 1.500 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.939 1.441 jit_kernel_multiply 5 15.7 0.773 1.420 0.773 1.420 wfi_extrapolate 11 7.9 0.001 0.001 1.384 1.384 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.324 1.343 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.313 1.314 cp_fm_diag_elpa_base 50 14.0 1.227 1.240 1.312 1.312 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.067 1.290 mp_alltoall_d11v 2130 13.8 1.226 1.255 1.226 1.255 potential_pw2rs 119 12.3 0.014 0.014 1.210 1.211 make_images_data 4572 15.5 0.046 0.049 0.995 1.188 acc_transpose_blocks_sync 27432 16.5 1.124 1.146 1.124 1.146 cp_fm_cholesky_invert 11 10.9 1.120 1.127 1.120 1.127 acc_transpose_blocks_kernels 9144 16.5 0.120 0.122 1.087 1.115 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.093 1.112 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.988 1.016 jit_kernel_transpose 5 15.6 0.967 0.994 0.967 0.994 qs_create_task_list 11 7.9 0.000 0.000 0.929 0.941 generate_qs_task_list 11 8.9 0.368 0.389 0.929 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.932 mp_alltoall_z22v 1201 15.6 0.856 0.885 0.856 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.919000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.272727, yerr=13.632121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 496.615424E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 652010. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.034 0.056 85.718 85.718 qs_mol_dyn_low 1 2.0 0.003 0.004 84.464 84.473 qs_forces 11 3.9 0.007 0.009 84.367 84.368 qs_energies 11 4.9 0.002 0.002 81.475 81.496 scf_env_do_scf 11 5.9 0.000 0.001 72.241 72.242 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.644 66.645 dbcsr_multiply_generic 2055 12.4 0.107 0.110 53.652 53.934 qs_scf_new_mos 99 7.5 0.000 0.001 49.904 50.082 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.903 50.081 ot_scf_mini 99 9.5 0.002 0.002 47.453 47.595 multiply_cannon 2055 13.4 0.181 0.187 44.217 45.297 multiply_cannon_loop 2055 14.4 1.809 1.884 43.273 44.329 velocity_verlet 10 3.0 0.015 0.035 42.332 42.333 ot_mini 99 10.5 0.001 0.001 28.856 28.974 qs_ot_get_derivative 99 11.5 0.031 0.245 22.039 22.143 multiply_cannon_multrec 49320 15.4 11.149 11.852 18.259 19.136 rebuild_ks_matrix 110 8.3 0.000 0.000 14.737 14.849 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.035 14.736 14.848 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.950 13.048 mp_waitall_1 220248 16.4 11.568 12.693 11.568 12.693 qs_ot_get_p 110 10.4 0.001 0.001 9.956 10.085 multiply_cannon_sync_h2d 49320 15.4 9.348 10.036 9.348 10.036 multiply_cannon_metrocomm3 49320 15.4 0.080 0.086 7.087 8.437 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.689 8.232 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.194 7.769 apply_single 110 13.6 0.000 0.001 7.194 7.769 dbcsr_mm_accdrv_process 87628 16.1 3.179 3.414 6.983 7.513 init_scf_run 11 5.9 0.000 0.001 7.148 7.149 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.148 7.148 sum_up_and_integrate 110 10.3 0.001 0.003 6.799 6.821 integrate_v_rspace 110 11.3 0.003 0.003 6.774 6.801 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.710 6.752 ot_diis_step 99 11.5 0.005 0.006 6.561 6.561 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.382 6.476 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.087 6.259 calculate_rho_elec 110 8.6 0.020 0.024 6.087 6.258 mp_sum_l 6594 12.7 5.074 5.982 5.074 5.982 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.720 5.720 init_scf_loop 11 6.9 0.000 0.000 5.566 5.566 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.245 5.247 cp_fm_redistribute_end 48 14.0 2.664 5.221 2.674 5.224 cp_fm_diag_elpa_base 48 14.0 2.540 5.102 2.548 5.115 calculate_dm_sparse 110 9.5 0.001 0.001 4.063 4.176 multiply_cannon_metrocomm1 49320 15.4 0.062 0.066 2.684 4.005 make_m2s 4110 13.4 0.061 0.066 3.903 4.001 wfi_extrapolate 11 7.9 0.001 0.001 3.927 3.927 make_images 4110 14.4 0.178 0.190 3.808 3.909 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.756 3.758 jit_kernel_multiply 13 15.9 3.525 3.668 3.525 3.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.417 3.464 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.371 3.407 grid_integrate_task_list 110 12.3 3.258 3.402 3.258 3.402 density_rs2pw 110 9.6 0.004 0.004 3.106 3.281 calculate_first_density_matrix 1 7.0 0.000 0.000 3.123 3.126 prepare_preconditioner 11 7.9 0.000 0.000 3.021 3.037 make_preconditioner 11 8.9 0.000 0.000 3.021 3.037 acc_transpose_blocks 49320 15.4 0.213 0.236 2.783 2.959 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.789 2.834 fft_wrap_pw1pw2 1111 11.6 0.014 0.019 2.625 2.705 mp_alltoall_d11v 2046 13.8 2.159 2.629 2.159 2.629 potential_pw2rs 110 12.3 0.006 0.006 2.406 2.434 fft_wrap_pw1pw2_140 451 12.1 0.092 0.100 2.236 2.315 grid_collocate_task_list 110 9.6 2.158 2.286 2.158 2.286 mp_sum_d 3893 11.9 1.633 2.245 1.633 2.245 fft3d_ps 1111 13.6 0.731 0.821 2.164 2.238 transfer_rs2pw 451 10.6 0.006 0.006 1.889 2.046 mp_waitany 14300 13.8 1.853 2.014 1.853 2.014 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.929 1.947 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.903 1.941 make_images_data 4110 15.4 0.043 0.046 1.774 1.931 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.543 1.808 transfer_pw2rs 451 13.1 0.006 0.007 1.709 1.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.718000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.181818, yerr=2.289032 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 590.999552E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 1935329. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.040 0.099 74.395 74.397 qs_mol_dyn_low 1 2.0 0.082 0.320 71.698 73.868 qs_forces 11 3.9 0.003 0.003 68.593 68.597 qs_energies 11 4.9 0.002 0.002 65.268 65.275 scf_env_do_scf 11 5.9 0.000 0.001 56.378 56.379 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.463 49.464 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.450 38.624 velocity_verlet 10 3.0 0.014 0.050 37.762 37.826 qs_scf_new_mos 99 7.5 0.001 0.001 33.866 34.023 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.866 34.022 multiply_cannon 2055 13.4 0.224 0.245 31.850 33.101 ot_scf_mini 99 9.5 0.003 0.003 32.193 32.355 multiply_cannon_loop 2055 14.4 1.155 1.181 30.708 31.715 ot_mini 99 10.5 0.001 0.001 18.989 19.166 multiply_cannon_multrec 24660 15.4 6.956 8.282 14.446 15.945 rebuild_ks_matrix 110 8.3 0.000 0.000 13.467 13.583 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.467 13.582 qs_ot_get_derivative 99 11.5 0.001 0.001 13.212 13.375 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.847 11.946 mp_waitall_1 176588 16.5 7.688 10.220 7.688 10.220 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.314 8.529 dbcsr_mm_accdrv_process 52282 16.1 5.614 6.759 7.320 7.660 multiply_cannon_sync_h2d 24660 15.4 6.395 7.458 6.395 7.458 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.488 7.218 apply_single 110 13.6 0.000 0.001 6.488 7.218 init_scf_loop 11 6.9 0.000 0.000 6.877 6.878 qs_ot_get_p 110 10.4 0.001 0.001 6.565 6.766 init_scf_run 11 5.9 0.000 0.001 6.386 6.387 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.386 6.387 sum_up_and_integrate 110 10.3 0.001 0.002 6.228 6.239 integrate_v_rspace 110 11.3 0.003 0.003 6.201 6.211 ot_diis_step 99 11.5 0.010 0.010 5.728 5.729 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.650 5.662 calculate_rho_elec 110 8.6 0.039 0.047 5.650 5.662 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.875 5.627 prepare_preconditioner 11 7.9 0.000 0.000 4.879 4.909 make_preconditioner 11 8.9 0.000 0.000 4.879 4.909 md_write_output 11 3.9 0.019 0.490 0.339 4.902 md_output 10 3.0 0.000 0.000 0.105 4.862 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.682 4.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.488 4.591 make_m2s 4110 13.4 0.057 0.059 4.178 4.513 make_images 4110 14.4 0.405 0.468 4.067 4.397 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.223 4.223 write_trajectory 44 4.9 0.001 0.006 0.086 4.100 write_particle_coordinates 11 5.9 0.085 4.093 0.085 4.093 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.684 3.684 cp_fm_redistribute_end 48 14.0 1.862 3.655 1.867 3.658 cp_fm_diag_elpa_base 48 14.0 1.728 3.511 1.784 3.587 wfi_extrapolate 11 7.9 0.001 0.001 3.421 3.421 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.303 3.393 grid_integrate_task_list 110 12.3 3.173 3.342 3.173 3.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.264 3.265 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.080 3.233 density_rs2pw 110 9.6 0.004 0.004 2.904 3.110 calculate_dm_sparse 110 9.5 0.001 0.001 2.993 3.027 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.896 2.937 calculate_first_density_matrix 1 7.0 0.000 0.000 2.871 2.874 mp_sum_dm 438 4.9 2.695 2.782 2.695 2.782 fft_wrap_pw1pw2_140 451 12.1 0.110 0.116 2.622 2.776 update_particle_set 20 4.0 0.000 0.000 2.649 2.726 make_images_data 4110 15.4 0.048 0.052 2.279 2.634 fft3d_ps 1111 13.6 1.069 1.280 2.459 2.599 mp_sum_l 6594 12.7 1.801 2.574 1.801 2.574 hybrid_alltoall_any 4261 16.3 0.105 0.451 1.954 2.555 mp_max_l 33 2.8 2.405 2.530 2.405 2.530 grid_collocate_task_list 110 9.6 2.173 2.391 2.173 2.391 cp_fm_cholesky_invert 11 10.9 2.364 2.378 2.364 2.378 jit_kernel_multiply 9 16.4 1.339 2.196 1.339 2.196 acc_transpose_blocks 24660 15.4 0.115 0.119 2.028 2.154 potential_pw2rs 110 12.3 0.008 0.009 2.110 2.118 mp_alltoall_d11v 2046 13.8 1.817 2.023 1.817 2.023 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.017 2.021 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 1.994 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.773 1.