=== 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: a85683755d81adf0cf830f47158f5f9130b730a2 ################# 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 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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_prefixed_scalapack.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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/01 job id: 43241613 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/02 job id: 43241614 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/03 job id: 43241615 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/04 job id: 43241616 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/05 job id: 43241617 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/06 job id: 43241618 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/07 job id: 43241619 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/08 job id: 43241620 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/09 job id: 43241621 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/10 job id: 43241623 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/11 job id: 43241624 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/12 job id: 43241625 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/13 job id: 43241626 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/14 job id: 43241627 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/15 job id: 43241628 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/16 job id: 43241630 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/17 job id: 43241633 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/18 job id: 43241634 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/19 job id: 43241635 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/20 job id: 43241636 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/21 job id: 43241637 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/22 job id: 43241639 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/23 job id: 43241642 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/24 job id: 43241643 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/25 job id: 43241644 --- 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/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/26 job id: 43241645 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. 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 0.023 0.036 137.439 137.440 farming_run 1 2.0 136.444 136.456 137.407 137.410 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459667E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 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 0.009 0.024 116.203 116.203 qs_energies 1 2.0 0.000 0.000 115.983 115.984 mp2_main 1 3.0 0.000 0.000 114.044 114.045 mp2_gpw_main 1 4.0 0.061 0.080 113.182 113.183 mp2_ri_gpw_compute_in 1 5.0 0.186 0.208 93.589 93.743 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.389 55.818 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.174 41.675 47.027 get_2c_integrals 1 6.0 0.000 0.001 37.254 37.722 integrate_v_rspace 273 8.0 0.439 0.457 25.090 30.175 pw_transfer 6555 10.6 0.373 0.401 27.363 28.158 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.015 26.696 grid_integrate_task_list 273 9.0 20.907 26.477 20.907 26.477 fft_wrap_pw1pw2_100 2178 12.4 1.195 1.407 23.583 24.268 compute_2c_integrals 1 7.0 0.002 0.003 19.635 19.645 rpa_ri_compute_en 1 5.0 0.000 0.000 19.444 19.554 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.912 19.335 mp2_eri_2c_integrate_gpw 1 9.0 2.370 2.420 18.909 19.334 cp_fm_cholesky_decompose 12 8.2 17.629 18.118 17.629 18.118 cholesky_decomp 1 7.0 0.000 0.000 16.464 16.967 fft3d_s 5443 13.4 16.154 16.664 16.176 16.686 ao_to_mo_and_store_B_mult_1 272 7.0 10.867 15.606 10.867 15.606 calculate_wavefunction 272 8.0 5.385 5.544 12.504 13.113 rpa_num_int 1 6.0 0.004 0.069 11.043 11.043 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.566 10.837 calc_mat_Q 8 8.0 0.000 0.000 9.355 9.694 contract_S_to_Q 8 9.0 0.000 0.000 8.778 9.113 parallel_gemm_fm 14 9.1 0.000 0.000 8.363 8.634 parallel_gemm_fm_cosma 14 10.1 8.363 8.634 8.363 8.634 calc_potential_gpw 544 9.5 0.006 0.006 8.246 8.601 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.234 8.534 potential_pw2rs 545 10.0 0.107 0.110 7.706 8.424 create_integ_mat 1 6.0 0.295 0.297 8.149 8.149 collocate_single_gaussian 272 10.0 0.040 0.043 7.481 7.738 array2fm 1 7.0 0.000 0.000 6.717 7.324 pw_scatter_s 2720 13.7 4.415 4.612 4.415 4.612 pw_gather_s 2722 13.2 3.887 4.223 3.887 4.223 array2fm_buffer_send 1 8.0 2.969 3.152 2.969 3.152 pw_poisson_solve 545 10.5 1.124 1.172 2.180 2.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.181928, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. 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 0.041 0.068 400.582 400.584 farming_run 1 2.0 399.408 399.415 400.515 400.533 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227158E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 20437349. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.051 209.377 209.378 qs_energies 1 2.0 0.000 0.000 209.150 209.160 scf_env_do_scf 1 3.0 0.000 0.000 106.185 106.185 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.321 105.329 rebuild_ks_matrix 4 6.0 0.000 0.000 105.320 105.328 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.320 105.328 hfx_ks_matrix 4 8.0 0.001 0.001 104.949 104.952 integrate_four_center 4 9.0 0.144 0.457 104.948 104.951 mp2_main 1 3.0 0.000 0.000 102.673 102.683 mp2_gpw_main 1 4.0 0.041 0.066 101.812 101.824 integrate_four_center_main 4 10.0 0.127 0.538 96.577 99.318 integrate_four_center_bin 265 11.0 96.451 99.314 96.451 99.314 init_scf_loop 1 4.0 0.000 0.000 92.097 92.097 mp2_ri_gpw_compute_in 1 5.0 0.071 0.087 74.897 76.110 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.016 54.647 55.867 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 42.336 46.961 integrate_v_rspace 95 8.0 0.400 0.567 28.702 33.238 pw_transfer 2240 10.6 0.146 0.160 29.817 30.183 ao_to_mo_and_store_B_mult_1 91 7.0 10.618 29.660 10.618 29.660 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.846 29.254 grid_integrate_task_list 95 9.0 23.995 28.774 23.995 28.774 mp2_ri_gpw_compute_en 1 5.0 0.068 0.098 26.733 28.625 fft_wrap_pw1pw2_100 730 12.4 1.234 1.384 26.589 27.007 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.847 1.925 24.850 24.860 get_2c_integrals 1 6.0 0.008 0.072 20.142 20.193 compute_2c_integrals 1 7.0 0.009 0.012 19.104 19.127 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 18.725 18.962 mp2_eri_2c_integrate_gpw 1 9.0 1.747 1.835 18.723 18.961 fft3d_s 1823 13.4 18.408 18.761 18.421 18.774 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.087 14.087 calculate_wavefunction 91 8.0 2.020 2.052 9.742 9.983 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.577 8.783 9.408 potential_pw2rs 186 10.0 0.033 0.035 8.593 9.138 local_gemm 172 8.0 8.229 8.831 8.229 8.831 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.156 8.491 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.537 7.830 8.139 calc_potential_gpw 182 9.5 0.002 0.002 7.865 8.133 collocate_single_gaussian 91 10.0 0.017 0.030 7.823 8.053 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.448 6.345 6.448 mp_sendrecv_dm3 2068 8.0 5.847 6.136 5.847 6.136 mp_sync 38 10.4 3.064 5.659 3.064 5.659 pw_gather_s 912 13.2 4.931 5.385 4.931 5.385 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.812164, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 451.440640E+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 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.070 0.132 60.429 60.432 qs_mol_dyn_low 1 2.0 0.014 0.093 59.945 59.953 qs_forces 11 3.9 0.002 0.002 59.500 59.503 qs_energies 11 4.9 0.002 0.005 58.015 58.032 scf_env_do_scf 11 5.9 0.000 0.001 46.087 46.087 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.569 43.569 dbcsr_multiply_generic 2286 12.5 0.100 0.103 38.702 39.219 qs_scf_new_mos 108 7.5 0.000 0.000 33.121 33.380 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.120 33.380 multiply_cannon 2286 13.5 0.182 0.189 30.667 32.357 multiply_cannon_loop 2286 14.5 1.493 1.571 29.875 31.618 ot_scf_mini 108 9.5 0.002 0.002 31.425 31.599 velocity_verlet 10 3.0 0.001 0.001 27.313 27.318 ot_mini 108 10.5 0.001 0.001 20.216 20.462 mp_waitall_1 267858 16.1 10.937 18.461 10.937 18.461 qs_ot_get_derivative 108 11.5 0.001 0.001 17.139 17.325 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 7.766 16.519 multiply_cannon_multrec 54864 15.5 4.193 6.531 8.591 12.534 init_scf_run 11 5.9 0.000 0.001 10.467 10.468 scf_env_initial_rho_setup 11 6.9 0.005 0.036 10.467 10.468 rebuild_ks_matrix 119 8.3 0.000 0.000 8.288 8.421 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.288 8.421 calculate_first_density_matrix 1 7.0 0.001 0.007 8.168 8.180 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.386 7.517 mp_sum_l 7207 12.9 5.800 7.516 5.800 7.516 multiply_cannon_sync_h2d 54864 15.5 5.803 7.425 5.803 7.425 qs_ot_get_p 119 10.4 0.001 0.001 6.496 6.803 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.047 6.464 dbcsr_mm_accdrv_process 76910 16.1 1.164 1.829 4.322 6.191 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.561 5.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.482 5.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.045 5.105 sum_up_and_integrate 119 10.3 0.012 0.014 4.685 4.692 integrate_v_rspace 119 11.3 0.002 0.002 4.674 4.681 qs_rho_update_rho_low 119 7.7 0.002 0.016 4.375 4.442 calculate_rho_elec 119 8.7 0.011 0.017 4.373 4.426 calculate_dm_sparse 119 9.5 0.000 0.000 4.269 4.398 jit_kernel_multiply 13 15.8 3.095 4.272 3.095 4.272 acc_transpose_blocks 54864 15.5 0.230 0.248 3.473 4.102 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.480 3.560 multiply_cannon_metrocomm1 54864 15.5 0.051 0.058 2.049 3.329 rs_pw_transfer 974 11.9 0.011 0.012 3.019 3.169 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.921 3.133 apply_single 119 13.6 0.000 0.000 2.921 3.133 acc_transpose_blocks_kernels 54864 16.5 0.247 0.381 2.487 2.955 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.747 2.769 ot_diis_step 108 11.5 0.006 0.006 2.722 2.723 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.686 2.687 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.530 2.619 jit_kernel_transpose 5 15.5 2.240 2.579 2.240 2.579 density_rs2pw 119 9.7 0.004 0.004 2.404 2.510 init_scf_loop 11 6.9 0.000 0.001 2.487 2.488 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.415 2.416 cp_fm_redistribute_end 50 14.0 2.184 2.384 2.189 2.386 cp_fm_diag_elpa_base 50 14.0 0.195 2.340 0.196 2.349 mp_sum_d 4125 12.0 1.696 2.298 1.696 2.298 wfi_extrapolate 11 7.9 0.001 0.001 2.182 2.182 grid_integrate_task_list 119 12.3 2.055 2.159 2.055 2.159 parallel_gemm_fm 81 9.0 0.000 0.000 2.051 2.055 parallel_gemm_fm_cosma 81 10.0 2.051 2.055 2.051 2.055 potential_pw2rs 119 12.3 0.004 0.004 2.013 2.033 make_m2s 4572 13.5 0.054 0.056 1.959 2.011 pw_transfer 1439 11.6 0.051 0.055 1.897 1.965 make_images 4572 14.5 0.132 0.139 1.876 1.927 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.822 1.893 make_basis_sm 11 9.8 0.000 0.000 1.667 1.669 fft3d_ps 1201 14.6 0.358 0.463 1.603 1.664 mp_alltoall_d11v 2130 13.8 1.494 1.617 1.494 1.617 fft_wrap_pw1pw2_140 487 13.2 0.078 0.093 1.383 1.455 mp_waitany 12084 13.8 1.284 1.455 1.284 1.455 grid_collocate_task_list 119 9.7 1.286 1.342 1.286 1.342 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.899 1.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.432000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 487.403520E+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 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1045778. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.093 0.110 45.168 45.169 qs_mol_dyn_low 1 2.0 0.003 0.004 44.645 44.662 qs_forces 11 3.9 0.002 0.002 44.580 44.582 qs_energies 11 4.9 0.017 0.102 42.850 42.853 scf_env_do_scf 11 5.9 0.000 0.001 31.971 31.972 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.385 29.385 dbcsr_multiply_generic 2286 12.5 0.108 0.129 25.781 26.219 multiply_cannon 2286 13.5 0.208 0.217 20.549 22.130 multiply_cannon_loop 2286 14.5 0.903 0.976 19.352 20.973 qs_scf_new_mos 108 7.5 0.001 0.001 20.151 20.399 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.150 20.398 ot_scf_mini 108 9.5 0.002 0.003 19.235 19.407 velocity_verlet 10 3.0 0.001 0.001 18.724 18.725 mp_waitall_1 217478 16.2 7.900 14.251 7.900 14.251 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 5.747 12.429 ot_mini 108 10.5 0.001 0.001 12.040 12.276 multiply_cannon_multrec 27432 15.5 1.971 4.307 6.891 10.388 qs_ot_get_derivative 108 11.5 0.001 0.001 9.543 9.717 init_scf_run 11 5.9 0.000 0.001 9.541 9.550 scf_env_initial_rho_setup 11 6.9 0.008 0.060 9.541 9.550 calculate_first_density_matrix 1 7.0 0.000 0.002 8.001 8.019 rebuild_ks_matrix 119 8.3 0.000 0.000 7.328 7.463 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.327 7.463 dbcsr_mm_accdrv_process 47894 16.0 3.708 6.673 4.851 7.230 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.445 6.567 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.684 4.686 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.665 4.514 qs_ot_get_p 119 10.4 0.001 0.001 4.267 4.