=== 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: d3f223ab808785118aa8f0a47e0c6d072683d8a1 ################# 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 1.16.2, LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 25.01.2024 # # \ 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; \ ./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; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_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 SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/01 job id: 51843407 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/02 job id: 51843408 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/03 job id: 51843409 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/04 job id: 51843410 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/05 job id: 51843411 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/06 job id: 51843413 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/07 job id: 51843414 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/08 job id: 51843416 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/09 job id: 51843418 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/10 job id: 51843421 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/11 job id: 51843422 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/12 job id: 51843423 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/13 job id: 51843424 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/14 job id: 51843426 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/15 job id: 51843427 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/16 job id: 51843429 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/17 job id: 51843430 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/18 job id: 51843431 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/19 job id: 51843432 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/20 job id: 51843433 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/21 job id: 51843434 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/22 job id: 51843435 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/23 job id: 51843436 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/24 job id: 51843437 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/25 job id: 51843438 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/26 job id: 51843439 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/27 job id: 51843440 --- 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/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 0.022 0.039 139.184 139.185 farming_run 1 2.0 138.472 138.473 139.136 139.136 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.537331E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 119.969 119.969 qs_energies 1 2.0 0.103 0.163 119.772 119.776 mp2_main 1 3.0 0.000 0.000 117.572 117.641 mp2_gpw_main 1 4.0 0.049 0.060 116.481 116.484 mp2_ri_gpw_compute_in 1 5.0 0.174 0.185 97.466 97.858 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.324 57.714 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.166 43.773 48.910 get_2c_integrals 1 6.0 0.044 0.056 39.491 39.968 integrate_v_rspace 273 8.0 0.444 0.459 25.549 30.084 fft_wrap_pw1pw2 5465 10.4 0.067 0.070 28.044 28.614 grid_integrate_task_list 273 9.0 20.784 26.263 20.784 26.263 fft_wrap_pw1pw2_100 2178 11.4 1.322 1.449 25.495 26.080 compute_2c_integrals 1 7.0 0.002 0.002 21.357 21.382 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.829 21.088 mp2_eri_2c_integrate_gpw 1 9.0 2.404 2.429 20.827 21.084 rpa_ri_compute_en 1 5.0 0.019 0.023 18.888 19.029 cp_fm_cholesky_decompose 12 8.2 18.084 18.573 18.084 18.573 fft3d_s 5443 12.4 16.809 17.477 16.830 17.498 cholesky_decomp 1 7.0 0.000 0.000 16.937 17.430 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.304 10.766 15.304 calculate_wavefunction 272 8.0 5.468 5.611 12.921 13.811 calc_potential_gpw 544 9.5 0.004 0.004 10.816 11.600 rpa_num_int 1 6.0 0.003 0.049 10.801 10.811 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.688 10.718 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.782 10.496 potential_pw2rs 545 10.0 0.106 0.108 8.484 10.353 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.598 contract_S_to_Q 8 9.0 0.000 0.000 8.920 9.027 parallel_gemm_fm 14 9.1 0.000 0.000 8.497 8.660 parallel_gemm_fm_cosma 14 10.1 8.497 8.659 8.497 8.659 collocate_single_gaussian 272 10.0 0.039 0.042 7.832 8.603 create_integ_mat 1 6.0 0.024 0.038 7.603 7.613 array2fm 1 7.0 0.000 0.000 6.663 7.042 pw_gather_s 2722 12.2 3.992 4.920 3.992 4.920 pw_poisson_solve 545 10.5 0.010 0.010 4.623 4.812 pw_scatter_s 2720 12.7 4.445 4.728 4.445 4.728 pw_poisson_set 548 11.5 0.018 0.018 3.321 3.497 array2fm_buffer_send 1 8.0 2.934 3.097 2.934 3.097 pw_copy 4911 11.6 2.591 2.762 2.591 2.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.484453, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 0.028 0.049 529.178 529.191 farming_run 1 2.0 528.212 528.218 529.040 529.043 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.265738E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 324.531 324.533 qs_energies 1 2.0 0.005 0.014 324.274 324.275 mp2_main 1 3.0 0.000 0.000 209.519 209.525 mp2_gpw_main 1 4.0 0.049 0.073 208.611 208.618 mp2_ri_gpw_compute_en 1 5.0 0.038 0.040 128.554 130.714 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.744 1.860 126.824 126.830 scf_env_do_scf 1 3.0 0.000 0.000 114.461 114.462 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.520 113.527 rebuild_ks_matrix 4 6.0 0.000 0.000 113.518 113.525 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.062 113.518 113.525 hfx_ks_matrix 4 8.0 0.001 0.001 113.161 113.165 integrate_four_center 4 9.0 0.154 0.471 113.160 113.164 mp2_ri_gpw_compute_en_expansio 172 7.0 0.463 0.496 110.934 112.027 local_gemm 172 8.0 110.470 111.556 110.470 111.556 integrate_four_center_main 4 10.0 0.129 0.596 102.233 104.591 integrate_four_center_bin 265 11.0 102.104 104.583 102.104 104.583 init_scf_loop 1 4.0 0.000 0.000 96.511 96.512 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 79.907 81.002 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.702 57.800 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 44.547 49.376 integrate_v_rspace 95 8.0 0.400 0.566 28.182 32.847 fft_wrap_pw1pw2 1868 10.4 0.029 0.037 30.509 31.069 ao_to_mo_and_store_B_mult_1 91 7.0 10.492 29.143 10.492 29.143 fft_wrap_pw1pw2_100 730 11.4 0.645 0.767 28.223 28.736 grid_integrate_task_list 95 9.0 23.481 28.370 23.481 28.370 get_2c_integrals 1 6.0 0.002 0.005 23.115 23.141 compute_2c_integrals 1 7.0 0.002 0.003 22.093 22.100 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.748 21.961 mp2_eri_2c_integrate_gpw 1 9.0 1.722 1.844 21.747 21.960 fft3d_s 1823 12.4 19.214 19.967 19.228 19.981 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.948 17.948 calc_potential_gpw 182 9.5 0.002 0.002 12.323 13.326 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.717 11.339 calculate_wavefunction 91 8.0 2.032 2.069 10.051 10.865 potential_pw2rs 186 10.0 0.033 0.034 9.085 10.441 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.522 8.041 9.015 collocate_single_gaussian 91 10.0 0.018 0.022 8.325 8.961 mp_sendrecv_dm3 2068 8.0 6.089 7.089 6.089 7.089 integrate_four_center_load 4 10.0 0.000 0.000 6.749 6.753 hfx_load_balance 1 11.0 0.000 0.000 6.749 6.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.610211, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1519.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 446.816256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.089 56.619 56.650 qs_mol_dyn_low 1 2.0 0.006 0.011 56.079 56.094 qs_forces 11 3.9 0.002 0.003 55.991 55.995 qs_energies 11 4.9 0.003 0.011 54.497 54.515 scf_env_do_scf 11 5.9 0.000 0.001 48.330 48.330 scf_env_do_scf_inner_loop 108 6.5 0.004 0.009 46.141 46.142 qs_scf_new_mos 108 7.5 0.000 0.001 35.788 36.138 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.787 36.138 dbcsr_multiply_generic 2286 12.5 0.095 0.099 35.104 35.740 ot_scf_mini 108 9.5 0.002 0.002 34.052 34.244 multiply_cannon 2286 13.5 0.187 0.193 26.790 28.416 multiply_cannon_loop 2286 14.5 1.815 1.914 26.061 27.695 velocity_verlet 10 3.0 0.003 0.006 26.729 26.730 ot_mini 108 10.5 0.001 0.001 20.766 21.032 qs_ot_get_derivative 108 11.5 0.001 0.001 17.690 17.902 mp_waitall_1 245248 16.5 9.127 15.349 9.127 15.349 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.369 13.242 multiply_cannon_multrec 54864 15.5 3.617 5.637 7.578 10.928 qs_ot_get_p 119 10.4 0.001 0.002 8.585 8.866 rebuild_ks_matrix 119 8.3 0.000 0.000 8.054 8.206 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 8.053 8.205 mp_sum_l 7287 12.8 6.235 8.082 6.235 8.082 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.104 7.239 multiply_cannon_sync_h2d 54864 15.5 5.115 6.896 5.115 6.896 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.329 6.436 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.924 6.365 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.576 5.617 dbcsr_mm_accdrv_process 76910 16.1 1.845 2.885 3.876 5.538 init_scf_run 11 5.9 0.002 0.013 4.826 4.827 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.825 4.827 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.679 4.679 sum_up_and_integrate 119 10.3 0.001 0.003 4.551 4.559 integrate_v_rspace 119 11.3 0.002 0.003 4.540 4.550 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.309 4.445 calculate_rho_elec 119 8.7 0.011 0.018 4.309 4.445 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.439 4.439 cp_fm_redistribute_end 50 14.0 2.261 4.404 2.269 4.409 cp_fm_diag_elpa_base 50 14.0 2.132 4.293 2.137 4.296 multiply_cannon_metrocomm1 54864 15.5 0.058 0.065 2.001 3.234 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.943 3.186 apply_single 119 13.6 0.000 0.000 2.943 3.186 calculate_dm_sparse 119 9.5 0.000 0.000 3.001 3.176 acc_transpose_blocks 54864 15.5 0.227 0.245 2.269 2.806 ot_diis_step 108 11.5 0.006 0.006 2.765 2.766 jit_kernel_multiply 13 15.8 1.969 2.586 1.969 2.586 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.509 2.577 calculate_first_density_matrix 1 7.0 0.001 0.005 2.552 2.557 density_rs2pw 119 9.7 0.003 0.004 2.273 2.361 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.295 2.297 wfi_extrapolate 11 7.9 0.002 0.015 2.178 2.178 init_scf_loop 11 6.9 0.001 0.008 2.166 2.170 grid_integrate_task_list 119 12.3 2.023 2.121 2.023 2.121 mp_sum_d 4139 12.0 1.484 2.090 1.484 2.090 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.979 2.033 potential_pw2rs 119 12.3 0.004 0.004 1.895 1.912 make_m2s 4572 13.5 0.053 0.055 1.836 1.896 make_images 4572 14.5 0.133 0.138 1.755 1.814 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.691 1.769 transfer_rs2pw 487 10.6 0.006 0.006 1.477 1.597 mp_alltoall_d11v 2130 13.8 1.446 1.593 1.446 1.593 fft3d_ps 1201 13.6 0.369 0.470 1.467 1.536 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.433 1.462 acc_transpose_blocks_sync 164592 16.5 1.209 1.440 1.209 1.440 grid_collocate_task_list 119 9.7 1.351 1.425 1.351 1.425 transfer_pw2rs 487 13.2 0.005 0.006 1.402 1.416 mp_waitany 12084 13.8 1.286 1.415 1.286 1.415 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.301 1.382 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.815 1.223 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.650000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=426.181818, yerr=1.028519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 484.990976E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 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 0.024 0.088 40.174 40.174 qs_mol_dyn_low 1 2.0 0.014 0.102 39.866 39.873 qs_forces 11 3.9 0.003 0.005 39.711 39.714 qs_energies 11 4.9 0.002 0.002 38.019 38.027 scf_env_do_scf 11 5.9 0.000 0.001 32.472 32.472 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.953 29.953 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.616 22.969 qs_scf_new_mos 108 7.5 0.001 0.001 21.122 21.343 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.121 21.342 ot_scf_mini 108 9.5 0.002 0.003 20.182 20.347 velocity_verlet 10 3.0 0.012 0.032 18.803 18.809 multiply_cannon 2286 13.5 0.210 0.217 17.234 18.681 multiply_cannon_loop 2286 14.5 1.187 1.250 16.041 17.543 ot_mini 108 10.5 0.001 0.001 12.471 12.704 mp_waitall_1 200699 16.5 5.782 11.082 5.782 11.082 qs_ot_get_derivative 108 11.5 0.001 0.001 10.088 10.257 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.097 9.750 multiply_cannon_multrec 27432 15.5 1.832 4.140 6.263 9.212 rebuild_ks_matrix 119 8.3 0.000 0.000 6.900 7.035 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.900 7.035 dbcsr_mm_accdrv_process 47894 16.0 3.551 5.741 4.350 6.497 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.072 6.195 qs_ot_get_p 119 10.4 0.002 0.003 4.759 4.984 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.698 4.616 init_scf_run 11 5.9 0.000 0.001 4.303 4.304 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.303 4.304 mp_sum_l 7287 12.8 2.137 4.139 2.137 4.139 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.080 4.070 apply_single 119 13.6 0.000 0.000 3.080 4.069 sum_up_and_integrate 119 10.3 0.001 0.001 3.764 3.771 integrate_v_rspace 119 11.3 0.002 0.003 3.750 3.757 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.723 3.750 calculate_rho_elec 119 8.7 0.021 0.024 3.722 3.749 make_m2s 4572 13.5 0.052 0.053 2.791 3.110 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.027 3.046 make_images 4572 14.5 0.206 0.244 2.702 3.023 calculate_first_density_matrix 1 7.0 0.000 0.000 2.719 2.720 jit_kernel_multiply 13 16.3 0.742 2.623 0.742 2.623 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.615 2.616 init_scf_loop 11 6.9 0.000 0.000 2.497 2.498 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.401 2.487 ot_diis_step 108 11.5 0.011 0.011 2.335 2.336 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.258 2.259 calculate_dm_sparse 119 9.5 0.000 0.001 2.151 2.236 cp_fm_redistribute_end 50 14.0 1.145 2.229 1.148 2.231 multiply_cannon_sync_h2d 27432 15.5 1.685 2.204 1.685 2.204 cp_fm_diag_elpa_base 50 14.0 1.052 2.134 1.079 2.172 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.147 2.150 acc_transpose_blocks 27432 15.5 0.117 0.123 1.713 2.086 density_rs2pw 119 9.7 0.003 0.004 2.018 2.081 grid_integrate_task_list 119 12.3 1.839 1.958 1.839 1.958 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.865 1.904 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.750 1.781 make_images_data 4572 15.5 0.047 0.053 1.276 1.653 prepare_preconditioner 11 7.9 0.000 0.000 1.594 1.620 make_preconditioner 11 8.9 0.000 0.000 1.594 1.