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.779 1.796 mp_irecv_dv 57340 16.2 0.640 1.646 0.640 1.646 transfer_rs2pw 451 10.6 0.007 0.008 1.404 1.625 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.555 1.571 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.390 1.496 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.397000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.818182, yerr=6.685373 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 667.181056E+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 3480 77281. 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 0.033 61.561 61.562 qs_mol_dyn_low 1 2.0 0.003 0.004 61.259 61.269 qs_forces 11 3.9 0.003 0.003 61.181 61.182 qs_energies 11 4.9 0.002 0.002 57.971 57.974 scf_env_do_scf 11 5.9 0.000 0.001 48.612 48.613 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.020 41.021 velocity_verlet 10 3.0 0.005 0.019 31.878 31.879 dbcsr_multiply_generic 2055 12.4 0.110 0.114 30.198 30.461 qs_scf_new_mos 99 7.5 0.001 0.001 26.258 26.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.257 26.345 ot_scf_mini 99 9.5 0.002 0.002 25.000 25.105 multiply_cannon 2055 13.4 0.210 0.219 23.692 24.968 multiply_cannon_loop 2055 14.4 0.816 0.839 22.497 23.514 ot_mini 99 10.5 0.001 0.001 14.262 14.364 rebuild_ks_matrix 110 8.3 0.000 0.000 12.305 12.425 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.305 12.425 multiply_cannon_multrec 16440 15.4 3.709 4.842 10.825 12.065 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.831 10.940 mp_waitall_1 139946 16.5 7.108 10.134 7.108 10.134 qs_ot_get_derivative 99 11.5 0.001 0.001 9.846 9.951 init_scf_loop 11 6.9 0.000 0.000 7.545 7.547 dbcsr_mm_accdrv_process 34862 16.1 6.115 6.683 6.959 7.230 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.309 7.167 init_scf_run 11 5.9 0.000 0.001 6.964 6.964 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.964 6.964 sum_up_and_integrate 110 10.3 0.001 0.002 6.330 6.345 integrate_v_rspace 110 11.3 0.003 0.003 6.303 6.319 qs_ot_get_p 110 10.4 0.002 0.004 5.866 5.989 prepare_preconditioner 11 7.9 0.000 0.000 5.767 5.787 make_preconditioner 11 8.9 0.000 0.000 5.767 5.787 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.720 5.727 calculate_rho_elec 110 8.6 0.058 0.058 5.720 5.726 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.224 5.470 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.904 5.367 apply_single 110 13.6 0.000 0.000 4.904 5.367 make_m2s 4110 13.4 0.049 0.051 4.408 4.778 make_images 4110 14.4 0.399 0.527 4.292 4.661 ot_diis_step 99 11.5 0.010 0.011 4.387 4.387 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.104 4.108 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.331 3.951 calculate_first_density_matrix 1 7.0 0.000 0.000 3.909 3.910 multiply_cannon_sync_h2d 16440 15.4 3.238 3.779 3.238 3.779 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.726 3.727 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.442 3.443 grid_integrate_task_list 110 12.3 3.194 3.375 3.194 3.375 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.216 3.238 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.077 3.128 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.117 3.127 cp_fm_diag_elpa_base 48 14.0 3.071 3.089 3.115 3.124 density_rs2pw 110 9.6 0.004 0.004 2.947 3.119 wfi_extrapolate 11 7.9 0.001 0.001 2.960 2.960 make_images_data 4110 15.4 0.045 0.049 2.404 2.888 calculate_dm_sparse 110 9.5 0.001 0.001 2.774 2.812 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.125 2.800 fft_wrap_pw1pw2_140 451 12.1 0.127 0.133 2.719 2.743 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.500 2.555 fft3d_ps 1111 13.6 1.087 1.103 2.527 2.551 cp_fm_cholesky_invert 11 10.9 2.436 2.448 2.436 2.448 grid_collocate_task_list 110 9.6 2.219 2.435 2.219 2.435 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.870 2.392 mp_alltoall_d11v 2046 13.8 1.969 2.294 1.969 2.294 mp_irecv_dv 48980 15.7 0.794 2.258 0.794 2.258 potential_pw2rs 110 12.3 0.010 0.010 2.138 2.156 mp_sum_l 6594 12.7 1.474 2.066 1.474 2.066 qs_energies_init_hamiltonians 11 5.9 0.008 0.021 1.961 1.962 acc_transpose_blocks 16440 15.4 0.076 0.078 1.745 1.913 dbcsr_complete_redistribute 325 12.2 0.295 0.332 1.438 1.750 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.712 1.725 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.690 transfer_rs2pw 451 10.6 0.005 0.006 1.392 1.563 mp_allgather_i34 2055 14.4 0.508 1.546 0.508 1.546 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.362 1.474 make_basis_sm 11 9.8 0.000 0.000 1.426 1.429 cp_fm_cholesky_decompose 22 10.9 1.355 1.392 1.355 1.392 parallel_gemm_fm 81 9.0 0.000 0.000 1.356 1.364 parallel_gemm_fm_cosma 81 10.0 1.356 1.363 1.356 1.363 mp_waitany 17072 13.8 1.135 1.321 1.135 1.321 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.297 1.308 transfer_pw2rs 451 13.1 0.005 0.005 1.255 1.272 mp_alltoall_z22v 1111 15.6 1.227 1.253 1.227 1.253 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.562000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.909091, yerr=8.479436 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 743.751680E+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 3480 77279. 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 0.062 0.088 65.975 65.976 qs_mol_dyn_low 1 2.0 0.004 0.009 64.920 65.116 qs_forces 11 3.9 0.003 0.004 64.803 64.805 qs_energies 11 4.9 0.003 0.004 61.394 61.400 scf_env_do_scf 11 5.9 0.000 0.001 51.884 51.889 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.233 42.233 velocity_verlet 10 3.0 0.002 0.003 35.297 35.301 dbcsr_multiply_generic 2055 12.4 0.137 0.144 31.760 31.989 qs_scf_new_mos 99 7.5 0.001 0.001 27.831 27.939 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.831 27.938 ot_scf_mini 99 9.5 0.002 0.003 26.123 26.228 multiply_cannon 2055 13.4 0.237 0.255 24.445 25.760 multiply_cannon_loop 2055 14.4 1.378 1.490 23.002 23.742 ot_mini 99 10.5 0.001 0.001 15.166 15.298 multiply_cannon_multrec 24660 15.4 4.096 6.899 13.419 14.605 rebuild_ks_matrix 110 8.3 0.000 0.000 12.094 12.191 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.094 12.191 qs_ot_get_derivative 99 11.5 0.001 0.001 11.010 11.119 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.704 10.792 dbcsr_mm_accdrv_process 52304 16.0 8.056 9.432 9.167 10.145 init_scf_loop 11 6.9 0.000 0.000 9.600 9.601 prepare_preconditioner 11 7.9 0.000 0.000 7.715 7.734 make_preconditioner 11 8.9 0.000 0.000 7.715 7.734 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.570 7.391 init_scf_run 11 5.9 0.000 0.001 6.651 6.652 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.651 6.652 mp_waitall_1 121746 16.5 4.497 6.552 4.497 6.552 sum_up_and_integrate 110 10.3 0.001 0.002 6.106 6.120 integrate_v_rspace 110 11.3 0.003 0.003 6.080 6.093 make_m2s 4110 13.4 0.060 0.062 5.559 5.938 qs_ot_get_p 110 10.4 0.001 0.002 5.786 5.924 make_images 4110 14.4 0.575 0.697 5.418 5.793 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.616 5.627 calculate_rho_elec 110 8.6 0.077 0.081 5.615 5.627 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.042 4.119 apply_single 110 13.6 0.000 0.000 4.042 4.119 ot_diis_step 99 11.5 0.011 0.011 4.116 4.116 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.915 3.974 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.889 3.908 calculate_first_density_matrix 1 7.0 0.000 0.000 3.572 3.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.496 3.503 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.375 3.434 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.589 3.428 grid_integrate_task_list 110 12.3 3.266 3.423 3.266 3.423 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.317 3.317 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.199 3.225 calculate_dm_sparse 110 9.5 0.001 0.001 3.145 3.187 hybrid_alltoall_any 4261 16.3 0.122 0.463 2.332 3.149 make_images_data 4110 15.4 0.048 0.052 2.779 3.102 acc_transpose_blocks 24660 15.4 0.109 0.113 2.718 3.012 dbcsr_complete_redistribute 325 12.2 0.437 0.543 2.303 3.001 wfi_extrapolate 11 7.9 0.001 0.001 2.984 2.984 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.933 2.964 density_rs2pw 110 9.6 0.004 0.004 2.741 2.956 cp_fm_upper_to_full 70 14.2 2.083 2.941 2.083 2.941 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.757 2.769 cp_fm_diag_elpa_base 48 14.0 2.692 2.717 2.754 2.765 fft_wrap_pw1pw2_140 451 12.1 0.146 0.149 2.728 2.757 cp_fm_cholesky_invert 11 10.9 2.495 2.505 2.495 2.505 fft3d_ps 1111 13.6 1.130 1.174 2.468 2.486 multiply_cannon_sync_h2d 24660 15.4 2.349 2.468 2.349 2.468 grid_collocate_task_list 110 9.6 2.266 2.430 2.266 2.430 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.714 2.397 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.353 2.355 mp_alltoall_d11v 2046 13.8 1.889 2.174 1.889 2.174 potential_pw2rs 110 12.3 0.012 0.013 1.907 1.910 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.807 1.842 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.136 1.804 mp_alltoall_i22 605 13.7 1.052 1.780 1.052 1.780 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.695 1.713 mp_sum_l 6594 12.7 1.082 1.712 1.082 1.712 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.711 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.649 mp_allgather_i34 2055 14.4 0.536 1.606 0.536 1.606 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.860 1.576 acc_transpose_blocks_sync 73980 16.4 1.378 1.528 1.378 1.528 mp_irecv_dv 62702 16.1 0.756 1.493 0.756 1.493 cp_fm_cholesky_decompose 22 10.9 1.451 1.487 1.451 1.487 transfer_rs2pw 451 10.6 0.005 0.006 1.196 1.478 jit_kernel_multiply 6 15.6 0.766 1.465 0.766 1.465 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.289 1.382 mp_waitany 13376 13.8 1.023 1.369 1.023 1.369 acc_transpose_blocks_kernels 24660 16.4 0.308 0.317 1.190 1.