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.441 4.481 sum_up_and_integrate 119 10.3 0.024 0.028 4.344 4.351 integrate_v_rspace 119 11.3 0.002 0.002 4.320 4.329 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.066 4.154 apply_single 119 13.6 0.000 0.000 3.066 4.154 mp_sum_l 7207 12.9 2.121 4.070 2.121 4.070 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.744 3.781 calculate_rho_elec 119 8.7 0.021 0.024 3.743 3.780 calculate_dm_sparse 119 9.5 0.000 0.000 3.446 3.523 acc_transpose_blocks 27432 15.5 0.111 0.116 2.762 3.174 jit_kernel_multiply 10 16.1 1.090 3.058 1.090 3.058 make_m2s 4572 13.5 0.053 0.054 2.655 2.906 rs_pw_transfer 974 11.9 0.010 0.011 2.784 2.893 make_images 4572 14.5 0.200 0.235 2.566 2.815 multiply_cannon_sync_h2d 27432 15.5 2.174 2.789 2.174 2.789 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.731 2.752 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.653 2.661 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 2.222 2.570 init_scf_loop 11 6.9 0.000 0.000 2.552 2.552 ot_diis_step 108 11.5 0.010 0.011 2.434 2.435 jit_kernel_transpose 5 15.5 2.040 2.309 2.040 2.309 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.293 2.293 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.189 2.281 density_rs2pw 119 9.7 0.004 0.004 2.092 2.171 potential_pw2rs 119 12.3 0.006 0.006 2.075 2.088 grid_integrate_task_list 119 12.3 1.840 1.943 1.840 1.943 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.927 1.927 cp_fm_redistribute_end 50 14.0 1.583 1.899 1.586 1.901 cp_fm_diag_elpa_base 50 14.0 0.305 1.844 0.312 1.875 pw_transfer 1439 11.6 0.063 0.066 1.831 1.855 parallel_gemm_fm 81 9.0 0.000 0.000 1.771 1.787 parallel_gemm_fm_cosma 81 10.0 1.771 1.787 1.771 1.787 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.742 1.768 make_basis_sm 11 9.8 0.000 0.000 1.749 1.751 prepare_preconditioner 11 7.9 0.000 0.000 1.642 1.670 make_preconditioner 11 8.9 0.000 0.000 1.642 1.670 make_images_data 4572 15.5 0.045 0.051 1.183 1.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.544 1.602 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.034 1.495 wfi_extrapolate 11 7.9 0.001 0.001 1.477 1.477 fft3d_ps 1201 14.6 0.499 0.551 1.451 1.473 multiply_cannon_metrocomm1 27432 15.5 0.031 0.035 0.300 1.426 mp_allgather_i34 2286 14.5 0.641 1.410 0.641 1.410 mp_alltoall_d11v 2130 13.8 1.273 1.380 1.273 1.380 fft_wrap_pw1pw2_140 487 13.2 0.074 0.082 1.304 1.332 grid_collocate_task_list 119 9.7 1.241 1.322 1.241 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.245 mp_sum_d 4125 12.0 0.746 1.199 0.746 1.199 qs_energies_init_hamiltonians 11 5.9 0.001 0.009 0.968 0.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.939 0.953 make_images_sizes 4572 15.5 0.005 0.005 0.657 0.934 mp_alltoall_i44 4572 16.5 0.653 0.929 0.653 0.929 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.531 0.913 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.169000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.909091, yerr=1.676281 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 523.382784E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 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 0.050 0.091 39.199 39.201 qs_mol_dyn_low 1 2.0 0.003 0.004 38.514 38.529 qs_forces 11 3.9 0.001 0.002 38.456 38.458 qs_energies 11 4.9 0.003 0.016 36.882 36.886 scf_env_do_scf 11 5.9 0.001 0.002 26.713 26.713 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.712 23.718 dbcsr_multiply_generic 2286 12.5 0.106 0.112 20.135 20.280 multiply_cannon 2286 13.5 0.201 0.243 16.195 17.177 multiply_cannon_loop 2286 14.5 0.640 0.666 15.333 16.321 velocity_verlet 10 3.0 0.001 0.002 16.168 16.169 qs_scf_new_mos 108 7.5 0.001 0.001 15.498 15.519 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.497 15.519 ot_scf_mini 108 9.5 0.002 0.003 14.741 14.758 ot_mini 108 10.5 0.001 0.001 9.180 9.200 init_scf_run 11 5.9 0.000 0.001 8.802 8.803 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.802 8.802 multiply_cannon_multrec 18288 15.5 1.909 2.877 7.969 8.278 qs_ot_get_derivative 108 11.5 0.001 0.001 7.627 7.642 calculate_first_density_matrix 1 7.0 0.001 0.004 7.057 7.059 dbcsr_mm_accdrv_process 38222 16.0 5.223 6.443 5.976 6.919 rebuild_ks_matrix 119 8.3 0.000 0.000 6.609 6.630 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.608 6.629 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.868 5.888 mp_waitall_1 169478 16.3 3.538 4.833 3.538 4.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.590 4.592 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.366 4.401 sum_up_and_integrate 119 10.3 0.031 0.032 4.090 4.095 integrate_v_rspace 119 11.3 0.002 0.003 4.059 4.068 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.817 3.828 calculate_rho_elec 119 8.7 0.030 0.031 3.816 3.828 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.934 3.562 qs_ot_get_p 119 10.4 0.001 0.001 3.314 3.342 jit_kernel_multiply 10 16.3 0.701 3.086 0.701 3.086 acc_transpose_blocks 18288 15.5 0.077 0.079 2.965 2.987 init_scf_loop 11 6.9 0.000 0.001 2.975 2.979 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.591 2.820 rs_pw_transfer 974 11.9 0.009 0.010 2.637 2.755 acc_transpose_blocks_kernels 18288 16.5 0.210 0.220 2.518 2.532 density_rs2pw 119 9.7 0.004 0.004 2.399 2.483 make_m2s 4572 13.5 0.045 0.046 2.259 2.439 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.071 2.391 apply_single 119 13.6 0.000 0.000 2.071 2.391 make_images 4572 14.5 0.225 0.474 2.172 2.352 jit_kernel_transpose 5 15.6 2.308 2.314 2.308 2.314 calculate_dm_sparse 119 9.5 0.000 0.000 2.281 2.294 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.165 2.173 pw_transfer 1439 11.6 0.063 0.066 2.162 2.172 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.071 2.082 prepare_preconditioner 11 7.9 0.000 0.000 2.001 2.003 make_preconditioner 11 8.9 0.000 0.001 2.001 2.003 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.846 1.929 potential_pw2rs 119 12.3 0.007 0.008 1.876 1.883 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.882 1.882 grid_integrate_task_list 119 12.3 1.802 1.877 1.802 1.877 fft3d_ps 1201 14.6 0.503 0.520 1.760 1.769 mp_sum_l 7207 12.9 1.361 1.760 1.361 1.760 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.649 1.654 parallel_gemm_fm 81 9.0 0.000 0.000 1.649 1.653 parallel_gemm_fm_cosma 81 10.0 1.649 1.653 1.649 1.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.628 1.633 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.608 1.609 make_basis_sm 11 9.8 0.000 0.000 1.590 1.593 cp_fm_redistribute_end 50 14.0 1.190 1.582 1.192 1.583 cp_fm_diag_elpa_base 50 14.0 0.374 1.517 0.389 1.556 ot_diis_step 108 11.5 0.011 0.011 1.523 1.523 multiply_cannon_sync_h2d 18288 15.5 1.335 1.476 1.335 1.476 make_images_data 4572 15.5 0.045 0.049 0.954 1.314 wfi_extrapolate 11 7.9 0.001 0.001 1.277 1.277 grid_collocate_task_list 119 9.7 1.204 1.271 1.204 1.271 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.241 1.255 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.818 1.209 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.096 1.099 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.436 1.095 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.948 0.971 mp_sum_d 4125 12.0 0.717 0.933 0.717 0.933 mp_alltoall_d11v 2130 13.8 0.767 0.885 0.767 0.885 cp_fm_cholesky_invert 11 10.9 0.873 0.877 0.873 0.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.813 0.815 make_images_sizes 4572 15.5 0.005 0.005 0.564 0.801 mp_alltoall_i44 4572 16.5 0.559 0.796 0.559 0.796 mp_waitany 9880 13.7 0.653 0.794 0.653 0.794 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.201000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=2.479669 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 544.313344E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.472 0.551 41.865 41.866 qs_mol_dyn_low 1 2.0 0.003 0.003 40.323 40.331 qs_forces 11 3.9 0.002 0.002 40.260 40.264 qs_energies 11 4.9 0.171 0.463 38.523 38.532 scf_env_do_scf 11 5.9 0.004 0.011 29.438 29.440 scf_env_do_scf_inner_loop 108 6.5 0.005 0.016 25.404 25.408 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.779 20.894 velocity_verlet 10 3.0 0.001 0.001 18.779 18.780 multiply_cannon 2286 13.5 0.226 0.265 16.717 17.181 qs_scf_new_mos 108 7.5 0.001 0.001 17.093 17.149 qs_scf_loop_do_ot 108 8.5 0.013 0.060 17.093 17.148 ot_scf_mini 108 9.5 0.003 0.006 16.126 16.200 multiply_cannon_loop 2286 14.5 0.942 0.970 15.691 16.052 ot_mini 108 10.5 0.001 0.001 10.070 10.155 multiply_cannon_multrec 27432 15.5 2.390 3.095 9.295 9.767 qs_ot_get_derivative 108 11.5 0.001 0.001 8.240 8.319 dbcsr_mm_accdrv_process 47916 15.9 5.456 8.002 6.811 8.276 init_scf_run 11 5.9 0.000 0.001 7.082 7.082 scf_env_initial_rho_setup 11 6.9 0.001 0.004 7.081 7.082 rebuild_ks_matrix 119 8.3 0.000 0.000 6.700 6.755 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.700 6.754 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.961 6.010 calculate_first_density_matrix 1 7.0 0.001 0.004 5.608 5.610 init_scf_loop 11 6.9 0.009 0.036 4.006 4.007 sum_up_and_integrate 119 10.3 0.035 0.038 3.897 3.902 integrate_v_rspace 119 11.3 0.002 0.002 3.861 3.867 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.230 3.718 qs_ot_get_p 119 10.4 0.001 0.001 3.392 3.481 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.433 3.434 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.380 3.415 calculate_rho_elec 119 8.7 0.040 0.046 3.379 3.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.162 3.177 mp_waitall_1 145218 16.4 2.523 3.124 2.523 3.124 prepare_preconditioner 11 7.9 0.000 0.000 3.010 3.018 make_preconditioner 11 8.9 0.014 0.041 3.010 3.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.588 2.938 acc_transpose_blocks 27432 15.5 0.112 0.114 2.886 2.906 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.146 2.607 apply_single 119 13.6 0.000 0.000 2.145 2.607 rs_pw_transfer 974 11.9 0.009 0.009 2.359 2.496 make_m2s 4572 13.5 0.055 0.057 2.404 2.482 calculate_dm_sparse 119 9.5 0.000 0.000 2.392 2.447 make_images 4572 14.5 0.272 0.332 2.295 2.372 jit_kernel_multiply 10 15.7 1.293 2.276 1.293 2.276 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 2.256 2.272 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.052 2.063 jit_kernel_transpose 5 15.6 1.988 1.996 1.988 1.996 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.956 1.984 density_rs2pw 119 9.7 0.003 0.004 1.870 1.961 grid_integrate_task_list 119 12.3 1.822 1.912 1.822 1.912 mp_sum_l 7207 12.9 1.237 1.860 1.237 1.860 pw_transfer 1439 11.6 0.062 0.066 1.765 1.795 ot_diis_step 108 11.5 0.012 0.012 1.787 1.788 parallel_gemm_fm 81 9.0 0.000 0.000 1.758 1.762 parallel_gemm_fm_cosma 81 10.0 1.758 1.762 1.758 1.762 make_basis_sm 11 9.8 0.000 0.000 1.752 1.753 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.729 1.730 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.676 1.708 potential_pw2rs 119 12.3 0.008 0.009 1.691 1.695 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.565 1.576 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.483 1.484 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.850 1.459 cp_fm_redistribute_end 50 14.0 0.973 1.455 0.975 1.455 cp_fm_diag_elpa_base 50 14.0 0.458 1.396 0.478 1.436 wfi_extrapolate 11 7.9 0.001 0.001 1.407 1.407 fft3d_ps 1201 14.6 0.530 0.584 1.358 1.383 fft_wrap_pw1pw2_140 487 13.2 0.084 0.094 1.319 1.353 grid_collocate_task_list 119 9.7 1.222 1.301 1.222 1.301 qs_energies_init_hamiltonians 11 5.9 0.004 0.009 1.259 1.263 dbcsr_complete_redistribute 329 12.2 0.141 0.181 0.955 1.233 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.146 1.168 cp_fm_upper_to_full 72 13.5 0.817 1.138 0.817 1.138 multiply_cannon_sync_h2d 27432 15.5 0.976 1.076 0.976 1.076 make_images_data 4572 15.5 0.045 0.048 0.933 1.037 cp_fm_cholesky_invert 11 10.9 0.991 0.995 0.991 0.995 mp_alltoall_d11v 2130 13.8 0.815 0.979 0.815 0.979 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.695 0.975 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.805 0.947 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.838 0.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.866000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.727273, yerr=2.956992 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 601.931776E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 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 0.015 0.030 32.181 32.182 qs_mol_dyn_low 1 2.0 0.003 0.004 31.959 31.970 qs_forces 11 3.9 0.001 0.002 31.899 31.900 qs_energies 11 4.9 0.001 0.001 30.187 30.191 scf_env_do_scf 11 5.9 0.000 0.001 23.104 23.104 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.313 20.314 dbcsr_multiply_generic 2286 12.5 0.093 0.096 15.452 15.535 velocity_verlet 10 3.0 0.001 0.001 14.517 14.520 multiply_cannon 2286 13.5 0.231 0.241 12.479 13.033 qs_scf_new_mos 108 7.5 0.001 0.001 12.417 12.439 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.417 12.438 ot_scf_mini 108 9.5 0.002 0.002 11.704 11.726 multiply_cannon_loop 2286 14.5 0.331 0.344 11.462 11.663 multiply_cannon_multrec 9144 15.5 1.792 2.098 6.427 6.696 ot_mini 108 10.5 0.001 0.001 6.465 6.491 rebuild_ks_matrix 119 8.3 0.000 0.000 6.186 6.211 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.186 6.211 init_scf_run 11 5.9 0.000 0.001 5.650 5.