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.492 1.548 wfi_extrapolate 11 7.9 0.001 0.001 1.521 1.521 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.142 1.516 potential_pw2rs 119 12.3 0.006 0.006 1.498 1.507 fft3d_ps 1201 13.6 0.512 0.566 1.469 1.500 mp_alltoall_d11v 2130 13.8 1.288 1.414 1.288 1.414 fft_wrap_pw1pw2_140 487 12.2 0.048 0.050 1.345 1.377 grid_collocate_task_list 119 9.7 1.288 1.340 1.288 1.340 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.327 1.336 transfer_rs2pw 487 10.6 0.005 0.006 1.235 1.297 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.224 1.273 mp_sum_d 4139 12.0 0.559 1.047 0.559 1.047 mp_allgather_i34 2286 14.5 0.614 1.015 0.614 1.015 acc_transpose_blocks_kernels 27432 16.5 0.193 0.283 0.750 1.003 transfer_pw2rs 487 13.2 0.004 0.005 0.988 0.999 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.962 0.965 acc_transpose_blocks_sync 82296 16.5 0.820 0.942 0.820 0.942 make_images_sizes 4572 15.5 0.005 0.005 0.649 0.915 mp_alltoall_i44 4572 16.5 0.644 0.910 0.644 0.910 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.899 mp_alltoall_z22v 1201 15.6 0.735 0.808 0.735 0.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.174000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.636364, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 514.244608E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.081 34.526 34.546 qs_mol_dyn_low 1 2.0 0.009 0.053 33.648 33.656 qs_forces 11 3.9 0.002 0.003 33.295 33.297 qs_energies 11 4.9 0.011 0.078 31.668 31.723 scf_env_do_scf 11 5.9 0.001 0.002 26.612 26.612 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.006 24.007 dbcsr_multiply_generic 2286 12.5 0.093 0.097 17.258 17.319 qs_scf_new_mos 108 7.5 0.001 0.001 16.099 16.114 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.099 16.114 velocity_verlet 10 3.0 0.064 0.168 15.986 16.003 ot_scf_mini 108 9.5 0.002 0.004 15.327 15.339 multiply_cannon 2286 13.5 0.192 0.197 13.842 14.517 multiply_cannon_loop 2286 14.5 0.861 0.907 13.030 13.735 ot_mini 108 10.5 0.001 0.001 9.310 9.320 qs_ot_get_derivative 108 11.5 0.001 0.001 7.808 7.818 multiply_cannon_multrec 18288 15.5 1.897 2.868 6.990 7.280 rebuild_ks_matrix 119 8.3 0.000 0.000 6.156 6.171 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.155 6.171 dbcsr_mm_accdrv_process 38222 16.0 4.935 5.724 4.999 5.790 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.433 5.449 mp_waitall_1 158411 16.6 2.844 3.930 2.844 3.930 qs_ot_get_p 119 10.4 0.002 0.004 3.738 3.765 init_scf_run 11 5.9 0.001 0.007 3.704 3.705 scf_env_initial_rho_setup 11 6.9 0.003 0.013 3.703 3.705 sum_up_and_integrate 119 10.3 0.001 0.001 3.666 3.672 integrate_v_rspace 119 11.3 0.003 0.003 3.654 3.661 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.989 3.534 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.354 3.358 calculate_rho_elec 119 8.7 0.031 0.031 3.354 3.358 init_scf_loop 11 6.9 0.001 0.004 2.587 2.588 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.523 2.530 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.485 2.397 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.961 2.362 apply_single 119 13.6 0.000 0.000 1.961 2.362 calculate_first_density_matrix 1 7.0 0.001 0.005 2.309 2.313 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.212 2.213 make_m2s 4572 13.5 0.044 0.045 1.989 2.146 make_images 4572 14.5 0.192 0.206 1.904 2.061 density_rs2pw 119 9.7 0.004 0.006 1.911 1.980 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.935 1.947 cp_fm_diag_elpa_base 50 14.0 1.908 1.925 1.932 1.944 grid_integrate_task_list 119 12.3 1.811 1.901 1.811 1.901 calculate_dm_sparse 119 9.5 0.000 0.001 1.868 1.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.829 1.842 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.799 1.803 prepare_preconditioner 11 7.9 0.000 0.000 1.797 1.800 make_preconditioner 11 8.9 0.000 0.003 1.797 1.800 acc_transpose_blocks 18288 15.5 0.081 0.083 1.706 1.765 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.739 1.753 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.652 1.733 mp_sum_l 7287 12.8 1.192 1.635 1.192 1.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.567 1.573 ot_diis_step 108 11.5 0.011 0.012 1.486 1.486 potential_pw2rs 119 12.3 0.007 0.007 1.444 1.454 fft3d_ps 1201 13.6 0.536 0.554 1.424 1.440 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.345 1.358 grid_collocate_task_list 119 9.7 1.234 1.347 1.234 1.347 wfi_extrapolate 11 7.9 0.012 0.095 1.332 1.332 transfer_rs2pw 487 10.6 0.005 0.005 1.116 1.232 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.197 1.201 multiply_cannon_sync_h2d 18288 15.5 1.033 1.166 1.033 1.166 make_images_data 4572 15.5 0.047 0.051 0.841 1.034 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.972 0.996 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.994 0.996 hybrid_alltoall_any 4725 16.4 0.059 0.117 0.727 0.957 transfer_pw2rs 487 13.2 0.004 0.004 0.936 0.945 acc_transpose_blocks_kernels 18288 16.5 0.222 0.228 0.860 0.894 mp_alltoall_d11v 2130 13.8 0.739 0.817 0.739 0.817 acc_transpose_blocks_sync 54864 16.5 0.745 0.803 0.745 0.803 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.795 0.798 mp_alltoall_z22v 1201 15.6 0.724 0.794 0.724 0.794 cp_fm_cholesky_invert 11 10.9 0.768 0.772 0.768 0.772 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.677 0.750 dbcsr_complete_redistribute 329 12.2 0.098 0.108 0.583 0.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.546000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 553.512960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 37.468 37.479 qs_mol_dyn_low 1 2.0 0.006 0.038 36.738 36.746 qs_forces 11 3.9 0.002 0.002 36.521 36.523 qs_energies 11 4.9 0.009 0.063 34.651 34.714 scf_env_do_scf 11 5.9 0.000 0.001 29.468 29.469 scf_env_do_scf_inner_loop 108 6.5 0.004 0.017 26.037 26.038 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.925 20.054 velocity_verlet 10 3.0 0.025 0.104 18.591 18.597 qs_scf_new_mos 108 7.5 0.001 0.001 18.022 18.078 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.021 18.077 ot_scf_mini 108 9.5 0.002 0.004 16.985 17.040 multiply_cannon 2286 13.5 0.217 0.225 16.093 16.658 multiply_cannon_loop 2286 14.5 1.519 1.603 15.129 15.545 ot_mini 108 10.5 0.001 0.001 10.389 10.454 multiply_cannon_multrec 27432 15.5 2.452 3.104 9.029 9.321 qs_ot_get_derivative 108 11.5 0.001 0.001 8.516 8.572 dbcsr_mm_accdrv_process 47916 15.9 6.035 7.701 6.474 7.764 rebuild_ks_matrix 119 8.3 0.000 0.000 6.305 6.356 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.305 6.356 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.587 5.632 qs_ot_get_p 119 10.4 0.002 0.002 3.723 3.802 init_scf_run 11 5.9 0.000 0.001 3.703 3.704 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.703 3.704 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.082 3.559 sum_up_and_integrate 119 10.3 0.001 0.001 3.453 3.460 integrate_v_rspace 119 11.3 0.003 0.003 3.441 3.449 init_scf_loop 11 6.9 0.001 0.005 3.411 3.412 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.323 3.359 calculate_rho_elec 119 8.7 0.040 0.047 3.323 3.359 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.186 2.664 apply_single 119 13.6 0.000 0.000 2.185 2.664 prepare_preconditioner 11 7.9 0.000 0.000 2.575 2.583 make_preconditioner 11 8.9 0.000 0.002 2.575 2.583 acc_transpose_blocks 27432 15.5 0.120 0.125 2.426 2.550 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.184 2.513 make_m2s 4572 13.5 0.054 0.055 2.317 2.434 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.370 2.398 make_images 4572 14.5 0.274 0.334 2.211 2.327 calculate_first_density_matrix 1 7.0 0.000 0.002 2.293 2.294 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.276 2.289 calculate_dm_sparse 119 9.5 0.000 0.000 2.185 2.249 mp_waitall_1 137007 16.6 1.743 2.245 1.743 2.245 density_rs2pw 119 9.7 0.004 0.007 1.766 1.929 grid_integrate_task_list 119 12.3 1.833 1.924 1.833 1.924 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.870 1.870 ot_diis_step 108 11.5 0.012 0.012 1.831 1.831 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.806 1.806 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.740 1.771 mp_sum_l 7287 12.8 1.171 1.764 1.171 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.634 1.646 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.610 1.620 cp_fm_diag_elpa_base 50 14.0 1.576 1.594 1.608 1.617 acc_transpose_blocks_sync 82296 16.5 1.426 1.547 1.426 1.547 fft_wrap_pw1pw2_140 487 12.2 0.071 0.075 1.429 1.466 fft3d_ps 1201 13.6 0.578 0.642 1.401 1.426 wfi_extrapolate 11 7.9 0.001 0.001 1.350 1.350 grid_collocate_task_list 119 9.7 1.248 1.328 1.248 1.328 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.318 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.279 1.291 potential_pw2rs 119 12.3 0.008 0.009 1.252 1.254 transfer_rs2pw 487 10.6 0.005 0.005 0.998 1.179 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.692 1.178 cp_fm_upper_to_full 72 14.2 0.825 1.172 0.825 1.172 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.158 1.160 jit_kernel_multiply 6 16.3 0.375 1.102 0.375 1.102 dbcsr_complete_redistribute 329 12.2 0.115 0.152 0.769 1.050 make_images_data 4572 15.5 0.047 0.052 0.862 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.013 0.101 0.816 0.935 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.737 0.930 mp_alltoall_d11v 2130 13.8 0.753 0.894 0.753 0.894 acc_transpose_blocks_kernels 27432 16.5 0.278 0.285 0.851 0.877 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.814 transfer_pw2rs 487 13.2 0.004 0.004 0.762 0.764 mp_sum_d 4137 12.0 0.618 0.762 0.618 0.762 cp_fm_cholesky_invert 11 10.9 0.758 0.761 0.758 0.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.479000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.363636, yerr=2.226809 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 622.231552E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.028 30.468 30.468 qs_mol_dyn_low 1 2.0 0.003 0.006 30.059 30.078 qs_forces 11 3.9 0.006 0.007 29.998 29.999 qs_energies 11 4.9 0.001 0.001 28.248 28.253 scf_env_do_scf 11 5.9 0.000 0.001 23.266 23.266 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 20.623 20.624 velocity_verlet 10 3.0 0.007 0.009 15.445 15.448 dbcsr_multiply_generic 2286 12.5 0.093 0.097 13.797 13.910 qs_scf_new_mos 108 7.5 0.001 0.001 12.688 12.716 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.687 12.715 ot_scf_mini 108 9.5 0.002 0.002 11.946 11.965 multiply_cannon 2286 13.5 0.224 0.234 10.871 11.393 multiply_cannon_loop 2286 14.5 0.645 0.666 9.877 10.126 ot_mini 108 10.5 0.001 0.001 6.817 6.841 multiply_cannon_multrec 9144 15.5 1.741 1.958 6.159 6.466 rebuild_ks_matrix 119 8.3 0.000 0.000 6.017 6.037 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.017 6.037 qs_ot_get_derivative 108 11.5 0.001 0.001 5.461 5.478 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.354 5.372 dbcsr_mm_accdrv_process 12550 15.8 3.697 4.349 4.307 4.390 sum_up_and_integrate 119 10.3 0.001 0.001 3.604 3.611 integrate_v_rspace 119 11.3 0.003 0.003 3.594 3.600 init_scf_run 11 5.9 0.000 0.001 3.500 3.501 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.500 3.500 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.448 3.458 calculate_rho_elec 119 8.7 0.060 0.061 3.447 3.458 qs_ot_get_p 119 10.4 0.003 0.003 3.161 3.193 init_scf_loop 11 6.9 0.000 0.000 2.618 2.620 calculate_first_density_matrix 1 7.0 0.000 0.000 2.296 2.297 make_m2s 4572 13.5 0.034 0.035 2.024 2.169 mp_waitall_1 115863 16.7 1.714 2.153 1.714 2.153 make_images 4572 14.5 0.269 0.300 1.934 2.079 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.055 2.057 grid_integrate_task_list 119 12.3 1.888 2.049 1.888 2.049 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.915 1.928 calculate_dm_sparse 119 9.5 0.000 0.000 1.898 1.917 density_rs2pw 