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.976000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.909091, yerr=11.492721 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 857.112576E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.024 0.072 56.200 56.200 qs_mol_dyn_low 1 2.0 0.003 0.003 55.867 55.902 qs_forces 11 3.9 0.003 0.003 55.800 55.801 qs_energies 11 4.9 0.002 0.002 52.126 52.133 scf_env_do_scf 11 5.9 0.000 0.001 42.680 42.681 scf_env_do_scf_inner_loop 99 6.5 0.006 0.010 35.399 35.400 velocity_verlet 10 3.0 0.005 0.007 30.135 30.138 dbcsr_multiply_generic 2055 12.4 0.107 0.110 24.257 24.436 qs_scf_new_mos 99 7.5 0.001 0.001 21.197 21.269 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.196 21.269 ot_scf_mini 99 9.5 0.002 0.002 19.935 19.982 multiply_cannon 2055 13.4 0.239 0.247 18.693 19.812 multiply_cannon_loop 2055 14.4 0.605 0.617 17.440 17.733 rebuild_ks_matrix 110 8.3 0.000 0.000 11.585 11.638 qs_ks_build_kohn_sham_matrix 110 9.3 0.028 0.033 11.584 11.638 ot_mini 99 10.5 0.001 0.001 10.998 11.040 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.318 10.366 multiply_cannon_multrec 8220 15.4 3.208 4.907 8.231 9.466 mp_waitall_1 103326 16.6 5.880 7.413 5.880 7.413 qs_ot_get_derivative 99 11.5 0.001 0.001 7.213 7.259 init_scf_loop 11 6.9 0.000 0.000 7.234 7.235 init_scf_run 11 5.9 0.000 0.001 6.372 6.372 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.372 6.372 sum_up_and_integrate 110 10.3 0.001 0.002 6.047 6.058 integrate_v_rspace 110 11.3 0.003 0.004 6.019 6.031 dbcsr_mm_accdrv_process 17442 15.9 3.426 4.525 4.883 5.943 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.711 5.729 calculate_rho_elec 110 8.6 0.115 0.116 5.710 5.728 prepare_preconditioner 11 7.9 0.000 0.000 5.463 5.476 make_preconditioner 11 8.9 0.000 0.000 5.462 5.476 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.055 5.128 qs_ot_get_p 110 10.4 0.005 0.006 4.999 5.057 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.104 4.505 make_m2s 4110 13.4 0.038 0.039 4.175 4.357 make_images 4110 14.4 0.648 0.703 4.045 4.226 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.804 3.860 apply_single 110 13.6 0.000 0.000 3.804 3.860 calculate_first_density_matrix 1 7.0 0.000 0.000 3.770 3.770 ot_diis_step 99 11.5 0.012 0.012 3.759 3.759 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.588 3.591 grid_integrate_task_list 110 12.3 3.380 3.516 3.380 3.516 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.339 3.348 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.241 3.242 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.122 3.125 fft_wrap_pw1pw2_140 451 12.1 0.192 0.194 2.879 2.889 density_rs2pw 110 9.6 0.004 0.004 2.691 2.771 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.727 2.752 calculate_dm_sparse 110 9.5 0.001 0.001 2.697 2.739 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.659 2.660 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.644 2.651 cp_fm_diag_elpa_base 48 14.0 2.588 2.615 2.642 2.649 cp_fm_cholesky_invert 11 10.9 2.589 2.597 2.589 2.597 hybrid_alltoall_any 4261 16.3 0.202 0.857 2.219 2.597 multiply_cannon_sync_h2d 8220 15.4 2.357 2.532 2.357 2.532 fft3d_ps 1111 13.6 1.277 1.290 2.497 2.509 wfi_extrapolate 11 7.9 0.001 0.001 2.507 2.507 make_images_data 4110 15.4 0.041 0.047 2.208 2.492 grid_collocate_task_list 110 9.6 2.369 2.462 2.369 2.462 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.102 2.146 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.980 2.003 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.999 mp_alltoall_d11v 2046 13.8 1.772 1.957 1.772 1.957 potential_pw2rs 110 12.3 0.015 0.015 1.837 1.841 acc_transpose_blocks 8220 15.4 0.039 0.040 1.704 1.730 jit_kernel_multiply 7 15.6 1.142 1.694 1.142 1.694 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.649 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.597 1.602 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.483 1.600 dbcsr_complete_redistribute 325 12.2 0.541 0.571 1.469 1.566 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.478 1.482 cp_fm_cholesky_decompose 22 10.9 1.448 1.464 1.448 1.464 mp_allgather_i34 2055 14.4 0.420 1.454 0.420 1.454 mp_irecv_dv 24056 15.7 0.451 1.437 0.451 1.437 make_basis_sm 11 9.8 0.000 0.000 1.321 1.323 qs_create_task_list 11 7.9 0.000 0.000 1.205 1.307 generate_qs_task_list 11 8.9 0.374 0.442 1.205 1.307 copy_dbcsr_to_fm 151 11.3 0.002 0.003 1.204 1.231 transfer_rs2pw 451 10.6 0.005 0.006 1.117 1.211 mp_waitany 9240 13.8 1.049 1.140 1.049 1.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.200000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.363636, yerr=9.984285 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 1.396736E+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 3469 77990. 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 0.068 0.163 80.419 80.420 qs_mol_dyn_low 1 2.0 0.003 0.003 79.927 79.941 qs_forces 11 3.9 0.003 0.003 79.861 79.862 qs_energies 11 4.9 0.002 0.002 75.780 75.782 scf_env_do_scf 11 5.9 0.000 0.001 64.534 64.534 velocity_verlet 10 3.0 0.002 0.002 47.886 47.892 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.021 44.023 dbcsr_multiply_generic 2055 12.4 0.122 0.126 31.064 31.184 qs_scf_new_mos 99 7.5 0.001 0.001 27.108 27.144 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.107 27.143 ot_scf_mini 99 9.5 0.002 0.002 25.305 25.330 multiply_cannon 2055 13.4 0.335 0.376 23.901 24.685 multiply_cannon_loop 2055 14.4 0.829 0.852 22.098 22.410 init_scf_loop 11 6.9 0.000 0.000 20.439 20.442 prepare_preconditioner 11 7.9 0.000 0.000 18.269 18.276 make_preconditioner 11 8.9 0.000 0.000 18.269 18.276 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.462 17.787 ot_mini 99 10.5 0.001 0.001 14.623 14.650 rebuild_ks_matrix 110 8.3 0.000 0.001 13.465 13.500 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.017 13.464 13.500 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.229 12.261 cp_fm_upper_to_full 70 14.2 8.233 11.660 8.233 11.660 multiply_cannon_multrec 8220 15.4 4.258 4.468 10.638 10.811 qs_ot_get_derivative 99 11.5 0.001 0.001 10.043 10.068 mp_waitall_1 84994 16.7 7.649 8.609 7.649 8.609 dbcsr_complete_redistribute 325 12.2 0.939 0.964 5.346 7.366 init_scf_run 11 5.9 0.000 0.001 6.973 6.973 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.972 6.973 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.835 6.873 calculate_rho_elec 110 8.6 0.227 0.227 6.834 6.872 sum_up_and_integrate 110 10.3 0.002 0.002 6.563 6.581 integrate_v_rspace 110 11.3 0.004 0.004 6.535 6.553 dbcsr_mm_accdrv_process 11614 15.7 3.855 4.158 6.232 6.471 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.339 6.355 make_m2s 4110 13.4 0.043 0.044 5.353 5.850 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.791 5.805 make_images 4110 14.4 0.888 0.941 5.164 5.661 qs_ot_get_p 110 10.4 0.015 0.026 5.491 5.544 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.015 5.516 mp_alltoall_i22 605 13.7 3.425 5.424 3.425 5.424 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.738 5.193 apply_single 110 13.6 0.000 0.000 4.738 5.192 ot_diis_step 99 11.5 0.015 0.015 4.563 4.563 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.481 4.495 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 3.911 3.925 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.378 3.849 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.791 3.798 cp_fm_cholesky_invert 11 10.9 3.759 3.769 3.759 3.769 grid_integrate_task_list 110 12.3 3.670 3.731 3.670 3.731 qs_energies_init_hamiltonians 11 5.9 0.015 0.036 3.689 3.690 calculate_first_density_matrix 1 7.0 0.000 0.000 3.688 3.689 calculate_dm_sparse 110 9.5 0.001 0.001 3.594 3.617 hybrid_alltoall_any 4261 16.3 0.264 0.562 2.806 3.563 make_images_data 4110 15.4 0.045 0.047 2.802 3.431 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.375 3.378 fft3d_ps 1111 13.6 1.841 1.871 3.339 3.355 density_rs2pw 110 9.6 0.004 0.004 3.316 3.329 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.275 3.276 wfi_extrapolate 11 7.9 0.001 0.001 3.175 3.175 multiply_cannon_sync_h2d 8220 15.4 3.114 3.134 3.114 3.134 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.893 2.906 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.665 2.678 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.673 2.673 cp_fm_diag_elpa_base 48 14.0 2.372 2.447 2.671 2.671 grid_collocate_task_list 110 9.6 2.660 2.669 2.660 2.669 jit_kernel_multiply 10 15.2 2.182 2.432 2.182 2.432 acc_transpose_blocks 8220 15.4 0.041 0.041 2.139 2.350 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.178 2.241 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.106 2.199 potential_pw2rs 110 12.3 0.021 0.021 2.108 2.112 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.093 2.108 mp_alltoall_d11v 2046 13.8 2.017 2.064 2.017 2.064 qs_create_task_list 11 7.9 0.000 0.000 1.884 1.925 generate_qs_task_list 11 8.9 0.732 0.787 1.884 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.837 1.841 cp_fm_cholesky_decompose 22 10.9 1.740 1.767 1.740 1.767 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.722 1.742 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.622 1.631 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=80.420000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1251.000000, yerr=67.268796 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/15/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 1.094965E+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 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 630.951936E+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 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4027 66905. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 3206174. 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 0.055 0.077 205.577 205.578 qs_mol_dyn_low 1 2.0 0.003 0.004 204.807 204.821 qs_forces 11 3.9 0.023 0.035 204.674 204.676 qs_energies 11 4.9 0.002 0.003 198.926 198.964 scf_env_do_scf 11 5.9 0.001 0.001 182.072 182.088 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.201 163.203 dbcsr_multiply_generic 2507 12.6 0.177 0.181 126.587 127.421 qs_scf_new_mos 117 7.6 0.001 0.001 124.201 124.461 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.200 124.460 velocity_verlet 10 3.0 0.008 0.046 121.879 121.880 ot_scf_mini 117 9.6 0.003 0.003 117.388 117.607 multiply_cannon 2507 13.6 0.238 0.248 102.437 104.858 multiply_cannon_loop 2507 14.6 2.369 2.429 100.277 102.486 ot_mini 117 10.6 0.001 0.001 66.662 66.849 multiply_cannon_multrec 60168 15.6 31.938 33.510 41.928 43.593 qs_ot_get_derivative 117 11.6 0.001 0.001 41.740 41.947 rebuild_ks_matrix 128 8.3 0.001 0.001 33.701 34.068 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.700 34.068 mp_waitall_1 267128 16.5 29.290 33.032 29.290 33.032 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.290 30.612 multiply_cannon_sync_h2d 60168 15.6 26.407 28.658 26.407 28.658 qs_ot_get_p 128 10.4 0.001 0.001 28.369 28.618 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.500 25.141 apply_single 128 13.6 0.001 0.001 24.499 25.140 ot_diis_step 117 11.6 0.007 0.008 24.658 24.660 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.846 21.943 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.977 20.136 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.993 18.994 init_scf_loop 11 6.9 0.000 0.001 18.788 18.789 multiply_cannon_metrocomm3 60168 15.6 0.119 0.125 16.367 18.695 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.991 16.026 cp_fm_diag_elpa_base 83 14.4 15.943 15.978 15.985 16.019 prepare_preconditioner 11 7.9 0.000 0.000 14.146 14.211 make_preconditioner 11 8.9 0.000 0.000 14.146 14.211 make_m2s 5014 13.6 0.105 0.114 13.783 14.200 make_images 5014 14.6 0.400 0.418 13.598 14.027 sum_up_and_integrate 128 10.3 0.002 0.004 13.696 13.712 integrate_v_rspace 128 11.3 0.004 0.004 13.636 13.655 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.425 13.633 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.546 12.681 calculate_rho_elec 128 8.7 0.045 0.064 12.545 12.680 init_scf_run 11 5.9 0.000 0.002 12.640 12.641 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.640 12.640 mp_sum_l 7950 12.9 9.107 10.556 9.107 10.556 dbcsr_mm_accdrv_process 124484 16.2 4.705 4.897 9.561 10.034 calculate_dm_sparse 128 9.5 0.001 0.001 8.792 8.940 wfi_extrapolate 11 7.9 0.001 0.001 8.902 8.903 multiply_cannon_metrocomm1 60168 15.6 0.092 0.096 6.474 8.784 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.195 8.376 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.152 8.308 cp_fm_cholesky_invert 11 10.9 8.083 8.101 8.083 8.101 make_images_data 5014 15.6 0.066 0.071 6.756 7.750 grid_integrate_task_list 128 12.3 7.036 7.394 7.036 7.394 hybrid_alltoall_any 5200 16.5 0.296 2.276 5.928 7.173 density_rs2pw 128 9.7 0.005 0.006 6.162 6.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.822 6.834 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.387 6.657 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.902 5.995 fft_wrap_pw1pw2_140 523 12.2 0.233 0.244 5.554 5.729 fft3d_ps 1291 13.7 2.093 2.596 5.211 5.420 mp_alltoall_d11v 2415 14.1 4.474 5.420 4.474 5.420 grid_collocate_task_list 128 9.7 4.844 5.263 4.844 5.263 mp_sum_d 4483 12.1 3.597 4.238 3.597 4.238 potential_pw2rs 128 12.3 0.009 0.010 4.214 4.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.578000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.727273, yerr=5.971004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 847.626240E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67821. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6543719. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.098 0.233 186.151 186.152 qs_mol_dyn_low 1 2.0 0.006 0.029 185.246 185.262 qs_forces 11 3.9 0.005 0.012 184.720 184.722 qs_energies 11 4.9 0.276 0.488 177.924 177.938 scf_env_do_scf 11 5.9 0.003 0.022 159.665 159.682 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.962 132.976 velocity_verlet 10 3.0 0.060 0.412 112.820 112.825 dbcsr_multiply_generic 2507 12.6 0.198 0.216 99.064 100.302 qs_scf_new_mos 117 7.6 0.001 0.001 94.910 95.422 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.909 95.421 ot_scf_mini 117 9.6 0.004 0.004 90.105 90.682 multiply_cannon 2507 13.6 0.506 0.567 78.738 83.398 multiply_cannon_loop 2507 14.6 1.562 1.627 75.353 78.300 ot_mini 117 10.6 0.001 0.001 51.326 51.893 mp_waitall_1 214728 16.6 24.046 39.799 24.046 39.799 multiply_cannon_multrec 30084 15.6 20.893 26.313 32.604 38.440 rebuild_ks_matrix 128 8.3 0.001 0.001 32.343 33.114 qs_ks_build_kohn_sham_matrix 128 9.3 0.035 0.109 32.342 33.114 qs_ot_get_derivative 117 11.6 0.001 0.002 29.341 29.910 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.089 29.798 multiply_cannon_metrocomm3 30084 15.6 0.095 0.100 15.805 29.409 init_scf_loop 11 6.9 0.001 0.004 26.590 26.600 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.088 23.216 apply_single 128 13.6 0.001 0.001 22.087 23.216 qs_ot_get_p 128 10.4 0.003 0.018 21.299 21.939 prepare_preconditioner 11 7.9 0.000 0.000 21.719 21.807 make_preconditioner 11 8.9 0.000 0.001 21.718 21.807 ot_diis_step 117 11.6 0.014 0.015 21.801 21.802 multiply_cannon_sync_h2d 30084 15.6 17.947 21.319 17.947 21.319 make_full_inverse_cholesky 11 9.9 0.000 0.001 20.579 20.959 qs_ot_p2m_diag 83 11.4 0.188 0.217 15.866 15.903 make_m2s 5014 13.6 0.088 0.093 14.190 15.868 make_images 5014 14.6 1.176 1.377 13.975 15.650 cp_dbcsr_syevd 83 12.4 0.010 0.039 14.674 14.674 sum_up_and_integrate 128 10.3 0.002 0.004 13.644 13.678 integrate_v_rspace 128 11.3 0.003 0.004 13.584 13.620 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.997 13.041 calculate_rho_elec 128 8.7 0.088 0.106 12.997 13.040 init_scf_run 11 5.9 0.000 0.001 12.808 12.810 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.808 12.810 cp_fm_cholesky_invert 11 10.9 11.879 11.915 11.879 11.915 dbcsr_mm_accdrv_process 62242 16.2 5.740 6.713 11.161 11.771 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.320 11.755 cp_fm_diag_elpa 83 13.4 0.001 0.004 11.195 11.229 cp_fm_diag_elpa_base 83 14.4 10.993 11.029 11.158 11.185 multiply_cannon_metrocomm4 27577 15.6 0.106 0.118 3.883 10.921 mp_irecv_dv 69486 16.3 3.671 10.512 3.671 10.512 make_images_data 5014 15.6 0.066 0.075 8.289 10.441 hybrid_alltoall_any 5200 16.5 0.359 1.546 7.312 10.059 wfi_extrapolate 11 7.9 0.001 0.001 8.284 8.285 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.432 7.537 grid_integrate_task_list 128 12.3 7.177 7.536 7.177 7.536 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.255 7.274 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.424 7.262 density_rs2pw 128 9.7 0.006 0.007 6.568 6.976 calculate_dm_sparse 128 9.5 0.001 0.001 6.596 6.733 fft_wrap_pw1pw2_140 523 12.2 0.249 0.260 6.444 6.524 mp_sum_l 7950 12.9 4.373 6.402 4.373 6.402 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.156 6.339 fft3d_ps 1291 13.7 2.763 2.941 5.885 5.968 cp_fm_cholesky_decompose 22 10.9 5.623 5.705 5.623 5.705 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.374 5.450 grid_collocate_task_list 128 9.7 4.998 5.365 4.998 5.365 mp_alltoall_d11v 2415 14.1 4.864 5.342 4.864 5.342 mp_allgather_i34 2507 14.6 1.894 4.890 1.894 4.890 mp_sum_d 4469 12.1 2.822 4.401 2.822 4.401 calculate_first_density_matrix 1 7.0 0.002 0.019 4.203 4.205 potential_pw2rs 128 12.3 0.014 0.016 4.137 4.154 dbcsr_complete_redistribute 395 12.7 0.738 0.909 3.302 3.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.668 3.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=186.152000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=808.363636, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.033970E+09 # 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 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67815. MP_Allreduce 11057 1000. 