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.650 5.650 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.548 5.567 qs_ot_get_derivative 108 11.5 0.001 0.001 5.120 5.141 dbcsr_mm_accdrv_process 12550 15.8 3.344 4.219 4.535 4.608 calculate_first_density_matrix 1 7.0 0.000 0.000 4.504 4.505 sum_up_and_integrate 119 10.3 0.037 0.041 3.735 3.755 integrate_v_rspace 119 11.3 0.002 0.003 3.698 3.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.658 3.660 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.500 3.508 qs_ot_get_p 119 10.4 0.001 0.001 3.369 3.402 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.325 3.332 calculate_rho_elec 119 8.7 0.059 0.061 3.324 3.332 acc_transpose_blocks 9144 15.5 0.039 0.040 2.845 2.883 mp_waitall_1 121218 16.5 2.251 2.778 2.251 2.778 init_scf_loop 11 6.9 0.000 0.000 2.768 2.769 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 2.602 2.644 jit_kernel_transpose 5 15.6 2.486 2.531 2.486 2.531 make_m2s 4572 13.5 0.035 0.036 2.031 2.240 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.200 2.202 make_images 4572 14.5 0.268 0.301 1.940 2.148 calculate_dm_sparse 119 9.5 0.000 0.000 2.006 2.025 rs_pw_transfer 974 11.9 0.008 0.008 1.874 1.964 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.961 1.962 grid_integrate_task_list 119 12.3 1.844 1.935 1.844 1.935 prepare_preconditioner 11 7.9 0.000 0.000 1.863 1.867 make_preconditioner 11 8.9 0.000 0.000 1.863 1.867 jit_kernel_multiply 8 15.9 1.153 1.859 1.153 1.859 pw_transfer 1439 11.6 0.063 0.065 1.792 1.822 density_rs2pw 119 9.7 0.003 0.003 1.734 1.820 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.750 1.777 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.702 1.732 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.599 1.613 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.588 1.590 potential_pw2rs 119 12.3 0.010 0.010 1.516 1.539 cp_fm_redistribute_end 50 14.0 0.782 1.537 0.783 1.539 cp_fm_diag_elpa_base 50 14.0 0.706 1.450 0.753 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.392 1.400 fft3d_ps 1201 14.6 0.539 0.551 1.369 1.394 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.374 1.387 grid_collocate_task_list 119 9.7 1.271 1.370 1.271 1.370 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.332 1.363 ot_diis_step 108 11.5 0.012 0.013 1.325 1.325 make_images_data 4572 15.5 0.039 0.042 0.952 1.219 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.211 1.212 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.181 1.200 apply_single 119 13.6 0.000 0.000 1.181 1.199 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.889 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.100 1.100 mp_alltoall_d11v 2130 13.8 0.859 0.995 0.859 0.995 cp_fm_cholesky_invert 11 10.9 0.951 0.954 0.951 0.954 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.916 mp_allgather_i34 2286 14.5 0.350 0.839 0.350 0.839 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.807 0.814 multiply_cannon_sync_h2d 9144 15.5 0.709 0.774 0.709 0.774 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.762 0.768 mp_alltoall_z22v 1201 16.6 0.704 0.735 0.704 0.735 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.405 0.718 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.668 0.717 arnoldi_extremal 119 11.4 0.001 0.001 0.629 0.647 arnoldi_normal_ev 119 12.4 0.007 0.014 0.628 0.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.182000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.636364, yerr=8.160578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 756.408320E+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 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.023 0.075 46.031 46.032 qs_mol_dyn_low 1 2.0 0.003 0.004 45.191 45.390 qs_forces 11 3.9 0.002 0.003 45.093 45.094 qs_energies 11 4.9 0.001 0.001 43.156 43.162 scf_env_do_scf 11 5.9 0.001 0.001 33.864 33.881 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.689 25.691 velocity_verlet 10 3.0 0.001 0.001 23.530 23.536 dbcsr_multiply_generic 2286 12.5 0.164 0.204 20.704 20.885 multiply_cannon 2286 13.5 0.306 0.313 16.067 17.312 qs_scf_new_mos 108 7.5 0.001 0.001 16.735 16.833 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.735 16.832 multiply_cannon_loop 2286 14.5 0.345 0.351 14.753 16.042 ot_scf_mini 108 9.5 0.002 0.002 15.630 15.732 ot_mini 108 10.5 0.001 0.001 9.495 9.612 multiply_cannon_multrec 9144 15.5 3.592 5.171 8.875 9.168 init_scf_loop 11 6.9 0.000 0.000 8.142 8.148 qs_ot_get_derivative 108 11.5 0.001 0.001 7.464 7.565 init_scf_run 11 5.9 0.000 0.001 7.209 7.209 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.209 7.209 prepare_preconditioner 11 7.9 0.000 0.000 7.134 7.147 make_preconditioner 11 8.9 0.000 0.000 7.134 7.147 rebuild_ks_matrix 119 8.3 0.000 0.000 6.894 7.033 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.893 7.032 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.707 7.026 dbcsr_mm_accdrv_process 12550 15.8 3.988 5.709 5.159 6.808 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.247 6.373 calculate_first_density_matrix 1 7.0 0.000 0.000 5.738 5.768 cp_fm_upper_to_full 72 14.2 3.178 4.546 3.178 4.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.253 4.254 mp_waitall_1 97218 16.6 2.837 3.977 2.837 3.977 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.938 3.948 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.741 3.787 calculate_rho_elec 119 8.7 0.118 0.121 3.740 3.787 sum_up_and_integrate 119 10.3 0.064 0.065 3.718 3.725 integrate_v_rspace 119 11.3 0.003 0.003 3.654 3.662 qs_ot_get_p 119 10.4 0.001 0.001 3.389 3.528 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.649 3.396 acc_transpose_blocks 9144 15.5 0.038 0.039 2.980 2.988 make_m2s 4572 13.5 0.039 0.039 2.594 2.834 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.504 2.821 apply_single 119 13.6 0.000 0.000 2.504 2.821 dbcsr_complete_redistribute 329 12.2 0.290 0.294 2.011 2.813 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 2.728 2.734 make_images 4572 14.5 0.353 0.384 2.471 2.711 jit_kernel_transpose 5 15.6 2.610 2.617 2.610 2.617 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.677 2.468 calculate_dm_sparse 119 9.5 0.000 0.000 2.344 2.363 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.337 2.294 mp_sum_l 7207 12.9 1.481 2.278 1.481 2.278 mp_alltoall_i22 627 13.8 1.449 2.250 1.449 2.250 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.421 2.207 grid_integrate_task_list 119 12.3 2.007 2.027 2.007 2.027 pw_transfer 1439 11.6 0.066 0.066 2.016 2.019 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.012 2.013 ot_diis_step 108 11.5 0.014 0.015 2.001 2.001 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.921 1.925 jit_kernel_multiply 10 15.5 1.144 1.919 1.144 1.919 density_rs2pw 119 9.7 0.003 0.004 1.821 1.839 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.737 1.787 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.760 1.761 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.728 1.729 cp_fm_cholesky_invert 11 10.9 1.691 1.694 1.691 1.694 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.582 1.585 fft3d_ps 1201 14.6 0.567 0.579 1.555 1.558 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.261 1.511 rs_pw_transfer 974 11.9 0.009 0.009 1.484 1.511 grid_collocate_task_list 119 9.7 1.449 1.463 1.449 1.463 make_images_data 4572 15.5 0.042 0.045 1.215 1.452 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.399 1.420 make_basis_sm 11 9.8 0.000 0.000 1.417 1.418 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.415 1.415 cp_fm_diag_elpa_base 50 14.0 1.271 1.323 1.413 1.413 wfi_extrapolate 11 7.9 0.001 0.001 1.383 1.383 potential_pw2rs 119 12.3 0.014 0.014 1.256 1.259 mp_alltoall_d11v 2130 13.8 1.218 1.257 1.218 1.257 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.184 1.203 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.097 1.113 parallel_gemm_fm 81 9.0 0.000 0.000 1.067 1.068 parallel_gemm_fm_cosma 81 10.0 1.067 1.068 1.067 1.068 multiply_cannon_sync_h2d 9144 15.5 1.037 1.041 1.037 1.041 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.980 1.035 qs_create_task_list 11 7.9 0.001 0.001 0.948 0.957 generate_qs_task_list 11 8.9 0.374 0.393 0.948 0.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.032000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.363636, yerr=13.479093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 503.308288E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 405669. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.266 0.332 90.951 90.953 qs_mol_dyn_low 1 2.0 0.004 0.011 90.000 90.010 qs_forces 11 3.9 0.002 0.002 89.013 89.015 qs_energies 11 4.9 0.002 0.012 85.722 85.737 scf_env_do_scf 11 5.9 0.004 0.017 72.605 72.608 scf_env_do_scf_inner_loop 99 6.5 0.006 0.019 66.787 66.791 dbcsr_multiply_generic 2055 12.4 0.108 0.112 56.118 56.471 qs_scf_new_mos 99 7.5 0.000 0.001 49.700 49.812 qs_scf_loop_do_ot 99 8.5 0.003 0.010 49.699 49.812 ot_scf_mini 99 9.5 0.006 0.019 47.240 47.369 multiply_cannon 2055 13.4 0.176 0.182 46.113 47.163 multiply_cannon_loop 2055 14.4 1.550 1.588 45.082 46.159 velocity_verlet 10 3.0 0.003 0.018 43.484 43.494 ot_mini 99 10.5 0.001 0.001 28.895 29.009 qs_ot_get_derivative 99 11.5 0.002 0.010 21.966 22.061 multiply_cannon_multrec 49320 15.4 12.140 12.810 18.196 18.736 rebuild_ks_matrix 110 8.3 0.000 0.001 14.599 14.736 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.598 14.736 mp_waitall_1 241148 16.1 12.546 13.700 12.546 13.700 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.862 12.982 init_scf_run 11 5.9 0.000 0.001 10.866 10.867 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.866 10.867 multiply_cannon_sync_h2d 49320 15.4 9.877 10.522 9.877 10.522 qs_ot_get_p 110 10.4 0.001 0.001 9.762 9.899 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.384 8.745 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.403 7.896 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.665 7.822 calculate_rho_elec 110 8.6 0.027 0.076 7.665 7.822 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.198 7.668 apply_single 110 13.6 0.000 0.001 7.197 7.667 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.097 7.150 sum_up_and_integrate 110 10.3 0.036 0.043 7.077 7.092 integrate_v_rspace 110 11.3 0.005 0.023 7.041 7.064 ot_diis_step 99 11.5 0.005 0.006 6.603 6.603 mp_sum_l 6514 12.8 5.598 6.391 5.598 6.391 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.263 6.283 dbcsr_mm_accdrv_process 87628 16.1 2.078 2.173 5.936 6.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.802 5.806 init_scf_loop 11 6.9 0.004 0.016 5.783 5.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.467 5.526 calculate_first_density_matrix 1 7.0 0.000 0.001 5.336 5.341 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.205 5.206 density_rs2pw 110 9.6 0.080 0.611 4.770 4.935 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.633 4.635 cp_fm_redistribute_end 48 14.0 4.024 4.599 4.028 4.600 cp_fm_diag_elpa_base 48 14.0 0.565 4.525 0.568 4.550 pw_transfer 1331 11.6 0.054 0.066 4.273 4.438 calculate_dm_sparse 110 9.5 0.000 0.001 4.269 4.385 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 4.186 4.349 acc_transpose_blocks 49320 15.4 0.223 0.233 4.064 4.135 rs_pw_transfer 902 11.9 0.012 0.013 3.970 4.120 make_m2s 4110 13.4 0.061 0.066 4.017 4.117 wfi_extrapolate 11 7.9 0.001 0.001 4.025 4.025 make_images 4110 14.4 0.177 0.191 3.920 4.023 fft3d_ps 1111 14.6 0.758 0.872 3.724 3.875 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 2.949 3.803 jit_kernel_multiply 13 15.9 3.574 3.736 3.574 3.736 prepare_preconditioner 11 7.9 0.000 0.000 3.501 3.520 make_preconditioner 11 8.9 0.000 0.001 3.501 3.520 grid_integrate_task_list 110 12.3 3.207 3.440 3.207 3.440 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.321 3.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.300 3.344 fft_wrap_pw1pw2_140 451 13.1 0.168 0.189 2.921 3.052 acc_transpose_blocks_kernels 49320 16.4 0.445 0.466 2.954 3.003 potential_pw2rs 110 12.3 0.006 0.007 2.742 2.770 mp_sum_d 3879 11.9 2.026 2.587 2.026 2.587 jit_kernel_transpose 5 15.6 2.509 2.545 2.509 2.545 mp_alltoall_d11v 2046 13.8 2.061 2.509 2.061 2.509 mp_alltoall_z22v 1111 16.6 2.060 2.297 2.060 2.297 grid_collocate_task_list 110 9.6 2.089 2.202 2.089 2.202 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.093 2.128 mp_waitany 14300 13.8 1.901 2.098 1.901 2.098 cp_fm_cholesky_invert 11 10.9 1.933 1.938 1.933 1.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.881 1.909 make_images_data 4110 15.4 0.042 0.045 1.736 1.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=90.953000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.545455, yerr=2.571297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 585.510912E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2152495. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.093 0.122 76.539 76.541 qs_mol_dyn_low 1 2.0 0.003 0.004 76.149 76.166 qs_forces 11 3.9 0.002 0.002 75.950 75.954 qs_energies 11 4.9 0.008 0.048 72.470 72.477 scf_env_do_scf 11 5.9 0.019 0.074 60.675 60.734 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 51.067 51.074 dbcsr_multiply_generic 2055 12.4 0.117 0.121 40.867 41.066 velocity_verlet 10 3.0 0.053 0.127 38.086 38.088 multiply_cannon 2055 13.4 0.222 0.249 33.469 34.904 qs_scf_new_mos 99 7.5 0.001 0.001 34.669 34.797 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.669 34.796 ot_scf_mini 99 9.5 0.003 0.004 32.973 33.096 multiply_cannon_loop 2055 14.4 0.927 0.948 31.994 33.003 ot_mini 99 10.5 0.001 0.001 19.227 19.348 multiply_cannon_multrec 24660 15.4 7.525 9.110 13.960 15.616 rebuild_ks_matrix 110 8.3 0.000 0.001 14.740 14.841 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.044 14.739 14.841 qs_ot_get_derivative 99 11.5 0.003 0.018 13.328 13.450 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.157 13.285 mp_waitall_1 186928 16.3 8.677 11.508 8.677 11.508 init_scf_loop 11 6.9 0.045 0.229 9.548 9.570 init_scf_run 11 5.9 0.000 0.001 9.024 9.025 scf_env_initial_rho_setup 11 6.9 0.001 0.002 9.024 9.025 multiply_cannon_metrocomm3 24660 15.4 0.068 0.071 5.526 8.521 multiply_cannon_sync_h2d 24660 15.4 6.990 7.942 6.990 7.942 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.646 7.415 apply_single 110 13.6 0.000 0.001 6.646 7.415 qs_ot_get_p 110 10.4 0.008 0.029 7.112 7.266 sum_up_and_integrate 110 10.3 0.052 0.060 6.926 6.962 integrate_v_rspace 110 11.3 0.002 0.002 6.874 6.912 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.683 6.694 calculate_rho_elec 110 8.6 0.039 0.047 6.682 6.693 dbcsr_mm_accdrv_process 52282 16.1 4.921 5.887 6.276 6.582 prepare_preconditioner 11 7.9 0.000 0.000 6.426 6.451 make_preconditioner 11 8.9 0.000 0.002 6.426 6.451 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.996 6.163 ot_diis_step 99 11.5 0.010 0.010 5.842 5.842 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.965 5.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.209 5.218 make_m2s 4110 13.4 0.057 0.060 4.571 5.115 make_images 4110 14.4 0.402 0.445 4.460 5.000 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.754 4.808 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.762 4.785 calculate_first_density_matrix 1 7.0 0.002 0.014 4.617 4.633 pw_transfer 1331 11.6 0.065 0.070 4.216 4.379 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.109 4.274 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.259 4.259 density_rs2pw 110 9.6 0.004 0.004 3.940 4.107 wfi_extrapolate 11 7.9 0.001 0.001 3.676 3.676 fft3d_ps 1111 14.6 1.077 1.285 3.469 3.623 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.522 3.592 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.573 3.576 cp_fm_redistribute_end 48 14.0 2.645 3.522 2.648 3.522 rs_pw_transfer 902 11.9 0.012 0.013 3.340 3.499 cp_fm_diag_elpa_base 48 14.0 0.837 3.396 0.870 3.480 acc_transpose_blocks 24660 15.4 0.112 0.116 3.366 3.421 grid_integrate_task_list 110 12.3 3.147 3.324 3.147 3.324 calculate_dm_sparse 110 9.5 0.001 0.001 3.225 3.249 cp_fm_cholesky_invert 11 10.9 3.095 3.104 3.095 3.104 make_images_data 4110 15.4 0.047 0.050 2.489 3.036 mp_sum_l 6514 12.8 2.222 3.023 2.222 3.023 hybrid_alltoall_any 4261 16.3 0.103 0.443 2.166 2.955 fft_wrap_pw1pw2_140 451 13.1 0.199 0.222 2.770 2.930 potential_pw2rs 110 12.3 0.008 0.009 2.787 2.827 acc_transpose_blocks_kernels 24660 16.4 0.310 0.317 2.741 2.799 jit_kernel_transpose 5 15.6 2.431 2.486 2.431 2.486 mp_sum_d 3879 11.9 1.667 2.293 1.667 2.293 grid_collocate_task_list 110 9.6 2.122 2.290 2.122 2.290 mp_alltoall_d11v 2046 13.8 1.957 2.251 1.957 2.251 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.053 2.070 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.984 1.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.933 mp_allgather_i34 2055 14.4 0.805 1.913 0.805 1.913 jit_kernel_multiply 9 16.4 1.002 1.815 1.002 1.815 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.775 1.792 cp_fm_cholesky_decompose 22 10.9 1.646 1.653 1.646 1.653 dbcsr_complete_redistribute 325 12.2 0.274 0.337 1.363 1.626 mp_alltoall_z22v 1111 16.6 1.517 1.616 1.517 1.616 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=76.541000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.545455, yerr=6.485840 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 658.382848E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 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 0.072 0.167 66.971 66.976 qs_mol_dyn_low 1 2.0 0.003 0.004 66.415 66.428 qs_forces 11 3.9 0.002 0.002 66.184 66.188 qs_energies 11 4.9 0.002 0.003 62.914 62.920 scf_env_do_scf 11 5.9 0.001 0.001 52.002 52.003 scf_env_do_scf_inner_loop 99 6.5 0.007 0.045 41.077 41.081 velocity_verlet 10 3.0 0.001 0.002 33.837 33.844 dbcsr_multiply_generic 2055 12.4 0.109 0.112 31.877 32.199 multiply_cannon 2055 13.4 0.211 0.223 25.083 26.408 qs_scf_new_mos 99 7.5 0.001 0.001 26.276 26.393 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.275 26.392 ot_scf_mini 99 9.5 0.002 0.003 25.037 25.163 multiply_cannon_loop 2055 14.4 0.617 0.634 23.786 24.978 ot_mini 99 10.5 0.001 0.001 14.166 14.295 rebuild_ks_matrix 110 8.3 0.000 0.000 13.312 13.441 qs_ks_build_kohn_sham_matrix 110 9.3 0.050 0.244 13.311 13.441 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.254 12.436 multiply_cannon_multrec 16440 15.4 3.885 4.985 10.610 11.636 init_scf_loop 11 6.9 0.071 0.564 10.880 10.905 mp_waitall_1 146766 16.3 7.523 10.362 7.523 10.362 qs_ot_get_derivative 99 11.5 0.001 0.003 9.718 9.842 init_scf_run 11 5.9 0.000 0.001 8.420 8.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.419 8.420 prepare_preconditioner 11 7.9 0.000 0.000 7.505 7.523 make_preconditioner 11 8.9 0.002 0.018 7.505 7.523 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.822 7.207 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.347 7.102 dbcsr_mm_accdrv_process 34862 16.1 5.443 6.185 6.582 6.790 sum_up_and_integrate 110 10.3 0.060 0.060 6.627 6.650 integrate_v_rspace 110 11.3 0.002 0.003 6.567 6.590 qs_ot_get_p 110 10.4 0.001 0.001 6.081 6.212 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.760 5.777 calculate_rho_elec 110 8.6 0.058 0.058 5.760 5.776 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.006 5.458 apply_single 110 13.6 0.000 0.000 5.006 5.457 calculate_first_density_matrix 1 7.0 0.000 0.002 5.238 5.239 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.057 5.059 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.686 4.732 make_m2s 4110 13.4 0.050 0.052 4.370 4.722 make_images 4110 14.4 0.395 0.514 4.254 4.606 ot_diis_step 99 11.5 0.010 0.011 4.411 4.412 multiply_cannon_sync_h2d 16440 15.4 3.710 4.258 3.710 4.258 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.200 4.208 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.321 3.974 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.837 3.838 grid_integrate_task_list 110 12.3 3.208 3.362 3.208 3.362 pw_transfer 1331 11.6 0.064 0.071 3.302 3.326 density_rs2pw 110 9.6 0.004 0.004 3.009 3.268 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.248 3.251 rs_pw_transfer 902 11.9 0.010 0.011 2.946 3.234 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.197 3.222 cp_fm_redistribute_end 48 14.0 2.015 3.212 2.018 3.213 cp_fm_diag_elpa_base 48 14.0 1.129 3.062 1.189 3.175 cp_fm_cholesky_invert 11 10.9 3.134 3.141 3.134 3.141 acc_transpose_blocks 16440 15.4 0.073 0.075 3.115 3.140 calculate_dm_sparse 110 9.5 0.001 0.001 3.107 3.134 wfi_extrapolate 11 7.9 0.001 0.001 3.059 3.059 make_images_data 4110 15.4 0.043 0.047 2.495 2.963 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.226 2.920 acc_transpose_blocks_kernels 16440 16.4 0.209 0.214 2.687 2.718 fft_wrap_pw1pw2_140 451 13.1 0.208 0.215 2.668 2.694 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.548 2.607 mp_sum_l 6514 12.8 1.910 2.595 1.910 2.595 fft3d_ps 1111 14.6 1.064 1.075 2.514 2.539 jit_kernel_transpose 5 15.6 2.478 2.509 2.478 2.509 potential_pw2rs 110 12.3 0.011 0.011 2.442 2.464 grid_collocate_task_list 110 9.6 2.184 2.373 2.184 2.373 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.855 2.365 mp_sum_d 3877 11.9 1.815 2.336 1.815 2.336 mp_irecv_dv 48980 15.7 0.787 2.242 0.787 2.242 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.059 2.074 mp_alltoall_d11v 2046 13.8 1.888 2.043 1.888 2.043 dbcsr_complete_redistribute 325 12.2 0.315 0.331 1.545 1.999 qs_energies_init_hamiltonians 11 5.9 0.024 0.041 1.990 1.994 cp_fm_upper_to_full 70 13.6 1.407 1.862 1.407 1.862 jit_kernel_multiply 8 16.4 0.748 1.856 0.748 1.856 cp_fm_cholesky_decompose 22 10.9 1.828 1.851 1.828 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.717 1.731 mp_allgather_i34 2055 14.4 0.624 1.706 0.624 1.706 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.048 1.506 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.464 mp_waitany 17072 13.8 1.191 1.425 1.191 1.425 qs_vxc_create 110 10.3 0.058 0.264 1.359 1.400 xc_vxc_pw_create 110 11.3 0.114 0.187 1.301 1.398 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.336 1.344 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=66.976000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.272727, yerr=8.550289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 738.336768E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 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 0.231 0.868 72.263 72.264 qs_mol_dyn_low 1 2.0 0.003 0.004 70.545 70.613 qs_forces 11 3.9 0.002 0.002 70.472 70.473 qs_energies 11 4.9 0.007 0.038 67.058 67.062 scf_env_do_scf 11 5.9 0.003 0.014 54.552 54.567 scf_env_do_scf_inner_loop 99 6.5 0.006 0.025 41.654 41.657 velocity_verlet 10 3.0 0.001 0.001 37.300 37.302 dbcsr_multiply_generic 2055 12.4 0.116 0.122 32.300 32.562 qs_scf_new_mos 99 7.5 0.001 0.001 27.037 27.170 qs_scf_loop_do_ot 99 8.5 0.001 0.004 27.037 27.169 multiply_cannon 2055 13.4 0.282 0.579 25.055 26.049 ot_scf_mini 99 9.5 0.003 0.004 25.400 25.506 multiply_cannon_loop 2055 14.4 0.888 0.908 23.540 24.114 ot_mini 99 10.5 0.001 0.001 14.373 14.498 multiply_cannon_multrec 24660 15.4 4.213 6.915 13.015 14.244 init_scf_loop 11 6.9 0.017 0.094 12.846 12.865 rebuild_ks_matrix 110 8.3 0.000 0.000 12.519 12.725 qs_ks_build_kohn_sham_matrix 110 9.3 0.032 0.164 12.518 12.724 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.170 11.300 prepare_preconditioner 11 7.9 0.000 0.000 10.421 10.450 make_preconditioner 11 8.9 0.019 0.045 10.421 10.450 qs_ot_get_derivative 99 11.5 0.001 0.001 10.200 10.314 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.624 10.071 dbcsr_mm_accdrv_process 52304 16.0 6.835 8.234 8.656 9.665 init_scf_run 11 5.9 0.001 0.004 9.188 9.189 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.187 9.188 sum_up_and_integrate 110 10.3 0.068 0.071 6.473 6.529 integrate_v_rspace 110 11.3 0.002 0.003 6.405 6.460 mp_waitall_1 126806 16.4 4.667 6.422 4.667 6.422 qs_ot_get_p 110 10.4 0.003 0.018 6.125 6.253 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.941 5.958 calculate_rho_elec 110 8.6 0.077 0.081 5.940 5.958 make_m2s 4110 13.4 0.060 0.063 5.563 5.914 calculate_first_density_matrix 1 7.0 0.011 0.083 5.842 5.908 make_images 4110 14.4 0.578 0.698 5.421 5.768 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.432 5.434 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.738 5.028 cp_fm_upper_to_full 70 13.8 3.256 4.634 3.256 4.634 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.033 4.139 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.124 4.139 apply_single 110 13.6 0.000 0.000 4.033 4.139 ot_diis_step 99 11.5 0.011 0.011 4.138 4.139 dbcsr_complete_redistribute 325 12.2 0.432 0.503 2.763 3.888 acc_transpose_blocks 24660 15.4 0.105 0.109 3.706 3.757 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.633 3.634 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.482 3.533 pw_transfer 1331 11.6 0.064 0.072 3.384 3.454 grid_integrate_task_list 110 12.3 3.277 3.433 3.277 3.433 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.278 3.352 multiply_cannon_sync_h2d 24660 15.4 3.168 3.331 3.168 3.331 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.202 3.320 rs_pw_transfer 902 11.9 0.010 0.011 2.994 3.254 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.469 3.235 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.400 3.205 make_images_data 4110 15.4 0.046 0.050 2.816 3.165 density_rs2pw 110 9.6 0.004 0.004 2.963 3.160 mp_alltoall_i22 605 13.7 1.947 3.147 1.947 3.147 calculate_dm_sparse 110 9.5 0.001 0.001 3.120 3.140 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.069 3.128 acc_transpose_blocks_kernels 24660 16.4 0.303 0.308 3.060 3.097 wfi_extrapolate 11 7.9 0.001 0.001 3.040 3.040 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.017 3.020 cp_fm_cholesky_invert 11 10.9 2.988 2.998 2.988 2.998 cp_fm_redistribute_end 48 14.0 1.500 2.984 1.501 2.985 cp_fm_diag_elpa_base 48 14.0 1.400 2.851 1.480 2.962 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.768 2.867 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.746 2.819 jit_kernel_transpose 5 15.6 2.756 2.791 2.756 2.791 