=== 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: c5db040a230440454a8e3bd22f6bdea107af8992 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/01 job id: 50865512 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/02 job id: 50865513 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/03 job id: 50865514 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/04 job id: 50865515 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/05 job id: 50865516 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/06 job id: 50865517 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/07 job id: 50865518 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/08 job id: 50865519 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/09 job id: 50865520 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/10 job id: 50865521 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/11 job id: 50865522 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/12 job id: 50865523 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/13 job id: 50865524 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/14 job id: 50865526 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/15 job id: 50865527 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/16 job id: 50865528 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/17 job id: 50865529 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/18 job id: 50865530 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/19 job id: 50865531 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/20 job id: 50865532 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/21 job id: 50865533 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/22 job id: 50865534 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/23 job id: 50865535 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/24 job id: 50865536 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/25 job id: 50865537 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/26 job id: 50865538 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/27 job id: 50865539 --- 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/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 133.758 133.759 farming_run 1 2.0 133.305 133.307 133.731 133.734 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.493709E+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 0.025 115.548 115.548 qs_energies 1 2.0 0.000 0.000 115.327 115.328 mp2_main 1 3.0 0.000 0.000 113.198 113.200 mp2_gpw_main 1 4.0 0.028 0.034 112.173 112.174 mp2_ri_gpw_compute_in 1 5.0 0.175 0.202 92.975 93.391 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.608 55.024 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.163 41.027 46.190 get_2c_integrals 1 6.0 0.008 0.009 36.948 38.192 integrate_v_rspace 273 8.0 0.440 0.452 24.614 29.513 pw_transfer 6555 10.6 0.378 0.385 26.787 27.386 fft_wrap_pw1pw2 5465 11.4 0.046 0.047 25.411 26.114 grid_integrate_task_list 273 9.0 20.515 25.871 20.515 25.871 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.954 23.634 rpa_ri_compute_en 1 5.0 0.020 0.040 19.092 19.259 cp_fm_cholesky_decompose 12 8.2 17.822 19.083 17.822 19.083 compute_2c_integrals 1 7.0 0.002 0.002 19.067 19.069 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.567 18.796 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.435 18.566 18.795 cholesky_decomp 1 7.0 0.000 0.000 16.728 17.987 fft3d_s 5443 13.4 16.108 16.507 16.130 16.529 ao_to_mo_and_store_B_mult_1 272 7.0 10.757 15.308 10.757 15.308 calculate_wavefunction 272 8.0 5.397 5.543 12.319 12.971 rpa_num_int 1 6.0 0.001 0.011 10.670 10.679 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.618 10.625 calc_mat_Q 8 8.0 0.000 0.000 9.473 9.563 contract_S_to_Q 8 9.0 0.000 0.000 8.898 8.987 calc_potential_gpw 544 9.5 0.004 0.005 8.253 8.679 parallel_gemm_fm 14 9.1 0.000 0.000 8.492 8.587 parallel_gemm_fm_cosma 14 10.1 8.492 8.587 8.492 8.587 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.060 8.300 potential_pw2rs 545 10.0 0.106 0.108 7.456 8.121 create_integ_mat 1 6.0 0.038 0.050 7.896 7.905 collocate_single_gaussian 272 10.0 0.039 0.041 7.305 7.502 array2fm 1 7.0 0.000 0.000 6.763 7.309 pw_scatter_s 2720 13.7 4.355 4.617 4.355 4.617 pw_gather_s 2722 13.2 3.417 3.816 3.417 3.816 array2fm_buffer_send 1 8.0 2.992 3.154 2.992 3.154 pw_poisson_solve 545 10.5 1.119 1.171 2.199 2.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.172050, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2812.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 0.024 0.037 419.740 419.741 farming_run 1 2.0 418.941 418.947 419.698 419.700 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.236840E+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 26646217. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 216.553 216.554 qs_energies 1 2.0 0.000 0.001 216.320 216.335 scf_env_do_scf 1 3.0 0.000 0.000 114.400 114.401 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.512 113.520 rebuild_ks_matrix 4 6.0 0.000 0.000 113.511 113.519 qs_ks_build_kohn_sham_matrix 4 7.0 0.053 0.061 113.511 113.519 hfx_ks_matrix 4 8.0 0.001 0.001 113.168 113.173 integrate_four_center 4 9.0 0.153 0.470 113.167 113.172 integrate_four_center_main 4 10.0 0.098 0.438 101.839 104.776 integrate_four_center_bin 267 11.0 101.741 104.771 101.741 104.771 mp2_main 1 3.0 0.001 0.004 101.630 101.643 mp2_gpw_main 1 4.0 0.053 0.107 100.778 100.792 init_scf_loop 1 4.0 0.000 0.000 96.346 96.347 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.860 74.886 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.780 54.806 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 41.417 46.504 integrate_v_rspace 95 8.0 0.395 0.561 27.894 32.822 pw_transfer 2240 10.6 0.146 0.165 29.450 29.800 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.472 28.836 mp2_ri_gpw_compute_en 1 5.0 0.060 0.109 26.738 28.680 grid_integrate_task_list 95 9.0 23.318 28.456 23.318 28.456 ao_to_mo_and_store_B_mult_1 91 7.0 10.695 27.061 10.695 27.061 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.246 26.619 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.879 25.010 25.022 get_2c_integrals 1 6.0 0.000 0.000 19.996 20.015 compute_2c_integrals 1 7.0 0.002 0.003 18.979 18.985 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.562 18.842 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.854 18.561 18.841 fft3d_s 1823 13.4 18.468 18.825 18.482 18.839 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.052 18.052 calculate_wavefunction 91 8.0 2.014 2.043 9.614 9.825 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.821 9.343 potential_pw2rs 186 10.0 0.033 0.034 8.408 8.895 local_gemm 172 8.0 8.264 8.780 8.264 8.780 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.108 8.428 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.523 7.979 8.394 calc_potential_gpw 182 9.5 0.002 0.002 7.863 8.118 collocate_single_gaussian 91 10.0 0.016 0.019 7.729 7.957 integrate_four_center_load 4 10.0 0.000 0.000 6.782 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.782 6.786 mp2_ri_gpw_compute_en_ener 172 7.0 6.327 6.422 6.327 6.422 mp_sendrecv_dm3 2068 8.0 6.018 6.415 6.018 6.415 mp_sync 37 10.5 3.272 5.898 3.272 5.898 pw_gather_s 912 13.2 4.499 4.848 4.499 4.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.777936, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1496.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.201920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 54.856 54.857 qs_mol_dyn_low 1 2.0 0.003 0.004 54.659 54.668 qs_forces 11 3.9 0.002 0.003 54.577 54.577 qs_energies 11 4.9 0.002 0.002 53.023 53.044 scf_env_do_scf 11 5.9 0.010 0.025 47.152 47.152 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 45.013 45.013 qs_scf_new_mos 108 7.5 0.000 0.001 35.019 35.298 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.019 35.297 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.433 34.969 ot_scf_mini 108 9.5 0.002 0.002 33.300 33.475 multiply_cannon 2286 13.5 0.195 0.207 26.532 28.406 multiply_cannon_loop 2286 14.5 1.811 1.915 25.863 27.766 velocity_verlet 10 3.0 0.002 0.004 27.031 27.042 ot_mini 108 10.5 0.001 0.001 20.386 20.633 qs_ot_get_derivative 108 11.5 0.001 0.001 17.352 17.564 mp_waitall_1 245248 16.5 8.915 15.053 8.915 15.053 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.322 13.304 multiply_cannon_multrec 54864 15.5 3.625 5.723 7.526 10.848 qs_ot_get_p 119 10.4 0.003 0.007 8.320 8.639 rebuild_ks_matrix 119 8.3 0.000 0.000 7.846 7.979 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.846 7.978 mp_sum_l 7287 12.8 5.936 7.670 5.936 7.670 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.871 6.989 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.964 6.424 multiply_cannon_sync_h2d 54864 15.5 5.131 6.109 5.131 6.109 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.811 5.934 dbcsr_mm_accdrv_process 76910 16.1 1.819 2.888 3.814 5.482 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.328 5.362 init_scf_run 11 5.9 0.000 0.001 4.601 4.602 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.601 4.601 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.518 4.518 sum_up_and_integrate 119 10.3 0.001 0.002 4.428 4.447 integrate_v_rspace 119 11.3 0.002 0.002 4.417 4.437 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.132 4.332 calculate_rho_elec 119 8.7 0.011 0.016 4.132 4.332 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.322 4.322 cp_fm_redistribute_end 50 14.0 2.207 4.296 2.214 4.299 cp_fm_diag_elpa_base 50 14.0 2.080 4.184 2.084 4.192 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.949 3.367 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.951 3.145 apply_single 119 13.6 0.000 0.000 2.951 3.145 calculate_dm_sparse 119 9.5 0.000 0.000 2.931 3.095 acc_transpose_blocks 54864 15.5 0.222 0.251 2.206 2.766 ot_diis_step 108 11.5 0.006 0.006 2.752 2.753 jit_kernel_multiply 13 15.8 1.933 2.506 1.933 2.506 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.413 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.353 2.363 density_rs2pw 119 9.7 0.004 0.004 2.156 2.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.219 2.221 wfi_extrapolate 11 7.9 0.001 0.001 2.175 2.175 init_scf_loop 11 6.9 0.000 0.000 2.111 2.120 grid_integrate_task_list 119 12.3 2.015 2.117 2.015 2.117 mp_sum_d 4137 12.0 1.396 1.998 1.396 1.998 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.911 1.987 potential_pw2rs 119 12.3 0.004 0.004 1.803 1.825 pw_transfer 1439 11.6 0.052 0.056 1.665 1.748 make_m2s 4572 13.5 0.053 0.055 1.648 1.693 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.590 1.675 make_images 4572 14.5 0.133 0.139 1.566 1.610 mp_alltoall_d11v 2130 13.8 1.314 1.583 1.314 1.583 transfer_rs2pw 487 10.6 0.005 0.006 1.369 1.481 acc_transpose_blocks_sync 164592 16.5 1.202 1.456 1.202 1.456 mp_waitany 12084 13.8 1.241 1.425 1.241 1.425 grid_collocate_task_list 119 9.7 1.353 1.421 1.353 1.421 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.379 1.415 fft3d_ps 1201 14.6 0.368 0.477 1.302 1.383 fft_wrap_pw1pw2_140 487 13.2 0.108 0.115 1.241 1.327 transfer_pw2rs 487 13.2 0.006 0.006 1.307 1.316 mp_alltoall_i22 627 13.8 0.873 1.184 0.873 1.184 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.759 1.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.857000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.628224E+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 1482100. 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 0.028 0.056 38.653 38.654 qs_mol_dyn_low 1 2.0 0.003 0.003 38.413 38.421 qs_forces 11 3.9 0.003 0.004 38.353 38.353 qs_energies 11 4.9 0.002 0.007 36.669 36.673 scf_env_do_scf 11 5.9 0.001 0.001 31.654 31.655 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.259 29.259 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.951 22.293 qs_scf_new_mos 108 7.5 0.001 0.001 20.621 20.850 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.621 20.850 ot_scf_mini 108 9.5 0.002 0.003 19.708 19.874 multiply_cannon 2286 13.5 0.208 0.215 16.898 18.410 velocity_verlet 10 3.0 0.002 0.004 18.252 18.253 multiply_cannon_loop 2286 14.5 1.183 1.245 15.744 17.114 ot_mini 108 10.5 0.001 0.001 12.121 12.354 mp_waitall_1 200699 16.5 5.586 10.622 5.586 10.622 qs_ot_get_derivative 108 11.5 0.001 0.001 9.758 9.929 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.080 9.382 multiply_cannon_multrec 27432 15.5 1.815 4.077 6.124 8.893 rebuild_ks_matrix 119 8.3 0.000 0.000 6.759 6.894 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 6.758 6.893 dbcsr_mm_accdrv_process 47894 16.0 3.570 5.751 4.228 6.284 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.945 6.068 qs_ot_get_p 119 10.4 0.001 0.002 4.655 4.879 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.743 4.553 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.962 4.049 apply_single 119 13.6 0.000 0.000 2.961 4.049 mp_sum_l 7287 12.8 2.008 3.888 2.008 3.888 init_scf_run 11 5.9 0.000 0.001 3.780 3.781 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.780 3.781 sum_up_and_integrate 119 10.3 0.001 0.003 3.684 3.689 integrate_v_rspace 119 11.3 0.002 0.002 3.670 3.675 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.632 3.660 calculate_rho_elec 119 8.7 0.021 0.024 3.631 3.659 qs_ot_p2m_diag 50 11.0 0.009 0.017 3.015 3.032 make_m2s 4572 13.5 0.051 0.053 2.598 2.920 make_images 4572 14.5 0.205 0.243 2.508 2.834 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.614 2.615 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.324 2.412 init_scf_loop 11 6.9 0.001 0.004 2.376 2.376 ot_diis_step 108 11.5 0.011 0.011 2.315 2.