119 9.7 0.003 0.003 1.818 1.898 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.870 1.883 prepare_preconditioner 11 7.9 0.000 0.000 1.847 1.852 make_preconditioner 11 8.9 0.000 0.000 1.847 1.852 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.789 1.791 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.732 1.760 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.577 1.578 fft_wrap_pw1pw2_140 487 12.2 0.094 0.095 1.510 1.524 fft3d_ps 1201 13.6 0.638 0.652 1.481 1.494 acc_transpose_blocks 9144 15.5 0.042 0.044 1.465 1.494 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.474 1.483 cp_fm_diag_elpa_base 50 14.0 1.446 1.464 1.472 1.481 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.451 1.459 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.399 1.407 grid_collocate_task_list 119 9.7 1.312 1.403 1.312 1.403 jit_kernel_multiply 8 15.5 0.571 1.378 0.571 1.378 ot_diis_step 108 11.5 0.013 0.013 1.337 1.338 potential_pw2rs 119 12.3 0.010 0.010 1.293 1.298 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.264 1.286 apply_single 119 13.6 0.000 0.000 1.264 1.286 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.259 1.264 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.252 1.255 wfi_extrapolate 11 7.9 0.001 0.001 1.149 1.150 make_images_data 4572 15.5 0.042 0.046 0.894 1.109 hybrid_alltoall_any 4725 16.4 0.066 0.178 0.850 1.062 transfer_rs2pw 487 10.6 0.004 0.005 0.955 1.035 mp_alltoall_d11v 2130 13.8 0.927 0.991 0.927 0.991 cp_fm_cholesky_invert 11 10.9 0.961 0.964 0.961 0.964 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.880 0.928 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.866 0.873 mp_allgather_i34 2286 14.5 0.307 0.799 0.307 0.799 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.380 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.785 0.787 mp_alltoall_z22v 1201 15.6 0.718 0.754 0.718 0.754 transfer_pw2rs 487 13.2 0.003 0.004 0.749 0.751 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.698 0.745 acc_transpose_blocks_sync 27432 16.5 0.721 0.745 0.721 0.745 acc_transpose_blocks_kernels 9144 16.5 0.120 0.122 0.686 0.701 mp_waitany 5200 13.7 0.522 0.620 0.522 0.620 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.468000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.272727, yerr=6.823634 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 791.470080E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.084 44.538 44.539 qs_mol_dyn_low 1 2.0 0.008 0.046 44.160 44.168 qs_forces 11 3.9 0.004 0.014 44.097 44.102 qs_energies 11 4.9 0.004 0.009 42.083 42.092 scf_env_do_scf 11 5.9 0.000 0.001 35.446 35.446 scf_env_do_scf_inner_loop 108 6.5 0.004 0.008 27.444 27.445 velocity_verlet 10 3.0 0.002 0.002 24.360 24.366 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.875 20.101 qs_scf_new_mos 108 7.5 0.001 0.001 17.955 18.038 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.955 18.037 ot_scf_mini 108 9.5 0.002 0.002 16.764 16.857 multiply_cannon 2286 13.5 0.295 0.299 15.717 16.635 multiply_cannon_loop 2286 14.5 0.861 0.877 14.379 15.319 ot_mini 108 10.5 0.001 0.001 10.079 10.185 multiply_cannon_multrec 9144 15.5 3.417 4.812 9.089 9.268 qs_ot_get_derivative 108 11.5 0.001 0.001 7.971 8.064 init_scf_loop 11 6.9 0.000 0.000 7.971 7.973 rebuild_ks_matrix 119 8.3 0.000 0.000 7.172 7.302 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.171 7.302 prepare_preconditioner 11 7.9 0.000 0.000 7.005 7.019 make_preconditioner 11 8.9 0.000 0.000 7.005 7.019 dbcsr_mm_accdrv_process 12550 15.8 4.792 6.892 5.535 6.925 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.559 6.893 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.461 6.578 cp_fm_upper_to_full 72 14.2 3.189 4.572 3.189 4.572 init_scf_run 11 5.9 0.000 0.001 4.484 4.484 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.484 4.484 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.065 4.080 calculate_rho_elec 119 8.7 0.118 0.122 4.064 4.079 sum_up_and_integrate 119 10.3 0.001 0.001 3.813 3.820 integrate_v_rspace 119 11.3 0.003 0.003 3.803 3.809 qs_ot_get_p 119 10.4 0.003 0.003 3.659 3.789 mp_waitall_1 94719 16.7 2.585 3.557 2.585 3.557 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.918 3.308 calculate_first_density_matrix 1 7.0 0.000 0.000 2.989 2.991 dbcsr_complete_redistribute 329 12.2 0.285 0.288 2.027 2.853 make_m2s 4572 13.5 0.037 0.037 2.566 2.777 make_images 4572 14.5 0.352 0.385 2.444 2.655 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.243 2.548 apply_single 119 13.6 0.000 0.000 2.243 2.547 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.691 2.510 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.497 2.506 calculate_dm_sparse 119 9.5 0.000 0.000 2.475 2.493 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.420 2.338 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.440 2.251 mp_alltoall_i22 627 13.8 1.423 2.242 1.423 2.242 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.205 2.207 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.155 2.203 grid_integrate_task_list 119 12.3 2.083 2.126 2.083 2.126 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.116 2.117 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.092 2.103 density_rs2pw 119 9.7 0.003 0.003 2.069 2.098 ot_diis_step 108 11.5 0.014 0.014 2.082 2.082 acc_transpose_blocks 9144 15.5 0.045 0.046 1.996 2.022 fft3d_ps 1201 13.6 0.859 0.882 1.949 1.956 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.865 1.902 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.842 1.850 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.829 1.830 mp_sum_l 7287 12.8 1.049 1.762 1.049 1.762 cp_fm_cholesky_invert 11 10.9 1.562 1.565 1.562 1.565 grid_collocate_task_list 119 9.7 1.524 1.548 1.524 1.548 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.504 1.504 cp_fm_diag_elpa_base 50 14.0 1.356 1.410 1.502 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.447 1.457 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.159 1.436 make_images_data 4572 15.5 0.045 0.049 1.134 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.419 1.419 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.358 1.381 potential_pw2rs 119 12.3 0.013 0.014 1.291 1.294 mp_alltoall_d11v 2130 13.8 1.246 1.270 1.246 1.270 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.166 1.186 jit_kernel_multiply 6 15.6 0.714 1.180 0.714 1.180 acc_transpose_blocks_sync 27432 16.5 1.097 1.121 1.097 1.121 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.003 1.046 mp_alltoall_z22v 1201 15.6 0.955 0.989 0.955 0.989 qs_create_task_list 11 7.9 0.006 0.007 0.936 0.952 generate_qs_task_list 11 8.9 0.367 0.387 0.930 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.923 0.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.539000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=741.636364, yerr=18.167267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 497.770496E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1632366. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.097 82.822 82.824 qs_mol_dyn_low 1 2.0 0.006 0.029 82.188 82.232 qs_forces 11 3.9 0.004 0.005 82.051 82.054 qs_energies 11 4.9 0.006 0.009 79.173 79.192 scf_env_do_scf 11 5.9 0.001 0.002 70.077 70.079 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.437 64.438 dbcsr_multiply_generic 2055 12.4 0.105 0.108 50.764 51.004 qs_scf_new_mos 99 7.5 0.000 0.001 47.488 47.618 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.488 47.618 ot_scf_mini 99 9.5 0.002 0.002 45.123 45.229 velocity_verlet 10 3.0 0.009 0.056 43.169 43.171 multiply_cannon 2055 13.4 0.184 0.189 42.355 43.163 multiply_cannon_loop 2055 14.4 1.781 1.818 41.351 42.208 ot_mini 99 10.5 0.001 0.001 26.635 26.740 qs_ot_get_derivative 99 11.5 0.001 0.001 19.826 19.933 multiply_cannon_multrec 49320 15.4 11.369 12.114 17.429 18.180 rebuild_ks_matrix 110 8.3 0.000 0.000 14.551 14.671 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.551 14.670 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.760 12.871 mp_waitall_1 220248 16.4 10.603 11.386 10.603 11.386 qs_ot_get_p 110 10.4 0.001 0.002 10.117 10.282 multiply_cannon_sync_h2d 49320 15.4 9.547 10.048 9.547 10.048 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.665 8.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.214 7.764 apply_single 110 13.6 0.000 0.001 7.214 7.764 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.345 7.463 sum_up_and_integrate 110 10.3 0.002 0.003 6.974 6.988 integrate_v_rspace 110 11.3 0.003 0.003 6.949 6.968 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.880 6.893 init_scf_run 11 5.9 0.000 0.001 6.891 6.891 scf_env_initial_rho_setup 11 6.9 0.002 0.004 6.890 6.891 ot_diis_step 99 11.5 0.006 0.006 6.578 6.578 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.203 6.354 calculate_rho_elec 110 8.6 0.020 0.024 6.203 6.353 dbcsr_mm_accdrv_process 87628 16.1 3.084 3.160 5.929 6.198 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.035 6.035 init_scf_loop 11 6.9 0.000 0.000 5.608 5.609 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.397 5.469 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.428 5.459 cp_fm_diag_elpa_base 48 14.0 5.411 5.444 5.426 5.457 mp_sum_l 6594 12.7 3.969 4.709 3.969 4.709 make_m2s 4110 13.4 0.060 0.064 4.018 4.121 wfi_extrapolate 11 7.9 0.001 0.001 4.071 4.071 make_images 4110 14.4 0.178 0.190 3.922 4.029 calculate_dm_sparse 110 9.5 0.001 0.001 3.752 3.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.608 3.612 density_rs2pw 110 9.6 0.004 0.004 3.259 3.457 prepare_preconditioner 11 7.9 0.000 0.000 3.435 3.450 make_preconditioner 11 8.9 0.000 0.000 3.435 3.450 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.358 3.425 grid_integrate_task_list 110 12.3 3.268 3.420 3.268 3.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.211 3.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.184 3.238 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.097 3.138 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 2.748 2.816 calculate_first_density_matrix 1 7.0 0.000 0.001 2.712 2.720 acc_transpose_blocks 49320 15.4 0.213 0.221 2.578 2.641 jit_kernel_multiply 13 15.9 2.567 2.629 2.567 2.629 potential_pw2rs 110 12.3 0.005 0.006 2.560 2.582 mp_alltoall_d11v 2046 13.8 2.080 2.539 2.080 2.539 fft3d_ps 1111 13.6 0.730 0.813 2.287 2.357 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.279 2.347 grid_collocate_task_list 110 9.6 2.159 2.284 2.159 2.284 transfer_rs2pw 451 10.6 0.006 0.006 2.044 2.253 mp_waitany 14300 13.8 1.852 2.122 1.852 2.122 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.930 1.945 mp_sum_d 3893 11.9 1.383 1.918 1.383 1.918 cp_fm_cholesky_invert 11 10.9 1.903 1.907 1.903 1.907 make_images_data 4110 15.4 0.043 0.046 1.776 1.893 transfer_pw2rs 451 13.1 0.006 0.007 1.806 1.820 hybrid_alltoall_any 4261 16.3 0.085 0.483 1.539 1.785 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.668 1.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.824000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=473.272727, yerr=2.799055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 591.917056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3093199. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.048 69.131 69.131 qs_mol_dyn_low 1 2.0 0.003 0.003 68.878 68.900 qs_forces 11 3.9 0.003 0.004 68.781 68.782 qs_energies 11 4.9 0.003 0.007 65.447 65.452 scf_env_do_scf 11 5.9 0.000 0.001 56.600 56.604 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.024 49.024 dbcsr_multiply_generic 2055 12.4 0.114 0.119 38.251 38.412 velocity_verlet 10 3.0 0.001 0.002 35.776 35.777 qs_scf_new_mos 99 7.5 0.001 0.001 33.332 33.505 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.331 33.504 multiply_cannon 2055 13.4 0.225 0.243 31.692 32.809 ot_scf_mini 99 9.5 0.003 0.003 31.660 31.808 multiply_cannon_loop 2055 14.4 1.159 1.179 30.452 31.423 ot_mini 99 10.5 0.001 0.001 18.402 18.565 multiply_cannon_multrec 24660 15.4 6.963 8.297 14.183 15.539 rebuild_ks_matrix 110 8.3 0.000 0.000 13.512 13.617 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.512 13.617 qs_ot_get_derivative 99 11.5 0.001 0.001 12.632 12.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.892 11.981 mp_waitall_1 176588 16.5 7.612 10.063 7.612 10.063 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.150 7.617 init_scf_loop 11 6.9 0.000 0.000 7.540 7.541 multiply_cannon_sync_h2d 24660 15.4 6.340 7.405 6.340 7.405 dbcsr_mm_accdrv_process 52282 16.1 5.534 6.367 7.047 7.315 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.473 7.110 apply_single 110 13.6 0.000 0.001 6.473 7.109 qs_ot_get_p 110 10.4 0.003 0.008 6.624 6.771 init_scf_run 11 5.9 0.000 0.001 6.480 6.481 scf_env_initial_rho_setup 11 6.9 0.001 0.005 6.480 6.480 sum_up_and_integrate 110 10.3 0.001 0.003 6.272 6.287 integrate_v_rspace 110 11.3 0.002 0.003 6.245 6.262 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.716 5.726 calculate_rho_elec 110 8.6 0.039 0.047 5.716 5.726 ot_diis_step 99 11.5 0.010 0.011 5.722 5.722 prepare_preconditioner 11 7.9 0.000 0.000 5.544 5.564 make_preconditioner 11 8.9 0.000 0.000 5.544 5.564 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.729 5.453 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.140 5.292 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.715 4.738 make_m2s 4110 13.4 0.057 0.059 4.198 4.640 make_images 4110 14.4 0.405 0.465 4.086 4.524 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.249 4.252 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.663 3.675 cp_fm_diag_elpa_base 48 14.0 3.615 3.629 3.660 3.