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 2.748 2.749 fft3d_ps 1111 14.6 1.065 1.098 2.583 2.644 jit_kernel_multiply 12 15.7 1.487 2.496 1.487 2.496 grid_collocate_task_list 110 9.6 2.290 2.420 2.290 2.420 potential_pw2rs 110 12.3 0.012 0.013 2.244 2.306 mp_alltoall_d11v 2046 13.8 1.979 2.166 1.979 2.166 mp_waitany 13376 13.8 1.511 1.842 1.511 1.842 cp_fm_cholesky_decompose 22 10.9 1.775 1.819 1.775 1.819 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.659 1.766 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.716 1.748 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.723 1.734 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.598 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.660 1.673 mp_sum_l 6514 12.8 1.018 1.615 1.018 1.615 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.847 1.573 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.396 1.501 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.175 1.224 1.499 mp_irecv_dv 62702 16.1 0.748 1.496 0.748 1.496 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.264000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.090909, yerr=14.304828 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 823.713792E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 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 0.028 0.155 59.256 59.257 qs_mol_dyn_low 1 2.0 0.003 0.004 58.377 58.510 qs_forces 11 3.9 0.002 0.002 58.296 58.301 qs_energies 11 4.9 0.001 0.001 54.684 54.693 scf_env_do_scf 11 5.9 0.000 0.001 44.713 44.718 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.669 35.669 velocity_verlet 10 3.0 0.001 0.001 31.226 31.229 dbcsr_multiply_generic 2055 12.4 0.105 0.107 24.810 25.003 qs_scf_new_mos 99 7.5 0.001 0.001 21.525 21.572 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.525 21.571 multiply_cannon 2055 13.4 0.246 0.264 19.139 20.423 ot_scf_mini 99 9.5 0.002 0.002 20.245 20.272 multiply_cannon_loop 2055 14.4 0.322 0.335 17.766 18.090 rebuild_ks_matrix 110 8.3 0.000 0.000 12.358 12.379 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.357 12.378 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.111 11.131 ot_mini 99 10.5 0.001 0.001 10.640 10.657 init_scf_loop 11 6.9 0.000 0.000 8.989 8.990 multiply_cannon_multrec 8220 15.4 3.203 4.424 7.542 8.449 mp_waitall_1 106626 16.5 6.356 8.192 6.356 8.192 init_scf_run 11 5.9 0.000 0.001 6.844 6.844 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.843 6.843 qs_ot_get_derivative 99 11.5 0.001 0.001 6.797 6.821 prepare_preconditioner 11 7.9 0.000 0.000 6.479 6.485 make_preconditioner 11 8.9 0.000 0.000 6.479 6.485 sum_up_and_integrate 110 10.3 0.079 0.080 6.360 6.462 integrate_v_rspace 110 11.3 0.003 0.004 6.281 6.384 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.062 6.136 qs_ot_get_p 110 10.4 0.001 0.001 5.618 5.643 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.624 5.640 calculate_rho_elec 110 8.6 0.115 0.115 5.623 5.640 dbcsr_mm_accdrv_process 17442 15.9 2.826 3.882 4.208 5.121 make_m2s 4110 13.4 0.039 0.041 4.324 4.553 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.070 4.432 make_images 4110 14.4 0.637 0.695 4.193 4.420 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.200 4.201 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.186 4.190 calculate_first_density_matrix 1 7.0 0.000 0.000 4.039 4.040 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.872 3.872 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.828 3.839 ot_diis_step 99 11.5 0.012 0.012 3.813 3.814 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.718 3.749 apply_single 110 13.6 0.000 0.000 3.718 3.749 grid_integrate_task_list 110 12.3 3.354 3.532 3.354 3.532 pw_transfer 1331 11.6 0.064 0.068 3.307 3.397 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.201 3.294 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.285 3.286 cp_fm_cholesky_invert 11 10.9 3.250 3.254 3.250 3.254 cp_fm_redistribute_end 48 14.0 0.832 3.244 0.837 3.245 cp_fm_diag_elpa_base 48 14.0 2.218 3.036 2.398 3.203 multiply_cannon_sync_h2d 8220 15.4 2.922 3.045 2.922 3.045 density_rs2pw 110 9.6 0.004 0.004 2.700 2.832 make_images_data 4110 15.4 0.038 0.043 2.435 2.830 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.691 2.784 hybrid_alltoall_any 4261 16.3 0.199 0.863 2.347 2.756 wfi_extrapolate 11 7.9 0.001 0.001 2.719 2.719 calculate_dm_sparse 110 9.5 0.001 0.001 2.614 2.665 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.652 2.657 acc_transpose_blocks 8220 15.4 0.036 0.037 2.557 2.585 fft3d_ps 1111 14.6 1.114 1.143 2.460 2.552 rs_pw_transfer 902 11.9 0.010 0.010 2.310 2.468 grid_collocate_task_list 110 9.6 2.328 2.452 2.328 2.452 acc_transpose_blocks_kernels 8220 16.4 0.109 0.112 2.322 2.344 jit_kernel_transpose 5 15.6 2.213 2.236 2.213 2.236 potential_pw2rs 110 12.3 0.015 0.015 2.099 2.190 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 1.991 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.975 1.988 mp_alltoall_d11v 2046 13.8 1.660 1.977 1.660 1.977 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.899 1.908 jit_kernel_multiply 10 15.8 1.069 1.865 1.069 1.865 cp_fm_cholesky_decompose 22 10.9 1.799 1.813 1.799 1.813 mp_allgather_i34 2055 14.4 0.564 1.664 0.564 1.664 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.498 1.620 dbcsr_complete_redistribute 325 12.2 0.556 0.579 1.506 1.593 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.580 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.551 qs_vxc_create 110 10.3 0.001 0.002 1.403 1.500 xc_vxc_pw_create 110 11.3 0.236 0.319 1.401 1.499 qs_create_task_list 11 7.9 0.001 0.001 1.230 1.329 generate_qs_task_list 11 8.9 0.380 0.449 1.229 1.329 mp_waitany 9240 13.8 1.089 1.236 1.089 1.236 mp_alltoall_z22v 1111 16.6 1.114 1.204 1.114 1.204 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.189 1.198 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.257000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=773.363636, yerr=11.780359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 1.335890E+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 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 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 0.028 0.053 89.345 89.345 qs_mol_dyn_low 1 2.0 0.003 0.004 88.752 88.778 qs_forces 11 3.9 0.002 0.002 88.149 88.153 qs_energies 11 4.9 0.002 0.003 84.129 84.134 scf_env_do_scf 11 5.9 0.000 0.001 72.002 72.003 velocity_verlet 10 3.0 0.001 0.001 55.303 55.376 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.546 43.547 dbcsr_multiply_generic 2055 12.4 0.119 0.121 31.455 31.640 init_scf_loop 11 6.9 0.000 0.000 28.384 28.386 qs_scf_new_mos 99 7.5 0.001 0.001 27.166 27.253 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.165 27.253 prepare_preconditioner 11 7.9 0.000 0.000 26.364 26.377 make_preconditioner 11 8.9 0.000 0.000 26.364 26.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.680 25.824 ot_scf_mini 99 9.5 0.002 0.002 25.387 25.457 multiply_cannon 2055 13.4 0.349 0.379 23.719 24.276 multiply_cannon_loop 2055 14.4 0.342 0.344 21.872 22.492 cp_fm_upper_to_full 70 14.2 12.508 17.901 12.508 17.901 ot_mini 99 10.5 0.001 0.001 14.344 14.424 rebuild_ks_matrix 110 8.3 0.001 0.001 13.164 13.269 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.163 13.269 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.988 12.083 dbcsr_complete_redistribute 325 12.2 1.026 1.050 7.301 10.403 multiply_cannon_multrec 8220 15.4 4.374 4.542 9.947 10.275 qs_ot_get_derivative 99 11.5 0.001 0.001 9.683 9.754 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.247 9.349 mp_waitall_1 87304 16.6 7.904 9.041 7.904 9.041 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.668 8.747 mp_alltoall_i22 605 13.7 5.316 8.388 5.316 8.388 init_scf_run 11 5.9 0.000 0.001 7.880 7.880 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.880 7.880 sum_up_and_integrate 110 10.3 0.150 0.151 6.634 6.653 integrate_v_rspace 110 11.3 0.003 0.003 6.484 6.503 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.477 6.502 calculate_rho_elec 110 8.6 0.227 0.227 6.476 6.501 qs_ot_get_p 110 10.4 0.001 0.001 5.872 5.966 make_m2s 4110 13.4 0.044 0.045 5.377 5.945 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.170 5.822 apply_single 110 13.6 0.000 0.000 5.170 5.822 make_images 4110 14.4 0.881 0.929 5.188 5.756 dbcsr_mm_accdrv_process 11614 15.7 3.305 3.765 5.431 5.650 cp_fm_cholesky_invert 11 10.9 5.578 5.582 5.578 5.582 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.745 5.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.896 4.899 ot_diis_step 99 11.5 0.015 0.016 4.633 4.633 calculate_first_density_matrix 1 7.0 0.000 0.000 4.526 4.527 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.359 4.389 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.221 4.228 multiply_cannon_sync_h2d 8220 15.4 3.946 3.956 3.946 3.956 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.018 3.955 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.775 3.776 pw_transfer 1331 11.6 0.073 0.073 3.733 3.744 hybrid_alltoall_any 4261 16.3 0.255 0.551 2.965 3.720 grid_integrate_task_list 110 12.3 3.662 3.717 3.662 3.717 make_images_data 4110 15.4 0.041 0.044 2.963 3.668 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 3.644 3.649 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.618 3.629 calculate_dm_sparse 110 9.5 0.001 0.001 3.330 3.346 wfi_extrapolate 11 7.9 0.001 0.001 3.247 3.247 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.149 3.150 cp_fm_diag_elpa_base 48 14.0 2.618 2.813 3.148 3.148 fft_wrap_pw1pw2_140 451 13.1 0.216 0.218 3.129 3.143 density_rs2pw 110 9.6 0.004 0.004 2.980 3.002 fft3d_ps 1111 14.6 1.267 1.280 2.827 2.836 acc_transpose_blocks 8220 15.4 0.035 0.035 2.702 2.733 grid_collocate_task_list 110 9.6 2.629 2.652 2.629 2.652 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.448 2.491 acc_transpose_blocks_kernels 8220 16.4 0.109 0.110 2.464 2.490 jit_kernel_transpose 5 15.6 2.354 2.381 2.354 2.381 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.176 2.241 rs_pw_transfer 902 11.9 0.010 0.010 2.172 2.229 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.168 jit_kernel_multiply 10 15.4 1.921 2.103 1.921 2.103 mp_alltoall_d11v 2046 13.8 2.048 2.083 2.048 2.083 potential_pw2rs 110 12.3 0.021 0.021 2.075 2.081 cp_fm_cholesky_decompose 22 10.9 2.041 2.057 2.041 2.057 qs_create_task_list 11 7.9 0.000 0.001 1.900 1.943 generate_qs_task_list 11 8.9 0.740 0.794 1.899 1.942 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.850 1.865 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.821 1.859 mp_sum_l 6514 12.8 1.475 1.796 1.475 1.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.781 1.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.345000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1203.454545, yerr=62.001999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/15/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 1.094965E+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 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 626.892800E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3677846. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 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 0.029 0.067 209.874 209.876 qs_mol_dyn_low 1 2.0 0.005 0.010 209.110 209.120 qs_forces 11 3.9 0.003 0.003 208.989 209.018 qs_energies 11 4.9 0.001 0.002 203.475 203.510 scf_env_do_scf 11 5.9 0.001 0.001 184.649 184.653 scf_env_do_scf_inner_loop 117 6.6 0.004 0.013 162.467 162.469 dbcsr_multiply_generic 2507 12.6 0.177 0.182 127.728 128.947 qs_scf_new_mos 117 7.6 0.001 0.001 123.624 123.897 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.623 123.896 velocity_verlet 10 3.0 0.001 0.001 123.816 123.818 ot_scf_mini 117 9.6 0.003 0.003 117.036 117.309 multiply_cannon 2507 13.6 0.236 0.244 103.992 106.323 multiply_cannon_loop 2507 14.6 2.115 2.200 101.753 104.246 ot_mini 117 10.6 0.001 0.001 66.753 67.029 multiply_cannon_multrec 60168 15.6 33.386 35.613 42.121 44.268 qs_ot_get_derivative 117 11.6 0.001 0.001 41.552 41.818 rebuild_ks_matrix 128 8.3 0.001 0.001 34.343 34.770 qs_ks_build_kohn_sham_matrix 128 9.3 0.034 0.168 34.343 34.770 mp_waitall_1 291448 16.2 29.110 32.360 29.110 32.360 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.998 31.371 multiply_cannon_sync_h2d 60168 15.6 27.475 29.571 27.475 29.571 qs_ot_get_p 128 10.4 0.001 0.001 28.292 28.559 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.630 25.323 apply_single 128 13.6 0.001 0.001 24.629 25.323 ot_diis_step 117 11.6 0.007 0.008 24.794 24.795 init_scf_loop 11 6.9 0.000 0.000 22.106 22.107 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.444 21.558 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.318 19.497 multiply_cannon_metrocomm3 60168 15.6 0.116 0.120 16.207 19.131 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.666 18.667 prepare_preconditioner 11 7.9 0.000 0.000 16.498 16.546 make_preconditioner 11 8.9 0.000 0.000 16.498 16.546 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.710 15.895 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.758 15.764 cp_fm_redistribute_end 83 14.4 12.418 15.682 12.432 15.686 cp_fm_diag_elpa_base 83 14.4 3.206 15.442 3.241 15.571 init_scf_run 11 5.9 0.000 0.001 14.585 14.