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.292 2.293 calculate_first_density_matrix 1 7.0 0.000 0.002 2.289 2.291 cp_fm_redistribute_end 50 14.0 1.162 2.266 1.167 2.269 multiply_cannon_sync_h2d 27432 15.5 1.695 2.213 1.695 2.213 cp_fm_diag_elpa_base 50 14.0 1.070 2.171 1.099 2.211 calculate_dm_sparse 119 9.5 0.000 0.001 2.052 2.133 density_rs2pw 119 9.7 0.004 0.004 1.927 2.027 acc_transpose_blocks 27432 15.5 0.114 0.119 1.619 1.954 grid_integrate_task_list 119 12.3 1.845 1.951 1.845 1.951 pw_transfer 1439 11.6 0.064 0.069 1.853 1.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.858 1.860 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.761 1.800 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.696 make_images_data 4572 15.5 0.048 0.054 1.227 1.575 prepare_preconditioner 11 7.9 0.000 0.000 1.496 1.523 make_preconditioner 11 8.9 0.000 0.001 1.496 1.523 jit_kernel_multiply 8 16.4 0.601 1.459 0.601 1.459 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.399 1.456 potential_pw2rs 119 12.3 0.006 0.006 1.435 1.443 fft_wrap_pw1pw2_140 487 13.2 0.128 0.133 1.402 1.440 wfi_extrapolate 11 7.9 0.001 0.001 1.432 1.432 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.090 1.432 fft3d_ps 1201 14.6 0.521 0.580 1.392 1.426 mp_alltoall_d11v 2130 13.8 1.177 1.366 1.177 1.366 grid_collocate_task_list 119 9.7 1.293 1.356 1.293 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.257 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.203 1.249 transfer_rs2pw 487 10.6 0.005 0.005 1.117 1.196 mp_sum_d 4137 12.0 0.570 1.060 0.570 1.060 mp_allgather_i34 2286 14.5 0.578 0.976 0.578 0.976 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.968 0.968 acc_transpose_blocks_sync 82296 16.5 0.818 0.945 0.818 0.945 transfer_pw2rs 487 13.2 0.004 0.005 0.902 0.909 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.887 acc_transpose_blocks_kernels 27432 16.5 0.191 0.283 0.659 0.865 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.701 0.775 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.654000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.534912E+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 0.006 0.027 31.904 31.905 qs_mol_dyn_low 1 2.0 0.003 0.003 31.704 31.712 qs_forces 11 3.9 0.002 0.003 31.560 31.561 qs_energies 11 4.9 0.002 0.002 30.024 30.026 scf_env_do_scf 11 5.9 0.000 0.001 25.364 25.364 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.969 22.969 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.996 17.074 qs_scf_new_mos 108 7.5 0.001 0.001 15.361 15.379 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.361 15.379 velocity_verlet 10 3.0 0.002 0.002 15.112 15.116 ot_scf_mini 108 9.5 0.002 0.002 14.602 14.623 multiply_cannon 2286 13.5 0.193 0.198 13.818 14.505 multiply_cannon_loop 2286 14.5 0.853 0.902 13.045 13.760 ot_mini 108 10.5 0.001 0.001 9.065 9.086 qs_ot_get_derivative 108 11.5 0.001 0.001 7.561 7.582 multiply_cannon_multrec 18288 15.5 1.872 2.895 7.032 7.370 rebuild_ks_matrix 119 8.3 0.000 0.000 5.927 5.944 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.926 5.943 dbcsr_mm_accdrv_process 38222 16.0 4.961 5.780 5.067 5.845 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.221 5.236 mp_waitall_1 158411 16.6 2.688 3.810 2.688 3.810 sum_up_and_integrate 119 10.3 0.001 0.001 3.514 3.519 init_scf_run 11 5.9 0.000 0.001 3.510 3.510 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.510 3.510 integrate_v_rspace 119 11.3 0.002 0.003 3.501 3.508 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.914 3.474 qs_ot_get_p 119 10.4 0.001 0.002 3.286 3.313 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.222 3.226 calculate_rho_elec 119 8.7 0.031 0.031 3.221 3.226 multiply_cannon_metrocomm3 18288 15.5 0.048 0.051 1.531 2.529 init_scf_loop 11 6.9 0.000 0.000 2.378 2.379 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.986 2.357 apply_single 119 13.6 0.000 0.000 1.986 2.356 calculate_first_density_matrix 1 7.0 0.000 0.000 2.302 2.304 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.193 2.198 make_m2s 4572 13.5 0.044 0.045 1.793 1.953 grid_integrate_task_list 119 12.3 1.809 1.939 1.809 1.939 calculate_dm_sparse 119 9.5 0.000 0.001 1.915 1.922 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.899 1.899 density_rs2pw 119 9.7 0.004 0.004 1.790 1.888 make_images 4572 14.5 0.192 0.204 1.707 1.867 pw_transfer 1439 11.6 0.065 0.069 1.829 1.837 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.737 1.746 acc_transpose_blocks 18288 15.5 0.079 0.084 1.626 1.724 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.694 1.705 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.703 1.704 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.659 1.672 cp_fm_diag_elpa_base 50 14.0 1.636 1.648 1.657 1.670 prepare_preconditioner 11 7.9 0.000 0.000 1.637 1.640 make_preconditioner 11 8.9 0.000 0.000 1.637 1.639 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.493 1.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.553 1.560 ot_diis_step 108 11.5 0.011 0.012 1.486 1.486 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.410 1.420 mp_sum_l 7287 12.8 0.996 1.400 0.996 1.400 grid_collocate_task_list 119 9.7 1.234 1.344 1.234 1.344 potential_pw2rs 119 12.3 0.007 0.008 1.312 1.316 fft3d_ps 1201 14.6 0.546 0.564 1.292 1.302 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.254 multiply_cannon_sync_h2d 18288 15.5 1.048 1.203 1.048 1.203 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 transfer_rs2pw 487 10.6 0.005 0.005 0.942 1.058 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.946 0.963 make_images_data 4572 15.5 0.048 0.052 0.777 0.942 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.933 0.933 hybrid_alltoall_any 4725 16.4 0.059 0.119 0.676 0.885 jit_kernel_multiply 4 16.0 0.048 0.843 0.048 0.843 acc_transpose_blocks_sync 54864 16.5 0.736 0.834 0.736 0.834 acc_transpose_blocks_kernels 18288 16.5 0.220 0.226 0.790 0.798 transfer_pw2rs 487 13.2 0.004 0.004 0.792 0.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.776 0.778 mp_alltoall_d11v 2130 13.8 0.649 0.746 0.649 0.746 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.670 0.732 cp_fm_cholesky_invert 11 10.9 0.696 0.700 0.696 0.700 multiply_cannon_metrocomm1 18288 15.5 0.030 0.033 0.429 0.696 mp_alltoall_z22v 1201 16.6 0.599 0.671 0.599 0.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.905000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.727273, yerr=2.004128 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 560.971776E+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 0.007 0.027 36.509 36.510 qs_mol_dyn_low 1 2.0 0.003 0.005 36.264 36.271 qs_forces 11 3.9 0.003 0.004 36.183 36.183 qs_energies 11 4.9 0.002 0.002 34.482 34.488 scf_env_do_scf 11 5.9 0.000 0.001 29.347 29.348 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.969 25.970 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.035 20.155 velocity_verlet 10 3.0 0.002 0.002 18.382 18.384 qs_scf_new_mos 108 7.5 0.001 0.001 17.898 17.955 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.898 17.955 multiply_cannon 2286 13.5 0.218 0.224 16.340 17.118 ot_scf_mini 108 9.5 0.002 0.003 16.870 16.929 multiply_cannon_loop 2286 14.5 1.480 1.571 15.390 16.030 ot_mini 108 10.5 0.001 0.001 10.438 10.504 multiply_cannon_multrec 27432 15.5 2.526 3.239 9.244 9.548 qs_ot_get_derivative 108 11.5 0.001 0.001 8.563 8.625 dbcsr_mm_accdrv_process 47916 15.9 6.209 7.758 6.616 8.026 rebuild_ks_matrix 119 8.3 0.000 0.000 6.303 6.349 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.302 6.349 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.584 5.626 init_scf_run 11 5.9 0.000 0.001 3.779 3.780 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.779 3.779 qs_ot_get_p 119 10.4 0.001 0.001 3.598 3.679 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.177 3.561 sum_up_and_integrate 119 10.3 0.001 0.001 3.476 3.486 integrate_v_rspace 119 11.3 0.002 0.003 3.465 3.474 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.374 3.413 calculate_rho_elec 119 8.7 0.040 0.046 3.374 3.413 init_scf_loop 11 6.9 0.000 0.000 3.355 3.356 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.115 2.664 apply_single 119 13.6 0.000 0.000 2.115 2.664 acc_transpose_blocks 27432 15.5 0.119 0.122 2.427 2.599 prepare_preconditioner 11 7.9 0.000 0.000 2.529 2.537 make_preconditioner 11 8.9 0.000 0.000 2.529 2.537 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.137 2.464 calculate_first_density_matrix 1 7.0 0.000 0.000 2.377 2.380 make_m2s 4572 13.5 0.054 0.056 2.278 2.379 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.316 2.345 mp_waitall_1 137007 16.6 1.778 2.323 1.778 2.323 make_images 4572 14.5 0.272 0.335 2.170 2.269 calculate_dm_sparse 119 9.5 0.000 0.000 2.176 2.236 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.194 2.204 pw_transfer 1439 11.6 0.064 0.069 2.006 2.040 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.914 1.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.932 1.933 grid_integrate_task_list 119 12.3 1.830 1.928 1.830 1.928 density_rs2pw 119 9.7 0.004 0.004 1.835 1.924 ot_diis_step 108 11.5 0.012 0.012 1.832 1.833 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.795 1.795 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.757 1.767 mp_sum_l 7287 12.8 0.987 1.745 0.987 1.745 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.583 1.621 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.560 1.569 cp_fm_diag_elpa_base 50 14.0 1.527 1.544 1.558 1.568 acc_transpose_blocks_sync 82296 16.5 1.378 1.539 1.378 1.539 fft3d_ps 1201 14.6 0.591 0.651 1.402 1.425 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.814 1.363 wfi_extrapolate 11 7.9 0.001 0.001 1.347 1.347 grid_collocate_task_list 119 9.7 1.248 1.343 1.248 1.343 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.280 1.297 potential_pw2rs 119 12.3 0.009 0.009 1.286 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.277 1.287 jit_kernel_multiply 6 16.1 0.343 1.209 0.343 1.209 cp_fm_upper_to_full 72 14.2 0.818 1.164 0.818 1.164 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.097 1.097 dbcsr_complete_redistribute 329 12.2 0.118 0.141 0.757 1.031 transfer_rs2pw 487 10.6 0.004 0.005 0.901 0.991 make_images_data 4572 15.5 0.049 0.053 0.857 0.969 acc_transpose_blocks_kernels 27432 16.5 0.274 0.282 0.900 0.926 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.749 0.898 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.874 mp_alltoall_d11v 2130 13.8 0.735 0.830 0.735 0.830 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.812 cp_fm_cholesky_invert 11 10.9 0.737 0.740 0.737 0.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.510000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.363636, yerr=4.995866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 622.710784E+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 0.022 0.041 28.447 28.448 qs_mol_dyn_low 1 2.0 0.003 0.003 28.258 28.275 qs_forces 11 3.9 0.002 0.002 28.184 28.184 qs_energies 11 4.9 0.002 0.002 26.486 26.488 scf_env_do_scf 11 5.9 0.000 0.001 21.841 21.841 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.425 19.426 velocity_verlet 10 3.0 0.002 0.002 14.551 14.555 dbcsr_multiply_generic 2286 12.5 0.095 0.099 13.077 13.198 qs_scf_new_mos 108 7.5 0.001 0.001 11.817 11.842 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.816 11.841 ot_scf_mini 108 9.5 0.002 0.002 11.114 11.141 multiply_cannon 2286 13.5 0.222 0.227 10.552 11.092 multiply_cannon_loop 2286 14.5 0.641 0.661 9.641 9.871 ot_mini 108 10.5 0.001 0.001 6.500 6.531 multiply_cannon_multrec 9144 15.5 1.671 1.939 5.993 6.252 rebuild_ks_matrix 119 8.3 0.000 0.000 5.737 5.762 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.736 5.762 qs_ot_get_derivative 108 11.5 0.001 0.001 5.220 5.246 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.103 5.126 dbcsr_mm_accdrv_process 12550 15.8 3.544 4.240 4.215 4.292 sum_up_and_integrate 119 10.3 0.001 0.001 3.396 3.401 integrate_v_rspace 119 11.3 0.003 0.003 3.386 3.390 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.333 3.340 calculate_rho_elec 119 8.7 0.060 0.061 3.332 3.339 init_scf_run 11 5.9 0.000 0.001 3.199 3.199 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.199 3.199 qs_ot_get_p 119 10.4 0.002 0.002 2.743 2.777 init_scf_loop 11 6.9 0.000 0.000 2.397 2.398 calculate_first_density_matrix 1 7.0 0.000 0.000 2.084 2.084 pw_transfer 1439 11.6 0.065 0.068 2.074 2.082 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.981 1.990 grid_integrate_task_list 119 12.3 1.869 1.927 1.869 1.927 mp_waitall_1 115863 16.7 1.419 1.924 1.419 1.924 make_m2s 4572 13.5 0.033 0.034 1.753 1.915 make_images 4572 14.5 0.269 0.299 1.663 1.824 density_rs2pw 119 9.7 0.003 0.003 1.715 1.809 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.782 1.784 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.745 1.764 calculate_dm_sparse 119 9.5 0.000 0.000 1.739 1.757 prepare_preconditioner 11 7.9 0.000 0.000 1.671 1.677 make_preconditioner 11 8.9 0.000 0.000 1.671 1.677 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.654 1.