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 0.010 0.032 169.317 169.318 qs_mol_dyn_low 1 2.0 0.003 0.004 168.062 168.075 qs_forces 11 3.9 0.004 0.006 167.895 167.897 qs_energies 11 4.9 0.003 0.009 161.381 161.394 scf_env_do_scf 11 5.9 0.001 0.001 144.540 144.541 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 115.951 115.952 velocity_verlet 10 3.0 0.023 0.046 105.473 105.477 dbcsr_multiply_generic 2507 12.6 0.184 0.188 83.094 84.270 qs_scf_new_mos 117 7.6 0.001 0.001 79.449 79.760 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.448 79.759 ot_scf_mini 117 9.6 0.003 0.004 75.246 75.643 multiply_cannon 2507 13.6 0.509 0.542 63.465 68.683 multiply_cannon_loop 2507 14.6 1.132 1.188 60.453 63.287 ot_mini 117 10.6 0.001 0.001 42.431 42.817 mp_waitall_1 170520 16.6 24.767 34.286 24.767 34.286 rebuild_ks_matrix 128 8.3 0.001 0.001 30.283 30.733 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 30.283 30.732 init_scf_loop 11 6.9 0.000 0.000 28.489 28.490 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.340 27.755 multiply_cannon_multrec 20056 15.6 13.044 15.867 22.925 25.755 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 15.441 24.862 prepare_preconditioner 11 7.9 0.000 0.000 24.423 24.467 make_preconditioner 11 8.9 0.000 0.000 24.423 24.467 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.489 23.412 qs_ot_get_derivative 117 11.6 0.001 0.002 22.960 23.346 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.561 20.542 apply_single 128 13.6 0.001 0.001 19.560 20.541 ot_diis_step 117 11.6 0.018 0.018 19.363 19.363 qs_ot_get_p 128 10.4 0.001 0.002 18.200 18.687 make_m2s 5014 13.6 0.078 0.084 14.580 15.779 multiply_cannon_sync_h2d 20056 15.6 13.673 15.548 13.673 15.548 make_images 5014 14.6 1.138 1.236 14.344 15.538 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.921 13.929 sum_up_and_integrate 128 10.3 0.002 0.003 13.576 13.600 integrate_v_rspace 128 11.3 0.003 0.004 13.515 13.541 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.870 12.907 calculate_rho_elec 128 8.7 0.132 0.146 12.870 12.907 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.877 12.877 cp_fm_cholesky_invert 11 10.9 11.666 11.686 11.666 11.686 init_scf_run 11 5.9 0.000 0.001 11.679 11.679 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.678 11.679 make_images_data 5014 15.6 0.062 0.070 8.663 10.261 hybrid_alltoall_any 5200 16.5 0.451 2.036 7.542 10.023 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.624 9.642 cp_fm_diag_elpa_base 83 14.4 9.381 9.451 9.620 9.639 dbcsr_mm_accdrv_process 41502 16.2 6.164 6.663 9.332 9.581 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.058 9.334 multiply_cannon_metrocomm4 17549 15.6 0.066 0.076 3.487 9.040 mp_irecv_dv 50230 16.2 3.355 8.779 3.355 8.779 grid_integrate_task_list 128 12.3 7.311 7.728 7.311 7.728 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.484 7.600 wfi_extrapolate 11 7.9 0.001 0.001 7.358 7.359 density_rs2pw 128 9.7 0.005 0.006 6.334 7.005 fft_wrap_pw1pw2_140 523 12.2 0.289 0.296 6.610 6.729 calculate_dm_sparse 128 9.5 0.001 0.001 6.478 6.572 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.076 6.083 cp_fm_cholesky_decompose 22 10.9 5.905 5.982 5.905 5.982 fft3d_ps 1291 13.7 2.781 3.031 5.778 5.877 grid_collocate_task_list 128 9.7 5.145 5.492 5.145 5.492 dbcsr_complete_redistribute 395 12.7 1.041 1.109 4.220 5.364 mp_alltoall_d11v 2415 14.1 4.455 5.343 4.455 5.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.179 5.310 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.611 5.268 mp_sum_l 7950 12.9 3.214 4.915 3.214 4.915 cp_fm_upper_to_full 105 14.8 3.913 4.879 3.913 4.879 mp_allgather_i34 2507 14.6 1.475 4.550 1.475 4.550 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.073 4.225 calculate_first_density_matrix 1 7.0 0.000 0.000 4.136 4.139 potential_pw2rs 128 12.3 0.020 0.022 4.005 4.012 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.900 3.940 qs_energies_init_hamiltonians 11 5.9 0.091 0.133 3.898 3.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.453 3.496 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=169.318000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=981.727273, yerr=4.047242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 1.112740E+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 3999 67812. 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 0.036 0.055 178.286 178.287 qs_mol_dyn_low 1 2.0 0.011 0.063 177.483 177.707 qs_forces 11 3.9 0.023 0.035 177.011 177.022 qs_energies 11 4.9 0.002 0.002 169.806 169.829 scf_env_do_scf 11 5.9 0.001 0.001 152.214 152.231 velocity_verlet 10 3.0 0.015 0.061 115.272 115.288 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.139 115.140 dbcsr_multiply_generic 2507 12.6 0.205 0.211 81.447 82.115 qs_scf_new_mos 117 7.6 0.001 0.001 79.593 79.948 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.592 79.947 ot_scf_mini 117 9.6 0.003 0.003 75.042 75.359 multiply_cannon 2507 13.6 0.551 0.581 56.506 59.128 multiply_cannon_loop 2507 14.6 1.795 1.907 52.948 54.648 ot_mini 117 10.6 0.001 0.001 42.870 43.180 init_scf_loop 11 6.9 0.000 0.000 36.952 36.953 prepare_preconditioner 11 7.9 0.000 0.000 32.864 32.894 make_preconditioner 11 8.9 0.000 0.000 32.864 32.894 multiply_cannon_multrec 30084 15.6 13.724 18.610 27.546 32.117 make_full_inverse_cholesky 11 9.9 0.017 0.022 28.076 31.674 rebuild_ks_matrix 128 8.3 0.001 0.001 29.182 29.456 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.181 29.456 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.335 26.573 mp_waitall_1 147882 16.7 17.184 26.407 17.184 26.407 qs_ot_get_derivative 117 11.6 0.001 0.002 23.159 23.479 make_m2s 5014 13.6 0.093 0.097 20.651 21.624 make_images 5014 14.6 1.989 2.240 20.343 21.317 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.254 19.805 apply_single 128 13.6 0.001 0.001 19.254 19.805 ot_diis_step 117 11.6 0.017 0.020 19.581 19.582 qs_ot_get_p 128 10.4 0.001 0.002 18.253 18.628 multiply_cannon_metrocomm3 30084 15.6 0.049 0.053 6.237 14.510 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.194 14.248 dbcsr_mm_accdrv_process 62264 16.2 8.573 9.470 13.387 13.895 sum_up_and_integrate 128 10.3 0.002 0.003 13.548 13.574 integrate_v_rspace 128 11.3 0.003 0.004 13.488 13.517 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.061 13.115 calculate_rho_elec 128 8.7 0.174 0.190 13.060 13.114 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.791 12.793 make_images_data 5014 15.6 0.064 0.070 11.105 12.495 cp_fm_upper_to_full 105 14.8 8.786 12.494 8.786 12.494 cp_fm_cholesky_invert 11 10.9 12.038 12.060 12.038 12.060 init_scf_run 11 5.9 0.000 0.001 11.712 11.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.712 11.713 hybrid_alltoall_any 5200 16.5 0.529 2.221 10.035 11.643 multiply_cannon_sync_h2d 30084 15.6 10.429 11.106 10.429 11.106 dbcsr_complete_redistribute 395 12.7 1.408 1.507 7.652 10.488 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.720 9.956 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.479 9.501 cp_fm_diag_elpa_base 83 14.4 8.996 9.134 9.472 9.488 copy_fm_to_dbcsr 209 11.7 0.001 0.002 6.281 9.104 grid_integrate_task_list 128 12.3 7.470 7.790 7.470 7.790 fft_wrap_pw1pw2 1291 11.7 0.018 0.024 7.699 7.759 transfer_fm_to_dbcsr 11 9.9 0.003 0.011 4.767 7.541 wfi_extrapolate 11 7.9 0.001 0.001 7.399 7.399 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.785 7.308 mp_irecv_dv 76098 16.2 2.631 7.025 2.631 7.025 fft_wrap_pw1pw2_140 523 12.2 0.325 0.335 6.818 6.897 mp_alltoall_i22 716 14.1 4.120 6.897 4.120 6.897 density_rs2pw 128 9.7 0.005 0.005 6.162 6.761 calculate_dm_sparse 128 9.5 0.001 0.001 6.467 6.541 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.126 6.159 cp_fm_cholesky_decompose 22 10.9 5.988 6.037 5.988 6.037 fft3d_ps 1291 13.7 2.941 3.007 5.895 5.954 grid_collocate_task_list 128 9.7 5.292 5.614 5.292 5.614 mp_alltoall_d11v 2415 14.1 4.994 5.542 4.994 5.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.028 5.104 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.460 4.543 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.411 4.427 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.254 4.325 calculate_first_density_matrix 1 7.0 0.000 0.000 4.118 4.120 potential_pw2rs 128 12.3 0.022 0.022 3.882 3.891 dbcsr_special_finalize 12535 15.6 0.066 0.068 3.521 3.696 dbcsr_merge_single_wm 12535 16.6 0.324 0.338 3.430 3.604 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=178.287000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1048.000000, yerr=15.521246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 1.442148E+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 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 759871. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 67759. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828162. 