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.584 14.585 make_m2s 5014 13.6 0.104 0.112 13.875 14.331 make_images 5014 14.6 0.399 0.416 13.693 14.159 sum_up_and_integrate 128 10.3 0.089 0.108 14.028 14.058 integrate_v_rspace 128 11.3 0.003 0.004 13.939 13.970 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.711 12.842 calculate_rho_elec 128 8.7 0.045 0.063 12.710 12.841 mp_sum_l 7870 13.0 8.614 9.828 8.614 9.828 cp_fm_cholesky_invert 11 10.9 9.495 9.504 9.495 9.504 wfi_extrapolate 11 7.9 0.001 0.001 9.187 9.187 dbcsr_mm_accdrv_process 124484 16.2 3.294 3.504 8.296 9.058 calculate_dm_sparse 128 9.5 0.001 0.001 8.592 8.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 8.681 8.689 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.903 8.139 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.874 7.973 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.830 7.963 multiply_cannon_metrocomm1 60168 15.6 0.085 0.090 5.989 7.842 make_images_data 5014 15.6 0.067 0.072 6.772 7.707 grid_integrate_task_list 128 12.3 7.038 7.474 7.038 7.474 pw_transfer 1547 11.6 0.074 0.107 6.911 7.184 hybrid_alltoall_any 5200 16.5 0.293 2.262 5.922 7.143 density_rs2pw 128 9.7 0.006 0.007 6.475 6.990 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.707 6.947 rs_pw_transfer 1046 11.9 0.016 0.018 5.669 6.161 fft_wrap_pw1pw2_140 523 13.2 0.446 0.512 5.715 5.888 fft3d_ps 1291 14.7 2.102 2.569 5.496 5.684 mp_alltoall_d11v 2415 14.1 4.358 5.565 4.358 5.565 mp_sum_d 4455 12.2 4.404 5.300 4.404 5.300 calculate_first_density_matrix 1 7.0 0.000 0.001 5.241 5.247 acc_transpose_blocks 60168 15.6 0.305 0.323 4.952 5.141 grid_collocate_task_list 128 9.7 4.725 5.138 4.725 5.138 cp_fm_cholesky_decompose 22 10.9 4.779 4.794 4.779 4.794 potential_pw2rs 128 12.3 0.009 0.011 4.503 4.534 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.876000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.454545, yerr=6.213821 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 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 828.592128E+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 1473826487232 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4123 56687. MP_Allreduce 11347 943. MP_Sync 170 MP_Alltoall 1983 6185519. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 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 0.074 0.147 199.338 199.340 qs_mol_dyn_low 1 2.0 0.003 0.004 198.353 198.377 qs_forces 11 3.9 0.003 0.003 198.259 198.259 qs_energies 11 4.9 0.010 0.021 191.491 191.505 scf_env_do_scf 11 5.9 0.009 0.028 171.353 171.363 scf_env_do_scf_inner_loop 118 6.6 0.017 0.053 137.263 137.279 velocity_verlet 10 3.0 0.003 0.010 122.717 122.719 dbcsr_multiply_generic 2527 12.6 0.188 0.197 100.340 101.591 qs_scf_new_mos 118 7.6 0.001 0.001 97.850 98.351 qs_scf_loop_do_ot 118 8.6 0.007 0.019 97.849 98.350 ot_scf_mini 118 9.6 0.007 0.014 92.998 93.616 multiply_cannon 2527 13.6 0.479 0.536 79.809 84.178 multiply_cannon_loop 2527 14.6 1.262 1.298 76.362 79.375 ot_mini 118 10.6 0.001 0.001 50.572 51.107 mp_waitall_1 228564 16.4 25.507 40.170 25.507 40.170 multiply_cannon_multrec 30324 15.6 22.041 27.092 31.910 37.173 init_scf_loop 11 6.9 0.008 0.025 33.989 34.024 rebuild_ks_matrix 129 8.3 0.001 0.001 33.387 33.912 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.036 33.387 33.911 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.091 30.575 prepare_preconditioner 11 7.9 0.000 0.000 29.420 29.480 make_preconditioner 11 8.9 0.000 0.002 29.420 29.480 multiply_cannon_metrocomm3 30324 15.6 0.092 0.097 16.044 29.234 qs_ot_get_derivative 118 11.6 0.001 0.002 28.389 29.009 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.049 28.585 qs_ot_get_p 129 10.4 0.001 0.002 24.004 24.686 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.215 23.275 apply_single 129 13.6 0.001 0.001 22.215 23.275 ot_diis_step 118 11.6 0.014 0.015 21.999 22.001 multiply_cannon_sync_h2d 30324 15.6 19.345 21.424 19.345 21.424 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.967 19.005 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.761 17.762 cp_fm_cholesky_invert 11 10.9 17.208 17.223 17.208 17.223 make_m2s 5054 13.6 0.090 0.095 14.499 16.098 make_images 5054 14.6 1.173 1.369 14.287 15.884 init_scf_run 11 5.9 0.000 0.001 14.924 14.925 scf_env_initial_rho_setup 11 6.9 0.001 0.002 14.923 14.925 sum_up_and_integrate 129 10.3 0.117 0.133 14.688 14.718 integrate_v_rspace 129 11.3 0.003 0.003 14.571 14.603 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.502 14.512 cp_fm_redistribute_end 83 14.4 8.432 14.422 8.445 14.424 cp_fm_diag_elpa_base 83 14.4 5.738 13.941 5.958 14.308 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.280 13.322 calculate_rho_elec 129 8.7 0.088 0.106 13.279 13.321 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.095 11.518 multiply_cannon_metrocomm4 27797 15.6 0.098 0.114 3.811 10.675 make_images_data 5054 15.6 0.065 0.072 8.700 10.587 mp_irecv_dv 70031 16.3 3.613 10.285 3.613 10.285 dbcsr_mm_accdrv_process 62734 16.2 4.628 5.361 9.336 9.944 hybrid_alltoall_any 5240 16.5 0.347 1.528 7.329 9.936 wfi_extrapolate 11 7.9 0.001 0.001 8.458 8.458 pw_transfer 1559 11.6 0.085 0.103 7.986 8.061 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.829 7.838 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.760 7.825 grid_integrate_task_list 129 12.3 7.233 7.611 7.233 7.611 density_rs2pw 129 9.7 0.006 0.007 6.997 7.572 cp_fm_cholesky_decompose 22 10.9 7.300 7.386 7.300 7.386 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.442 7.313 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.962 7.088 fft_wrap_pw1pw2_140 527 13.2 0.471 0.514 6.687 6.756 calculate_dm_sparse 129 9.5 0.001 0.001 6.528 6.684 rs_pw_transfer 1054 12.0 0.014 0.016 5.916 6.609 mp_sum_l 7930 13.1 4.306 6.571 4.306 6.571 calculate_first_density_matrix 1 7.0 0.002 0.005 6.261 6.268 fft3d_ps 1301 14.7 2.800 2.949 6.127 6.173 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.418 5.493 grid_collocate_task_list 129 9.7 4.961 5.351 4.961 5.351 potential_pw2rs 129 12.3 0.015 0.017 5.063 5.085 mp_allgather_i34 2527 14.6 2.014 5.061 2.014 5.061 mp_alltoall_d11v 2423 14.1 4.258 4.763 4.258 4.763 mp_sum_d 4497 12.2 2.746 4.170 2.746 4.170 dbcsr_complete_redistribute 395 12.7 0.771 0.845 3.277 4.129 make_basis_sm 11 9.8 0.000 0.000 4.064 4.074 parallel_gemm_fm 81 9.0 0.000 0.000 4.014 4.041 parallel_gemm_fm_cosma 81 10.0 4.014 4.041 4.014 4.041 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=199.340000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.363636, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 942.288896E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 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 0.039 0.073 183.313 183.315 qs_mol_dyn_low 1 2.0 0.003 0.004 182.357 182.370 qs_forces 11 3.9 0.003 0.003 182.259 182.261 qs_energies 11 4.9 0.007 0.031 175.733 175.746 scf_env_do_scf 11 5.9 0.001 0.001 157.268 157.272 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.739 120.780 velocity_verlet 10 3.0 0.001 0.002 114.303 114.306 dbcsr_multiply_generic 2507 12.6 0.181 0.186 84.480 85.609 qs_scf_new_mos 117 7.6 0.001 0.001 83.990 84.334 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.989 84.333 ot_scf_mini 117 9.6 0.004 0.005 79.844 80.229 multiply_cannon 2507 13.6 0.500 0.517 64.073 68.437 multiply_cannon_loop 2507 14.6 0.858 0.886 60.690 63.189 ot_mini 117 10.6 0.001 0.001 43.477 43.888 init_scf_loop 11 6.9 0.002 0.007 36.429 36.476 mp_waitall_1 178456 16.5 25.728 35.626 25.728 35.626 prepare_preconditioner 11 7.9 0.000 0.000 31.913 31.974 make_preconditioner 11 8.9 0.004 0.025 31.913 31.974 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.517 30.948 rebuild_ks_matrix 128 8.3 0.001 0.001 30.256 30.746 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.054 30.256 30.745 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.310 27.757 multiply_cannon_multrec 20056 15.6 13.296 16.611 22.165 25.609 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.424 25.327 qs_ot_get_derivative 117 11.6 0.001 0.002 23.668 24.049 qs_ot_get_p 128 10.4 0.005 0.017 21.923 22.330 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.333 21.545 apply_single 128 13.6 0.001 0.001 20.333 21.545 ot_diis_step 117 11.6 0.018 0.018 19.705 19.705 qs_ot_p2m_diag 83 11.4 0.266 0.272 17.230 17.243 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.191 16.193 make_m2s 5014 13.6 0.081 0.086 14.819 15.974 multiply_cannon_sync_h2d 20056 15.6 14.344 15.830 14.344 15.830 make_images 5014 14.6 1.181 1.270 14.585 15.736 cp_fm_cholesky_invert 11 10.9 14.996 15.007 14.996 15.007 sum_up_and_integrate 128 10.3 0.134 0.145 14.213 14.246 integrate_v_rspace 128 11.3 0.003 0.004 14.079 14.114 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.019 13.058 calculate_rho_elec 128 8.7 0.132 0.147 13.019 13.057 init_scf_run 11 5.9 0.000 0.001 13.000 13.002 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.000 13.002 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.972 12.974 cp_fm_redistribute_end 83 14.4 4.869 12.888 4.888 12.891 cp_fm_diag_elpa_base 83 14.4 7.555 12.297 7.977 12.779 make_images_data 5014 15.6 0.060 0.066 9.092 10.563 hybrid_alltoall_any 5200 16.5 0.432 1.953 7.845 9.601 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.081 9.345 multiply_cannon_metrocomm4 17549 15.6 0.059 0.070 3.476 9.207 mp_irecv_dv 50230 16.2 3.356 8.954 3.356 8.954 dbcsr_mm_accdrv_process 41502 16.2 4.507 5.531 8.327 8.870 grid_integrate_task_list 128 12.3 7.341 7.763 7.341 7.763 pw_transfer 1547 11.6 0.084 0.103 7.629 7.740 cp_fm_cholesky_decompose 22 10.9 7.652 7.685 7.652 7.685 cp_fm_upper_to_full 105 14.5 5.847 7.563 5.847 7.563 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.407 7.522 wfi_extrapolate 11 7.9 0.001 0.001 7.513 7.513 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.140 7.145 density_rs2pw 128 9.7 0.005 0.006 6.479 6.783 fft_wrap_pw1pw2_140 523 13.2 0.477 0.526 6.456 6.582 dbcsr_complete_redistribute 395 12.7 1.167 1.195 4.615 6.432 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.228 6.359 calculate_dm_sparse 128 9.5 0.001 0.001 5.937 6.011 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.796 5.912 fft3d_ps 1291 14.7 2.696 2.905 5.674 5.744 grid_collocate_task_list 128 9.7 5.072 5.593 5.072 5.593 rs_pw_transfer 1046 11.9 0.013 0.014 5.212 5.507 calculate_first_density_matrix 1 7.0 0.006 0.018 5.298 5.305 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.385 5.204 mp_sum_l 7870 13.0 3.798 5.204 3.798 5.204 mp_allgather_i34 2507 14.6 1.917 4.974 1.917 4.974 mp_alltoall_d11v 2415 14.1 4.215 4.738 4.215 4.738 potential_pw2rs 128 12.3 0.020 0.022 4.602 4.623 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.115 4.143 qs_energies_init_hamiltonians 11 5.9 0.112 0.300 4.125 4.132 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.373 4.108 mp_alltoall_i22 716 14.1 1.981 3.920 1.981 3.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=183.315000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=894.454545, yerr=9.642365 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.149313E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.152 0.252 195.472 195.476 qs_mol_dyn_low 1 2.0 0.003 0.003 193.948 194.195 qs_forces 11 3.9 0.003 0.003 193.801 193.809 qs_energies 11 4.9 0.022 0.051 186.650 186.666 scf_env_do_scf 11 5.9 0.004 0.026 166.121 166.130 velocity_verlet 10 3.0 0.002 0.007 126.087 126.090 scf_env_do_scf_inner_loop 117 6.6 0.006 0.033 118.315 118.319 qs_scf_new_mos 117 7.6 0.001 0.001 82.957 83.259 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.957 83.259 dbcsr_multiply_generic 2507 12.6 0.195 0.218 81.479 82.512 ot_scf_mini 117 9.6 0.003 0.005 78.480 78.749 multiply_cannon 2507 13.6 0.551 0.578 56.446 59.855 multiply_cannon_loop 2507 14.6 1.186 1.222 52.619 55.067 init_scf_loop 11 6.9 0.000 0.001 47.657 47.658 prepare_preconditioner 11 7.9 0.000 0.000 43.627 43.655 make_preconditioner 11 8.9 0.000 0.005 43.627 43.655 ot_mini 117 10.6 0.001 0.001 42.693 42.947 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.057 42.135 multiply_cannon_multrec 30084 15.6 14.255 19.133 26.275 30.617 rebuild_ks_matrix 128 8.3 0.001 0.001 28.906 29.274 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 28.905 29.274 mp_waitall_1 153770 16.5 17.682 27.406 17.682 27.406 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.135 26.462 qs_ot_get_derivative 117 11.6 0.011 0.078 22.905 23.172 qs_ot_get_p 128 10.4 0.001 0.001 21.997 22.277 make_m2s 5014 13.6 0.095 0.100 20.460 21.661 make_images 5014 14.6 1.968 2.165 20.154 21.353 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.134 19.825 apply_single 128 13.6 0.001 0.001 19.134 19.825 ot_diis_step 117 11.6 0.018 0.019 19.650 19.652 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.772 17.825 cp_fm_cholesky_invert 11 10.9 16.882 16.892 16.882 16.892 cp_fm_upper_to_full 105 14.7 11.143 16.570 11.143 16.570 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.467 16.469 