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.564 1.590 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.542 1.543 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.526 1.527 acc_transpose_blocks 9144 15.5 0.043 0.044 1.430 1.461 jit_kernel_multiply 8 15.7 0.632 1.386 0.632 1.386 grid_collocate_task_list 119 9.7 1.297 1.377 1.297 1.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.352 1.361 fft3d_ps 1201 14.6 0.649 0.663 1.333 1.342 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.318 1.332 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.264 1.272 cp_fm_diag_elpa_base 50 14.0 1.237 1.255 1.263 1.271 ot_diis_step 108 11.5 0.013 0.013 1.269 1.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.212 1.247 apply_single 119 13.6 0.000 0.000 1.212 1.247 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.238 1.239 potential_pw2rs 119 12.3 0.010 0.010 1.214 1.216 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.132 1.138 wfi_extrapolate 11 7.9 0.001 0.001 1.065 1.065 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.754 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.929 make_images_data 4572 15.5 0.042 0.046 0.756 0.929 transfer_rs2pw 487 10.6 0.004 0.005 0.755 0.851 mp_alltoall_d11v 2130 13.8 0.774 0.831 0.774 0.831 cp_fm_cholesky_invert 11 10.9 0.824 0.826 0.824 0.826 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.808 0.815 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.436 0.811 acc_transpose_blocks_sync 27432 16.5 0.726 0.756 0.726 0.756 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.755 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.686 0.732 acc_transpose_blocks_kernels 9144 16.5 0.119 0.122 0.645 0.647 mp_allgather_i34 2286 14.5 0.228 0.637 0.228 0.637 transfer_pw2rs 487 13.2 0.003 0.004 0.627 0.628 mp_alltoall_z22v 1201 16.6 0.558 0.590 0.558 0.590 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.448000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.727273, yerr=4.091919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 780.529664E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 43.421 43.422 qs_mol_dyn_low 1 2.0 0.003 0.003 43.127 43.135 qs_forces 11 3.9 0.002 0.002 43.028 43.028 qs_energies 11 4.9 0.003 0.004 40.996 40.999 scf_env_do_scf 11 5.9 0.001 0.001 35.074 35.074 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.046 27.048 velocity_verlet 10 3.0 0.002 0.002 24.583 24.593 dbcsr_multiply_generic 2286 12.5 0.103 0.105 19.059 19.221 qs_scf_new_mos 108 7.5 0.001 0.001 17.273 17.376 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.272 17.375 ot_scf_mini 108 9.5 0.002 0.002 16.102 16.209 multiply_cannon 2286 13.5 0.298 0.308 15.109 15.965 multiply_cannon_loop 2286 14.5 0.862 0.870 13.840 14.743 ot_mini 108 10.5 0.001 0.001 9.787 9.908 multiply_cannon_multrec 9144 15.5 3.379 4.818 8.806 8.899 init_scf_loop 11 6.9 0.000 0.000 8.001 8.004 qs_ot_get_derivative 108 11.5 0.001 0.001 7.722 7.825 rebuild_ks_matrix 119 8.3 0.000 0.000 7.335 7.483 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.334 7.483 prepare_preconditioner 11 7.9 0.000 0.000 7.014 7.028 make_preconditioner 11 8.9 0.000 0.000 7.014 7.028 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.454 6.908 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.609 6.744 dbcsr_mm_accdrv_process 12550 15.8 4.557 6.337 5.296 6.594 cp_fm_upper_to_full 72 14.2 3.229 4.735 3.229 4.735 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.303 4.370 calculate_rho_elec 119 8.7 0.117 0.120 4.302 4.370 sum_up_and_integrate 119 10.3 0.001 0.001 3.896 3.903 integrate_v_rspace 119 11.3 0.003 0.003 3.885 3.893 init_scf_run 11 5.9 0.000 0.001 3.787 3.787 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.787 3.787 qs_ot_get_p 119 10.4 0.009 0.009 3.373 3.518 mp_waitall_1 94719 16.7 2.335 3.343 2.335 3.343 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.829 3.220 pw_transfer 1439 11.6 0.068 0.069 3.068 3.071 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.109 2.980 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.969 2.972 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.789 2.646 make_m2s 4572 13.5 0.037 0.037 2.383 2.543 fft_wrap_pw1pw2_140 487 13.2 0.651 0.653 2.531 2.536 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.199 2.492 apply_single 119 13.6 0.000 0.000 2.199 2.492 mp_alltoall_i22 627 13.8 1.562 2.431 1.562 2.431 make_images 4572 14.5 0.352 0.385 2.263 2.421 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.555 2.407 calculate_dm_sparse 119 9.5 0.000 0.000 2.344 2.406 calculate_first_density_matrix 1 7.0 0.000 0.000 2.214 2.283 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.381 2.278 density_rs2pw 119 9.7 0.003 0.003 2.257 2.278 grid_integrate_task_list 119 12.3 2.081 2.111 2.081 2.111 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.057 2.111 ot_diis_step 108 11.5 0.014 0.014 2.037 2.038 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.969 1.971 fft3d_ps 1201 14.6 0.870 0.889 1.888 1.891 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.872 1.874 qs_energies_init_hamiltonians 11 5.9 0.005 0.005 1.847 1.847 acc_transpose_blocks 9144 15.5 0.045 0.045 1.819 1.831 mp_sum_l 7287 12.8 1.021 1.789 1.021 1.789 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.618 1.618 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.583 1.604 grid_collocate_task_list 119 9.7 1.522 1.554 1.522 1.554 cp_fm_cholesky_invert 11 10.9 1.447 1.451 1.447 1.451 wfi_extrapolate 11 7.9 0.001 0.001 1.442 1.442 potential_pw2rs 119 12.3 0.014 0.014 1.408 1.410 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.336 1.400 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.320 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.337 1.337 cp_fm_diag_elpa_base 50 14.0 1.191 1.245 1.335 1.335 mp_alltoall_d11v 2130 13.8 1.241 1.316 1.241 1.316 hybrid_alltoall_any 4725 16.4 0.092 0.152 1.076 1.273 make_images_data 4572 15.5 0.047 0.050 1.012 1.205 qs_env_update_s_mstruct 11 6.9 0.056 0.057 1.171 1.191 acc_transpose_blocks_sync 27432 16.5 1.110 1.120 1.110 1.120 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.994 1.043 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.935 0.949 qs_create_task_list 11 7.9 0.000 0.001 0.928 0.938 generate_qs_task_list 11 8.9 0.368 0.387 0.927 0.937 mp_alltoall_z22v 1201 16.6 0.882 0.902 0.882 0.902 jit_kernel_multiply 5 15.4 0.710 0.878 0.710 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.422000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.181818, yerr=19.158754 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.628352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.027 79.994 79.994 qs_mol_dyn_low 1 2.0 0.003 0.004 79.752 79.761 qs_forces 11 3.9 0.004 0.004 79.685 79.686 qs_energies 11 4.9 0.002 0.002 76.859 76.874 scf_env_do_scf 11 5.9 0.000 0.001 68.099 68.102 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 62.750 62.750 dbcsr_multiply_generic 2055 12.4 0.107 0.109 50.169 50.508 qs_scf_new_mos 99 7.5 0.000 0.001 45.993 46.140 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.992 46.139 ot_scf_mini 99 9.5 0.002 0.002 43.639 43.765 multiply_cannon 2055 13.4 0.184 0.188 42.263 43.205 multiply_cannon_loop 2055 14.4 1.769 1.830 41.338 42.257 velocity_verlet 10 3.0 0.001 0.003 42.083 42.083 ot_mini 99 10.5 0.001 0.001 25.872 25.996 qs_ot_get_derivative 99 11.5 0.001 0.001 19.047 19.159 multiply_cannon_multrec 49320 15.4 11.476 12.190 17.453 18.252 rebuild_ks_matrix 110 8.3 0.000 0.001 14.342 14.492 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.342 14.492 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.572 12.699 mp_waitall_1 220248 16.4 10.270 11.108 10.270 11.108 multiply_cannon_sync_h2d 49320 15.4 9.746 10.269 9.746 10.269 qs_ot_get_p 110 10.4 0.001 0.001 9.371 9.525 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.240 7.770 apply_single 110 13.6 0.000 0.001 7.240 7.770 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.243 7.755 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.289 7.509 sum_up_and_integrate 110 10.3 0.002 0.003 6.872 6.884 integrate_v_rspace 110 11.3 0.002 0.003 6.846 6.866 init_scf_run 11 5.9 0.000 0.001 6.740 6.740 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.740 6.740 ot_diis_step 99 11.5 0.006 0.006 6.607 6.607 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.152 6.301 calculate_rho_elec 110 8.6 0.020 0.025 6.152 6.300 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.266 6.298 dbcsr_mm_accdrv_process 87628 16.1 2.996 3.138 5.845 6.163 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.476 5.476 init_scf_loop 11 6.9 0.000 0.000 5.323 5.324 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.197 5.269 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.927 4.949 cp_fm_diag_elpa_base 48 14.0 4.915 4.937 4.926 4.948 mp_sum_l 6594 12.7 3.610 4.518 3.610 4.518 wfi_extrapolate 11 7.9 0.001 0.001 3.989 3.989 make_m2s 4110 13.4 0.060 0.065 3.833 3.925 calculate_dm_sparse 110 9.5 0.001 0.001 3.738 3.863 make_images 4110 14.4 0.178 0.191 3.738 3.832 multiply_cannon_metrocomm1 49320 15.4 0.067 0.070 2.238 3.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.579 3.583 grid_integrate_task_list 110 12.3 3.260 3.422 3.260 3.422 density_rs2pw 110 9.6 0.004 0.004 3.203 3.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.254 3.297 prepare_preconditioner 11 7.9 0.000 0.000 3.178 3.202 make_preconditioner 11 8.9 0.000 0.000 3.177 3.202 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.090 3.130 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.981 3.040 pw_transfer 1331 11.6 0.055 0.067 2.960 3.037 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.873 2.953 calculate_first_density_matrix 1 7.0 0.000 0.000 2.653 2.657 acc_transpose_blocks 49320 15.4 0.207 0.218 2.535 2.624 jit_kernel_multiply 13 15.9 2.572 2.620 2.572 2.620 fft_wrap_pw1pw2_140 451 13.1 0.309 0.331 2.477 2.564 potential_pw2rs 110 12.3 0.005 0.006 2.474 2.486 mp_alltoall_d11v 2046 13.8 1.999 2.370 1.999 2.370 grid_collocate_task_list 110 9.6 2.159 2.295 2.159 2.295 fft3d_ps 1111 14.6 0.798 0.896 2.166 2.228 mp_waitany 14300 13.8 1.804 2.008 1.804 2.008 transfer_rs2pw 451 10.6 0.005 0.006 1.807 1.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.898 1.919 mp_sum_d 3891 11.9 1.331 1.915 1.331 1.915 make_images_data 4110 15.4 0.044 0.048 1.747 1.871 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.517 1.772 cp_fm_cholesky_invert 11 10.9 1.719 1.723 1.719 1.723 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.657 1.682 transfer_pw2rs 451 13.1 0.006 0.007 1.675 1.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=79.994000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.712079 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 591.970304E+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 1212689. 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 0.032 68.275 68.276 qs_mol_dyn_low 1 2.0 0.003 0.003 68.060 68.070 qs_forces 11 3.9 0.004 0.006 67.938 67.938 qs_energies 11 4.9 0.002 0.002 64.637 64.640 scf_env_do_scf 11 5.9 0.000 0.001 56.125 56.128 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.684 48.684 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.908 38.106 velocity_verlet 10 3.0 0.002 0.004 35.676 35.678 qs_scf_new_mos 99 7.5 0.001 0.001 32.977 33.099 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.976 33.098 multiply_cannon 2055 13.4 0.224 0.241 31.430 32.380 ot_scf_mini 99 9.5 0.003 0.003 31.314 31.423 multiply_cannon_loop 2055 14.4 1.156 1.179 30.314 31.133 ot_mini 99 10.5 0.001 0.001 18.395 18.514 multiply_cannon_multrec 24660 15.4 6.954 8.759 14.209 15.945 rebuild_ks_matrix 110 8.3 0.000 0.000 13.439 13.561 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.439 13.561 qs_ot_get_derivative 99 11.5 0.001 0.001 12.654 12.776 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.822 11.927 mp_waitall_1 176588 16.5 7.548 9.835 7.548 9.835 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.186 8.199 multiply_cannon_sync_h2d 24660 15.4 6.403 7.448 6.403 7.448 init_scf_loop 11 6.9 0.000 0.000 7.408 7.409 dbcsr_mm_accdrv_process 52282 16.1 5.634 6.627 7.084 7.406 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.432 7.053 apply_single 110 13.6 0.000 0.001 6.432 7.053 qs_ot_get_p 110 10.4 0.001 0.002 6.310 6.461 sum_up_and_integrate 110 10.3 0.002 0.003 6.310 6.320 integrate_v_rspace 110 11.3 0.003 0.010 6.282 6.295 init_scf_run 11 5.9 0.000 0.001 6.110 6.111 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.110 6.110 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.804 5.813 calculate_rho_elec 110 8.6 0.039 0.047 5.804 5.813 ot_diis_step 99 11.5 0.010 0.010 5.694 5.695 prepare_preconditioner 11 7.9 0.000 0.000 5.420 5.442 make_preconditioner 11 8.9 0.000 0.000 5.420 5.442 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.681 5.401 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.989 5.146 make_m2s 4110 13.4 0.056 0.060 4.158 4.479 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.457 4.477 make_images 4110 14.4 0.405 0.465 4.046 4.362 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.997 3.997 pw_transfer 1331 11.6 0.066 0.071 3.495 3.638 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.388 3.534 wfi_extrapolate 11 7.9 0.001 0.001 3.464 3.464 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.445 3.457 cp_fm_diag_elpa_base 48 14.0 3.399 3.422 3.442 3.454 grid_integrate_task_list 110 12.3 3.169 3.329 3.169 3.329 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.240 3.310 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.295 3.297 density_rs2pw 110 9.6 0.004 0.004 3.066 3.253 fft_wrap_pw1pw2_140 451 13.1 0.358 0.379 2.923 3.