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.516 3.517 wfi_extrapolate 11 7.9 0.001 0.001 3.474 3.474 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.303 3.386 grid_integrate_task_list 110 12.3 3.175 3.330 3.175 3.330 fft_wrap_pw1pw2 1111 11.6 0.015 0.017 3.110 3.247 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.181 3.211 calculate_dm_sparse 110 9.5 0.001 0.001 3.096 3.126 density_rs2pw 110 9.6 0.004 0.004 2.958 3.120 calculate_first_density_matrix 1 7.0 0.000 0.000 2.910 2.913 fft_wrap_pw1pw2_140 451 12.1 0.110 0.117 2.629 2.764 hybrid_alltoall_any 4261 16.3 0.106 0.463 2.012 2.726 make_images_data 4110 15.4 0.048 0.053 2.277 2.722 fft3d_ps 1111 13.6 1.069 1.273 2.488 2.617 cp_fm_cholesky_invert 11 10.9 2.565 2.573 2.565 2.573 mp_sum_l 6594 12.7 1.740 2.453 1.740 2.453 grid_collocate_task_list 110 9.6 2.169 2.313 2.169 2.313 jit_kernel_multiply 11 16.3 1.146 2.262 1.146 2.262 acc_transpose_blocks 24660 15.4 0.118 0.123 2.155 2.255 potential_pw2rs 110 12.3 0.008 0.008 2.148 2.162 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.963 1.985 mp_alltoall_d11v 2046 13.8 1.757 1.917 1.757 1.917 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.855 1.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.796 transfer_rs2pw 451 10.6 0.007 0.007 1.455 1.694 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.657 1.668 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.790 1.565 mp_allgather_i34 2055 14.4 0.545 1.549 0.545 1.549 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.394 1.504 mp_irecv_dv 57340 16.2 0.661 1.441 0.661 1.441 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.131000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.272727, yerr=7.123468 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 663.568384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.125 60.246 60.258 qs_mol_dyn_low 1 2.0 0.004 0.009 59.636 59.663 qs_forces 11 3.9 0.003 0.003 59.565 59.567 qs_energies 11 4.9 0.002 0.006 56.397 56.402 scf_env_do_scf 11 5.9 0.001 0.002 48.335 48.336 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 39.764 39.764 velocity_verlet 10 3.0 0.001 0.005 32.318 32.320 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.748 29.037 qs_scf_new_mos 99 7.5 0.001 0.001 25.389 25.514 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.388 25.514 ot_scf_mini 99 9.5 0.003 0.004 24.127 24.267 multiply_cannon 2055 13.4 0.209 0.223 22.454 23.551 multiply_cannon_loop 2055 14.4 0.814 0.836 21.397 22.323 ot_mini 99 10.5 0.001 0.001 13.788 13.928 rebuild_ks_matrix 110 8.3 0.000 0.000 12.079 12.258 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.078 12.257 multiply_cannon_multrec 16440 15.4 3.565 4.614 9.820 10.818 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.635 10.795 mp_waitall_1 139946 16.5 6.994 10.102 6.994 10.102 qs_ot_get_derivative 99 11.5 0.001 0.001 9.354 9.497 init_scf_loop 11 6.9 0.001 0.003 8.536 8.538 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.442 7.451 prepare_preconditioner 11 7.9 0.000 0.000 6.774 6.791 make_preconditioner 11 8.9 0.000 0.002 6.774 6.791 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.103 6.458 dbcsr_mm_accdrv_process 34862 16.1 5.323 5.713 6.098 6.325 sum_up_and_integrate 110 10.3 0.001 0.002 6.144 6.157 integrate_v_rspace 110 11.3 0.003 0.003 6.118 6.131 qs_ot_get_p 110 10.4 0.001 0.003 5.517 5.675 init_scf_run 11 5.9 0.002 0.018 5.590 5.591 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.588 5.591 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.526 5.534 calculate_rho_elec 110 8.6 0.058 0.058 5.525 5.533 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.935 5.378 apply_single 110 13.6 0.000 0.000 4.934 5.378 make_m2s 4110 13.4 0.049 0.051 4.164 4.440 ot_diis_step 99 11.5 0.011 0.011 4.400 4.401 make_images 4110 14.4 0.400 0.520 4.047 4.325 multiply_cannon_sync_h2d 16440 15.4 3.206 3.893 3.206 3.893 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.862 3.866 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.159 3.781 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.490 3.490 grid_integrate_task_list 110 12.3 3.194 3.372 3.194 3.372 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.071 3.082 density_rs2pw 110 9.6 0.004 0.006 2.771 3.027 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.945 2.956 cp_fm_diag_elpa_base 48 14.0 2.877 2.907 2.941 2.952 wfi_extrapolate 11 7.9 0.001 0.001 2.927 2.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.873 2.878 make_images_data 4110 15.4 0.045 0.049 2.333 2.669 cp_fm_cholesky_invert 11 10.9 2.648 2.654 2.648 2.654 fft_wrap_pw1pw2_140 451 12.1 0.127 0.131 2.627 2.639 calculate_dm_sparse 110 9.5 0.001 0.001 2.551 2.577 calculate_first_density_matrix 1 7.0 0.001 0.006 2.566 2.570 hybrid_alltoall_any 4261 16.3 0.109 0.382 2.009 2.563 multiply_cannon_metrocomm4 14385 15.4 0.047 0.052 0.908 2.545 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.460 2.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.467 2.523 mp_irecv_dv 48980 15.7 0.832 2.406 0.832 2.406 grid_collocate_task_list 110 9.6 2.219 2.398 2.219 2.398 fft3d_ps 1111 13.6 1.087 1.099 2.383 2.396 mp_sum_l 6594 12.7 1.596 2.278 1.596 2.278 mp_alltoall_d11v 2046 13.8 1.747 2.081 1.747 2.081 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.033 2.035 potential_pw2rs 110 12.3 0.010 0.011 2.006 2.011 dbcsr_complete_redistribute 325 12.2 0.339 0.420 1.500 1.960 cp_fm_upper_to_full 70 14.2 1.410 1.774 1.410 1.774 acc_transpose_blocks 16440 15.4 0.078 0.082 1.604 1.668 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.662 cp_fm_cholesky_decompose 22 10.9 1.597 1.622 1.597 1.622 transfer_rs2pw 451 10.6 0.005 0.007 1.318 1.570 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.031 1.484 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.473 1.483 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.381 1.480 mp_waitany 17072 13.8 1.164 1.452 1.164 1.452 mp_allgather_i34 2055 14.4 0.373 1.410 0.373 1.410 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.295 1.310 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.258000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.818182, yerr=7.661215 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 733.196288E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.051 66.360 66.360 qs_mol_dyn_low 1 2.0 0.003 0.003 65.871 65.880 qs_forces 11 3.9 0.003 0.003 65.804 65.808 qs_energies 11 4.9 0.001 0.002 62.405 62.412 scf_env_do_scf 11 5.9 0.000 0.001 53.932 53.935 scf_env_do_scf_inner_loop 99 6.5 0.005 0.021 42.203 42.203 velocity_verlet 10 3.0 0.001 0.002 37.321 37.323 dbcsr_multiply_generic 2055 12.4 0.115 0.119 30.958 31.157 qs_scf_new_mos 99 7.5 0.001 0.001 27.742 27.860 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.741 27.859 ot_scf_mini 99 9.5 0.002 0.003 26.045 26.157 multiply_cannon 2055 13.4 0.237 0.255 23.640 25.145 multiply_cannon_loop 2055 14.4 1.409 1.473 22.254 22.845 ot_mini 99 10.5 0.001 0.001 14.884 15.021 multiply_cannon_multrec 24660 15.4 4.103 6.893 13.194 14.313 rebuild_ks_matrix 110 8.3 0.000 0.000 12.007 12.116 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.007 12.116 init_scf_loop 11 6.9 0.000 0.000 11.689 11.690 qs_ot_get_derivative 99 11.5 0.001 0.001 10.714 10.831 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.610 10.707 prepare_preconditioner 11 7.9 0.000 0.000 9.935 9.954 make_preconditioner 11 8.9 0.000 0.000 9.935 9.954 dbcsr_mm_accdrv_process 52304 16.0 7.970 9.416 8.932 9.895 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.113 9.608 mp_waitall_1 121746 16.5 4.290 6.213 4.290 6.213 sum_up_and_integrate 110 10.3 0.001 0.002 6.146 6.158 qs_ot_get_p 110 10.4 0.002 0.003 5.989 6.141 integrate_v_rspace 110 11.3 0.003 0.003 6.119 6.132 make_m2s 4110 13.4 0.059 0.060 5.626 5.927 make_images 4110 14.4 0.575 0.697 5.487 5.786 init_scf_run 11 5.9 0.000 0.001 5.707 5.708 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.707 5.708 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.636 5.648 calculate_rho_elec 110 8.6 0.077 0.081 5.635 5.648 cp_fm_upper_to_full 70 14.2 3.443 4.983 3.443 4.983 ot_diis_step 99 11.5 0.011 0.011 4.131 4.131 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.122 apply_single 110 13.6 0.000 0.000 4.057 4.122 qs_ot_p2m_diag 48 11.0 0.055 0.065 4.062 4.079 dbcsr_complete_redistribute 325 12.2 0.414 0.447 2.691 3.855 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.533 3.594 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.490 3.491 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.391 3.452 grid_integrate_task_list 110 12.3 3.295 3.435 3.295 3.435 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.164 3.308 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.217 3.241 make_images_data 4110 15.4 0.048 0.052 2.833 3.142 calculate_dm_sparse 110 9.5 0.001 0.001 3.098 3.139 hybrid_alltoall_any 4261 16.3 0.122 0.458 2.323 3.064 wfi_extrapolate 11 7.9 0.001 0.001 2.996 2.996 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.346 2.995 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.939 2.951 cp_fm_diag_elpa_base 48 14.0 2.787 2.849 2.936 2.949 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.813 2.942 mp_alltoall_i22 605 13.7 1.686 2.885 1.686 2.885 density_rs2pw 110 9.6 0.004 0.004 2.748 2.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.813 2.815 fft_wrap_pw1pw2_140 451 12.1 0.145 0.148 2.736 2.766 cp_fm_cholesky_invert 11 10.9 2.649 2.658 2.649 2.658 calculate_first_density_matrix 1 7.0 0.000 0.000 2.615 2.617 multiply_cannon_sync_h2d 24660 15.4 2.389 2.617 2.389 2.617 acc_transpose_blocks 24660 15.4 0.112 0.114 2.430 2.571 fft3d_ps 1111 13.6 1.127 1.166 2.485 2.502 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.460 2.496 grid_collocate_task_list 110 9.6 2.267 2.441 2.267 2.441 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.252 2.257 potential_pw2rs 110 12.3 0.012 0.013 1.942 1.946 mp_alltoall_d11v 2046 13.8 1.766 1.923 1.766 1.923 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.806 1.840 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.621 1.718 mp_allgather_i34 2055 14.4 0.475 1.714 0.475 1.714 cp_fm_cholesky_decompose 22 10.9 1.640 1.689 1.640 1.689 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.645 1.665 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.650 mp_sum_l 6594 12.7 1.007 1.628 1.007 1.628 acc_transpose_blocks_sync 73980 16.4 1.399 1.554 1.399 1.554 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.837 1.472 jit_kernel_multiply 8 16.0 0.616 1.458 0.616 1.458 mp_irecv_dv 62702 16.1 0.735 1.387 0.735 1.387 transfer_rs2pw 451 10.6 0.005 0.006 1.205 1.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.360000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.000000, yerr=8.934102 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 863.993856E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.084 55.842 55.843 qs_mol_dyn_low 1 2.0 0.005 0.032 55.414 55.424 qs_forces 11 3.9 0.006 0.007 55.220 55.223 qs_energies 11 4.9 0.001 0.003 51.581 51.586 scf_env_do_scf 11 5.9 0.000 0.001 42.628 42.628 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.797 34.798 velocity_verlet 10 3.0 0.012 0.013 30.994 31.004 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.671 23.802 qs_scf_new_mos 99 7.5 0.001 0.001 20.650 20.691 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.650 20.691 multiply_cannon 2055 13.4 0.236 0.251 18.107 19.484 ot_scf_mini 99 9.5 0.002 0.002 19.384 19.407 multiply_cannon_loop 2055 14.4 0.604 0.623 16.832 17.140 rebuild_ks_matrix 110 8.3 0.000 0.000 11.608 11.634 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.608 11.634 ot_mini 99 10.5 0.001 0.001 10.733 10.747 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.341 10.362 multiply_cannon_multrec 8220 15.4 3.169 4.409 7.906 8.787 init_scf_loop 11 6.9 0.000 0.000 7.785 7.786 mp_waitall_1 103326 16.6 5.860 7.510 5.860 7.510 qs_ot_get_derivative 99 11.5 0.001 0.001 7.056 7.075 sum_up_and_integrate 110 10.3 0.001 0.002 6.055 6.070 prepare_preconditioner 11 7.9 0.000 0.000 6.045 6.052 make_preconditioner 11 8.9 0.000 0.000 6.045 6.052 integrate_v_rspace 110 11.3 0.003 0.003 6.028 6.044 init_scf_run 11 5.9 0.000 0.001 5.816 5.816 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.816 5.816 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.609 5.686 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.645 5.656 calculate_rho_elec 110 8.6 0.115 0.116 5.644 5.656 dbcsr_mm_accdrv_process 17442 15.9 3.334 4.631 4.596 5.567 qs_ot_get_p 110 10.4 0.002 0.003 4.696 4.716 make_m2s 4110 13.4 0.037 0.039 4.223 4.513 make_images 4110 14.4 0.645 0.706 4.093 4.383 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.964 4.241 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.724 3.767 apply_single 110 13.6 0.000 0.000 3.724 3.767 ot_diis_step 99 11.5 0.012 0.012 3.655 3.655 