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 0.012 0.031 165.350 165.351 qs_mol_dyn_low 1 2.0 0.003 0.010 164.735 164.766 qs_forces 11 3.9 0.008 0.009 164.629 164.633 qs_energies 11 4.9 0.002 0.002 157.143 157.153 scf_env_do_scf 11 5.9 0.001 0.001 138.547 138.559 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.618 107.620 velocity_verlet 10 3.0 0.011 0.021 105.318 105.322 dbcsr_multiply_generic 2485 12.5 0.212 0.219 73.173 73.615 qs_scf_new_mos 116 7.6 0.001 0.001 71.799 71.923 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.798 71.923 ot_scf_mini 116 9.6 0.003 0.003 67.420 67.545 multiply_cannon 2485 13.5 0.555 0.567 54.435 57.428 multiply_cannon_loop 2485 14.5 0.799 0.827 51.311 52.406 ot_mini 116 10.6 0.001 0.001 37.912 38.027 mp_waitall_1 124680 16.7 24.636 31.076 24.636 31.076 init_scf_loop 11 6.9 0.000 0.000 30.767 30.769 rebuild_ks_matrix 127 8.3 0.001 0.001 28.955 29.108 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.955 29.107 prepare_preconditioner 11 7.9 0.000 0.000 26.687 26.708 make_preconditioner 11 8.9 0.000 0.000 26.687 26.708 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.423 26.556 make_full_inverse_cholesky 11 9.9 0.022 0.028 24.808 25.056 multiply_cannon_multrec 9940 15.5 10.226 15.926 18.740 23.814 qs_ot_get_derivative 116 11.6 0.001 0.002 20.822 20.949 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.767 20.512 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.532 17.780 apply_single 127 13.6 0.001 0.001 17.532 17.780 qs_ot_get_p 127 10.4 0.002 0.002 16.919 17.047 ot_diis_step 116 11.6 0.019 0.022 17.015 17.015 make_m2s 4970 13.5 0.063 0.069 14.529 15.445 make_images 4970 14.5 2.143 2.564 14.226 15.133 cp_fm_cholesky_invert 11 10.9 14.691 14.702 14.691 14.702 sum_up_and_integrate 127 10.3 0.002 0.002 13.655 13.698 integrate_v_rspace 127 11.3 0.004 0.004 13.595 13.639 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.313 13.344 calculate_rho_elec 127 8.7 0.255 0.266 13.313 13.343 qs_ot_p2m_diag 82 11.4 0.490 0.495 13.218 13.234 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.067 12.068 init_scf_run 11 5.9 0.000 0.001 11.795 11.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.794 11.795 multiply_cannon_sync_h2d 9940 15.5 10.621 11.157 10.621 11.157 make_images_data 4970 15.5 0.055 0.065 8.308 9.697 hybrid_alltoall_any 5155 16.4 0.841 3.739 8.130 9.328 dbcsr_mm_accdrv_process 20590 16.1 3.357 5.045 8.171 9.260 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.177 9.188 cp_fm_diag_elpa_base 82 14.4 8.946 9.021 9.173 9.184 grid_integrate_task_list 127 12.3 7.708 8.122 7.708 8.122 fft_wrap_pw1pw2 1281 11.7 0.017 0.019 8.016 8.043 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.928 8.024 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.331 7.656 wfi_extrapolate 11 7.9 0.001 0.001 7.145 7.145 fft_wrap_pw1pw2_140 519 12.2 0.430 0.436 7.060 7.090 calculate_dm_sparse 127 9.5 0.001 0.001 6.613 6.698 cp_fm_cholesky_decompose 22 10.9 6.482 6.526 6.482 6.526 density_rs2pw 127 9.7 0.005 0.005 6.099 6.395 fft3d_ps 1281 13.7 3.091 3.198 6.001 6.026 grid_collocate_task_list 127 9.7 5.517 5.822 5.517 5.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.735 5.777 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.858 5.682 mp_irecv_dv 28618 15.9 1.817 5.592 1.817 5.592 dbcsr_complete_redistribute 393 12.7 2.000 2.036 5.095 5.422 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.252 5.253 mp_alltoall_d11v 2401 14.1 4.614 5.122 4.614 5.122 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.609 4.648 mp_allgather_i34 2485 14.5 1.319 4.592 1.319 4.592 calculate_first_density_matrix 1 7.0 0.000 0.000 4.422 4.424 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.734 4.114 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.597 3.902 potential_pw2rs 127 12.3 0.026 0.027 3.897 3.902 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.609 3.668 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.480 3.566 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.234 3.560 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.430 3.450 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=165.351000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1344.636364, yerr=30.314331 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.032797E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4047 67952. MP_Allreduce 11177 1500. MP_Sync 87 MP_Alltoall 1724 36993717. 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 0.019 0.042 246.287 246.288 qs_mol_dyn_low 1 2.0 0.003 0.004 245.704 245.720 qs_forces 11 3.9 0.004 0.005 245.569 245.572 qs_energies 11 4.9 0.002 0.003 236.881 236.889 scf_env_do_scf 11 5.9 0.001 0.001 213.306 213.314 velocity_verlet 10 3.0 0.039 0.045 170.382 170.390 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 130.931 130.932 qs_scf_new_mos 118 7.6 0.001 0.001 88.076 88.135 qs_scf_loop_do_ot 118 8.6 0.001 0.001 88.075 88.134 dbcsr_multiply_generic 2527 12.6 0.219 0.229 83.760 84.153 ot_scf_mini 118 9.6 0.004 0.004 83.059 83.100 init_scf_loop 11 6.9 0.000 0.000 82.080 82.083 prepare_preconditioner 11 7.9 0.000 0.000 77.291 77.314 make_preconditioner 11 8.9 0.000 0.000 77.290 77.314 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.568 74.560 multiply_cannon 2527 13.6 0.690 0.719 60.507 61.457 multiply_cannon_loop 2527 14.6 1.049 1.067 56.536 57.667 cp_fm_upper_to_full 105 14.8 34.084 48.177 34.084 48.177 ot_mini 118 10.6 0.001 0.001 44.608 44.651 rebuild_ks_matrix 129 8.3 0.001 0.001 33.809 33.851 qs_ks_build_kohn_sham_matrix 129 9.3 0.047 0.064 33.808 33.850 mp_waitall_1 104500 16.8 28.135 32.409 28.135 32.409 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.486 31.524 dbcsr_complete_redistribute 395 12.7 3.849 3.979 21.795 29.523 copy_fm_to_dbcsr 209 11.7 0.001 0.001 18.222 26.007 qs_ot_get_derivative 118 11.6 0.002 0.002 24.855 24.897 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 15.680 23.381 qs_ot_get_p 129 10.4 0.002 0.002 23.320 23.364 mp_alltoall_i22 718 14.1 13.454 21.483 13.454 21.483 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 19.309 20.422 make_m2s 5054 13.6 0.075 0.078 18.297 19.731 ot_diis_step 118 11.6 0.022 0.022 19.720 19.720 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.367 19.512 apply_single 129 13.6 0.001 0.001 19.366 19.511 qs_ot_p2m_diag 83 11.4 0.880 0.884 19.323 19.352 make_images 5054 14.6 3.103 3.312 17.812 19.248 multiply_cannon_multrec 10108 15.6 10.331 12.152 18.983 19.052 cp_fm_cholesky_invert 11 10.9 18.328 18.341 18.328 18.341 cp_dbcsr_syevd 83 12.4 0.006 0.006 17.530 17.531 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.880 15.900 calculate_rho_elec 129 8.7 0.486 0.487 15.879 15.899 sum_up_and_integrate 129 10.3 0.002 0.002 15.756 15.846 integrate_v_rspace 129 11.3 0.005 0.005 15.693 15.782 multiply_cannon_sync_h2d 10108 15.6 14.425 14.454 14.425 14.454 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.371 14.372 cp_fm_diag_elpa_base 83 14.4 11.545 12.541 14.367 14.367 init_scf_run 11 5.9 0.000 0.001 13.129 13.130 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.129 13.130 hybrid_alltoall_any 5240 16.5 1.325 3.096 10.333 12.401 make_images_data 5054 15.6 0.063 0.068 10.138 12.234 dbcsr_mm_accdrv_process 20926 16.1 4.426 6.155 8.399 10.280 fft_wrap_pw1pw2 1301 11.7 0.020 0.021 10.008 10.019 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.640 9.674 fft_wrap_pw1pw2_140 527 12.2 0.830 0.832 8.863 8.884 wfi_extrapolate 11 7.9 0.001 0.001 8.837 8.837 grid_integrate_task_list 129 12.3 8.615 8.817 8.615 8.817 qs_energies_init_hamiltonians 11 5.9 0.064 0.099 8.229 8.230 cp_fm_cholesky_decompose 22 10.9 7.402 7.447 7.402 7.447 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.327 7.401 mp_alltoall_d11v 2423 14.1 7.181 7.334 7.181 7.334 fft3d_ps 1301 13.7 4.017 4.036 7.289 7.306 calculate_dm_sparse 129 9.5 0.001 0.001 6.971 7.029 density_rs2pw 129 9.7 0.005 0.005 6.860 6.922 grid_collocate_task_list 129 9.7 6.457 6.546 6.457 6.546 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.244 6.437 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.912 5.634 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.270 5.310 qs_env_update_s_mstruct 11 6.9 0.000 0.000 4.968 5.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=246.288000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2683.545455, yerr=166.794574 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 1.280201E+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 1033 290947. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 33294403. 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 0.129 0.593 87.004 87.006 qs_energies 1 2.0 0.002 0.017 85.401 85.459 ls_scf 1 3.0 0.002 0.017 84.159 84.220 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.586 72.761 multiply_cannon 111 7.7 0.017 0.020 55.963 56.876 multiply_cannon_loop 111 8.7 0.229 0.251 52.564 53.665 ls_scf_main 1 4.0 0.004 0.032 52.187 52.200 density_matrix_trs4 2 5.0 0.002 0.003 46.650 46.715 ls_scf_init_scf 1 4.0 0.000 0.001 28.844 28.859 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.365 27.417 matrix_sqrt_Newton_Schulz 2 6.5 0.036 0.092 25.235 25.253 mp_waitall_1 11031 10.9 22.282 24.264 22.282 24.264 multiply_cannon_multrec 2664 9.7 8.157 8.881 15.545 17.351 multiply_cannon_sync_h2d 2664 9.7 13.661 15.181 13.661 15.181 make_m2s 222 7.7 0.008 0.011 13.048 13.571 make_images 222 8.7 0.099 0.110 13.026 13.550 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.594 12.834 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.452 8.361 make_images_data 222 9.7 0.004 0.005 7.579 8.160 hybrid_alltoall_any 227 10.6 0.216 1.825 6.544 8.137 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.722 7.005 8.034 dbcsr_mm_accdrv_process_sort 4760 11.4 6.201 7.106 6.201 7.106 calculate_norms 4752 9.8 5.514 6.102 5.514 6.102 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.034 5.184 mp_sum_l 887 5.1 3.078 4.459 3.078 4.459 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.057 3.857 mp_irecv_dv 6231 10.9 2.041 3.821 2.041 3.821 make_images_sizes 222 9.7 0.000 0.000 0.791 3.529 mp_alltoall_i44 222 10.7 0.791 3.529 0.791 3.529 arnoldi_extremal 4 6.8 0.000 0.000 3.271 3.326 arnoldi_normal_ev 4 7.8 0.015 0.054 3.271 3.326 ls_scf_post 1 4.0 0.005 0.066 3.126 3.188 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.251 3.179 build_subspace 16 8.4 0.009 0.012 3.080 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.313 2.713 dbcsr_merge_single_wm 555 10.7 0.455 0.592 2.304 2.704 make_images_pack 222 9.7 2.213 2.627 2.215 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.297 2.534 dbcsr_matrix_vector_mult_local 304 10.0 2.059 2.445 2.061 2.446 dbcsr_sort_data 658 11.4 2.107 2.446 2.107 2.446 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.297 2.387 buffer_matrices_ensure_size 222 8.7 1.745 2.073 1.745 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.045 rebuild_ks_matrix 3 7.3 0.000 0.000 2.021 2.035 qs_ks_build_kohn_sham_matrix 3 8.3 0.024 0.112 2.021 2.