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.962 3.024 calculate_dm_sparse 110 9.5 0.001 0.001 2.968 2.999 fft3d_ps 1111 14.6 1.115 1.356 2.479 2.637 make_images_data 4110 15.4 0.049 0.053 2.278 2.636 hybrid_alltoall_any 4261 16.3 0.106 0.447 1.990 2.571 calculate_first_density_matrix 1 7.0 0.000 0.000 2.551 2.554 cp_fm_cholesky_invert 11 10.9 2.486 2.493 2.486 2.493 grid_collocate_task_list 110 9.6 2.166 2.338 2.166 2.338 jit_kernel_multiply 11 16.2 1.086 2.334 1.086 2.334 mp_sum_l 6594 12.7 1.689 2.259 1.689 2.259 potential_pw2rs 110 12.3 0.008 0.009 2.180 2.188 acc_transpose_blocks 24660 15.4 0.115 0.118 1.977 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.955 1.977 mp_alltoall_d11v 2046 13.8 1.757 1.933 1.757 1.933 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.856 1.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.772 1.791 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.769 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.546 transfer_rs2pw 451 10.6 0.006 0.007 1.389 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.491 mp_allgather_i34 2055 14.4 0.425 1.460 0.425 1.460 mp_irecv_dv 57340 16.2 0.643 1.409 0.643 1.409 dbcsr_complete_redistribute 325 12.2 0.241 0.299 1.107 1.366 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.276000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.454545, yerr=6.624522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 669.876224E+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 0.007 0.027 59.426 59.426 qs_mol_dyn_low 1 2.0 0.003 0.003 59.114 59.123 qs_forces 11 3.9 0.003 0.003 59.051 59.052 qs_energies 11 4.9 0.002 0.002 55.892 55.895 scf_env_do_scf 11 5.9 0.001 0.001 48.057 48.057 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.588 39.588 velocity_verlet 10 3.0 0.001 0.002 32.249 32.251 dbcsr_multiply_generic 2055 12.4 0.110 0.113 28.469 28.806 qs_scf_new_mos 99 7.5 0.001 0.001 25.072 25.188 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.071 25.187 ot_scf_mini 99 9.5 0.002 0.002 23.823 23.958 multiply_cannon 2055 13.4 0.212 0.224 22.365 23.648 multiply_cannon_loop 2055 14.4 0.812 0.835 21.319 22.624 ot_mini 99 10.5 0.001 0.001 13.654 13.791 rebuild_ks_matrix 110 8.3 0.000 0.000 12.024 12.196 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.023 12.195 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.578 10.731 multiply_cannon_multrec 16440 15.4 3.616 4.592 9.799 10.707 mp_waitall_1 139946 16.5 6.802 9.902 6.802 9.902 qs_ot_get_derivative 99 11.5 0.001 0.001 9.229 9.364 init_scf_loop 11 6.9 0.000 0.000 8.435 8.435 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 4.396 7.564 prepare_preconditioner 11 7.9 0.000 0.000 6.717 6.738 make_preconditioner 11 8.9 0.000 0.000 6.717 6.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.037 6.392 sum_up_and_integrate 110 10.3 0.001 0.002 6.230 6.244 dbcsr_mm_accdrv_process 34862 16.1 5.301 5.810 6.027 6.227 integrate_v_rspace 110 11.3 0.003 0.003 6.203 6.218 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.720 5.729 calculate_rho_elec 110 8.6 0.059 0.059 5.719 5.729 qs_ot_get_p 110 10.4 0.001 0.001 5.393 5.562 init_scf_run 11 5.9 0.000 0.001 5.424 5.424 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.423 5.424 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.879 5.321 apply_single 110 13.6 0.000 0.000 4.879 5.321 ot_diis_step 99 11.5 0.011 0.011 4.393 4.394 make_m2s 4110 13.4 0.049 0.051 4.047 4.310 make_images 4110 14.4 0.398 0.521 3.930 4.191 multiply_cannon_sync_h2d 16440 15.4 3.223 3.863 3.223 3.863 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.181 3.782 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.752 3.756 pw_transfer 1331 11.6 0.065 0.071 3.592 3.599 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.485 3.495 grid_integrate_task_list 110 12.3 3.200 3.424 3.200 3.424 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.384 3.384 density_rs2pw 110 9.6 0.004 0.004 2.957 3.201 fft_wrap_pw1pw2_140 451 13.1 0.462 0.469 3.024 3.033 wfi_extrapolate 11 7.9 0.001 0.001 2.935 2.935 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.851 2.864 cp_fm_diag_elpa_base 48 14.0 2.786 2.813 2.849 2.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.729 2.730 cp_fm_cholesky_invert 11 10.9 2.629 2.635 2.629 2.635 make_images_data 4110 15.4 0.047 0.051 2.223 2.587 multiply_cannon_metrocomm4 14385 15.4 0.048 0.051 0.892 2.554 calculate_dm_sparse 110 9.5 0.001 0.001 2.530 2.553 hybrid_alltoall_any 4261 16.3 0.108 0.381 2.009 2.531 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.450 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.403 2.473 mp_irecv_dv 48980 15.7 0.816 2.415 0.816 2.415 fft3d_ps 1111 14.6 1.119 1.133 2.401 2.414 grid_collocate_task_list 110 9.6 2.221 2.406 2.221 2.406 calculate_first_density_matrix 1 7.0 0.000 0.000 2.399 2.401 mp_sum_l 6594 12.7 1.430 2.147 1.430 2.147 potential_pw2rs 110 12.3 0.011 0.011 2.091 2.096 mp_alltoall_d11v 2046 13.8 1.766 2.019 1.766 2.019 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.981 1.981 dbcsr_complete_redistribute 325 12.2 0.324 0.348 1.394 1.854 cp_fm_upper_to_full 70 14.2 1.407 1.778 1.407 1.778 acc_transpose_blocks 16440 15.4 0.079 0.082 1.602 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.656 cp_fm_cholesky_decompose 22 10.9 1.564 1.585 1.564 1.585 transfer_rs2pw 451 10.6 0.005 0.006 1.241 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.468 1.477 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.361 1.468 mp_waitany 17072 13.8 1.112 1.413 1.112 1.413 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.966 1.411 mp_allgather_i34 2055 14.4 0.359 1.328 0.359 1.328 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.283 1.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.426000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=634.727273, yerr=7.398905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 728.027136E+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 0.069 0.082 66.290 66.291 qs_mol_dyn_low 1 2.0 0.003 0.008 65.820 65.829 qs_forces 11 3.9 0.003 0.003 65.755 65.755 qs_energies 11 4.9 0.008 0.024 62.369 62.372 scf_env_do_scf 11 5.9 0.000 0.001 53.557 53.560 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.948 41.949 velocity_verlet 10 3.0 0.001 0.002 37.096 37.098 dbcsr_multiply_generic 2055 12.4 0.117 0.120 30.629 30.869 qs_scf_new_mos 99 7.5 0.001 0.001 27.381 27.496 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.381 27.496 ot_scf_mini 99 9.5 0.002 0.003 25.698 25.804 multiply_cannon 2055 13.4 0.238 0.255 23.565 24.417 multiply_cannon_loop 2055 14.4 1.429 1.478 22.160 22.852 ot_mini 99 10.5 0.001 0.001 14.716 14.843 multiply_cannon_multrec 24660 15.4 4.057 6.834 12.991 14.195 rebuild_ks_matrix 110 8.3 0.000 0.000 11.967 12.092 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.966 12.092 init_scf_loop 11 6.9 0.000 0.000 11.571 11.571 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.572 10.683 qs_ot_get_derivative 99 11.5 0.001 0.001 10.541 10.652 prepare_preconditioner 11 7.9 0.000 0.000 9.831 9.846 make_preconditioner 11 8.9 0.000 0.000 9.831 9.846 dbcsr_mm_accdrv_process 52304 16.0 7.736 9.127 8.775 9.745 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.029 9.538 sum_up_and_integrate 110 10.3 0.001 0.002 6.192 6.204 integrate_v_rspace 110 11.3 0.003 0.003 6.165 6.177 qs_ot_get_p 110 10.4 0.001 0.001 5.840 5.989 init_scf_run 11 5.9 0.000 0.001 5.951 5.952 scf_env_initial_rho_setup 11 6.9 0.027 0.058 5.951 5.951 mp_waitall_1 121746 16.5 4.188 5.849 4.188 5.849 make_m2s 4110 13.4 0.059 0.062 5.447 5.812 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.800 5.808 calculate_rho_elec 110 8.6 0.077 0.081 5.799 5.808 make_images 4110 14.4 0.572 0.694 5.306 5.667 cp_fm_upper_to_full 70 14.2 3.423 4.975 3.423 4.975 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.047 4.151 apply_single 110 13.6 0.000 0.000 4.046 4.151 ot_diis_step 99 11.5 0.011 0.011 4.138 4.138 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.960 3.973 dbcsr_complete_redistribute 325 12.2 0.423 0.468 2.644 3.812 pw_transfer 1331 11.6 0.065 0.073 3.727 3.759 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.621 3.655 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.490 3.546 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.340 3.404 grid_integrate_task_list 110 12.3 3.274 3.401 3.274 3.401 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.395 3.395 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.143 3.291 fft_wrap_pw1pw2_140 451 13.1 0.532 0.547 3.132 3.167 density_rs2pw 110 9.6 0.004 0.004 2.934 3.140 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.321 3.120 make_images_data 4110 15.4 0.049 0.053 2.733 3.086 calculate_dm_sparse 110 9.5 0.001 0.001 3.007 3.036 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.386 2.987 wfi_extrapolate 11 7.9 0.001 0.001 2.976 2.976 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.792 2.929 mp_alltoall_i22 605 13.7 1.696 2.889 1.696 2.889 calculate_first_density_matrix 1 7.0 0.000 0.000 2.852 2.880 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.863 2.875 cp_fm_diag_elpa_base 48 14.0 2.713 2.777 2.860 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.796 2.797 cp_fm_cholesky_invert 11 10.9 2.598 2.606 2.598 2.606 multiply_cannon_sync_h2d 24660 15.4 2.384 2.587 2.384 2.587 acc_transpose_blocks 24660 15.4 0.111 0.115 2.470 2.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.439 2.485 fft3d_ps 1111 14.6 1.150 1.187 2.446 2.461 grid_collocate_task_list 110 9.6 2.264 2.451 2.264 2.451 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.348 2.348 mp_alltoall_d11v 2046 13.8 1.704 2.094 1.704 2.094 potential_pw2rs 110 12.3 0.012 0.013 2.008 2.013 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.801 1.844 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.709 cp_fm_cholesky_decompose 22 10.9 1.620 1.667 1.620 1.667 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.633 1.649 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.566 1.584 acc_transpose_blocks_sync 73980 16.4 1.450 1.557 1.450 1.557 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.856 1.549 mp_sum_l 6594 12.7 0.925 1.479 0.925 1.479 mp_irecv_dv 62702 16.1 0.753 1.463 0.753 1.463 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.277 1.387 transfer_rs2pw 451 10.6 0.005 0.005 1.149 1.379 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.291000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.545455, yerr=6.555548 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 854.519808E+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 0.037 0.041 55.808 55.810 qs_mol_dyn_low 1 2.0 0.003 0.003 55.513 55.527 qs_forces 11 3.9 0.003 0.003 55.200 55.201 qs_energies 11 4.9 0.002 0.002 51.561 51.566 scf_env_do_scf 11 5.9 0.001 0.002 43.137 43.137 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.444 35.445 velocity_verlet 10 3.0 0.005 0.006 31.282 31.301 dbcsr_multiply_generic 2055 12.4 0.107 0.110 23.435 23.583 qs_scf_new_mos 99 7.5 0.001 0.001 20.784 20.837 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.784 20.837 ot_scf_mini 99 9.5 0.002 0.002 19.534 19.582 multiply_cannon 2055 13.4 0.238 0.248 17.933 19.179 multiply_cannon_loop 2055 14.4 0.601 0.622 16.674 16.924 rebuild_ks_matrix 110 8.3 0.000 0.000 11.655 11.715 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.654 11.715 ot_mini 99 10.5 0.001 0.001 10.780 10.826 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.367 10.423 multiply_cannon_multrec 8220 15.4 3.193 4.536 7.788 8.957 init_scf_loop 11 6.9 0.000 0.000 7.647 7.648 mp_waitall_1 103326 16.6 5.886 7.612 5.886 7.612 qs_ot_get_derivative 99 11.5 0.001 0.001 7.048 7.097 sum_up_and_integrate 110 10.3 0.001 0.002 6.225 6.239 integrate_v_rspace 110 11.3 0.003 0.003 6.198 6.212 prepare_preconditioner 11 7.9 0.000 0.000 6.002 6.011 make_preconditioner 11 8.9 0.000 0.000 6.002 6.011 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.980 5.999 calculate_rho_elec 110 8.6 0.112 0.113 5.980 5.998 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.578 5.652 dbcsr_mm_accdrv_process 17442 15.9 3.153 4.277 4.455 5.412 init_scf_run 11 5.9 0.000 0.001 5.290 5.290 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.289 5.290 qs_ot_get_p 110 10.4 0.002 0.003 4.789 4.857 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.066 4.494 make_m2s 4110 13.4 0.038 0.039 4.182 4.432 make_images 4110 14.4 0.643 0.705 4.051 4.297 pw_transfer 1331 11.6 0.065 0.071 4.076 4.089 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.968 3.985 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.765 3.813 apply_single 110 13.6 0.000 0.000 3.765 3.813 ot_diis_step 99 11.5 0.012 0.012 3.709 3.709 grid_integrate_task_list 110 12.3 3.370 3.531 3.370 3.531 fft_wrap_pw1pw2_140 451 13.1 0.718 0.733 3.445 3.458 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.423 3.426 density_rs2pw 110 9.6 0.004 0.004 3.012 3.117 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.082 3.083 cp_fm_cholesky_invert 11 10.9 2.881 2.884 2.881 2.884 wfi_extrapolate 11 7.9 0.001 0.001 2.703 2.703 qs_energies_init_hamiltonians 11 5.9 0.005 0.008 2.679 2.680 hybrid_alltoall_any 4261 16.3 0.200 0.855 2.238 2.