grid_integrate_task_list 110 12.3 3.376 3.526 3.376 3.526 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.337 3.342 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.323 3.327 calculate_first_density_matrix 1 7.0 0.000 0.000 3.032 3.035 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.983 2.984 cp_fm_cholesky_invert 11 10.9 2.892 2.896 2.892 2.896 fft_wrap_pw1pw2_140 451 12.1 0.191 0.194 2.873 2.882 density_rs2pw 110 9.6 0.003 0.004 2.677 2.793 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.764 2.765 hybrid_alltoall_any 4261 16.3 0.200 0.854 2.226 2.760 calculate_dm_sparse 110 9.5 0.001 0.001 2.679 2.715 wfi_extrapolate 11 7.9 0.001 0.001 2.688 2.688 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.680 2.681 make_images_data 4110 15.4 0.041 0.046 2.271 2.674 grid_collocate_task_list 110 9.6 2.369 2.549 2.369 2.549 fft3d_ps 1111 13.6 1.273 1.281 2.495 2.505 multiply_cannon_sync_h2d 8220 15.4 2.376 2.496 2.376 2.496 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.471 2.479 cp_fm_diag_elpa_base 48 14.0 2.414 2.440 2.470 2.477 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.401 2.413 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.041 2.075 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.794 2.006 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.935 1.947 potential_pw2rs 110 12.3 0.015 0.015 1.849 1.852 mp_alltoall_d11v 2046 13.8 1.620 1.789 1.620 1.789 cp_fm_cholesky_decompose 22 10.9 1.685 1.700 1.685 1.700 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.617 1.632 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.498 1.618 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.596 1.601 dbcsr_complete_redistribute 325 12.2 0.564 0.593 1.453 1.554 acc_transpose_blocks 8220 15.4 0.040 0.041 1.500 1.542 mp_allgather_i34 2055 14.4 0.442 1.528 0.442 1.528 qs_create_task_list 11 7.9 0.001 0.001 1.208 1.308 generate_qs_task_list 11 8.9 0.375 0.445 1.208 1.307 transfer_rs2pw 451 10.6 0.005 0.005 1.116 1.258 jit_kernel_multiply 6 15.5 0.942 1.207 0.942 1.207 mp_waitany 9240 13.8 1.047 1.175 1.047 1.175 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.145 1.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.843000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=812.909091, yerr=10.974802 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 1.432375E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.061 87.134 87.134 qs_mol_dyn_low 1 2.0 0.003 0.003 86.689 86.697 qs_forces 11 3.9 0.003 0.003 86.623 86.624 qs_energies 11 4.9 0.002 0.002 82.503 82.505 scf_env_do_scf 11 5.9 0.000 0.001 72.460 72.460 velocity_verlet 10 3.0 0.002 0.002 55.606 55.612 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.894 43.895 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.798 29.911 init_scf_loop 11 6.9 0.000 0.000 28.497 28.498 qs_scf_new_mos 99 7.5 0.001 0.001 26.962 27.013 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.961 27.013 prepare_preconditioner 11 7.9 0.000 0.000 26.521 26.529 make_preconditioner 11 8.9 0.000 0.000 26.521 26.529 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.542 25.979 ot_scf_mini 99 9.5 0.002 0.002 25.155 25.196 multiply_cannon 2055 13.4 0.334 0.354 22.821 23.488 multiply_cannon_loop 2055 14.4 0.828 0.838 21.065 21.381 cp_fm_upper_to_full 70 14.2 12.849 18.544 12.849 18.544 ot_mini 99 10.5 0.001 0.001 14.338 14.379 rebuild_ks_matrix 110 8.3 0.000 0.000 13.411 13.457 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.410 13.457 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.173 12.216 dbcsr_complete_redistribute 325 12.2 1.011 1.036 7.522 10.787 multiply_cannon_multrec 8220 15.4 4.087 4.289 9.868 9.995 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.525 9.783 qs_ot_get_derivative 99 11.5 0.001 0.001 9.720 9.764 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.962 9.185 mp_alltoall_i22 605 13.7 5.572 8.851 5.572 8.851 mp_waitall_1 84994 16.7 7.560 8.330 7.560 8.330 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.856 6.892 calculate_rho_elec 110 8.6 0.227 0.227 6.856 6.891 sum_up_and_integrate 110 10.3 0.002 0.002 6.580 6.594 integrate_v_rspace 110 11.3 0.004 0.004 6.551 6.565 dbcsr_mm_accdrv_process 11614 15.7 4.063 4.532 5.632 5.896 init_scf_run 11 5.9 0.000 0.001 5.778 5.779 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.778 5.778 make_m2s 4110 13.4 0.043 0.043 5.217 5.682 qs_ot_get_p 110 10.4 0.003 0.004 5.522 5.566 make_images 4110 14.4 0.886 0.942 5.029 5.491 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.040 5.376 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.775 5.330 apply_single 110 13.6 0.000 0.000 4.774 5.329 cp_fm_cholesky_invert 11 10.9 5.225 5.229 5.225 5.229 ot_diis_step 99 11.5 0.015 0.015 4.595 4.596 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.486 4.503 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.965 3.973 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.923 3.945 grid_integrate_task_list 110 12.3 3.679 3.737 3.679 3.737 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.656 3.657 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.161 3.638 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.448 3.448 hybrid_alltoall_any 4261 16.3 0.268 0.580 2.799 3.403 fft3d_ps 1111 13.6 1.839 1.873 3.345 3.365 density_rs2pw 110 9.6 0.004 0.004 3.314 3.335 make_images_data 4110 15.4 0.044 0.047 2.696 3.287 wfi_extrapolate 11 7.9 0.001 0.001 3.268 3.268 calculate_dm_sparse 110 9.5 0.001 0.001 3.210 3.239 multiply_cannon_sync_h2d 8220 15.4 3.120 3.137 3.120 3.137 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.982 2.986 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.870 2.870 cp_fm_diag_elpa_base 48 14.0 2.323 2.532 2.868 2.869 grid_collocate_task_list 110 9.6 2.668 2.681 2.668 2.681 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.651 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.487 2.502 calculate_first_density_matrix 1 7.0 0.000 0.000 2.399 2.399 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.169 2.229 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.109 2.213 potential_pw2rs 110 12.3 0.021 0.021 2.121 2.125 mp_alltoall_d11v 2046 13.8 2.026 2.065 2.026 2.065 cp_fm_cholesky_decompose 22 10.9 1.963 1.983 1.963 1.983 qs_create_task_list 11 7.9 0.001 0.001 1.879 1.923 generate_qs_task_list 11 8.9 0.731 0.785 1.879 1.922 acc_transpose_blocks 8220 15.4 0.042 0.042 1.839 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.837 1.842 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.732 1.774 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.134000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1290.818182, yerr=64.916475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 628.305920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 2530001. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.046 209.229 209.231 qs_mol_dyn_low 1 2.0 0.008 0.043 208.720 208.733 qs_forces 11 3.9 0.005 0.005 208.558 208.563 qs_energies 11 4.9 0.002 0.002 202.910 202.930 scf_env_do_scf 11 5.9 0.001 0.003 186.350 186.354 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 165.773 165.775 dbcsr_multiply_generic 2527 12.6 0.179 0.182 127.280 128.476 qs_scf_new_mos 118 7.6 0.001 0.001 126.250 126.513 qs_scf_loop_do_ot 118 8.6 0.001 0.001 126.250 126.512 velocity_verlet 10 3.0 0.008 0.058 125.027 125.030 ot_scf_mini 118 9.6 0.003 0.003 119.485 119.741 multiply_cannon 2527 13.6 0.241 0.250 102.896 105.421 multiply_cannon_loop 2527 14.6 2.418 2.473 100.586 103.178 ot_mini 118 10.6 0.001 0.001 67.347 67.658 multiply_cannon_multrec 60648 15.6 32.022 34.697 42.019 44.103 qs_ot_get_derivative 118 11.6 0.001 0.001 42.239 42.509 rebuild_ks_matrix 129 8.3 0.001 0.001 33.959 34.467 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.037 33.958 34.466 mp_waitall_1 269254 16.5 29.516 32.934 29.516 32.934 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.528 30.952 qs_ot_get_p 129 10.4 0.001 0.003 29.641 29.961 multiply_cannon_sync_h2d 60648 15.6 26.353 28.980 26.353 28.980 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.549 25.452 apply_single 129 13.6 0.001 0.001 24.549 25.452 ot_diis_step 118 11.6 0.008 0.008 24.875 24.876 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.932 22.993 init_scf_loop 11 6.9 0.000 0.001 20.500 20.502 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.140 20.333 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.028 20.028 multiply_cannon_metrocomm3 60648 15.6 0.118 0.123 16.457 18.720 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.016 17.048 cp_fm_diag_elpa_base 83 14.4 16.944 16.984 17.012 17.045 prepare_preconditioner 11 7.9 0.000 0.000 15.901 15.951 make_preconditioner 11 8.9 0.000 0.000 15.901 15.951 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.155 15.338 make_m2s 5054 13.6 0.106 0.114 13.950 14.339 make_images 5054 14.6 0.402 0.420 13.762 14.167 sum_up_and_integrate 129 10.3 0.002 0.004 13.852 13.880 integrate_v_rspace 129 11.3 0.004 0.004 13.793 13.825 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.844 13.012 calculate_rho_elec 129 8.7 0.045 0.064 12.843 13.011 init_scf_run 11 5.9 0.000 0.001 12.380 12.380 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.379 12.380 mp_sum_l 8010 12.9 9.176 10.498 9.176 10.498 dbcsr_mm_accdrv_process 125468 16.2 4.840 4.987 9.560 10.131 wfi_extrapolate 11 7.9 0.001 0.001 9.100 9.100 cp_fm_cholesky_invert 11 10.9 9.000 9.009 9.000 9.009 calculate_dm_sparse 129 9.5 0.001 0.001 8.580 8.691 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.439 8.609 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.255 8.345 multiply_cannon_metrocomm1 60648 15.6 0.096 0.100 6.497 8.155 make_images_data 5054 15.6 0.065 0.070 6.869 7.856 grid_integrate_task_list 129 12.3 7.120 7.504 7.120 7.504 hybrid_alltoall_any 5240 16.5 0.295 2.272 6.017 7.296 density_rs2pw 129 9.7 0.005 0.006 6.408 6.962 fft_wrap_pw1pw2 1301 11.7 0.018 0.026 6.611 6.905 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.703 6.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.816 5.962 fft_wrap_pw1pw2_140 527 12.2 0.234 0.244 5.689 5.865 fft3d_ps 1301 13.7 2.108 2.627 5.428 5.672 mp_alltoall_d11v 2423 14.1 4.330 5.580 4.330 5.580 grid_collocate_task_list 129 9.7 4.878 5.244 4.878 5.244 cp_fm_cholesky_decompose 22 10.9 4.711 4.725 4.711 4.725 mp_sum_d 4503 12.1 3.556 4.322 3.556 4.322 potential_pw2rs 129 12.3 0.009 0.010 4.289 4.316 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.231000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.727273, yerr=6.729729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 844.746752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 6385938. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.127 0.172 191.756 191.762 qs_mol_dyn_low 1 2.0 0.118 0.259 190.866 190.883 qs_forces 11 3.9 0.005 0.010 190.192 190.299 qs_energies 11 4.9 0.012 0.073 183.298 183.405 scf_env_do_scf 11 5.9 0.001 0.001 164.967 164.977 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 132.015 132.017 velocity_verlet 10 3.0 0.051 0.092 118.439 118.463 dbcsr_multiply_generic 2485 12.5 0.184 0.191 96.845 98.041 qs_scf_new_mos 116 7.6 0.001 0.001 93.984 94.603 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.983 94.602 ot_scf_mini 116 9.6 0.004 0.005 89.201 89.916 multiply_cannon 2485 13.5 0.501 0.562 76.746 81.709 multiply_cannon_loop 2485 14.5 1.549 1.611 73.367 76.217 ot_mini 116 10.6 0.001 0.001 49.852 50.478 mp_waitall_1 212858 16.6 23.865 37.241 23.865 37.241 multiply_cannon_multrec 29820 15.5 20.840 26.116 31.462 37.134 rebuild_ks_matrix 127 8.3 0.001 0.001 32.300 33.072 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 32.299 33.071 init_scf_loop 11 6.9 0.001 0.004 32.862 32.863 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.033 29.732 qs_ot_get_derivative 116 11.6 0.001 0.002 28.104 28.812 prepare_preconditioner 11 7.9 0.000 0.000 28.527 28.597 make_preconditioner 11 8.9 0.000 0.002 28.527 28.597 multiply_cannon_metrocomm3 29820 15.5 0.094 0.099 15.383 28.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.223 27.763 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.901 23.108 apply_single 127 13.6 0.001 0.001 21.900 23.107 qs_ot_get_p 127 10.4 0.002 0.003 21.477 22.115 ot_diis_step 116 11.6 0.014 0.015 21.577 21.579 multiply_cannon_sync_h2d 29820 15.5 17.900 20.108 17.900 20.108 cp_fm_cholesky_invert 11 10.9 16.651 16.664 16.651 16.664 qs_ot_p2m_diag 82 11.4 0.186 0.214 16.575 16.608 make_m2s 4970 13.5 0.086 0.090 14.167 15.819 make_images 4970 14.5 1.159 1.365 13.952 15.600 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.380 15.381 sum_up_and_integrate 127 10.3 0.002 0.004 13.704 13.744 integrate_v_rspace 127 11.3 0.003 0.004 13.644 13.689 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.835 12.871 calculate_rho_elec 127 8.7 0.087 0.105 12.834 12.871 init_scf_run 11 5.9 0.001 0.003 12.707 12.709 scf_env_initial_rho_setup 11 6.9 0.002 0.005 12.706 12.709 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.212 12.244 cp_fm_diag_elpa_base 82 14.4 11.957 12.044 12.206 12.234 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.080 11.589 multiply_cannon_metrocomm4 27335 15.5 0.104 0.117 3.824 10.728 dbcsr_mm_accdrv_process 61726 16.2 5.560 6.534 10.070 10.625 mp_irecv_dv 68888 16.3 3.617 10.328 3.617 10.328 make_images_data 4970 15.5 0.064 0.075 8.381 10.264 hybrid_alltoall_any 5155 16.4 0.346 1.526 7.108 9.842 wfi_extrapolate 11 7.9 0.003 0.018 8.338 8.338 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.363 7.438 grid_integrate_task_list 127 12.3 7.116 7.428 7.116 7.428 density_rs2pw 127 9.7 0.006 0.008 6.567 7.148 cp_fm_cholesky_decompose 22 10.9 7.004 7.084 7.004 7.084 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.281 7.041 calculate_dm_sparse 127 9.5 0.001 0.001 6.463 6.606 fft_wrap_pw1pw2_140 519 12.2 0.247 0.257 6.438 6.498 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.141 6.150 mp_sum_l 7884 12.9 4.181 6.110 4.181 6.110 fft3d_ps 1281 13.7 2.743 2.917 5.824 5.