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.259 15.201 init_scf_run 11 5.9 0.000 0.001 14.095 14.096 scf_env_initial_rho_setup 11 6.9 0.034 0.059 14.095 14.096 sum_up_and_integrate 128 10.3 0.140 0.154 14.007 14.029 integrate_v_rspace 128 11.3 0.003 0.004 13.867 13.894 dbcsr_complete_redistribute 395 12.7 1.666 1.905 9.866 13.716 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.219 13.222 cp_fm_redistribute_end 83 14.4 2.257 13.123 2.275 13.130 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.006 13.070 calculate_rho_elec 128 8.7 0.175 0.190 13.006 13.070 cp_fm_diag_elpa_base 83 14.4 10.221 12.492 10.807 12.983 multiply_cannon_sync_h2d 30084 15.6 11.756 12.781 11.756 12.781 make_images_data 5014 15.6 0.063 0.067 10.927 12.579 copy_fm_to_dbcsr 209 11.7 0.002 0.002 8.422 12.274 dbcsr_mm_accdrv_process 62264 16.2 7.400 8.406 11.595 12.264 hybrid_alltoall_any 5200 16.5 0.519 2.191 9.705 11.847 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.548 10.256 mp_alltoall_i22 716 14.1 5.970 9.848 5.970 9.848 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.450 9.643 cp_fm_cholesky_decompose 22 10.9 7.926 8.021 7.926 8.021 grid_integrate_task_list 128 12.3 7.477 7.850 7.477 7.850 pw_transfer 1547 11.6 0.084 0.098 7.655 7.713 wfi_extrapolate 11 7.9 0.001 0.001 7.539 7.539 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.430 7.493 calculate_dm_sparse 128 9.5 0.001 0.001 7.108 7.250 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.041 7.070 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.744 6.921 mp_irecv_dv 76098 16.2 2.600 6.652 2.600 6.652 density_rs2pw 128 9.7 0.005 0.006 6.219 6.644 fft_wrap_pw1pw2_140 523 13.2 0.480 0.489 6.507 6.586 calculate_first_density_matrix 1 7.0 0.045 0.361 6.337 6.376 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.669 5.763 fft3d_ps 1291 14.7 2.787 2.853 5.665 5.723 grid_collocate_task_list 128 9.7 5.205 5.656 5.205 5.656 mp_alltoall_d11v 2415 14.1 4.868 5.493 4.868 5.493 rs_pw_transfer 1046 11.9 0.013 0.014 4.830 5.243 qs_energies_init_hamiltonians 11 5.9 0.013 0.058 4.623 4.628 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.399 4.484 potential_pw2rs 128 12.3 0.022 0.023 4.301 4.316 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.244 4.311 mp_allgather_i34 2507 14.6 1.948 4.099 1.948 4.099 dbcsr_dot_sd 1318 12.0 1.349 1.461 3.183 3.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=195.476000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.909091, yerr=19.491045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.488380E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57658. MP_Allreduce 11099 1171. MP_Sync 88 MP_Alltoall 1724 18848021. MP_ISendRecv 7740 122880. MP_Wait 20114 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 0.151 0.166 173.630 173.630 qs_mol_dyn_low 1 2.0 0.008 0.081 171.869 171.882 qs_forces 11 3.9 0.003 0.003 171.742 171.775 qs_energies 11 4.9 0.002 0.002 164.490 164.531 scf_env_do_scf 11 5.9 0.001 0.001 146.243 146.255 velocity_verlet 10 3.0 0.001 0.001 112.031 112.035 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 110.528 110.530 dbcsr_multiply_generic 2535 12.6 0.184 0.189 75.748 76.156 qs_scf_new_mos 118 7.6 0.001 0.001 75.221 75.272 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.221 75.271 ot_scf_mini 118 9.6 0.004 0.004 70.797 70.851 multiply_cannon 2535 13.6 0.586 0.610 56.336 60.477 multiply_cannon_loop 2535 14.6 0.452 0.461 51.762 52.372 ot_mini 118 10.6 0.001 0.001 40.059 40.095 init_scf_loop 11 6.9 0.000 0.000 35.566 35.568 mp_waitall_1 131034 16.6 25.709 32.927 25.709 32.927 prepare_preconditioner 11 7.9 0.000 0.000 31.775 31.805 make_preconditioner 11 8.9 0.000 0.000 31.775 31.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.658 29.928 rebuild_ks_matrix 129 8.3 0.001 0.001 28.330 28.374 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.330 28.374 qs_ks_update_qs_env 129 7.6 0.001 0.001 25.877 25.918 multiply_cannon_multrec 10140 15.6 10.647 14.860 18.557 21.481 qs_ot_get_derivative 118 11.6 0.001 0.002 19.997 20.046 ot_diis_step 118 11.6 0.020 0.020 19.995 19.997 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.594 19.842 apply_single 129 13.6 0.001 0.001 19.593 19.841 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 12.463 19.406 cp_fm_cholesky_invert 11 10.9 18.181 18.186 18.181 18.186 make_m2s 5070 13.6 0.067 0.074 15.865 18.138 qs_ot_get_p 129 10.4 0.001 0.001 17.811 17.884 make_images 5070 14.6 2.342 2.900 15.556 17.832 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.147 14.163 sum_up_and_integrate 129 10.3 0.181 0.192 13.938 13.988 integrate_v_rspace 129 11.3 0.004 0.004 13.757 13.810 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.979 13.004 calculate_rho_elec 129 8.7 0.260 0.271 12.978 13.003 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.002 13.003 make_images_data 5070 15.6 0.053 0.061 9.552 12.033 multiply_cannon_sync_h2d 10140 15.6 11.750 12.020 11.750 12.020 hybrid_alltoall_any 5257 16.5 0.797 3.605 9.373 11.779 init_scf_run 11 5.9 0.000 0.001 11.430 11.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.430 11.431 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.927 9.928 cp_fm_diag_elpa_base 84 14.4 9.685 9.755 9.923 9.923 dbcsr_mm_accdrv_process 20958 16.1 2.993 4.032 7.536 8.311 grid_integrate_task_list 129 12.3 7.847 8.248 7.847 8.248 cp_fm_cholesky_decompose 22 10.9 7.976 8.070 7.976 8.070 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.891 7.921 pw_transfer 1559 11.6 0.083 0.091 7.447 7.470 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.226 7.247 wfi_extrapolate 11 7.9 0.001 0.001 7.196 7.196 multiply_cannon_metrocomm1 10140 15.6 0.028 0.029 4.230 6.977 mp_allgather_i34 2535 14.6 2.781 6.863 2.781 6.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.393 6.407 fft_wrap_pw1pw2_140 527 13.2 0.506 0.531 6.362 6.384 calculate_dm_sparse 129 9.5 0.001 0.001 6.112 6.203 density_rs2pw 129 9.7 0.005 0.006 5.776 6.049 grid_collocate_task_list 129 9.7 5.536 5.794 5.536 5.794 dbcsr_complete_redistribute 397 12.7 2.146 2.235 5.220 5.598 fft3d_ps 1301 14.7 2.745 2.834 5.376 5.419 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.285 5.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.273 5.290 mp_alltoall_d11v 2429 14.1 4.505 5.026 4.505 5.026 rs_pw_transfer 1054 12.0 0.012 0.013 4.127 4.392 calculate_first_density_matrix 1 7.0 0.000 0.000 4.046 4.047 multiply_cannon_metrocomm4 7605 15.6 0.023 0.025 1.866 4.021 potential_pw2rs 129 12.3 0.026 0.026 3.990 4.004 mp_irecv_dv 29150 15.9 1.829 3.954 1.829 3.954 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.855 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.732 3.756 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.443 3.738 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.494 3.560 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.481 3.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.630000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1389.909091, yerr=44.432696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 2.974982E+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 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 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 0.192 0.230 291.780 291.781 qs_mol_dyn_low 1 2.0 0.005 0.022 290.569 290.595 qs_forces 11 3.9 0.020 0.023 290.377 290.382 qs_energies 11 4.9 0.042 0.052 281.796 281.808 scf_env_do_scf 11 5.9 0.001 0.001 257.287 257.301 velocity_verlet 10 3.0 0.001 0.001 206.922 206.943 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.534 132.536 init_scf_loop 11 6.9 0.000 0.000 124.492 124.493 prepare_preconditioner 11 7.9 0.000 0.000 119.858 119.884 make_preconditioner 11 8.9 0.000 0.000 119.858 119.884 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.821 117.024 qs_scf_new_mos 116 7.6 0.001 0.001 91.436 91.598 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.436 91.597 ot_scf_mini 116 9.6 0.004 0.004 86.737 86.827 dbcsr_multiply_generic 2485 12.5 0.225 0.229 84.910 85.567 cp_fm_upper_to_full 104 14.8 52.736 75.715 52.736 75.715 multiply_cannon 2485 13.5 0.694 0.730 60.230 60.876 multiply_cannon_loop 2485 14.5 0.469 0.483 56.636 58.152 ot_mini 116 10.6 0.001 0.001 45.098 45.215 dbcsr_complete_redistribute 393 12.7 4.319 4.565 30.304 43.408 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.906 40.030 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.990 36.970 mp_alltoall_i22 712 14.1 21.822 35.084 21.822 35.084 cp_fm_cholesky_invert 11 10.9 33.771 33.777 33.771 33.777 rebuild_ks_matrix 127 8.3 0.001 0.001 32.864 32.941 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 32.863 32.941 mp_waitall_1 104546 16.7 27.566 31.159 27.566 31.159 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.632 30.696 qs_ot_get_p 127 10.4 0.001 0.001 26.476 26.589 qs_ot_get_derivative 116 11.6 0.002 0.002 25.177 25.269 qs_ot_p2m_diag 82 11.4 0.868 0.874 22.498 22.528 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.775 20.776 make_m2s 4970 13.5 0.076 0.079 19.591 20.502 make_images 4970 14.5 3.727 3.941 19.115 20.028 ot_diis_step 116 11.6 0.022 0.023 19.877 19.877 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 18.325 19.736 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.356 19.704 apply_single 127 13.6 0.001 0.001 19.355 19.703 multiply_cannon_multrec 9940 15.5 10.471 12.336 18.978 19.237 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.564 17.565 cp_fm_diag_elpa_base 82 14.4 13.268 14.834 17.560 17.560 sum_up_and_integrate 127 10.3 0.319 0.320 15.470 15.568 multiply_cannon_sync_h2d 9940 15.5 15.527 15.539 15.527 15.539 integrate_v_rspace 127 11.3 0.004 0.004 15.151 15.248 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.695 14.725 calculate_rho_elec 127 8.7 0.479 0.480 14.694 14.724 init_scf_run 11 5.9 0.000 0.001 14.449 14.449 scf_env_initial_rho_setup 11 6.9 0.043 0.054 14.449 14.449 hybrid_alltoall_any 5155 16.4 1.290 3.019 10.413 12.047 make_images_data 4970 15.5 0.059 0.062 10.306 11.901 dbcsr_mm_accdrv_process 20590 16.0 3.802 6.293 8.273 10.168 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.297 9.369 cp_fm_cholesky_decompose 22 10.9 9.004 9.038 9.004 9.038 grid_integrate_task_list 127 12.3 8.478 8.664 8.478 8.664 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.503 8.627 wfi_extrapolate 11 7.9 0.001 0.001 8.594 8.595 pw_transfer 1535 11.6 0.089 0.089 8.101 8.109 qs_energies_init_hamiltonians 11 5.9 0.017 0.027 7.984 7.984 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.870 7.878 calculate_dm_sparse 127 9.5 0.001 0.001 6.962 7.055 fft_wrap_pw1pw2_140 519 13.2 0.535 0.539 6.962 6.972 mp_alltoall_d11v 2401 14.1 6.777 6.923 6.777 6.923 grid_collocate_task_list 127 9.7 6.290 6.337 6.290 6.337 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.068 6.167 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.099 6.153 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.405 6.108 fft3d_ps 1281 14.7 2.735 2.748 5.928 5.937 density_rs2pw 127 9.7 0.005 0.005 5.886 5.914 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.781000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2628.818182, yerr=173.932494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 1.261392E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.070 0.160 87.013 87.014 qs_energies 1 2.0 0.000 0.000 86.457 86.472 ls_scf 1 3.0 0.005 0.038 85.262 85.279 dbcsr_multiply_generic 111 6.7 0.014 0.015 74.045 74.213 multiply_cannon 111 7.7 0.017 0.020 57.408 58.881 multiply_cannon_loop 111 8.7 0.213 0.227 53.974 55.605 ls_scf_main 1 4.0 0.000 0.008 52.083 52.088 density_matrix_trs4 2 5.0 0.002 0.003 46.628 46.709 ls_scf_init_scf 1 4.0 0.015 0.040 30.132 30.139 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.966 29.037 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.851 26.875 mp_waitall_1 11316 10.9 22.126 24.914 22.126 24.914 multiply_cannon_multrec 2664 9.7 8.187 8.914 15.553 17.264 multiply_cannon_sync_h2d 2664 9.7 13.793 15.316 13.793 15.316 make_m2s 222 7.7 0.008 0.011 12.912 13.486 make_images 222 8.7 0.099 0.110 12.890 13.466 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.362 12.397 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.598 8.725 make_images_data 222 9.7 0.004 0.005 7.487 8.050 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.631 6.983 7.941 hybrid_alltoall_any 227 10.6 0.215 1.827 6.413 7.680 dbcsr_mm_accdrv_process_sort 4760 11.4 6.272 7.182 6.272 7.182 calculate_norms 4752 9.8 5.537 6.198 5.537 6.198 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.983 5.175 mp_sum_l 807 5.4 3.187 4.636 3.187 4.636 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.370 3.564 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.062 3.466 mp_irecv_dv 6231 10.9 2.045 3.434 2.045 3.434 make_images_sizes 222 9.7 0.000 0.000 0.693 3.412 mp_alltoall_i44 222 10.7 0.693 3.412 0.693 3.412 arnoldi_extremal 4 6.8 0.000 0.000 3.187 3.216 arnoldi_normal_ev 4 7.8 0.001 0.003 3.187 3.216 build_subspace 16 8.4 0.009 0.012 3.084 3.086 ls_scf_post 1 4.0 0.002 0.015 3.042 3.057 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.923 2.927 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.375 2.772 dbcsr_merge_single_wm 555 10.7 0.466 0.596 2.367 2.764 make_images_pack 222 9.7 2.212 2.634 2.213 2.635 dbcsr_matrix_vector_mult 304 9.0 0.003 0.009 2.305 2.546 dbcsr_sort_data 658 11.4 2.158 2.481 2.158 2.481 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.454 2.065 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.216 2.362 buffer_matrices_ensure_size 222 8.7 1.756 2.109 1.756 2.109 acc_transpose_blocks 2664 9.7 0.020 0.021 1.773 1.928 acc_transpose_blocks_kernels 2664 10.7 0.036 0.040 1.663 1.814 jit_kernel_transpose 1 13.0 1.627 1.779 1.627 1.