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.600 2.602 calculate_dm_sparse 110 9.5 0.001 0.001 2.550 2.588 make_images_data 4110 15.4 0.041 0.047 2.236 2.558 fft3d_ps 1111 14.6 1.292 1.298 2.512 2.531 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.516 2.524 cp_fm_diag_elpa_base 48 14.0 2.461 2.483 2.514 2.522 multiply_cannon_sync_h2d 8220 15.4 2.350 2.510 2.350 2.510 grid_collocate_task_list 110 9.6 2.370 2.505 2.370 2.505 calculate_first_density_matrix 1 7.0 0.000 0.000 2.490 2.493 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.234 2.262 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.042 2.076 mp_alltoall_d11v 2046 13.8 1.687 2.039 1.687 2.039 potential_pw2rs 110 12.3 0.015 0.015 2.024 2.028 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.788 1.994 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.929 1.955 cp_fm_cholesky_decompose 22 10.9 1.645 1.662 1.645 1.662 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.614 1.619 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.499 1.613 dbcsr_complete_redistribute 325 12.2 0.561 0.589 1.463 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.504 1.514 mp_allgather_i34 2055 14.4 0.430 1.500 0.430 1.500 acc_transpose_blocks 8220 15.4 0.041 0.042 1.423 1.451 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.476 1.389 mp_irecv_dv 24056 15.7 0.449 1.341 0.449 1.341 qs_create_task_list 11 7.9 0.002 0.002 1.207 1.305 generate_qs_task_list 11 8.9 0.373 0.441 1.205 1.303 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.211 1.245 transfer_rs2pw 451 10.6 0.005 0.005 1.100 1.232 mp_waitany 9240 13.8 1.036 1.177 1.036 1.177 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.757 1.174 jit_kernel_multiply 6 15.6 0.986 1.159 0.986 1.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.810000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=801.090909, yerr=13.242182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 1.387106E+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 0.045 0.119 89.287 89.294 qs_mol_dyn_low 1 2.0 0.003 0.004 88.702 88.712 qs_forces 11 3.9 0.003 0.003 88.623 88.623 qs_energies 11 4.9 0.002 0.002 84.462 84.464 scf_env_do_scf 11 5.9 0.000 0.001 73.970 73.971 velocity_verlet 10 3.0 0.012 0.024 56.454 56.460 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.530 45.532 dbcsr_multiply_generic 2055 12.4 0.122 0.126 29.952 30.102 init_scf_loop 11 6.9 0.000 0.000 28.369 28.372 qs_scf_new_mos 99 7.5 0.001 0.001 27.446 27.526 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.445 27.526 prepare_preconditioner 11 7.9 0.000 0.000 26.310 26.324 make_preconditioner 11 8.9 0.000 0.000 26.310 26.324 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.532 25.773 ot_scf_mini 99 9.5 0.002 0.002 25.632 25.713 multiply_cannon 2055 13.4 0.333 0.355 22.640 23.309 multiply_cannon_loop 2055 14.4 0.826 0.835 20.816 21.138 cp_fm_upper_to_full 70 14.2 12.713 18.197 12.713 18.197 ot_mini 99 10.5 0.001 0.001 14.504 14.578 rebuild_ks_matrix 110 8.3 0.000 0.001 14.091 14.186 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.090 14.186 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.784 12.870 dbcsr_complete_redistribute 325 12.2 1.022 1.059 7.374 10.632 multiply_cannon_multrec 8220 15.4 4.063 4.330 9.810 9.920 qs_ot_get_derivative 99 11.5 0.001 0.001 9.715 9.796 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.327 9.593 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.762 8.992 mp_alltoall_i22 605 13.7 5.394 8.696 5.394 8.696 mp_waitall_1 84994 16.7 7.675 8.646 7.675 8.646 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.538 7.575 calculate_rho_elec 110 8.6 0.222 0.223 7.537 7.575 sum_up_and_integrate 110 10.3 0.002 0.002 6.996 7.011 integrate_v_rspace 110 11.3 0.003 0.003 6.968 6.983 init_scf_run 11 5.9 0.000 0.001 6.084 6.084 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.084 6.084 make_m2s 4110 13.4 0.043 0.043 5.409 5.936 qs_ot_get_p 110 10.4 0.002 0.002 5.817 5.910 pw_transfer 1331 11.6 0.075 0.075 5.847 5.853 dbcsr_mm_accdrv_process 11614 15.7 3.904 4.131 5.599 5.800 make_images 4110 14.4 0.879 0.934 5.220 5.747 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.730 5.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.897 5.413 apply_single 110 13.6 0.000 0.000 4.897 5.413 cp_fm_cholesky_invert 11 10.9 5.297 5.301 5.297 5.301 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.971 5.239 fft_wrap_pw1pw2_140 451 13.1 1.367 1.371 5.023 5.031 ot_diis_step 99 11.5 0.015 0.016 4.769 4.770 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.167 4.172 density_rs2pw 110 9.6 0.004 0.004 3.978 4.001 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.803 3.804 grid_integrate_task_list 110 12.3 3.700 3.795 3.700 3.795 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.644 3.644 hybrid_alltoall_any 4261 16.3 0.267 0.579 2.859 3.587 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.071 3.546 make_images_data 4110 15.4 0.046 0.049 2.864 3.523 fft3d_ps 1111 14.6 1.880 1.885 3.434 3.443 wfi_extrapolate 11 7.9 0.001 0.001 3.376 3.376 calculate_dm_sparse 110 9.5 0.001 0.001 3.233 3.264 multiply_cannon_sync_h2d 8220 15.4 3.118 3.156 3.118 3.156 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.038 3.039 cp_fm_diag_elpa_base 48 14.0 2.503 2.700 3.037 3.037 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.901 2.903 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.689 2.729 grid_collocate_task_list 110 9.6 2.685 2.723 2.685 2.723 calculate_first_density_matrix 1 7.0 0.000 0.000 2.591 2.592 potential_pw2rs 110 12.3 0.021 0.021 2.467 2.472 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.374 2.406 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.251 2.324 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.196 mp_alltoall_d11v 2046 13.8 2.098 2.150 2.098 2.150 cp_fm_cholesky_decompose 22 10.9 1.992 2.005 1.992 2.005 qs_create_task_list 11 7.9 0.000 0.001 1.886 1.932 generate_qs_task_list 11 8.9 0.733 0.790 1.886 1.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.903 1.913 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.798 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.294000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1227.090909, yerr=60.861175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.919744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2529087. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.042 205.166 205.173 qs_mol_dyn_low 1 2.0 0.003 0.004 204.650 204.691 qs_forces 11 3.9 0.005 0.006 204.549 204.550 qs_energies 11 4.9 0.003 0.009 198.982 198.997 scf_env_do_scf 11 5.9 0.001 0.002 182.337 182.341 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 162.148 162.150 dbcsr_multiply_generic 2507 12.6 0.182 0.188 125.232 125.610 qs_scf_new_mos 117 7.6 0.001 0.001 123.154 123.395 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.153 123.394 velocity_verlet 10 3.0 0.002 0.007 122.577 122.578 ot_scf_mini 117 9.6 0.003 0.003 116.506 116.766 multiply_cannon 2507 13.6 0.237 0.244 101.371 102.658 multiply_cannon_loop 2507 14.6 2.390 2.440 99.273 100.261 ot_mini 117 10.6 0.001 0.001 66.035 66.306 multiply_cannon_multrec 60168 15.6 31.778 33.532 41.511 43.292 qs_ot_get_derivative 117 11.6 0.001 0.001 41.296 41.541 rebuild_ks_matrix 128 8.3 0.001 0.001 33.292 33.505 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.026 33.291 33.504 mp_waitall_1 267128 16.5 28.563 31.013 28.563 31.013 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.887 30.070 qs_ot_get_p 128 10.4 0.001 0.002 28.270 28.594 multiply_cannon_sync_h2d 60168 15.6 26.364 28.069 26.364 28.069 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.258 24.897 apply_single 128 13.6 0.001 0.001 24.257 24.897 ot_diis_step 117 11.6 0.008 0.008 24.479 24.480 qs_ot_p2m_diag 83 11.4 0.079 0.092 21.786 21.846 init_scf_loop 11 6.9 0.000 0.001 20.116 20.117 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.796 19.996 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.014 19.014 multiply_cannon_metrocomm3 60168 15.6 0.121 0.126 16.084 17.571 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.094 16.124 cp_fm_diag_elpa_base 83 14.4 16.029 16.062 16.091 16.122 prepare_preconditioner 11 7.9 0.000 0.000 15.579 15.620 make_preconditioner 11 8.9 0.000 0.000 15.579 15.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.829 14.985 make_m2s 5014 13.6 0.104 0.113 13.673 14.040 make_images 5014 14.6 0.399 0.417 13.489 13.863 sum_up_and_integrate 128 10.3 0.002 0.005 13.796 13.815 integrate_v_rspace 128 11.3 0.003 0.004 13.738 13.758 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.937 13.059 calculate_rho_elec 128 8.7 0.045 0.064 12.936 13.059 init_scf_run 11 5.9 0.000 0.001 12.453 12.453 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.452 12.453 mp_sum_l 7950 12.9 8.927 10.035 8.927 10.035 dbcsr_mm_accdrv_process 124484 16.2 4.677 4.817 9.302 9.777 wfi_extrapolate 11 7.9 0.001 0.001 9.001 9.001 cp_fm_cholesky_invert 11 10.9 8.876 8.884 8.876 8.884 calculate_dm_sparse 128 9.5 0.001 0.001 8.481 8.572 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.067 8.199 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.101 8.188 multiply_cannon_metrocomm1 60168 15.6 0.097 0.101 6.210 7.727 pw_transfer 1547 11.6 0.074 0.086 7.463 7.634 make_images_data 5014 15.6 0.068 0.075 6.664 7.581 grid_integrate_task_list 128 12.3 7.064 7.429 7.064 7.429 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.262 7.424 density_rs2pw 128 9.7 0.006 0.006 6.569 7.184 hybrid_alltoall_any 5200 16.5 0.294 2.258 5.860 7.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.780 6.792 fft_wrap_pw1pw2_140 523 13.2 0.857 0.891 6.423 6.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.759 5.812 fft3d_ps 1291 14.7 2.193 2.836 5.269 5.581 mp_alltoall_d11v 2415 14.1 4.247 5.326 4.247 5.326 grid_collocate_task_list 128 9.7 4.838 5.282 4.838 5.282 cp_fm_cholesky_decompose 22 10.9 4.558 4.572 4.558 4.572 potential_pw2rs 128 12.3 0.009 0.010 4.327 4.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.173000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.454545, yerr=5.678726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 841.449472E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5023369. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.081 190.736 190.738 qs_mol_dyn_low 1 2.0 0.003 0.004 190.282 190.315 qs_forces 11 3.9 0.004 0.004 190.197 190.198 qs_energies 11 4.9 0.005 0.032 183.422 183.434 scf_env_do_scf 11 5.9 0.001 0.001 166.910 166.920 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 134.192 134.195 velocity_verlet 10 3.0 0.001 0.002 120.163 120.164 dbcsr_multiply_generic 2527 12.6 0.190 0.197 98.061 99.355 qs_scf_new_mos 118 7.6 0.001 0.001 95.221 95.685 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.220 95.684 ot_scf_mini 118 9.6 0.004 0.004 90.350 90.927 multiply_cannon 2527 13.6 0.506 0.568 77.769 82.684 multiply_cannon_loop 2527 14.6 1.569 1.634 74.361 76.988 ot_mini 118 10.6 0.001 0.001 50.591 51.147 mp_waitall_1 216438 16.6 24.191 38.296 24.191 38.296 multiply_cannon_multrec 30324 15.6 21.097 25.653 31.752 36.563 rebuild_ks_matrix 129 8.3 0.001 0.001 32.573 33.317 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.028 32.573 33.316 init_scf_loop 11 6.9 0.001 0.004 32.627 32.628 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.308 30.008 multiply_cannon_metrocomm3 30324 15.6 0.101 0.109 15.732 29.768 qs_ot_get_derivative 118 11.6 0.001 0.002 28.454 29.019 prepare_preconditioner 11 7.9 0.000 0.001 28.304 28.389 make_preconditioner 11 8.9 0.000 0.001 28.304 28.389 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.035 27.558 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.221 23.341 apply_single 129 13.6 0.001 0.001 22.220 23.341 qs_ot_get_p 129 10.4 0.001 0.003 21.601 22.248 ot_diis_step 118 11.6 0.015 0.015 21.958 21.960 multiply_cannon_sync_h2d 30324 15.6 18.146 20.785 18.146 20.785 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.601 16.635 cp_fm_cholesky_invert 11 10.9 16.607 16.619 16.607 16.619 make_m2s 5054 13.6 0.087 0.092 14.243 15.935 make_images 5054 14.6 1.170 1.379 14.025 15.716 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.390 15.390 sum_up_and_integrate 129 10.3 0.002 0.004 14.143 14.178 integrate_v_rspace 129 11.3 0.003 0.003 14.082 14.121 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.544 13.578 calculate_rho_elec 129 8.7 0.087 0.104 13.544 13.577 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.215 12.247 cp_fm_diag_elpa_base 83 14.4 11.960 12.047 12.209 12.237 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.248 11.675 init_scf_run 11 5.9 0.000 0.001 11.550 11.552 scf_env_initial_rho_setup 11 6.9 0.022 0.042 11.550 11.552 multiply_cannon_metrocomm4 27797 15.6 0.101 0.114 3.915 10.951 dbcsr_mm_accdrv_process 62734 16.2 5.532 6.454 10.098 10.662 make_images_data 5054 15.6 0.068 0.078 8.407 10.588 mp_irecv_dv 70031 16.3 3.712 10.553 3.712 10.553 hybrid_alltoall_any 5240 16.5 0.352 1.552 7.105 9.754 pw_transfer 1559 11.6 0.085 0.101 8.658 8.713 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.434 8.493 wfi_extrapolate 11 7.9 0.001 0.001 8.360 8.360 grid_integrate_task_list 129 12.3 7.240 7.593 7.240 7.593 density_rs2pw 129 9.7 0.006 0.006 7.170 7.527 fft_wrap_pw1pw2_140 527 13.2 0.939 0.956 7.437 7.518 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.399 7.204 cp_fm_cholesky_decompose 22 10.9 6.892 6.987 6.892 6.987 calculate_dm_sparse 129 9.5 0.001 0.001 6.522 6.672 mp_sum_l 8010 12.9 4.257 6.414 4.257 6.414 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.081 6.092 fft3d_ps 1301 14.7 2.847 3.040 5.975 6.032 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.433 5.501 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.192 5.374 grid_collocate_task_list 129 9.7 5.047 5.344 5.047 5.344 mp_allgather_i34 2527 14.6 1.924 5.040 1.924 5.040 mp_alltoall_d11v 2423 14.1 4.130 4.589 4.130 4.589 potential_pw2rs 129 12.3 0.016 0.018 4.556 4.574 mp_sum_d 4501 12.1 2.706 3.938 2.706 3.938 dbcsr_complete_redistribute 395 12.7 0.769 0.856 3.056 3.894 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.738000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.909091, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 941.879296E+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 0.033 0.044 173.827 173.833 qs_mol_dyn_low 1 2.0 0.003 0.004 173.145 173.243 qs_forces 11 3.9 0.004 0.004 173.040 173.042 qs_energies 11 4.9 0.014 0.027 166.519 166.531 scf_env_do_scf 11 5.9 0.001 0.001 150.875 150.876 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.269 116.270 velocity_verlet 10 3.0 0.001 0.002 111.342 111.344 dbcsr_multiply_generic 2507 12.6 0.189 0.195 81.708 82.935 qs_scf_new_mos 117 7.6 0.001 0.001 79.444 79.772 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.443 79.771 ot_scf_mini 117 9.6 0.003 0.004 75.217 75.608 multiply_cannon 2507 13.6 0.508 0.536 62.244 67.646 multiply_cannon_loop 2507 14.6 1.123 1.191 59.364 62.113 ot_mini 117 10.6 0.001 0.001 42.532 42.914 init_scf_loop 11 6.9 0.000 0.000 34.511 34.512 mp_waitall_1 170520 16.6 24.364 34.367 24.364 34.367 prepare_preconditioner 11 7.9 0.000 0.000 30.545 30.591 make_preconditioner 11 8.9 0.000 0.000 30.545 30.591 rebuild_ks_matrix 128 8.3 0.001 0.001 30.002 30.512 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.044 30.001 30.511 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.215 29.598 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.020 27.492 multiply_cannon_multrec 20056 15.6 13.084 16.076 22.446 25.394 multiply_cannon_metrocomm3 20056 15.6 0.062 0.067 15.239 24.901 qs_ot_get_derivative 117 11.6 0.001 0.002 23.032 23.415 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.557 20.654 apply_single 128 13.6 0.001 0.001 19.557 20.654 ot_diis_step 117 11.6 0.018 0.019 19.397 19.397 qs_ot_get_p 128 10.4 0.002 0.003 18.129 18.630 multiply_cannon_sync_h2d 20056 15.6 13.742 15.554 13.742 15.554 make_m2s 5014 13.6 0.078 0.083 14.361 15.548 make_images 5014 14.6 1.141 1.221 14.124 15.306 cp_fm_cholesky_invert 11 10.9 14.149 14.158 14.149 14.158 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.830 13.839 sum_up_and_integrate 128 10.3 0.002 0.003 13.813 13.836 integrate_v_rspace 128 11.3 0.003 0.003 13.753 13.778 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.463 13.499 calculate_rho_elec 128 8.7 0.129 0.143 13.462 13.499 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.785 12.786 init_scf_run 11 5.9 0.000 0.001 10.402 10.402 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.402 10.402 make_images_data 5014 15.6 0.065 0.075 8.553 10.136 hybrid_alltoall_any 5200 16.5 0.447 2.022 7.444 9.914 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.657 9.676 cp_fm_diag_elpa_base 83 14.4 9.251 9.408 9.654 9.673 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.098 9.379 multiply_cannon_metrocomm4 17549 15.6 0.066 0.074 3.477 9.161 dbcsr_mm_accdrv_process 41502 16.2 5.676 5.991 8.809 8.976 mp_irecv_dv 50230 16.2 3.347 8.901 3.347 8.901 pw_transfer 1547 11.6 0.084 0.102 8.704 8.806 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.482 8.588 grid_integrate_task_list 128 12.3 7.345 7.675 7.345 7.675 fft_wrap_pw1pw2_140 523 13.2 1.062 1.095 7.554 7.670 wfi_extrapolate 11 7.9 0.001 0.001 7.386 7.386 cp_fm_cholesky_decompose 22 10.9 7.268 7.291 7.268 7.291 cp_fm_upper_to_full 105 14.8 5.740 7.258 5.740 7.258 density_rs2pw 128 9.7 0.006 0.006 6.829 7.118 dbcsr_complete_redistribute 395 12.7 1.159 1.189 4.475 6.245 calculate_dm_sparse 128 9.5 0.001 0.001 5.831 5.910 fft3d_ps 1291 14.7 2.820 3.060 5.792 5.884 grid_collocate_task_list 128 9.7 5.150 5.461 5.150 5.461 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.410 5.417 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.635 5.292 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.350 5.116 mp_sum_l 7950 12.9 3.296 4.926 3.296 4.926 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.535 4.688 mp_alltoall_d11v 2415 14.1 4.198 4.602 4.198 4.602 mp_allgather_i34 2507 14.6 1.383 4.475 1.383 4.475 potential_pw2rs 128 12.3 0.020 0.022 4.245 4.251 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.311 4.043 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.917 3.947 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.932 3.933 mp_alltoall_i22 716 14.1 1.947 3.889 1.947 3.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.477 3.519 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.833000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.727273, yerr=7.688136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 1.148281E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509575. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.073 189.925 189.927 qs_mol_dyn_low 1 2.0 0.003 0.003 189.505 189.527 qs_forces 11 3.9 0.004 0.004 189.367 189.379 qs_energies 11 4.9 0.002 0.003 182.259 182.270 scf_env_do_scf 11 5.9 0.001 0.001 165.485 165.499 velocity_verlet 10 3.0 0.003 0.012 126.360 126.365 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 119.251 119.252 dbcsr_multiply_generic 2529 12.6 0.199 0.205 82.854 83.413 qs_scf_new_mos 118 7.6 0.001 0.001 82.686 83.025 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.685 83.024 ot_scf_mini 118 9.6 0.003 0.003 77.811 78.145 multiply_cannon 2529 13.6 0.560 0.593 56.647 59.039 multiply_cannon_loop 2529 14.6 1.857 1.938 52.429 53.908 init_scf_loop 11 6.9 0.000 0.000 46.114 46.115 ot_mini 118 10.6 0.001 0.001 44.103 44.447 prepare_preconditioner 11 7.9 0.000 0.000 42.053 42.086 make_preconditioner 11 8.9 0.000 0.000 42.053 42.086 make_full_inverse_cholesky 11 9.9 0.013 0.021 35.638 40.734 multiply_cannon_multrec 30348 15.6 13.827 18.858 26.813 31.929 rebuild_ks_matrix 129 8.3 0.001 0.001 29.369 29.609 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 29.368 29.608 mp_waitall_1 149172 16.7 17.189 27.206 17.189 27.206 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.495 26.710 qs_ot_get_derivative 118 11.6 0.001 0.002 24.092 24.437 make_m2s 5058 13.6 0.093 0.099 21.710 23.407 make_images 5058 14.6 1.999 2.771 21.399 23.098 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.453 19.937 apply_single 129 13.6 0.001 0.001 19.453 19.937 ot_diis_step 118 11.6 0.018 0.018 19.871 19.873 qs_ot_get_p 129 10.4 0.001 0.002 19.156 19.493 cp_fm_upper_to_full 106 14.8 11.676 17.171 11.676 17.171 cp_fm_cholesky_invert 11 10.9 15.826 15.834 15.826 15.834 qs_ot_p2m_diag 84 11.4 0.347 0.394 14.788 14.839 multiply_cannon_metrocomm3 30348 15.6 0.049 0.052 6.301 14.731 sum_up_and_integrate 129 10.3 0.002 0.003 13.995 14.024 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.953 14.010 calculate_rho_elec 129 8.7 0.171 0.186 13.952 14.009 integrate_v_rspace 129 11.3 0.003 0.004 13.935 13.964 make_images_data 5058 15.6 0.067 0.074 11.195 13.416 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.370 13.371 dbcsr_mm_accdrv_process 62780 16.2 8.410 9.155 12.546 13.054 dbcsr_complete_redistribute 397 12.7 1.499 1.625 9.121 12.995 hybrid_alltoall_any 5245 16.5 0.539 2.229 10.115 12.088 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.783 11.647 multiply_cannon_sync_h2d 30348 15.6 10.453 11.495 10.453 11.495 init_scf_run 11 5.9 0.000 0.001 10.843 10.844 scf_env_initial_rho_setup 11 6.9 0.002 0.003 10.843 10.844 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.187 10.438 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.152 10.170 cp_fm_diag_elpa_base 84 14.4 9.173 9.513 10.145 10.161 transfer_fm_to_dbcsr 11 9.9 0.002 0.007 6.392 10.118 mp_alltoall_i22 720 14.1 5.696 9.497 5.696 9.497 pw_transfer 1559 11.6 0.086 0.102 9.219 9.298 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.990 9.073 fft_wrap_pw1pw2_140 527 13.2 1.259 1.314 8.029 8.131 grid_integrate_task_list 129 12.3 7.534 7.855 7.534 7.855 wfi_extrapolate 11 7.9 0.001 0.001 7.749 7.749 multiply_cannon_metrocomm4 25290 15.6 0.081 0.091 2.814 7.709 cp_fm_cholesky_decompose 22 10.9 7.366 7.473 7.366 7.473 mp_irecv_dv 76751 16.2 2.660 7.427 2.660 7.427 density_rs2pw 129 9.7 0.005 0.006 6.822 7.112 calculate_dm_sparse 129 9.5 0.001 0.001 6.603 6.706 fft3d_ps 1301 14.7 3.024 3.104 6.011 6.067 mp_alltoall_d11v 2429 14.1 5.238 5.755 5.238 5.755 grid_collocate_task_list 129 9.7 5.328 5.651 5.328 5.651 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.532 5.595 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.690 4.773 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.622 4.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.389 4.514 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.461 4.462 potential_pw2rs 129 12.3 0.023 0.023 4.251 4.263 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.927000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1084.545455, yerr=16.870435 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.527050E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.069 169.453 169.467 qs_mol_dyn_low 1 2.0 0.003 0.003 168.960 168.982 qs_forces 11 3.9 0.017 0.022 168.508 168.511 qs_energies 11 4.9 0.002 0.002 161.067 161.088 scf_env_do_scf 11 5.9 0.001 0.001 143.813 143.823 velocity_verlet 10 3.0 0.014 0.019 111.835 111.861 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 108.355 108.356 dbcsr_multiply_generic 2527 12.6 0.186 0.194 72.292 72.797 qs_scf_new_mos 118 7.6 0.001 0.001 71.193 71.290 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.192 71.290 ot_scf_mini 118 9.6 0.003 0.004 66.775 66.880 multiply_cannon 2527 13.6 0.564 0.584 53.871 56.887 multiply_cannon_loop 2527 14.6 0.810 0.848 50.718 51.852 ot_mini 118 10.6 0.001 0.001 37.773 37.873 init_scf_loop 11 6.9 0.000 0.000 35.310 35.312 prepare_preconditioner 11 7.9 0.000 0.000 31.497 31.517 make_preconditioner 11 8.9 0.000 0.000 31.497 31.517 mp_waitall_1 126780 16.7 24.755 31.513 24.755 31.513 make_full_inverse_cholesky 11 9.9 0.019 0.028 29.452 29.710 rebuild_ks_matrix 129 8.3 0.001 0.001 29.211 29.348 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.210 29.347 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.669 26.794 multiply_cannon_multrec 10108 15.6 10.386 16.182 18.030 22.488 qs_ot_get_derivative 118 11.6 0.002 0.002 20.435 20.529 multiply_cannon_metrocomm3 10108 15.6 0.024 0.026 12.945 20.389 cp_fm_cholesky_invert 11 10.9 17.993 17.999 17.993 17.999 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.393 17.553 apply_single 129 13.6 0.001 0.001 17.393 17.553 ot_diis_step 118 11.6 0.020 0.021 17.268 17.268 qs_ot_get_p 129 10.4 0.002 0.002 16.242 16.389 make_m2s 5054 13.6 0.064 0.069 14.600 15.619 make_images 5054 14.6 2.172 2.605 14.289 15.302 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.254 14.287 calculate_rho_elec 129 8.7 0.255 0.267 14.253 14.287 sum_up_and_integrate 129 10.3 0.002 0.002 14.188 14.234 integrate_v_rspace 129 11.3 0.004 0.005 14.127 14.176 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.523 12.540 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.375 11.376 multiply_cannon_sync_h2d 10108 15.6 10.798 11.140 10.798 11.140 init_scf_run 11 5.9 0.000 0.001 10.421 10.421 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.420 10.421 pw_transfer 1559 11.6 0.086 0.096 9.871 9.916 make_images_data 5054 15.6 0.056 0.066 8.354 9.829 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.647 9.698 hybrid_alltoall_any 5240 16.5 0.846 3.804 8.191 9.458 fft_wrap_pw1pw2_140 527 13.2 1.686 1.713 8.568 8.622 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.545 8.557 cp_fm_diag_elpa_base 83 14.4 8.311 8.387 8.539 8.551 grid_integrate_task_list 129 12.3 7.830 8.196 7.830 8.196 dbcsr_mm_accdrv_process 20926 16.1 3.360 4.487 7.292 8.032 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.949 8.020 cp_fm_cholesky_decompose 22 10.9 7.857 7.998 7.857 7.998 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.266 7.310 wfi_extrapolate 11 7.9 0.001 0.001 7.281 7.281 density_rs2pw 129 9.7 0.005 0.005 6.924 7.183 calculate_dm_sparse 129 9.5 0.001 0.001 6.106 6.203 fft3d_ps 1301 14.7 3.163 3.272 6.074 6.109 multiply_cannon_metrocomm4 7581 15.6 0.026 0.030 1.888 5.985 mp_irecv_dv 29086 15.9 1.847 5.891 1.847 5.891 grid_collocate_task_list 129 9.7 5.611 5.837 5.611 5.837 dbcsr_complete_redistribute 395 12.7 2.100 2.147 5.082 5.357 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 5.281 5.296 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.177 5.188 mp_alltoall_d11v 2423 14.1 4.470 4.902 4.470 4.902 mp_allgather_i34 2527 14.6 1.334 4.554 1.334 4.554 potential_pw2rs 129 12.3 0.026 0.027 4.314 4.