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.240 5.448 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.337 5.407 grid_collocate_task_list 127 9.7 4.959 5.369 4.959 5.369 mp_allgather_i34 2485 14.5 1.902 5.134 1.902 5.134 mp_alltoall_d11v 2401 14.1 4.130 4.609 4.130 4.609 qs_energies_init_hamiltonians 11 5.9 0.014 0.052 4.206 4.313 transfer_rs2pw 519 10.6 0.014 0.062 3.486 4.286 potential_pw2rs 127 12.3 0.014 0.016 4.264 4.284 calculate_first_density_matrix 1 7.0 0.005 0.024 4.180 4.187 mp_sum_d 4458 12.1 2.754 4.015 2.754 4.015 dbcsr_complete_redistribute 393 12.7 0.758 0.855 3.133 3.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.762000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.545455, yerr=3.201239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 946.651136E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. 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 0.047 0.066 175.207 175.208 qs_mol_dyn_low 1 2.0 0.025 0.094 174.499 174.511 qs_forces 11 3.9 0.011 0.023 174.204 174.222 qs_energies 11 4.9 0.001 0.002 167.650 167.683 scf_env_do_scf 11 5.9 0.001 0.001 151.399 151.402 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.656 116.657 velocity_verlet 10 3.0 0.043 0.169 111.790 111.820 dbcsr_multiply_generic 2507 12.6 0.184 0.188 81.958 83.105 qs_scf_new_mos 117 7.6 0.001 0.001 80.003 80.292 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.003 80.291 ot_scf_mini 117 9.6 0.003 0.004 75.764 76.092 multiply_cannon 2507 13.6 0.506 0.536 62.436 67.550 multiply_cannon_loop 2507 14.6 1.127 1.195 59.484 62.151 ot_mini 117 10.6 0.001 0.001 42.570 42.895 init_scf_loop 11 6.9 0.000 0.000 34.644 34.645 mp_waitall_1 170520 16.6 24.640 34.487 24.640 34.487 prepare_preconditioner 11 7.9 0.000 0.000 30.668 30.721 make_preconditioner 11 8.9 0.000 0.000 30.668 30.721 rebuild_ks_matrix 128 8.3 0.001 0.001 30.246 30.620 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 30.245 30.619 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.306 29.768 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.267 27.611 multiply_cannon_multrec 20056 15.6 12.992 15.808 22.404 25.202 multiply_cannon_metrocomm3 20056 15.6 0.060 0.065 15.403 24.672 qs_ot_get_derivative 117 11.6 0.001 0.002 23.060 23.389 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.626 20.623 apply_single 128 13.6 0.001 0.001 19.626 20.623 ot_diis_step 117 11.6 0.018 0.018 19.400 19.400 qs_ot_get_p 128 10.4 0.002 0.003 18.554 18.976 make_m2s 5014 13.6 0.077 0.084 14.501 15.752 make_images 5014 14.6 1.150 1.247 14.265 15.511 multiply_cannon_sync_h2d 20056 15.6 13.707 15.421 13.707 15.421 qs_ot_p2m_diag 83 11.4 0.265 0.273 14.255 14.263 cp_fm_cholesky_invert 11 10.9 14.204 14.213 14.204 14.213 sum_up_and_integrate 128 10.3 0.002 0.003 13.629 13.652 integrate_v_rspace 128 11.3 0.003 0.004 13.570 13.594 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.203 13.204 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.024 13.057 calculate_rho_elec 128 8.7 0.132 0.146 13.023 13.057 init_scf_run 11 5.9 0.000 0.001 11.200 11.201 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.200 11.200 make_images_data 5014 15.6 0.061 0.072 8.625 10.296 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.047 10.066 cp_fm_diag_elpa_base 83 14.4 9.635 9.805 10.044 10.063 hybrid_alltoall_any 5200 16.5 0.449 2.040 7.433 9.737 multiply_cannon_metrocomm4 17549 15.6 0.065 0.076 3.479 9.391 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.115 9.347 mp_irecv_dv 50230 16.2 3.348 9.127 3.348 9.127 dbcsr_mm_accdrv_process 41502 16.2 5.690 5.985 8.866 8.997 grid_integrate_task_list 128 12.3 7.321 7.730 7.321 7.730 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.483 7.585 wfi_extrapolate 11 7.9 0.001 0.001 7.377 7.377 cp_fm_cholesky_decompose 22 10.9 7.281 7.334 7.281 7.334 cp_fm_upper_to_full 105 14.8 5.779 7.314 5.779 7.314 density_rs2pw 128 9.7 0.005 0.006 6.342 6.820 fft_wrap_pw1pw2_140 523 12.2 0.282 0.290 6.592 6.705 dbcsr_complete_redistribute 395 12.7 1.177 1.212 4.581 6.322 calculate_dm_sparse 128 9.5 0.001 0.001 5.880 5.991 fft3d_ps 1291 13.7 2.760 2.982 5.796 5.867 grid_collocate_task_list 128 9.7 5.153 5.521 5.153 5.521 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.491 5.500 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.602 5.286 mp_alltoall_d11v 2415 14.1 4.299 5.220 4.299 5.220 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.437 5.180 mp_sum_l 7950 12.9 3.197 4.722 3.197 4.722 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.578 4.705 mp_allgather_i34 2507 14.6 1.446 4.386 1.446 4.386 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.342 4.073 potential_pw2rs 128 12.3 0.020 0.022 4.040 4.048 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.927 3.962 mp_alltoall_i22 716 14.1 1.965 3.853 1.965 3.853 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.783 3.807 calculate_first_density_matrix 1 7.0 0.000 0.000 3.638 3.642 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.488 3.523 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.208000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.727273, yerr=7.398905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.176543E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509604. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.088 188.850 188.854 qs_mol_dyn_low 1 2.0 0.003 0.003 188.406 188.422 qs_forces 11 3.9 0.004 0.004 188.288 188.294 qs_energies 11 4.9 0.002 0.003 181.204 181.217 scf_env_do_scf 11 5.9 0.001 0.001 164.189 164.201 velocity_verlet 10 3.0 0.006 0.018 125.202 125.208 scf_env_do_scf_inner_loop 118 6.6 0.005 0.010 117.524 117.527 dbcsr_multiply_generic 2529 12.6 0.192 0.197 81.099 81.750 qs_scf_new_mos 118 7.6 0.001 0.001 81.380 81.736 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.379 81.735 ot_scf_mini 118 9.6 0.003 0.003 76.550 76.891 multiply_cannon 2529 13.6 0.555 0.583 56.266 58.067 multiply_cannon_loop 2529 14.6 1.861 1.960 52.749 54.358 init_scf_loop 11 6.9 0.000 0.000 46.544 46.546 ot_mini 118 10.6 0.001 0.001 43.039 43.367 prepare_preconditioner 11 7.9 0.000 0.000 42.478 42.501 make_preconditioner 11 8.9 0.000 0.000 42.478 42.501 make_full_inverse_cholesky 11 9.9 0.011 0.022 36.006 41.122 multiply_cannon_multrec 30348 15.6 13.760 19.201 27.099 32.298 rebuild_ks_matrix 129 8.3 0.001 0.001 29.441 29.672 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 29.440 29.671 mp_waitall_1 149172 16.7 16.977 26.943 16.977 26.943 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.613 26.819 qs_ot_get_derivative 118 11.6 0.001 0.002 23.593 23.942 make_m2s 5058 13.6 0.093 0.098 20.488 20.978 make_images 5058 14.6 1.967 2.307 20.187 20.680 qs_ot_get_p 129 10.4 0.016 0.024 19.585 19.972 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 18.884 19.409 apply_single 129 13.6 0.001 0.001 18.884 19.409 ot_diis_step 118 11.6 0.018 0.018 19.325 19.326 cp_fm_upper_to_full 106 14.8 11.709 17.238 11.709 17.238 cp_fm_cholesky_invert 11 10.9 16.062 16.071 16.062 16.071 qs_ot_p2m_diag 84 11.4 0.347 0.394 15.384 15.438 multiply_cannon_metrocomm3 30348 15.6 0.050 0.053 6.369 15.090 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.956 13.959 sum_up_and_integrate 129 10.3 0.002 0.003 13.798 13.823 integrate_v_rspace 129 11.3 0.003 0.004 13.737 13.764 dbcsr_mm_accdrv_process 62780 16.2 8.620 9.554 12.901 13.450 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.349 13.396 calculate_rho_elec 129 8.7 0.175 0.191 13.348 13.395 dbcsr_complete_redistribute 397 12.7 1.487 1.596 9.239 13.093 make_images_data 5058 15.6 0.065 0.071 10.800 11.831 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.855 11.691 hybrid_alltoall_any 5245 16.5 0.531 2.219 9.855 11.442 init_scf_run 11 5.9 0.000 0.001 11.124 11.125 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.124 11.125 multiply_cannon_sync_h2d 30348 15.6 10.448 10.991 10.448 10.991 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.672 10.690 cp_fm_diag_elpa_base 84 14.4 9.685 10.024 10.665 10.682 transfer_fm_to_dbcsr 11 9.9 0.002 0.011 6.450 10.218 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.963 10.213 mp_alltoall_i22 720 14.1 5.677 9.573 5.677 9.573 grid_integrate_task_list 129 12.3 7.582 7.949 7.582 7.949 fft_wrap_pw1pw2 1301 11.7 0.019 0.024 7.808 7.865 cp_fm_cholesky_decompose 22 10.9 7.510 7.591 7.510 7.591 wfi_extrapolate 11 7.9 0.001 0.001 7.572 7.572 multiply_cannon_metrocomm4 25290 15.6 0.079 0.090 2.822 7.478 mp_irecv_dv 76751 16.2 2.667 7.193 2.667 7.193 fft_wrap_pw1pw2_140 527 12.2 0.327 0.337 6.897 6.981 density_rs2pw 129 9.7 0.005 0.006 6.294 6.728 calculate_dm_sparse 129 9.5 0.001 0.001 6.517 6.587 fft3d_ps 1301 13.7 2.952 2.999 5.987 6.041 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.677 5.743 grid_collocate_task_list 129 9.7 5.327 5.714 5.327 5.714 mp_alltoall_d11v 2429 14.1 5.099 5.354 5.099 5.354 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.591 4.670 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.513 4.625 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.393 4.393 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.287 4.369 potential_pw2rs 129 12.3 0.022 0.022 4.002 4.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.854000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1095.181818, yerr=29.004132 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963540389888 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019185291264 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019185291264 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820203E+12 0.0% 0.0% 100.0% flops max/rank 5.821533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705859104 0.0% 0.0% 100.0% number of processed stacks 1948544 0.0% 0.0% 100.0% average stack size 0.0 0.0 3441.5 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 1.560527E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 239616 MPI messages size (bytes): total size 1.321658E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.515732E+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 113856 59693334528 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108641296 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8885 51. MP_Alltoall 9606 802511. MP_ISend 39892 2096308. MP_IRecv 39892 2095413. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 58940. MP_Allreduce 10955 1170. MP_Sync 86 MP_Alltoall 1700 18828285. 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 0.061 0.078 170.936 170.936 qs_mol_dyn_low 1 2.0 0.003 0.003 170.281 170.307 qs_forces 11 3.9 0.004 0.005 170.150 170.153 qs_energies 11 4.9 0.032 0.050 162.737 162.744 scf_env_do_scf 11 5.9 0.001 0.001 144.096 144.113 velocity_verlet 10 3.0 0.047 0.048 110.590 110.594 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.131 108.133 dbcsr_multiply_generic 2496 12.6 0.182 0.189 71.986 72.263 qs_scf_new_mos 116 7.6 0.001 0.001 72.012 72.107 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.012 72.106 ot_scf_mini 116 9.6 0.003 0.004 67.639 67.714 multiply_cannon 2496 13.6 0.556 0.589 53.453 56.941 multiply_cannon_loop 2496 14.6 0.804 0.840 50.461 51.126 ot_mini 116 10.6 0.001 0.001 37.263 37.337 init_scf_loop 11 6.9 0.000 0.000 35.814 35.816 prepare_preconditioner 11 7.9 0.000 0.000 31.949 31.976 make_preconditioner 11 8.9 0.000 0.000 31.949 31.976 mp_waitall_1 125208 16.7 24.462 30.816 24.462 30.816 make_full_inverse_cholesky 11 9.9 0.013 0.027 29.764 30.060 rebuild_ks_matrix 127 8.3 0.001 0.001 28.947 29.029 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.947 29.029 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.379 26.456 multiply_cannon_multrec 9984 15.6 10.356 14.289 18.308 21.190 qs_ot_get_derivative 116 11.6 0.001 0.002 20.423 20.502 multiply_cannon_metrocomm3 9984 15.6 0.025 0.026 12.387 19.324 cp_fm_cholesky_invert 11 10.9 18.203 18.209 18.203 18.209 qs_ot_get_p 127 10.4 0.046 0.058 17.777 17.841 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.876 17.009 apply_single 127 13.6 0.001 0.001 16.875 17.008 ot_diis_step 116 11.6 0.020 0.020 16.771 16.771 make_m2s 4992 13.6 0.064 0.069 14.742 15.552 make_images 4992 14.6 2.140 2.561 14.435 15.243 qs_ot_p2m_diag 82 11.4 0.490 0.495 13.948 13.987 sum_up_and_integrate 127 10.3 0.002 0.002 13.752 13.799 integrate_v_rspace 127 11.3 0.004 0.004 13.691 13.740 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.481 13.521 calculate_rho_elec 127 8.7 0.255 0.266 13.481 13.521 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.779 12.816 init_scf_run 11 5.9 0.000 0.001 11.741 11.741 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.741 11.741 multiply_cannon_sync_h2d 9984 15.6 10.843 11.175 10.843 11.175 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.894 9.906 cp_fm_diag_elpa_base 82 14.4 9.653 9.730 9.890 9.902 make_images_data 4992 15.6 0.055 0.064 8.508 9.804 hybrid_alltoall_any 5177 16.4 0.839 3.755 8.374 9.745 dbcsr_mm_accdrv_process 20634 16.1 3.271 4.284 7.604 8.348 fft_wrap_pw1pw2 1281 11.7 0.018 0.020 8.148 8.178 cp_fm_cholesky_decompose 22 10.9 8.017 8.114 8.017 8.114 grid_integrate_task_list 127 12.3 7.688 8.042 7.688 8.042 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.957 8.015 wfi_extrapolate 11 7.9 0.001 0.001 7.281 7.281 fft_wrap_pw1pw2_140 519 12.2 0.431 0.438 7.146 7.176 multiply_cannon_metrocomm1 9984 15.6 0.030 0.031 4.308 7.130 density_rs2pw 127 9.7 0.005 0.005 6.216 6.587 calculate_dm_sparse 127 9.5 0.001 0.001 6.143 6.218 fft3d_ps 1281 13.7 3.089 3.202 6.134 6.175 grid_collocate_task_list 127 9.7 5.533 5.849 5.533 5.849 dbcsr_complete_redistribute 393 12.7 2.172 2.230 5.450 5.