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.006000, 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/e39d2370f112907676398aed806e11d93bfbe80f_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 2.187678E+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 1033 301899. 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 0.047 0.064 90.038 90.039 qs_energies 1 2.0 0.000 0.000 89.587 89.593 ls_scf 1 3.0 0.000 0.000 88.022 88.028 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.983 74.297 multiply_cannon 111 7.7 0.027 0.042 52.687 56.875 ls_scf_main 1 4.0 0.000 0.000 54.251 54.281 multiply_cannon_loop 111 8.7 0.134 0.147 49.881 52.815 density_matrix_trs4 2 5.0 0.002 0.003 48.598 48.839 ls_scf_init_scf 1 4.0 0.000 0.000 30.092 30.124 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.860 28.958 mp_waitall_1 9105 10.9 20.659 28.640 20.659 28.640 multiply_cannon_multrec 1332 9.7 13.293 17.205 22.565 27.585 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.014 26.388 26.399 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.494 19.811 make_m2s 222 7.7 0.006 0.007 14.926 15.444 make_images 222 8.7 1.369 1.685 14.895 15.414 dbcsr_mm_accdrv_process 4041 10.4 0.339 0.545 8.865 10.484 dbcsr_mm_accdrv_process_sort 4041 11.4 8.382 9.939 8.382 9.939 make_images_data 222 9.7 0.004 0.005 8.666 9.552 hybrid_alltoall_any 227 10.6 0.542 2.548 8.081 9.453 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.215 7.680 mp_irecv_dv 3311 11.0 3.194 7.607 3.194 7.607 mp_sum_l 887 5.1 4.888 7.571 4.888 7.571 calculate_norms 2376 9.8 6.056 6.753 6.056 6.753 multiply_cannon_sync_h2d 1332 9.7 4.872 6.171 4.872 6.171 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.663 5.999 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.884 5.118 arnoldi_extremal 4 6.8 0.000 0.000 4.675 4.693 arnoldi_normal_ev 4 7.8 0.021 0.081 4.675 4.693 build_subspace 16 8.4 0.014 0.020 4.345 4.348 ls_scf_post 1 4.0 0.000 0.000 3.679 3.685 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.502 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.130 3.355 dbcsr_matrix_vector_mult_local 304 10.0 2.747 3.222 2.749 3.223 mp_allgather_i34 111 8.7 0.789 3.050 0.789 3.050 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.581 2.676 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.145 2.594 dbcsr_data_new 4174 10.1 2.123 2.413 2.123 2.413 make_images_pack 222 9.7 1.820 2.111 1.823 2.113 dbcsr_sort_data 436 11.2 1.801 2.022 1.801 2.022 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.924 1.957 rebuild_ks_matrix 3 7.3 0.000 0.000 1.911 1.944 qs_ks_build_kohn_sham_matrix 3 8.3 0.026 0.057 1.911 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.039000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1829.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 2.880176E+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 1033 300682. 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 0.051 0.103 93.497 93.497 qs_energies 1 2.0 0.000 0.000 92.094 92.117 ls_scf 1 3.0 0.000 0.000 90.663 90.686 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.143 75.503 ls_scf_main 1 4.0 0.000 0.000 56.464 56.481 multiply_cannon 111 7.7 0.031 0.045 51.880 56.274 multiply_cannon_loop 111 8.7 0.117 0.130 49.093 53.096 density_matrix_trs4 2 5.0 0.002 0.003 50.640 50.789 mp_waitall_1 7281 11.0 23.455 33.362 23.455 33.362 ls_scf_init_scf 1 4.0 0.000 0.000 30.571 30.582 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.264 29.350 matrix_sqrt_Newton_Schulz 2 6.5 0.014 0.036 26.859 26.878 multiply_cannon_multrec 888 9.7 12.691 15.200 21.261 24.532 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.945 23.035 make_m2s 222 7.7 0.006 0.008 16.521 17.269 make_images 222 8.7 1.587 1.858 16.483 17.228 make_images_data 222 9.7 0.004 0.004 9.617 10.725 hybrid_alltoall_any 227 10.6 0.640 2.960 9.156 10.408 dbcsr_mm_accdrv_process 3754 10.4 0.335 0.493 8.092 9.347 dbcsr_mm_accdrv_process_sort 3754 11.4 7.654 8.854 7.654 8.854 mp_sum_l 887 5.1 5.070 8.530 5.070 8.530 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.441 7.335 mp_irecv_dv 2335 11.1 2.425 7.280 2.425 7.280 multiply_cannon_sync_h2d 888 9.7 6.052 7.229 6.052 7.229 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.537 6.725 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.889 6.590 arnoldi_extremal 4 6.8 0.000 0.000 5.145 5.175 arnoldi_normal_ev 4 7.8 0.008 0.034 5.144 5.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.706 4.911 calculate_norms 1584 9.8 4.391 4.884 4.391 4.884 build_subspace 16 8.4 0.014 0.020 4.827 4.832 mp_allgather_i34 111 8.7 0.905 3.865 0.905 3.865 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.433 3.771 ls_scf_post 1 4.0 0.000 0.000 3.628 3.652 dbcsr_matrix_vector_mult_local 304 10.0 3.028 3.606 3.030 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.363 3.444 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.813 2.919 dbcsr_data_new 4116 9.9 2.111 2.457 2.111 2.457 dbcsr_sort_data 325 11.1 1.875 2.141 1.875 2.141 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.026 2.038 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.020 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.025 2.008 2.020 make_images_sizes 222 9.7 0.000 0.000 1.047 2.011 mp_alltoall_i44 222 10.7 1.047 2.010 1.047 2.010 dbcsr_finalize 304 7.8 0.026 0.032 1.597 1.901 make_images_pack 222 9.7 1.636 1.886 1.639 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.497000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2294.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 3.406287E+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 1033 301899. 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 0.081 0.142 96.928 96.928 qs_energies 1 2.0 0.000 0.000 96.118 96.136 ls_scf 1 3.0 0.000 0.000 94.453 94.471 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.077 78.360 ls_scf_main 1 4.0 0.000 0.000 58.447 58.449 multiply_cannon 111 7.7 0.059 0.100 51.354 56.021 density_matrix_trs4 2 5.0 0.002 0.003 52.365 52.507 multiply_cannon_loop 111 8.7 0.152 0.168 46.274 48.652 ls_scf_init_scf 1 4.0 0.000 0.000 32.778 32.780 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.577 31.633 matrix_sqrt_Newton_Schulz 2 6.5 0.034 0.059 29.061 29.073 mp_waitall_1 6369 11.0 22.415 29.013 22.415 29.013 multiply_cannon_multrec 1332 9.7 14.119 17.300 22.026 25.141 make_m2s 222 7.7 0.007 0.008 21.142 22.592 make_images 222 8.7 3.155 3.616 21.092 22.544 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.906 16.658 make_images_data 222 9.7 0.004 0.004 11.778 13.445 hybrid_alltoall_any 227 10.6 0.796 3.785 11.159 12.954 dbcsr_mm_accdrv_process 3641 10.4 0.279 0.482 7.540 9.064 dbcsr_mm_accdrv_process_sort 3641 11.4 7.098 8.562 7.098 8.562 mp_sum_l 887 5.1 4.005 7.239 4.005 7.239 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.091 6.095 mp_irecv_dv 3229 10.9 2.066 6.021 2.066 6.021 multiply_cannon_sync_h2d 1332 9.7 5.437 5.920 5.437 5.920 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.054 5.700 arnoldi_extremal 4 6.8 0.000 0.000 5.325 5.371 arnoldi_normal_ev 4 7.8 0.016 0.041 5.325 5.371 build_subspace 16 8.4 0.014 0.021 4.956 4.966 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.559 4.910 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.551 4.714 mp_allgather_i34 111 8.7 2.191 4.709 2.191 4.709 calculate_norms 2376 9.8 4.160 4.479 4.160 4.479 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.621 3.935 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.706 3.203 3.708 dbcsr_sort_data 658 11.4 3.083 3.416 3.083 3.416 ls_scf_post 1 4.0 0.000 0.000 3.227 3.248 dbcsr_special_finalize 555 9.7 0.006 0.007 2.837 3.236 dbcsr_merge_single_wm 555 10.7 0.541 0.661 2.828 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.993 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.964 3.020 dbcsr_data_release 10477 10.7 1.576 2.414 1.576 2.414 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.928000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 4.884308E+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 1033 300769. 