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.097 4.155 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.588 3.940 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.660 3.747 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.652 3.706 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.292 3.590 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.375 3.461 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.447 3.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.467000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1447.454545, yerr=15.587662 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 11.606414E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.068019E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108700816 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59121. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.128 289.765 289.768 qs_mol_dyn_low 1 2.0 0.003 0.003 288.970 288.987 qs_forces 11 3.9 0.005 0.005 288.876 288.878 qs_energies 11 4.9 0.010 0.011 279.868 279.878 scf_env_do_scf 11 5.9 0.001 0.001 257.203 257.209 velocity_verlet 10 3.0 0.002 0.002 209.105 209.113 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 133.370 133.372 init_scf_loop 11 6.9 0.000 0.000 123.551 123.554 prepare_preconditioner 11 7.9 0.000 0.000 118.694 118.712 make_preconditioner 11 8.9 0.000 0.000 118.694 118.712 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.898 115.866 qs_scf_new_mos 116 7.6 0.001 0.001 89.024 89.107 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.024 89.106 ot_scf_mini 116 9.6 0.004 0.004 84.195 84.241 dbcsr_multiply_generic 2485 12.5 0.216 0.227 81.154 81.555 cp_fm_upper_to_full 104 14.8 52.329 75.128 52.329 75.128 multiply_cannon 2485 13.5 0.673 0.717 58.675 59.711 multiply_cannon_loop 2485 14.5 1.036 1.054 54.754 55.823 ot_mini 116 10.6 0.001 0.001 43.975 44.024 dbcsr_complete_redistribute 393 12.7 3.972 4.008 29.744 42.011 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.228 38.465 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.754 35.835 rebuild_ks_matrix 127 8.3 0.001 0.001 34.441 34.480 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 34.441 34.480 mp_alltoall_i22 712 14.1 21.530 33.911 21.530 33.911 cp_fm_cholesky_invert 11 10.9 33.382 33.389 33.382 33.389 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.908 31.946 mp_waitall_1 102768 16.8 27.672 31.900 27.672 31.900 qs_ot_get_p 127 10.4 0.002 0.002 25.334 25.381 qs_ot_get_derivative 116 11.6 0.002 0.002 24.572 24.617 qs_ot_p2m_diag 82 11.4 0.869 0.879 21.352 21.386 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 19.042 20.098 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.571 19.573 ot_diis_step 116 11.6 0.022 0.022 19.368 19.368 make_m2s 4970 13.5 0.074 0.078 17.891 19.301 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.727 18.856 apply_single 127 13.6 0.001 0.001 18.726 18.855 make_images 4970 14.5 3.030 3.221 17.411 18.820 multiply_cannon_multrec 9940 15.5 10.215 12.004 17.940 18.051 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.163 17.179 calculate_rho_elec 127 8.7 0.475 0.475 17.162 17.178 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.427 16.428 cp_fm_diag_elpa_base 82 14.4 12.079 13.682 16.423 16.423 sum_up_and_integrate 127 10.3 0.002 0.002 16.322 16.411 integrate_v_rspace 127 11.3 0.004 0.004 16.260 16.350 multiply_cannon_sync_h2d 9940 15.5 14.193 14.223 14.193 14.223 pw_transfer 1535 11.6 0.094 0.095 13.125 13.137 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 12.888 12.900 init_scf_run 11 5.9 0.000 0.001 12.196 12.198 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.196 12.198 hybrid_alltoall_any 5155 16.4 1.300 3.021 10.104 12.132 make_images_data 4970 15.5 0.064 0.069 9.913 11.985 fft_wrap_pw1pw2_140 519 13.2 3.190 3.259 11.460 11.475 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.567 9.604 dbcsr_mm_accdrv_process 20590 16.0 4.208 6.159 7.478 9.381 wfi_extrapolate 11 7.9 0.001 0.001 9.003 9.003 cp_fm_cholesky_decompose 22 10.9 8.983 8.994 8.983 8.994 grid_integrate_task_list 127 12.3 8.495 8.692 8.495 8.692 density_rs2pw 127 9.7 0.005 0.005 8.293 8.340 qs_energies_init_hamiltonians 11 5.9 0.008 0.015 8.269 8.270 fft3d_ps 1281 14.7 3.949 3.963 7.394 7.440 mp_alltoall_d11v 2401 14.1 6.791 6.912 6.791 6.912 calculate_dm_sparse 127 9.5 0.001 0.001 6.691 6.767 grid_collocate_task_list 127 9.7 6.366 6.441 6.366 6.441 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.312 6.351 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.169 6.248 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.768000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2766.363636, yerr=151.748160 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 1.260933E+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 18743341. 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 0.025 0.115 85.159 85.161 qs_energies 1 2.0 0.000 0.000 84.475 84.488 ls_scf 1 3.0 0.002 0.020 83.572 83.585 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.392 72.576 multiply_cannon 111 7.7 0.017 0.020 55.693 56.930 multiply_cannon_loop 111 8.7 0.229 0.243 52.285 53.813 ls_scf_main 1 4.0 0.001 0.007 52.317 52.324 density_matrix_trs4 2 5.0 0.002 0.003 46.741 46.817 ls_scf_init_scf 1 4.0 0.004 0.005 28.230 28.240 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.917 26.972 mp_waitall_1 11031 10.9 22.136 25.079 22.136 25.079 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.834 24.855 multiply_cannon_multrec 2664 9.7 8.149 8.908 15.515 17.295 multiply_cannon_sync_h2d 2664 9.7 13.711 15.172 13.711 15.172 make_m2s 222 7.7 0.009 0.011 13.088 13.588 make_images 222 8.7 0.099 0.108 13.066 13.568 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.402 12.365 make_images_data 222 9.7 0.004 0.005 7.659 8.269 dbcsr_mm_accdrv_process 4760 10.4 0.586 0.691 6.983 7.930 hybrid_alltoall_any 227 10.6 0.217 1.854 6.542 7.872 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.415 7.525 dbcsr_mm_accdrv_process_sort 4760 11.4 6.199 7.097 6.199 7.097 calculate_norms 4752 9.8 5.523 6.151 5.523 6.151 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.028 5.171 mp_sum_l 887 5.1 3.151 4.618 3.151 4.618 make_images_sizes 222 9.7 0.000 0.000 0.761 3.671 mp_alltoall_i44 222 10.7 0.761 3.671 0.761 3.671 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.064 3.398 mp_irecv_dv 6231 10.9 2.048 3.373 2.048 3.373 arnoldi_extremal 4 6.8 0.000 0.000 3.205 3.239 arnoldi_normal_ev 4 7.8 0.001 0.003 3.205 3.239 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.277 3.231 build_subspace 16 8.4 0.009 0.012 3.109 3.111 ls_scf_post 1 4.0 0.000 0.001 3.021 3.036 ls_scf_store_result 1 5.0 0.000 0.000 2.840 2.878 dbcsr_special_finalize 555 9.7 0.005 0.006 2.313 2.786 dbcsr_merge_single_wm 555 10.7 0.459 0.599 2.305 2.777 make_images_pack 222 9.7 2.207 2.637 2.209 2.638 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.322 2.549 dbcsr_sort_data 658 11.4 2.105 2.519 2.105 2.519 dbcsr_matrix_vector_mult_local 304 10.0 2.079 2.458 2.081 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.326 2.426 buffer_matrices_ensure_size 222 8.7 1.748 2.081 1.748 2.081 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.832 1.842 rebuild_ks_matrix 3 7.3 0.000 0.000 1.822 1.833 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.038 1.822 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.161000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 2.173850E+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 0.057 0.112 89.329 89.330 qs_energies 1 2.0 0.003 0.021 88.656 88.723 ls_scf 1 3.0 0.000 0.003 87.319 87.369 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.421 73.769 multiply_cannon 111 7.7 0.027 0.036 52.226 56.250 ls_scf_main 1 4.0 0.001 0.013 53.935 53.942 multiply_cannon_loop 111 8.7 0.135 0.145 49.508 52.693 density_matrix_trs4 2 5.0 0.003 0.017 48.231 48.470 ls_scf_init_scf 1 4.0 0.001 0.004 29.668 29.677 mp_waitall_1 9105 10.9 20.070 29.129 20.070 29.129 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.417 28.517 multiply_cannon_multrec 1332 9.7 13.451 17.434 22.730 27.871 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.076 26.089 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 10.956 19.972 make_m2s 222 7.7 0.006 0.008 14.907 15.457 make_images 222 8.7 1.368 1.694 14.876 15.427 dbcsr_mm_accdrv_process 4041 10.4 0.331 0.504 8.877 10.444 dbcsr_mm_accdrv_process_sort 4041 11.4 8.394 9.908 8.394 9.908 make_images_data 222 9.7 0.004 0.005 8.600 9.393 hybrid_alltoall_any 227 10.6 0.541 2.552 7.988 9.282 mp_sum_l 887 5.1 4.836 7.678 4.836 7.678 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.197 7.648 mp_irecv_dv 3311 11.0 3.177 7.581 3.177 7.581 calculate_norms 2376 9.8 6.069 6.862 6.069 6.862 multiply_cannon_sync_h2d 1332 9.7 4.903 6.181 4.903 6.181 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.574 6.052 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.924 5.173 arnoldi_extremal 4 6.8 0.000 0.000 4.612 4.629 arnoldi_normal_ev 4 7.8 0.001 0.005 4.611 4.629 build_subspace 16 8.4 0.014 0.021 4.364 4.366 ls_scf_post 1 4.0 0.003 0.023 3.716 3.766 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.531 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.117 3.352 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.220 2.745 3.222 mp_allgather_i34 111 8.7 0.708 2.849 0.708 2.849 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.613 2.735 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.161 2.471 dbcsr_data_new 4174 10.1 2.115 2.409 2.115 2.409 make_images_pack 222 9.7 1.819 2.126 1.822 2.129 dbcsr_sort_data 436 11.2 1.827 2.048 1.827 2.048 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.862 1.874 rebuild_ks_matrix 3 7.3 0.000 0.000 1.850 1.861 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.016 1.850 1.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.330000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 2.801119E+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 0.061 0.173 93.259 93.261 qs_energies 1 2.0 0.000 0.002 92.071 92.092 ls_scf 1 3.0 0.001 0.005 90.636 90.660 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.227 75.519 ls_scf_main 1 4.0 0.000 0.003 56.579 56.586 multiply_cannon 111 7.7 0.033 0.072 51.998 56.356 multiply_cannon_loop 111 8.7 0.117 0.134 49.198 52.928 density_matrix_trs4 2 5.0 0.002 0.004 50.660 50.854 mp_waitall_1 7281 11.0 23.661 32.998 23.661 32.998 ls_scf_init_scf 1 4.0 0.001 0.003 30.413 30.425 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.014 29.132 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.686 26.701 multiply_cannon_multrec 888 9.7 12.740 15.271 21.302 24.421 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.953 22.487 make_m2s 222 7.7 0.006 0.007 16.564 17.374 make_images 222 8.7 1.584 1.866 16.526 17.333 make_images_data 222 9.7 0.004 0.004 9.708 10.810 hybrid_alltoall_any 227 10.6 0.640 2.956 9.190 10.449 dbcsr_mm_accdrv_process 3754 10.4 0.315 0.495 8.073 9.323 dbcsr_mm_accdrv_process_sort 3754 11.4 7.640 8.829 7.640 8.829 mp_sum_l 887 5.1 5.013 8.758 5.013 8.758 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.647 7.395 multiply_cannon_sync_h2d 888 9.7 6.030 6.983 6.030 6.983 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.454 6.855 mp_irecv_dv 2335 11.1 2.438 6.803 2.438 6.803 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.856 6.768 arnoldi_extremal 4 6.8 0.000 0.000 5.058 5.076 arnoldi_normal_ev 4 7.8 0.001 0.005 5.058 5.076 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.727 4.981 build_subspace 16 8.4 0.014 0.020 4.758 4.764 calculate_norms 1584 9.8 4.341 4.719 4.341 4.719 mp_allgather_i34 111 8.7 0.921 3.863 0.921 3.863 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.429 3.769 ls_scf_post 1 4.0 0.001 0.006 3.644 3.663 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.612 3.029 3.615 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.850 2.961 dbcsr_data_new 4116 9.9 2.100 2.442 2.100 2.442 make_images_sizes 222 9.7 0.000 0.000 1.010 2.149 mp_alltoall_i44 222 10.7 1.010 2.149 1.010 2.149 dbcsr_sort_data 325 11.1 1.890 2.106 1.890 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.017 2.028 rebuild_ks_matrix 3 7.3 0.000 0.000 1.999 2.010 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.032 1.999 2.010 make_images_pack 222 9.7 1.622 1.880 1.625 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.261000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2220.