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.685 5.695 qs_energies_init_hamiltonians 11 5.9 0.018 0.020 5.276 5.277 mp_alltoall_d11v 2401 14.1 4.729 5.232 4.729 5.232 mp_allgather_i34 2496 14.6 1.133 4.583 1.133 4.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.466 4.499 calculate_first_density_matrix 1 7.0 0.000 0.000 4.185 4.188 potential_pw2rs 127 12.3 0.026 0.027 3.969 3.976 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.590 3.935 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.602 3.891 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.715 3.773 multiply_cannon_metrocomm4 7488 15.6 0.025 0.027 1.686 3.658 mp_irecv_dv 28706 15.9 1.648 3.589 1.648 3.589 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.537 3.564 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.457 3.515 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.464 3.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.936000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1469.909091, yerr=28.820619 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 11.696234E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.146719E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350290592 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 58881. MP_Allreduce 11127 1502. MP_Sync 87 MP_Alltoall 1712 36974169. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.055 291.291 291.292 qs_mol_dyn_low 1 2.0 0.004 0.012 290.697 290.718 qs_forces 11 3.9 0.004 0.005 290.591 290.595 qs_energies 11 4.9 0.109 0.126 281.741 281.747 scf_env_do_scf 11 5.9 0.001 0.001 258.587 258.598 velocity_verlet 10 3.0 0.017 0.019 210.164 210.172 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 134.534 134.535 init_scf_loop 11 6.9 0.000 0.000 123.763 123.764 prepare_preconditioner 11 7.9 0.000 0.000 119.034 119.054 make_preconditioner 11 8.9 0.000 0.000 119.034 119.054 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.691 116.135 qs_scf_new_mos 117 7.6 0.001 0.001 91.429 91.531 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.428 91.530 ot_scf_mini 117 9.6 0.004 0.004 86.565 86.615 dbcsr_multiply_generic 2507 12.6 0.220 0.231 82.898 83.347 cp_fm_upper_to_full 105 14.8 52.726 75.000 52.726 75.000 multiply_cannon 2507 13.6 0.685 0.723 59.440 61.086 multiply_cannon_loop 2507 14.6 1.047 1.071 55.398 56.737 ot_mini 117 10.6 0.001 0.001 45.196 45.275 dbcsr_complete_redistribute 395 12.7 3.984 4.073 29.392 42.048 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.826 38.435 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.299 35.763 rebuild_ks_matrix 128 8.3 0.001 0.001 33.912 33.988 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 33.911 33.988 cp_fm_cholesky_invert 11 10.9 33.685 33.691 33.685 33.691 mp_alltoall_i22 716 14.1 21.056 33.505 21.056 33.505 mp_waitall_1 103674 16.8 28.467 32.321 28.467 32.321 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.540 31.607 qs_ot_get_p 128 10.4 0.003 0.003 26.426 26.571 qs_ot_get_derivative 117 11.6 0.002 0.002 24.980 25.028 qs_ot_p2m_diag 83 11.4 0.879 0.885 22.218 22.248 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.409 20.412 multiply_cannon_metrocomm3 10028 15.6 0.024 0.026 19.118 20.325 make_m2s 5014 13.6 0.074 0.076 18.729 20.192 ot_diis_step 117 11.6 0.022 0.022 20.175 20.175 make_images 5014 14.6 3.079 3.266 18.245 19.708 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.397 19.625 apply_single 128 13.6 0.001 0.001 19.397 19.625 multiply_cannon_multrec 10028 15.6 10.379 12.213 18.293 18.394 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.177 17.178 cp_fm_diag_elpa_base 83 14.4 12.786 14.390 17.173 17.173 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.019 16.030 calculate_rho_elec 128 8.7 0.482 0.483 16.019 16.029 sum_up_and_integrate 128 10.3 0.002 0.002 15.853 15.940 integrate_v_rspace 128 11.3 0.004 0.005 15.789 15.876 multiply_cannon_sync_h2d 10028 15.6 14.308 14.328 14.308 14.328 hybrid_alltoall_any 5200 16.5 1.316 3.055 10.836 12.917 make_images_data 5014 15.6 0.062 0.067 10.616 12.776 init_scf_run 11 5.9 0.000 0.001 12.645 12.646 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.645 12.646 fft_wrap_pw1pw2 1291 11.7 0.021 0.021 10.173 10.187 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.805 9.841 dbcsr_mm_accdrv_process 20762 16.1 4.257 6.201 7.662 9.479 cp_fm_cholesky_decompose 22 10.9 9.073 9.097 9.073 9.097 wfi_extrapolate 11 7.9 0.001 0.001 9.010 9.010 fft_wrap_pw1pw2_140 523 12.2 0.824 0.826 8.980 8.990 grid_integrate_task_list 128 12.3 8.554 8.758 8.554 8.758 qs_energies_init_hamiltonians 11 5.9 0.010 0.069 8.163 8.165 fft3d_ps 1291 13.7 3.987 3.998 7.473 7.499 mp_alltoall_d11v 2415 14.1 7.141 7.339 7.141 7.339 density_rs2pw 128 9.7 0.005 0.005 7.006 7.058 calculate_dm_sparse 128 9.5 0.001 0.001 6.768 6.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.541 6.618 grid_collocate_task_list 128 9.7 6.413 6.490 6.413 6.490 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.283 6.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.292000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2860.363636, yerr=138.682943 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 1.259184E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 25576241. 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 0.083 0.224 86.885 86.905 qs_energies 1 2.0 0.003 0.026 85.913 86.062 ls_scf 1 3.0 0.002 0.016 84.433 84.551 dbcsr_multiply_generic 111 6.7 0.017 0.038 72.704 72.868 multiply_cannon 111 7.7 0.018 0.027 55.904 57.089 multiply_cannon_loop 111 8.7 0.227 0.244 52.508 53.949 ls_scf_main 1 4.0 0.001 0.020 52.469 52.474 density_matrix_trs4 2 5.0 0.003 0.043 46.861 46.929 ls_scf_init_scf 1 4.0 0.002 0.007 28.902 28.928 ls_scf_init_matrix_S 1 5.0 0.001 0.011 27.486 27.548 mp_waitall_1 11031 10.9 22.344 25.371 22.344 25.371 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.011 25.011 25.033 multiply_cannon_multrec 2664 9.7 8.125 8.957 15.582 17.389 multiply_cannon_sync_h2d 2664 9.7 13.625 15.470 13.625 15.470 make_m2s 222 7.7 0.008 0.011 13.127 13.639 make_images 222 8.7 0.099 0.109 13.105 13.620 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.565 13.255 hybrid_alltoall_any 227 10.6 0.223 1.881 6.674 8.323 make_images_data 222 9.7 0.005 0.020 7.680 8.307 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.711 7.073 8.015 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.444 7.726 dbcsr_mm_accdrv_process_sort 4760 11.4 6.281 7.159 6.281 7.159 calculate_norms 4752 9.8 5.523 6.193 5.523 6.193 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.061 5.252 mp_sum_l 887 5.1 3.164 4.882 3.164 4.882 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.071 3.709 mp_irecv_dv 6231 10.9 2.056 3.683 2.056 3.683 make_images_sizes 222 9.7 0.000 0.000 0.810 3.679 mp_alltoall_i44 222 10.7 0.810 3.678 0.810 3.678 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.329 3.674 arnoldi_extremal 4 6.8 0.000 0.000 3.194 3.222 arnoldi_normal_ev 4 7.8 0.001 0.002 3.194 3.222 ls_scf_post 1 4.0 0.005 0.036 3.060 3.166 build_subspace 16 8.4 0.009 0.012 3.096 3.099 ls_scf_store_result 1 5.0 0.000 0.000 2.864 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.278 2.685 dbcsr_merge_single_wm 555 10.7 0.448 0.567 2.270 2.676 make_images_pack 222 9.7 2.212 2.627 2.214 2.629 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.309 2.554 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.457 2.065 2.459 dbcsr_sort_data 658 11.4 2.081 2.434 2.081 2.434 ls_scf_dm_to_ks 2 5.0 0.000 0.004 2.313 2.417 buffer_matrices_ensure_size 222 8.7 1.743 2.054 1.743 2.054 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.973 2.003 rebuild_ks_matrix 3 7.3 0.000 0.000 1.964 1.994 qs_ks_build_kohn_sham_matrix 3 8.3 0.016 0.086 1.964 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.905000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 2.170851E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.054 90.359 90.359 qs_energies 1 2.0 0.000 0.000 89.804 89.809 ls_scf 1 3.0 0.000 0.000 88.348 88.352 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.143 74.457 multiply_cannon 111 7.7 0.028 0.046 52.448 57.254 ls_scf_main 1 4.0 0.000 0.000 54.397 54.400 multiply_cannon_loop 111 8.7 0.135 0.146 49.650 53.491 density_matrix_trs4 2 5.0 0.002 0.003 48.739 48.976 ls_scf_init_scf 1 4.0 0.000 0.001 30.268 30.269 mp_waitall_1 9105 10.9 20.201 29.376 20.201 29.376 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.049 29.162 multiply_cannon_multrec 1332 9.7 13.421 17.061 22.791 27.538 matrix_sqrt_Newton_Schulz 2 6.5 0.053 0.078 26.563 26.572 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 10.998 20.795 make_m2s 222 7.7 0.006 0.008 15.103 15.645 make_images 222 8.7 1.370 1.697 15.072 15.616 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.543 8.960 10.532 dbcsr_mm_accdrv_process_sort 4041 11.4 8.474 9.989 8.474 9.989 hybrid_alltoall_any 227 10.6 0.543 2.591 8.090 9.716 make_images_data 222 9.7 0.004 0.005 8.729 9.595 mp_sum_l 887 5.1 5.120 8.288 5.120 8.288 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.205 7.672 mp_irecv_dv 3311 11.0 3.184 7.616 3.184 7.616 calculate_norms 2376 9.8 6.068 6.695 6.068 6.695 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.877 6.645 multiply_cannon_sync_h2d 1332 9.7 4.955 6.346 4.955 6.346 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.907 5.102 arnoldi_extremal 4 6.8 0.000 0.000 4.796 4.858 arnoldi_normal_ev 4 7.8 0.001 0.005 4.795 4.858 build_subspace 16 8.4 0.014 0.021 4.392 4.393 ls_scf_post 1 4.0 0.000 0.000 3.683 3.687 ls_scf_store_result 1 5.0 0.000 0.000 3.412 3.504 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.155 3.395 dbcsr_matrix_vector_mult_local 304 10.0 2.766 3.264 2.768 3.266 mp_allgather_i34 111 8.7 0.771 3.019 0.771 3.019 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.594 2.685 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.123 2.586 dbcsr_data_new 4174 10.1 2.123 2.413 2.123 2.413 make_images_pack 222 9.7 1.821 2.129 1.824 2.131 dbcsr_sort_data 436 11.2 1.845 2.094 1.845 2.094 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.891 1.893 rebuild_ks_matrix 3 7.3 0.000 0.000 1.878 1.880 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.015 1.878 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.359000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 2.938958E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.063 93.827 93.829 qs_energies 1 2.0 0.000 0.000 93.138 93.142 ls_scf 1 3.0 0.000 0.000 91.281 91.284 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.903 76.222 ls_scf_main 1 4.0 0.000 0.000 57.049 57.053 multiply_cannon 111 7.7 0.031 0.066 52.315 56.314 multiply_cannon_loop 111 8.7 0.116 0.130 49.555 54.217 density_matrix_trs4 2 5.0 0.002 0.003 51.120 51.371 mp_waitall_1 7281 11.0 23.745 34.296 23.745 34.296 ls_scf_init_scf 1 4.0 0.013 0.014 30.596 30.597 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.416 29.480 matrix_sqrt_Newton_Schulz 2 6.5 0.006 0.020 27.002 27.016 multiply_cannon_multrec 888 9.7 12.650 15.763 21.300 25.242 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.232 23.964 make_m2s 222 7.7 0.006 0.006 16.420 17.182 make_images 222 8.7 1.584 1.863 16.382 17.141 make_images_data 222 9.7 0.004 0.004 9.519 10.708 hybrid_alltoall_any 227 10.6 0.638 2.919 9.091 10.317 dbcsr_mm_accdrv_process 3754 10.4 0.324 0.493 8.166 9.413 mp_sum_l 887 5.1 5.490 9.195 5.490 9.195 dbcsr_mm_accdrv_process_sort 3754 11.4 7.730 8.921 7.730 8.921 multiply_cannon_sync_h2d 888 9.7 6.023 7.568 6.023 7.568 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.479 7.318 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.264 7.313 mp_irecv_dv 2335 11.1 2.463 7.258 2.463 7.258 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.635 6.652 arnoldi_extremal 4 6.8 0.000 0.000 5.122 5.138 arnoldi_normal_ev 4 7.8 0.001 0.004 5.122 5.138 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.798 5.042 calculate_norms 1584 9.8 4.406 4.837 4.406 4.837 build_subspace 16 8.4 0.014 0.020 4.820 4.825 mp_allgather_i34 111 8.7 0.876 3.804 0.876 3.804 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.461 3.797 ls_scf_post 1 4.0 0.000 0.000 3.637 3.642 dbcsr_matrix_vector_mult_local 304 10.0 3.048 3.638 3.050 3.639 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.838 2.938 dbcsr_data_new 4116 9.9 2.121 2.468 2.121 2.468 make_images_sizes 222 9.7 0.000 0.000 1.083 2.378 mp_alltoall_i44 222 10.7 1.083 2.378 1.083 2.378 dbcsr_sort_data 325 11.1 1.854 2.112 1.854 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.915 1.917 dbcsr_finalize 304 7.8 0.026 0.032 1.609 1.913 rebuild_ks_matrix 3 7.3 0.000 0.000 1.897 1.899 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.897 1.899 make_images_pack 222 9.7 1.630 1.884 1.633 1.887 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.829000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2252.