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 0.068 0.100 92.968 92.969 qs_energies 1 2.0 0.000 0.000 91.894 91.922 ls_scf 1 3.0 0.000 0.000 89.934 89.961 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.982 71.181 ls_scf_main 1 4.0 0.000 0.000 56.678 56.712 multiply_cannon 111 7.7 0.094 0.179 52.849 56.337 multiply_cannon_loop 111 8.7 0.088 0.096 50.267 51.645 density_matrix_trs4 2 5.0 0.002 0.003 49.781 49.847 ls_scf_init_scf 1 4.0 0.000 0.000 29.825 29.861 mp_waitall_1 5436 11.0 24.650 28.652 24.650 28.652 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.268 28.304 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.159 26.173 multiply_cannon_multrec 444 9.7 13.640 16.502 20.710 22.055 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.621 15.629 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.621 14.540 make_m2s 222 7.7 0.005 0.005 13.487 14.427 make_images 222 8.7 2.037 2.480 13.420 14.359 make_images_data 222 9.7 0.003 0.004 8.225 9.727 hybrid_alltoall_any 227 10.6 0.802 3.808 8.136 9.690 dbcsr_mm_accdrv_process 3003 10.4 0.348 0.411 6.766 7.911 multiply_cannon_sync_h2d 444 9.7 6.770 7.766 6.770 7.766 dbcsr_mm_accdrv_process_sort 3003 11.4 6.396 7.500 6.396 7.500 arnoldi_extremal 4 6.8 0.000 0.000 5.845 5.857 arnoldi_normal_ev 4 7.8 0.002 0.005 5.845 5.857 build_subspace 16 8.4 0.015 0.019 5.454 5.464 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.401 4.560 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.174 4.372 dbcsr_matrix_vector_mult_local 304 10.0 3.712 4.177 3.715 4.179 mp_sum_l 887 5.1 2.664 4.094 2.664 4.094 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.544 3.836 mp_irecv_dv 1241 11.2 1.530 3.800 1.530 3.800 mp_allgather_i34 111 8.7 1.158 3.737 1.158 3.737 calculate_norms 792 9.8 3.632 3.735 3.632 3.735 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.457 3.509 ls_scf_post 1 4.0 0.000 0.000 3.431 3.459 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.213 make_images_sizes 222 9.7 0.000 0.000 0.880 3.178 mp_alltoall_i44 222 10.7 0.880 3.178 0.880 3.178 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.804 2.952 dbcsr_finalize 304 7.8 0.061 0.077 2.202 2.292 dbcsr_data_new 4608 9.7 1.778 2.157 1.778 2.157 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.102 2.133 dbcsr_merge_all 275 8.9 0.479 0.522 2.059 2.123 rebuild_ks_matrix 3 7.3 0.000 0.000 2.069 2.099 qs_ks_build_kohn_sham_matrix 3 8.3 0.034 0.065 2.069 2.099 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.945 1.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.969000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 8.991334E+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 1033 319197. 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 0.099 0.259 109.275 109.276 qs_energies 1 2.0 0.000 0.000 107.227 107.238 ls_scf 1 3.0 0.000 0.000 104.191 104.202 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.997 78.170 ls_scf_main 1 4.0 0.000 0.000 63.679 63.697 density_matrix_trs4 2 5.0 0.002 0.003 54.888 54.964 multiply_cannon 111 7.7 0.126 0.217 51.602 53.729 multiply_cannon_loop 111 8.7 0.098 0.099 48.567 49.133 ls_scf_init_scf 1 4.0 0.000 0.000 36.672 36.679 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.147 35.160 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.717 31.725 mp_waitall_1 4527 11.1 22.864 26.690 22.864 26.690 make_m2s 222 7.7 0.005 0.005 22.689 23.797 make_images 222 8.7 3.579 3.882 22.581 23.689 multiply_cannon_multrec 444 9.7 17.853 18.464 22.937 23.519 hybrid_alltoall_any 227 10.6 1.652 3.607 12.837 15.796 make_images_data 222 9.7 0.003 0.004 13.048 15.217 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.048 13.937 multiply_cannon_sync_h2d 444 9.7 8.782 8.838 8.782 8.838 arnoldi_extremal 4 6.8 0.000 0.000 7.400 7.411 arnoldi_normal_ev 4 7.8 0.003 0.009 7.399 7.411 build_subspace 16 8.4 0.026 0.036 6.840 6.856 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.600 5.861 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.502 5.650 dbcsr_matrix_vector_mult_local 304 10.0 5.068 5.371 5.071 5.374 dbcsr_mm_accdrv_process 1814 10.4 0.674 1.057 4.907 5.319 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.904 4.993 dbcsr_mm_accdrv_process_sort 1814 11.4 4.116 4.240 4.116 4.240 ls_scf_post 1 4.0 0.000 0.000 3.840 3.850 mp_allgather_i34 111 8.7 1.136 3.689 1.136 3.689 make_images_sizes 222 9.7 0.000 0.000 1.498 3.677 mp_alltoall_i44 222 10.7 1.498 3.677 1.498 3.677 ls_scf_store_result 1 5.0 0.000 0.000 3.529 3.569 calculate_norms 792 9.8 3.239 3.311 3.239 3.311 dbcsr_finalize 304 7.8 0.082 0.089 3.075 3.161 compute_matrix_preconditioner 1 6.0 0.001 0.002 3.056 3.063 qs_energies_init_hamiltonians 1 3.0 0.010 0.033 3.006 3.006 dbcsr_merge_all 275 8.9 0.883 0.919 2.860 2.941 dbcsr_complete_redistribute 5 7.6 1.428 1.476 2.738 2.848 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.515 dbcsr_sort_data 325 11.1 2.436 2.488 2.436 2.488 dbcsr_new_transposed 4 7.5 0.249 0.274 2.362 2.382 dbcsr_data_new 6591 9.6 1.885 2.333 1.885 2.333 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.307 2.316 rebuild_ks_matrix 3 7.3 0.000 0.000 2.242 2.250 qs_ks_build_kohn_sham_matrix 3 8.3 0.019 0.026 2.242 2.250 dbcsr_frobenius_norm 74 6.6 2.054 2.131 2.187 2.218 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.204 dbcsr_add_anytype 103 7.2 0.859 0.891 2.123 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.276000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7218.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e39d2370f112907676398aed806e11d93bfbe80f_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 585.158656E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993462228. 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 0.157 0.441 266.071 266.076 qs_mol_dyn_low 1 2.0 0.007 0.062 263.722 263.756 qs_forces 5 3.8 0.004 0.006 263.552 263.553 qs_energies 5 4.8 0.004 0.124 260.240 260.291 scf_env_do_scf 5 5.8 0.000 0.002 244.418 244.419 scf_env_do_scf_inner_loop 105 6.6 0.007 0.350 216.964 216.966 qs_scf_new_mos 105 7.6 0.000 0.001 171.384 171.547 qs_scf_loop_do_ot 105 8.6 0.001 0.001 171.384 171.546 ot_scf_mini 105 9.6 0.003 0.004 161.353 161.585 dbcsr_multiply_generic 1445 12.2 0.159 0.178 136.230 136.678 multiply_cannon 1445 13.2 0.276 0.286 114.971 117.352 multiply_cannon_loop 1445 14.2 2.850 2.987 112.976 114.051 velocity_verlet 4 3.0 0.003 0.021 110.623 110.628 qs_ot_get_p 112 10.4 0.001 0.001 68.932 69.146 ot_mini 105 10.6 0.001 0.001 61.817 61.946 qs_ot_p2m_diag 40 11.0 0.020 0.030 57.161 57.247 cp_dbcsr_syevd 40 12.0 0.002 0.004 53.578 53.579 cp_fm_syevd 40 13.0 0.000 0.003 47.519 47.678 multiply_cannon_multrec 69360 15.2 29.665 35.129 40.002 45.715 mp_waitall_1 488190 16.1 36.231 44.053 36.231 44.053 cp_fm_redistribute_end 40 14.0 21.169 42.199 21.199 42.210 cp_fm_syevd_base 40 14.0 20.986 42.011 20.986 42.011 qs_ot_get_derivative 55 11.6 0.001 0.002 39.997 40.177 multiply_cannon_metrocomm3 69360 15.2 0.198 0.209 25.984 33.937 multiply_cannon_sync_h2d 69360 15.2 29.078 33.437 29.078 33.437 rebuild_ks_matrix 110 8.4 0.000 0.000 31.598 31.751 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.027 31.598 31.750 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.037 29.182 init_scf_loop 7 6.6 0.000 0.006 27.388 27.390 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.049 23.396 apply_single 62 13.6 0.000 0.000 23.049 23.396 prepare_preconditioner 7 7.6 0.000 0.000 21.848 21.882 make_preconditioner 7 8.6 0.000 0.002 21.848 21.881 ot_new_cg_direction 55 11.6 0.001 0.001 21.093 21.095 qs_rho_update_rho_low 110 7.6 0.000 0.001 20.657 21.054 calculate_rho_elec 110 8.6 0.032 0.137 20.656 21.053 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.264 15.449 density_rs2pw 110 9.6 0.007 0.217 14.086 14.488 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.263 14.359 mp_sum_l 4764 12.2 13.139 14.026 13.139 14.026 fft_wrap_pw1pw2 1425 12.5 0.018 0.025 13.615 13.900 make_full_inverse_cholesky 7 9.6 0.001 0.037 13.542 13.609 init_scf_run 5 5.8 0.001 0.021 13.168 13.169 scf_env_initial_rho_setup 5 6.8 0.024 0.040 13.167 13.169 calculate_dm_sparse 110 9.5 0.000 0.001 11.976 12.203 fft_wrap_pw1pw2_240 915 14.0 0.192 0.246 11.973 12.175 qs_vxc_create 110 10.4 0.002 0.003 11.368 11.386 dbcsr_mm_accdrv_process 154766 15.8 6.277 6.452 10.205 11.361 fft3d_pb 915 15.0 2.198 2.372 11.005 11.212 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.910 11.008 check_diag 80 13.5 8.653 8.925 10.448 10.603 transfer_rs2pw 445 10.6 0.008 0.010 10.023 10.448 sum_up_and_integrate 60 10.3 0.001 0.003 9.713 9.730 integrate_v_rspace 60 11.3 0.002 0.002 9.696 9.713 cp_fm_cholesky_invert 7 10.6 8.935 8.952 8.935 8.952 mp_alltoall_z22v 2340 16.7 8.310 8.587 8.310 8.587 make_m2s 2890 13.2 0.078 0.085 7.879 8.461 make_images 2890 14.2 0.242 0.262 7.772 8.354 xc_rho_set_and_dset_create 110 12.4 0.072 0.077 7.921 8.170 calculate_first_density_matrix 1 7.0 0.000 0.007 8.120 8.158 make_full_single_inverse 7 9.6 0.012 0.110 7.922 7.956 xc_vxc_pw_create 60 11.3 0.036 0.040 7.674 7.696 acc_transpose_blocks 69360 15.2 0.345 0.361 7.240 7.672 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.299 7.320 xc_pw_derive 510 13.4 0.005 0.006 7.133 7.203 multiply_cannon_metrocomm1 69360 15.2 0.095 0.100 4.733 7.134 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.593 6.654 mp_alltoall_d11v 1300 13.8 5.221 6.084 5.221 6.084 potential_pw2rs 60 12.3 0.002 0.003 5.645 5.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=266.076000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.400000, yerr=0.489898 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e39d2370f112907676398aed806e11d93bfbe80f Summary: empty Status: OK