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 3.356037E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.045 96.786 96.787 qs_energies 1 2.0 0.000 0.000 96.179 96.185 ls_scf 1 3.0 0.000 0.000 94.512 94.517 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.337 78.566 ls_scf_main 1 4.0 0.000 0.000 58.732 58.735 multiply_cannon 111 7.7 0.036 0.098 51.399 56.100 density_matrix_trs4 2 5.0 0.002 0.003 52.619 52.735 multiply_cannon_loop 111 8.7 0.152 0.166 46.376 49.837 ls_scf_init_scf 1 4.0 0.000 0.001 32.534 32.536 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.326 31.391 mp_waitall_1 6369 11.0 22.482 29.240 22.482 29.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.868 28.880 multiply_cannon_multrec 1332 9.7 14.146 17.783 22.035 25.601 make_m2s 222 7.7 0.007 0.008 21.273 22.709 make_images 222 8.7 3.142 3.602 21.222 22.660 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.009 17.326 make_images_data 222 9.7 0.004 0.004 11.811 13.617 hybrid_alltoall_any 227 10.6 0.798 3.801 11.141 12.824 dbcsr_mm_accdrv_process 3641 10.4 0.293 0.466 7.524 9.053 dbcsr_mm_accdrv_process_sort 3641 11.4 7.085 8.559 7.085 8.559 mp_sum_l 887 5.1 4.083 7.344 4.083 7.344 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.107 6.162 mp_irecv_dv 3229 10.9 2.082 6.075 2.082 6.075 multiply_cannon_sync_h2d 1332 9.7 5.449 5.925 5.449 5.925 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.152 5.718 arnoldi_extremal 4 6.8 0.000 0.000 5.185 5.199 arnoldi_normal_ev 4 7.8 0.001 0.005 5.185 5.199 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.495 5.111 build_subspace 16 8.4 0.014 0.021 4.852 4.862 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.546 4.767 calculate_norms 2376 9.8 4.186 4.568 4.186 4.568 mp_allgather_i34 111 8.7 2.138 4.437 2.138 4.437 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.576 3.876 dbcsr_matrix_vector_mult_local 304 10.0 3.188 3.693 3.190 3.695 dbcsr_sort_data 658 11.4 3.039 3.391 3.039 3.391 ls_scf_post 1 4.0 0.000 0.000 3.245 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.779 3.195 dbcsr_merge_single_wm 555 10.7 0.531 0.640 2.771 3.188 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.021 3.085 ls_scf_store_result 1 5.0 0.000 0.000 3.003 3.047 dbcsr_data_release 10477 10.7 1.580 2.517 1.580 2.517 dbcsr_finalize 304 7.8 0.049 0.061 1.802 1.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.787000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_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 4.762755E+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 0.045 0.055 91.623 91.625 qs_energies 1 2.0 0.000 0.000 90.891 90.897 ls_scf 1 3.0 0.000 0.000 88.923 88.929 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.408 70.589 ls_scf_main 1 4.0 0.000 0.000 56.324 56.325 multiply_cannon 111 7.7 0.071 0.111 52.121 55.627 multiply_cannon_loop 111 8.7 0.088 0.096 49.566 50.796 density_matrix_trs4 2 5.0 0.002 0.003 49.297 49.374 ls_scf_init_scf 1 4.0 0.000 0.001 29.210 29.211 mp_waitall_1 5436 11.0 24.058 28.764 24.058 28.764 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.941 27.983 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.838 25.846 multiply_cannon_multrec 444 9.7 13.766 16.447 20.817 23.459 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.512 15.176 make_m2s 222 7.7 0.005 0.005 13.618 14.550 make_images 222 8.7 2.038 2.475 13.550 14.481 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.037 13.957 hybrid_alltoall_any 227 10.6 0.802 3.838 8.105 9.798 make_images_data 222 9.7 0.003 0.004 8.326 9.584 multiply_cannon_sync_h2d 444 9.7 6.712 8.569 6.712 8.569 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.407 6.750 7.854 dbcsr_mm_accdrv_process_sort 3003 11.4 6.388 7.458 6.388 7.458 arnoldi_extremal 4 6.8 0.000 0.000 5.854 5.862 arnoldi_normal_ev 4 7.8 0.002 0.005 5.854 5.862 build_subspace 16 8.4 0.015 0.019 5.462 5.476 mp_sum_l 887 5.1 2.667 4.809 2.667 4.809 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.466 4.724 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.189 4.416 dbcsr_matrix_vector_mult_local 304 10.0 3.710 4.193 3.713 4.195 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.533 4.005 mp_irecv_dv 1241 11.2 1.520 3.977 1.520 3.977 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.861 3.836 calculate_norms 792 9.8 3.620 3.809 3.620 3.809 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.566 3.654 mp_allgather_i34 111 8.7 1.125 3.596 1.125 3.596 ls_scf_post 1 4.0 0.000 0.000 3.389 3.396 make_images_sizes 222 9.7 0.000 0.000 0.903 3.386 mp_alltoall_i44 222 10.7 0.902 3.385 0.902 3.385 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.216 dbcsr_data_new 4608 9.7 1.803 2.290 1.803 2.290 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.274 dbcsr_merge_all 275 8.9 0.479 0.530 2.051 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.040 2.041 rebuild_ks_matrix 3 7.3 0.000 0.000 2.009 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.008 2.009 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.952 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.625000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.976101E+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 0.091 0.108 108.199 108.200 qs_energies 1 2.0 0.000 0.000 106.657 106.664 ls_scf 1 3.0 0.000 0.000 103.650 103.656 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.327 77.467 ls_scf_main 1 4.0 0.000 0.000 63.795 63.796 density_matrix_trs4 2 5.0 0.002 0.003 54.758 54.819 multiply_cannon 111 7.7 0.180 0.288 51.122 53.535 multiply_cannon_loop 111 8.7 0.097 0.099 48.022 48.575 ls_scf_init_scf 1 4.0 0.000 0.000 36.082 36.083 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.527 34.551 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 31.831 31.838 mp_waitall_1 4527 11.1 22.228 25.709 22.228 25.709 make_m2s 222 7.7 0.005 0.005 22.618 23.816 make_images 222 8.7 3.568 3.861 22.511 23.708 multiply_cannon_multrec 444 9.7 17.804 18.421 22.426 23.061 hybrid_alltoall_any 227 10.6 1.650 3.605 12.791 15.670 make_images_data 222 9.7 0.004 0.004 13.014 15.250 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.622 13.071 multiply_cannon_sync_h2d 444 9.7 8.782 8.848 8.782 8.848 arnoldi_extremal 4 6.8 0.000 0.000 7.390 7.398 arnoldi_normal_ev 4 7.8 0.003 0.009 7.390 7.398 build_subspace 16 8.4 0.026 0.036 6.836 6.849 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.491 5.636 dbcsr_matrix_vector_mult_local 304 10.0 5.074 5.379 5.076 5.382 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.161 5.250 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.861 5.112 dbcsr_mm_accdrv_process 1814 10.4 0.219 0.346 4.438 4.559 dbcsr_mm_accdrv_process_sort 1814 11.4 4.106 4.237 4.106 4.237 mp_allgather_i34 111 8.7 1.219 3.923 1.219 3.923 ls_scf_post 1 4.0 0.000 0.000 3.773 3.780 make_images_sizes 222 9.7 0.000 0.000 1.483 3.615 mp_alltoall_i44 222 10.7 1.483 3.614 1.483 3.614 ls_scf_store_result 1 5.0 0.000 0.000 3.495 3.532 calculate_norms 792 9.8 3.240 3.290 3.240 3.290 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.258 3.264 dbcsr_finalize 304 7.8 0.082 0.089 3.064 3.153 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.976 2.976 dbcsr_merge_all 275 8.9 0.882 0.909 2.849 2.933 dbcsr_complete_redistribute 5 7.6 1.426 1.467 2.726 2.845 acc_transpose_blocks 444 9.7 0.003 0.003 2.473 2.798 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 2.390 2.715 jit_kernel_transpose 1 13.0 2.384 2.709 2.384 2.709 matrix_ls_to_qs 2 6.0 0.000 0.000 2.382 2.500 dbcsr_sort_data 325 11.1 2.435 2.493 2.435 2.493 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.457 2.459 rebuild_ks_matrix 3 7.3 0.000 0.000 2.391 2.393 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.391 2.393 dbcsr_data_new 6591 9.6 1.816 2.310 1.816 2.310 dbcsr_new_transposed 4 7.5 0.241 0.249 2.270 2.284 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.191 2.221 dbcsr_add_d 103 6.2 0.000 0.000 2.122 2.201 dbcsr_add_anytype 103 7.2 0.858 0.890 2.122 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.200000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7004.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c5db040a230440454a8e3bd22f6bdea107af8992_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.960960E+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 4801545. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.084 222.489 222.491 qs_mol_dyn_low 1 2.0 0.004 0.021 221.407 221.442 qs_forces 5 3.8 0.005 0.017 221.293 221.293 qs_energies 5 4.8 0.003 0.041 218.319 218.369 scf_env_do_scf 5 5.8 0.000 0.002 204.593 204.595 scf_env_do_scf_inner_loop 105 6.6 0.002 0.020 178.321 178.338 qs_scf_new_mos 105 7.6 0.000 0.001 139.487 139.686 qs_scf_loop_do_ot 105 8.6 0.001 0.001 139.486 139.685 dbcsr_multiply_generic 1445 12.2 0.129 0.137 132.547 132.993 ot_scf_mini 105 9.6 0.003 0.004 129.573 129.722 multiply_cannon 1445 13.2 0.274 0.287 113.612 115.942 multiply_cannon_loop 1445 14.2 2.819 2.973 112.015 113.307 velocity_verlet 4 3.0 0.001 0.023 106.305 106.306 ot_mini 105 10.6 0.001 0.003 60.178 60.281 multiply_cannon_multrec 69360 15.2 29.690 34.872 39.662 44.599 mp_waitall_1 488190 16.1 34.579 43.058 34.579 43.058 qs_ot_get_p 112 10.4 0.001 0.001 39.347 39.623 qs_ot_get_derivative 55 11.6 0.001 0.001 38.503 38.645 multiply_cannon_metrocomm3 69360 15.2 0.203 0.212 25.755 34.660 multiply_cannon_sync_h2d 69360 15.2 29.170 33.117 29.170 33.117 qs_ot_p2m_diag 40 11.0 0.020 0.030 28.352 28.461 rebuild_ks_matrix 110 8.4 0.000 0.000 28.280 28.450 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.023 28.280 28.449 init_scf_loop 7 6.6 0.000 0.004 26.243 26.244 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.955 26.112 cp_dbcsr_syevd 40 12.0 0.002 0.004 25.243 25.244 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.978 23.276 apply_single 62 13.6 0.000 0.000 22.978 23.276 prepare_preconditioner 7 7.6 0.000 0.000 21.444 21.480 make_preconditioner 7 8.6 0.000 0.002 21.444 21.480 ot_new_cg_direction 55 11.6 0.001 0.002 20.973 20.973 cp_fm_syevd 40 13.0 0.000 0.001 20.278 20.408 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.065 16.466 calculate_rho_elec 110 8.6 0.030 0.033 16.064 16.465 cp_fm_redistribute_end 40 14.0 7.813 15.588 7.818 15.589 cp_fm_syevd_base 40 14.0 7.766 15.544 7.766 15.544 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.886 15.041 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.401 14.469 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.331 14.439 mp_sum_l 4764 12.2 12.405 13.306 12.405 13.306 calculate_dm_sparse 110 9.5 0.000 0.001 11.534 11.670 pw_transfer 1645 12.4 0.080 0.096 11.222 11.392 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.083 11.257 init_scf_run 5 5.8 0.000 0.001 11.017 11.018 scf_env_initial_rho_setup 5 6.8 0.001 0.003 11.017 11.018 density_rs2pw 110 9.6 0.005 0.006 10.334 10.780 dbcsr_mm_accdrv_process 154766 15.8 6.209 6.395 9.838 10.632 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.199 10.262 qs_vxc_create 110 10.4 0.002 0.004 10.218 10.252 cp_fm_cholesky_invert 7 10.6 10.182 10.190 10.182 10.190 fft_wrap_pw1pw2_240 915 15.0 0.839 0.923 9.736 9.917 check_diag 80 13.5 8.685 8.936 9.357 9.486 sum_up_and_integrate 60 10.3 0.001 0.003 8.105 8.120 integrate_v_rspace 60 11.3 0.001 0.002 8.087 8.103 fft3d_pb 915 16.0 2.405 2.597 7.901 8.066 transfer_rs2pw 445 10.6 0.007 0.008 7.193 7.676 xc_rho_set_and_dset_create 110 12.4 0.076 0.095 7.171 7.416 acc_transpose_blocks 69360 15.2 0.357 0.372 6.787 7.337 multiply_cannon_metrocomm1 69360 15.2 0.099 0.104 4.713 7.212 xc_vxc_pw_create 60 11.3 0.039 0.050 6.810 6.844 make_full_single_inverse 7 9.6 0.001 0.002 6.779 6.815 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.637 6.653 make_m2s 2890 13.2 0.078 0.088 6.058 6.633 make_images 2890 14.2 0.238 0.258 5.950 6.525 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.217 6.279 calculate_first_density_matrix 1 7.0 0.000 0.002 6.257 6.274 xc_pw_derive 510 13.4 0.005 0.007 5.988 6.051 mp_alltoall_z22v 2340 17.7 4.881 5.162 4.881 5.162 mp_waitany 7680 13.5 4.383 5.015 4.383 5.015 multiply_cannon_metrocomm4 67915 15.2 0.189 0.208 2.016 4.881 wfi_extrapolate 5 7.8 0.000 0.001 4.459 4.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=222.491000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c5db040a230440454a8e3bd22f6bdea107af8992 Summary: empty Status: OK