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 3.430531E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.114 97.475 97.482 qs_energies 1 2.0 0.001 0.014 96.718 96.765 ls_scf 1 3.0 0.000 0.001 94.855 94.876 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.441 78.663 ls_scf_main 1 4.0 0.001 0.019 58.783 58.785 multiply_cannon 111 7.7 0.041 0.080 51.840 55.754 density_matrix_trs4 2 5.0 0.005 0.067 52.675 52.816 multiply_cannon_loop 111 8.7 0.151 0.168 46.831 49.739 ls_scf_init_scf 1 4.0 0.000 0.002 32.809 32.815 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.397 31.463 mp_waitall_1 6369 11.0 22.832 30.364 22.832 30.364 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.931 28.945 multiply_cannon_multrec 1332 9.7 14.168 17.147 22.147 24.908 make_m2s 222 7.7 0.007 0.007 21.183 22.534 make_images 222 8.7 3.146 3.611 21.132 22.486 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.340 16.894 make_images_data 222 9.7 0.004 0.004 11.847 13.405 hybrid_alltoall_any 227 10.6 0.799 3.787 11.241 12.979 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.478 7.620 9.155 dbcsr_mm_accdrv_process_sort 3641 11.4 7.179 8.657 7.179 8.657 mp_sum_l 887 5.1 3.877 6.835 3.877 6.835 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.084 6.086 mp_irecv_dv 3229 10.9 2.060 6.016 2.060 6.016 multiply_cannon_sync_h2d 1332 9.7 5.418 5.955 5.418 5.955 arnoldi_extremal 4 6.8 0.000 0.000 5.246 5.259 arnoldi_normal_ev 4 7.8 0.001 0.004 5.246 5.258 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.961 5.254 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.479 4.938 build_subspace 16 8.4 0.014 0.021 4.910 4.917 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.536 4.750 calculate_norms 2376 9.8 4.214 4.521 4.214 4.521 mp_allgather_i34 111 8.7 2.120 4.243 2.120 4.243 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.603 3.919 dbcsr_matrix_vector_mult_local 304 10.0 3.231 3.741 3.233 3.743 dbcsr_sort_data 658 11.4 3.077 3.444 3.077 3.444 ls_scf_post 1 4.0 0.002 0.017 3.262 3.285 dbcsr_special_finalize 555 9.7 0.006 0.007 2.837 3.266 dbcsr_merge_single_wm 555 10.7 0.541 0.664 2.828 3.258 ls_scf_dm_to_ks 2 5.0 0.001 0.010 2.984 3.030 ls_scf_store_result 1 5.0 0.000 0.000 2.970 3.025 dbcsr_data_release 10477 10.7 1.588 2.466 1.588 2.466 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.010 2.019 rebuild_ks_matrix 3 7.3 0.000 0.000 1.987 1.996 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.015 1.987 1.996 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.482000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 4.813738E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.077 91.791 91.791 qs_energies 1 2.0 0.000 0.000 91.014 91.019 ls_scf 1 3.0 0.000 0.000 89.064 89.069 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.595 70.769 ls_scf_main 1 4.0 0.000 0.000 56.299 56.299 multiply_cannon 111 7.7 0.074 0.117 52.242 55.839 multiply_cannon_loop 111 8.7 0.088 0.095 49.710 51.301 density_matrix_trs4 2 5.0 0.002 0.003 49.293 49.357 ls_scf_init_scf 1 4.0 0.000 0.001 29.340 29.345 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.111 28.146 mp_waitall_1 5436 11.0 23.917 27.986 23.917 27.986 matrix_sqrt_Newton_Schulz 2 6.5 0.041 0.055 26.029 26.039 multiply_cannon_multrec 444 9.7 13.808 16.120 20.978 23.749 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.325 15.315 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.110 14.823 make_m2s 222 7.7 0.005 0.005 13.574 14.472 make_images 222 8.7 2.038 2.480 13.507 14.402 hybrid_alltoall_any 227 10.6 0.802 3.838 8.106 9.724 make_images_data 222 9.7 0.003 0.004 8.299 9.539 multiply_cannon_sync_h2d 444 9.7 6.688 8.417 6.688 8.417 dbcsr_mm_accdrv_process 3003 10.4 0.369 0.420 6.849 7.960 dbcsr_mm_accdrv_process_sort 3003 11.4 6.481 7.565 6.481 7.565 arnoldi_extremal 4 6.8 0.000 0.000 5.882 5.913 arnoldi_normal_ev 4 7.8 0.001 0.005 5.882 5.913 build_subspace 16 8.4 0.015 0.020 5.451 5.460 mp_sum_l 887 5.1 2.787 5.210 2.787 5.210 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.492 4.711 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.195 4.393 dbcsr_matrix_vector_mult_local 304 10.0 3.734 4.187 3.736 4.189 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.923 4.151 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.582 3.907 mp_irecv_dv 1241 11.2 1.565 3.874 1.565 3.874 calculate_norms 792 9.8 3.613 3.732 3.613 3.732 mp_allgather_i34 111 8.7 1.098 3.563 1.098 3.563 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.487 3.559 ls_scf_post 1 4.0 0.000 0.000 3.426 3.431 make_images_sizes 222 9.7 0.000 0.000 0.883 3.351 mp_alltoall_i44 222 10.7 0.883 3.351 0.883 3.351 ls_scf_store_result 1 5.0 0.000 0.000 3.220 3.252 dbcsr_data_new 4608 9.7 1.810 2.269 1.810 2.269 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.261 dbcsr_merge_all 275 8.9 0.479 0.521 2.055 2.097 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.032 2.032 rebuild_ks_matrix 3 7.3 0.000 0.000 1.999 1.999 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.999 1.999 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 1.934 1.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.791000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3757.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 9.030803E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.084 0.104 104.927 104.928 qs_energies 1 2.0 0.000 0.000 103.419 103.422 ls_scf 1 3.0 0.000 0.000 100.465 100.468 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.343 74.491 ls_scf_main 1 4.0 0.000 0.000 63.122 63.123 density_matrix_trs4 2 5.0 0.002 0.003 54.330 54.390 multiply_cannon 111 7.7 0.146 0.187 48.251 50.327 multiply_cannon_loop 111 8.7 0.098 0.099 45.226 45.806 ls_scf_init_scf 1 4.0 0.001 0.001 33.544 33.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.129 32.152 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.432 29.441 mp_waitall_1 4527 11.1 21.369 25.167 21.369 25.167 make_m2s 222 7.7 0.005 0.005 22.435 23.555 make_images 222 8.7 3.567 3.884 22.327 23.445 multiply_cannon_multrec 444 9.7 17.834 18.489 22.515 23.204 hybrid_alltoall_any 227 10.6 1.653 3.617 12.607 15.376 make_images_data 222 9.7 0.003 0.004 12.788 14.961 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.001 10.313 multiply_cannon_sync_h2d 444 9.7 8.777 8.835 8.777 8.835 arnoldi_extremal 4 6.8 0.000 0.000 7.483 7.505 arnoldi_normal_ev 4 7.8 0.002 0.008 7.483 7.505 build_subspace 16 8.4 0.026 0.036 6.928 6.946 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.546 5.701 dbcsr_matrix_vector_mult_local 304 10.0 5.124 5.422 5.127 5.425 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.887 5.161 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.907 5.000 dbcsr_mm_accdrv_process 1814 10.4 0.248 0.362 4.502 4.627 dbcsr_mm_accdrv_process_sort 1814 11.4 4.163 4.298 4.163 4.298 ls_scf_post 1 4.0 0.000 0.000 3.798 3.801 make_images_sizes 222 9.7 0.000 0.000 1.498 3.638 mp_alltoall_i44 222 10.7 1.498 3.638 1.498 3.638 mp_allgather_i34 111 8.7 1.129 3.571 1.129 3.571 ls_scf_store_result 1 5.0 0.000 0.000 3.508 3.543 calculate_norms 792 9.8 3.246 3.312 3.246 3.312 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.185 dbcsr_merge_all 275 8.9 0.886 0.918 2.863 2.963 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.923 2.923 dbcsr_complete_redistribute 5 7.6 1.421 1.471 2.742 2.863 dbcsr_sort_data 325 11.1 2.437 2.513 2.437 2.513 matrix_ls_to_qs 2 6.0 0.000 0.000 2.389 2.506 dbcsr_new_transposed 4 7.5 0.255 0.279 2.325 2.349 dbcsr_data_new 6591 9.6 1.863 2.310 1.863 2.310 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.280 2.282 dbcsr_frobenius_norm 74 6.6 2.057 2.133 2.207 2.241 rebuild_ks_matrix 3 7.3 0.000 0.000 2.214 2.216 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.214 2.216 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.204 dbcsr_add_anytype 103 7.2 0.859 0.891 2.125 2.204 dbcsr_data_release 12724 10.6 1.989 2.179 1.989 2.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=104.928000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7197.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d3f223ab808785118aa8f0a47e0c6d072683d8a1_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 585.674752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993457479. 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 0.096 0.280 270.035 270.037 qs_mol_dyn_low 1 2.0 0.021 0.145 268.667 268.687 qs_forces 5 3.8 0.051 0.092 268.368 268.380 qs_energies 5 4.8 0.003 0.023 264.980 265.069 scf_env_do_scf 5 5.8 0.000 0.002 250.358 250.361 scf_env_do_scf_inner_loop 105 6.6 0.007 0.041 218.790 218.809 qs_scf_new_mos 105 7.6 0.000 0.001 172.632 172.860 qs_scf_loop_do_ot 105 8.6 0.001 0.001 172.632 172.860 ot_scf_mini 105 9.6 0.003 0.004 162.642 162.816 dbcsr_multiply_generic 1445 12.2 0.132 0.140 135.163 135.788 multiply_cannon 1445 13.2 0.275 0.304 114.362 116.110 velocity_verlet 4 3.0 0.022 0.145 114.309 114.351 multiply_cannon_loop 1445 14.2 2.851 2.995 112.078 113.607 qs_ot_get_p 112 10.4 0.001 0.005 69.923 70.216 ot_mini 105 10.6 0.001 0.001 62.348 62.488 qs_ot_p2m_diag 40 11.0 0.020 0.034 57.804 57.938 cp_dbcsr_syevd 40 12.0 0.002 0.013 54.229 54.231 cp_fm_syevd 40 13.0 0.001 0.013 48.234 48.372 multiply_cannon_multrec 69360 15.2 29.732 34.367 39.834 44.878 mp_waitall_1 488190 16.1 37.741 44.321 37.741 44.321 cp_fm_redistribute_end 40 14.0 21.403 42.704 21.411 42.709 cp_fm_syevd_base 40 14.0 21.279 42.583 21.279 42.583 qs_ot_get_derivative 55 11.6 0.001 0.001 40.457 40.605 multiply_cannon_metrocomm3 69360 15.2 0.200 0.209 26.291 33.191 multiply_cannon_sync_h2d 69360 15.2 28.900 32.817 28.900 32.817 rebuild_ks_matrix 110 8.4 0.000 0.000 31.421 31.616 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.089 31.421 31.615 init_scf_loop 7 6.6 0.002 0.068 31.516 31.519 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.827 29.011 prepare_preconditioner 7 7.6 0.000 0.000 25.936 25.975 make_preconditioner 7 8.6 0.002 0.060 25.936 25.975 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.151 23.339 apply_single 62 13.6 0.000 0.000 23.151 23.339 qs_rho_update_rho_low 110 7.6 0.000 0.001 21.521 21.937 calculate_rho_elec 110 8.6 0.029 0.044 21.521 21.937 ot_new_cg_direction 55 11.6 0.001 0.001 21.130 21.131 make_full_inverse_cholesky 7 9.6 0.000 0.000 17.582 17.646 density_rs2pw 110 9.6 0.004 0.026 15.190 15.776 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.065 15.256 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.291 14.396 mp_sum_l 4764 12.2 12.697 13.421 12.697 13.421 transfer_rs2pw 445 10.6 0.008 0.009 11.911 12.540 cp_fm_cholesky_invert 7 10.6 11.880 11.893 11.880 11.893 calculate_dm_sparse 110 9.5 0.000 0.001 11.711 11.858 fft_wrap_pw1pw2 1425 12.5 0.020 0.026 11.434 11.844 init_scf_run 5 5.8 0.000 0.001 11.736 11.737 scf_env_initial_rho_setup 5 6.8 0.001 0.003 11.736 11.737 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 11.071 11.152 dbcsr_mm_accdrv_process 154766 15.8 6.303 6.472 9.969 10.988 check_diag 80 13.5 8.684 8.935 10.835 10.974 sum_up_and_integrate 60 10.3 0.001 0.017 10.876 10.894 integrate_v_rspace 60 11.3 0.002 0.004 10.859 10.878 qs_vxc_create 110 10.4 0.002 0.009 10.163 10.187 fft_wrap_pw1pw2_240 915 14.0 0.204 0.231 9.870 10.183 fft3d_pb 915 15.0 2.267 2.516 8.820 9.166 make_m2s 2890 13.2 0.078 0.085 7.714 8.348 make_images 2890 14.2 0.237 0.258 7.607 8.238 make_full_single_inverse 7 9.6 0.001 0.040 8.002 8.041 xc_rho_set_and_dset_create 110 12.4 0.074 0.163 7.181 7.423 multiply_cannon_metrocomm1 69360 15.2 0.099 0.104 4.765 7.406 potential_pw2rs 60 12.3 0.002 0.003 6.889 6.942 xc_vxc_pw_create 60 11.3 0.036 0.051 6.850 6.876 acc_transpose_blocks 69360 15.2 0.359 0.377 6.353 6.757 mp_alltoall_z22v 2340 16.7 6.011 6.474 6.011 6.474 transfer_rs2pw_30 110 11.6 1.254 1.307 5.999 6.369 calculate_first_density_matrix 1 7.0 0.001 0.021 6.340 6.351 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.265 6.282 mp_sendrecv_dv 168740 12.6 5.888 6.082 5.888 6.082 xc_pw_derive 510 13.4 0.005 0.006 5.983 6.049 transfer_pw2rs 245 13.2 0.003 0.004 5.681 5.728 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.635 5.692 mp_alltoall_d11v 1300 13.8 4.470 5.440 4.470 5.440 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=270.037000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.200000, yerr=2.227106 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d3f223ab808785118aa8f0a47e0c6d072683d8a1 Summary: empty Status: OK