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.014000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 2.082521E+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 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.073 0.237 94.556 94.565 qs_energies 1 2.0 0.001 0.012 93.757 93.814 ls_scf 1 3.0 0.034 0.072 92.048 92.081 dbcsr_multiply_generic 111 6.7 0.015 0.019 77.192 77.510 multiply_cannon 111 7.7 0.028 0.041 55.227 58.782 multiply_cannon_loop 111 8.7 0.116 0.122 51.896 54.875 ls_scf_main 1 4.0 0.000 0.002 54.599 54.636 density_matrix_trs4 2 5.0 0.006 0.022 48.937 49.157 ls_scf_init_scf 1 4.0 0.002 0.004 33.231 33.240 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.794 31.906 mp_waitall_1 9246 10.9 21.468 30.336 21.468 30.336 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.013 29.119 29.169 multiply_cannon_multrec 1332 9.7 13.108 16.689 22.470 27.481 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.104 21.214 make_m2s 222 7.7 0.006 0.008 15.192 15.850 make_images 222 8.7 1.580 1.986 15.161 15.818 dbcsr_mm_accdrv_process 4041 10.4 0.350 0.781 8.967 10.434 dbcsr_mm_accdrv_process_sort 4041 11.4 8.468 10.013 8.468 10.013 make_images_data 222 9.7 0.004 0.004 8.541 9.459 mp_sum_l 807 5.4 5.497 8.894 5.497 8.894 hybrid_alltoall_any 227 10.6 0.524 2.497 7.973 8.755 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.227 7.682 mp_irecv_dv 3311 11.0 3.207 7.628 3.207 7.628 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.171 6.797 calculate_norms 2376 9.8 6.001 6.703 6.001 6.703 multiply_cannon_sync_h2d 1332 9.7 4.759 5.771 4.759 5.771 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.309 5.529 arnoldi_extremal 4 6.8 0.000 0.000 4.656 4.711 arnoldi_normal_ev 4 7.8 0.001 0.005 4.655 4.711 build_subspace 16 8.4 0.014 0.021 4.376 4.379 ls_scf_post 1 4.0 0.034 0.096 4.184 4.223 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.449 4.016 ls_scf_store_result 1 5.0 0.000 0.000 3.264 3.404 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.137 3.367 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.225 2.743 3.227 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.624 2.630 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.508 2.595 make_images_pack 222 9.7 2.053 2.450 2.055 2.452 mp_allgather_i34 111 8.7 1.032 2.285 1.032 2.285 dbcsr_sort_data 436 11.2 1.838 2.093 1.838 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.013 2.024 rebuild_ks_matrix 3 7.3 0.000 0.000 2.000 2.011 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 2.000 2.011 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.565000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 2.685338E+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 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.089 0.356 100.599 100.611 qs_energies 1 2.0 0.002 0.021 99.602 99.637 ls_scf 1 3.0 0.004 0.044 97.770 97.775 dbcsr_multiply_generic 111 6.7 0.021 0.050 78.426 78.795 ls_scf_main 1 4.0 0.007 0.052 60.748 60.799 multiply_cannon 111 7.7 0.081 0.391 54.751 59.109 multiply_cannon_loop 111 8.7 0.100 0.107 50.995 55.023 density_matrix_trs4 2 5.0 0.033 0.309 52.635 52.836 mp_waitall_1 7374 11.0 23.714 34.007 23.714 34.007 ls_scf_init_scf 1 4.0 0.000 0.002 33.428 33.748 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.418 31.487 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.035 28.966 28.982 multiply_cannon_multrec 888 9.7 12.558 15.132 21.251 24.337 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.175 23.669 make_m2s 222 7.7 0.006 0.009 16.916 18.251 make_images 222 8.7 1.984 2.276 16.876 18.213 hybrid_alltoall_any 227 10.6 0.621 2.864 9.062 10.726 make_images_data 222 9.7 0.004 0.004 9.353 10.475 dbcsr_mm_accdrv_process 3754 10.4 0.326 0.717 8.201 9.669 dbcsr_mm_accdrv_process_sort 3754 11.4 7.750 8.953 7.750 8.953 mp_sum_l 807 5.4 5.401 8.860 5.401 8.860 multiply_cannon_sync_h2d 888 9.7 6.029 7.189 6.029 7.189 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.479 6.984 mp_irecv_dv 2335 11.1 2.464 6.934 2.464 6.934 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.831 6.545 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.805 6.428 arnoldi_extremal 4 6.8 0.001 0.012 6.049 6.069 arnoldi_normal_ev 4 7.8 0.058 0.464 6.048 6.069 build_subspace 16 8.4 0.027 0.120 5.577 5.669 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.173 5.421 calculate_norms 1584 9.8 4.283 4.619 4.283 4.619 ls_scf_dm_to_ks 2 5.0 0.022 0.193 4.267 4.376 dbcsr_matrix_vector_mult 304 9.0 0.006 0.025 3.537 3.896 mp_allgather_i34 111 8.7 1.508 3.798 1.508 3.798 qs_ks_update_qs_env 3 6.3 0.000 0.001 3.382 3.636 rebuild_ks_matrix 3 7.3 0.001 0.007 3.327 3.604 dbcsr_matrix_vector_mult_local 304 10.0 3.026 3.596 3.028 3.598 qs_ks_build_kohn_sham_matrix 3 8.3 0.033 0.237 3.327 3.597 ls_scf_post 1 4.0 0.001 0.010 3.589 3.593 ls_scf_store_result 1 5.0 0.000 0.000 3.298 3.392 mp_sum_d 1051 6.0 2.174 3.019 2.174 3.019 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.681 2.688 make_images_sizes 222 9.7 0.000 0.000 1.254 2.336 mp_alltoall_i44 222 10.7 1.254 2.336 1.254 2.336 ls_scf_initial_guess 1 5.0 0.001 0.043 2.010 2.334 ls_scf_qs_atomic_guess 1 6.0 0.013 0.123 2.009 2.334 make_images_pack 222 9.7 1.825 2.141 1.828 2.144 dbcsr_sort_data 325 11.1 1.844 2.127 1.844 2.127 dbcsr_data_release 9322 10.9 1.312 2.026 1.312 2.026 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=100.611000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2168.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 3.319448E+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 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 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 0.401 1.038 101.264 101.269 qs_energies 1 2.0 0.000 0.000 98.429 98.453 ls_scf 1 3.0 0.000 0.001 96.767 96.791 dbcsr_multiply_generic 111 6.7 0.016 0.018 80.226 80.466 ls_scf_main 1 4.0 0.002 0.029 58.881 58.887 multiply_cannon 111 7.7 0.066 0.116 53.211 57.882 density_matrix_trs4 2 5.0 0.002 0.003 52.807 52.920 multiply_cannon_loop 111 8.7 0.114 0.125 48.049 51.692 ls_scf_init_scf 1 4.0 0.001 0.002 34.618 34.620 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.480 33.551 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.881 30.893 mp_waitall_1 6438 11.0 22.811 30.792 22.811 30.792 multiply_cannon_multrec 1332 9.7 14.240 17.424 22.170 25.242 make_m2s 222 7.7 0.007 0.008 21.117 22.665 make_images 222 8.7 3.140 3.620 21.066 22.618 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.379 16.648 make_images_data 222 9.7 0.004 0.004 11.796 13.465 hybrid_alltoall_any 227 10.6 0.797 3.776 10.950 13.046 dbcsr_mm_accdrv_process 3641 10.4 0.196 0.381 7.572 9.082 dbcsr_mm_accdrv_process_sort 3641 11.4 7.186 8.674 7.186 8.674 mp_sum_l 807 5.4 4.611 8.386 4.611 8.386 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.080 6.159 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.246 6.152 mp_irecv_dv 3229 10.9 2.057 6.080 2.057 6.080 multiply_cannon_sync_h2d 1332 9.7 5.507 6.062 5.507 6.062 arnoldi_extremal 4 6.8 0.000 0.000 5.371 5.698 arnoldi_normal_ev 4 7.8 0.069 0.363 5.371 5.698 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.465 5.086 build_subspace 16 8.4 0.014 0.021 4.965 4.974 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.775 4.946 mp_allgather_i34 111 8.7 2.282 4.806 2.282 4.806 calculate_norms 2376 9.8 4.191 4.515 4.191 4.515 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.632 3.936 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.692 3.203 3.694 dbcsr_sort_data 658 11.4 3.049 3.493 3.049 3.493 ls_scf_post 1 4.0 0.007 0.026 3.268 3.292 dbcsr_special_finalize 555 9.7 0.006 0.007 2.797 3.225 dbcsr_merge_single_wm 555 10.7 0.535 0.665 2.789 3.217 ls_scf_store_result 1 5.0 0.000 0.000 2.968 3.031 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.883 2.931 compute_matrix_preconditioner 1 6.0 0.001 0.001 2.750 2.757 dbcsr_data_release 10477 10.7 1.571 2.439 1.571 2.439 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=101.269000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 4.681601E+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 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.177 0.190 101.714 101.717 qs_energies 1 2.0 0.000 0.000 100.251 100.255 ls_scf 1 3.0 0.000 0.000 98.331 98.337 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.238 79.509 ls_scf_main 1 4.0 0.000 0.000 61.437 61.437 multiply_cannon 111 7.7 0.079 0.129 56.335 61.324 density_matrix_trs4 2 5.0 0.002 0.003 54.525 54.606 multiply_cannon_loop 111 8.7 0.069 0.076 51.887 54.526 ls_scf_init_scf 1 4.0 0.000 0.001 33.227 33.232 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.559 31.601 mp_waitall_1 5481 11.0 25.681 30.884 25.681 30.884 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.133 29.142 multiply_cannon_multrec 444 9.7 14.065 16.504 21.291 23.480 make_m2s 222 7.7 0.004 0.005 17.555 20.001 make_images 222 8.7 3.727 4.420 17.494 19.942 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.053 16.203 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.612 15.874 make_images_data 222 9.7 0.003 0.004 9.814 12.161 hybrid_alltoall_any 227 10.6 0.791 3.781 9.580 11.937 dbcsr_mm_accdrv_process 3003 10.4 0.178 0.353 6.930 8.223 multiply_cannon_sync_h2d 444 9.7 6.436 7.969 6.436 7.969 dbcsr_mm_accdrv_process_sort 3003 11.4 6.493 7.652 6.493 7.652 mp_allgather_i34 111 8.7 2.704 6.993 2.704 6.993 mp_sum_l 807 5.4 3.392 6.554 3.392 6.554 arnoldi_extremal 4 6.8 0.000 0.000 5.952 5.965 arnoldi_normal_ev 4 7.8 0.038 0.041 5.951 5.965 build_subspace 16 8.4 0.015 0.019 5.523 5.531 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.332 5.024 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.841 4.979 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.248 4.453 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.731 4.415 mp_irecv_dv 1241 11.2 1.710 4.382 1.710 4.382 dbcsr_matrix_vector_mult_local 304 10.0 3.731 4.206 3.733 4.208 ls_scf_post 1 4.0 0.000 0.000 3.667 3.671 calculate_norms 792 9.8 3.530 3.662 3.530 3.662 ls_scf_store_result 1 5.0 0.000 0.000 3.441 3.489 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.345 3.429 make_images_sizes 222 9.7 0.000 0.000 1.075 3.261 mp_alltoall_i44 222 10.7 1.075 3.261 1.075 3.261 compute_matrix_preconditioner 1 6.0 0.001 0.002 2.523 2.528 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.369 2.370 rebuild_ks_matrix 3 7.3 0.000 0.000 2.335 2.335 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.334 2.335 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.321 dbcsr_merge_all 275 8.9 0.474 0.535 2.055 2.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.717000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3654.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a85683755d81adf0cf830f47158f5f9130b730a2_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 8.695169E+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 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 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 0.482 0.521 109.740 109.782 qs_energies 1 2.0 0.000 0.000 107.497 107.508 ls_scf 1 3.0 0.000 0.000 104.534 104.545 dbcsr_multiply_generic 111 6.7 0.023 0.028 78.314 78.462 ls_scf_main 1 4.0 0.000 0.000 64.606 64.607 density_matrix_trs4 2 5.0 0.002 0.003 55.781 55.850 multiply_cannon 111 7.7 0.110 0.166 49.808 50.832 multiply_cannon_loop 111 8.7 0.068 0.071 46.509 48.573 ls_scf_init_scf 1 4.0 0.006 0.010 36.258 36.259 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.785 34.805 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.505 31.520 mp_waitall_1 4569 11.1 20.930 25.006 20.930 25.006 make_m2s 222 7.7 0.005 0.005 23.575 24.407 make_images 222 8.7 4.594 5.012 23.469 24.300 multiply_cannon_multrec 444 9.7 17.836 18.599 22.516 23.017 hybrid_alltoall_any 227 10.6 1.664 3.641 12.554 15.419 make_images_data 222 9.7 0.003 0.003 12.801 15.114 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.560 11.566 multiply_cannon_sync_h2d 444 9.7 8.844 8.896 8.844 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.533 7.562 arnoldi_normal_ev 4 7.8 0.010 0.021 7.533 7.562 build_subspace 16 8.4 0.026 0.037 6.969 6.979 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.615 5.766 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.417 5.679 dbcsr_matrix_vector_mult_local 304 10.0 5.118 5.435 5.120 5.437 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.776 4.865 dbcsr_mm_accdrv_process 1814 10.4 0.300 0.664 4.514 4.694 dbcsr_mm_accdrv_process_sort 1814 11.4 4.168 4.307 4.168 4.307 mp_sum_l 807 5.4 2.621 3.947 2.621 3.947 ls_scf_post 1 4.0 0.000 0.000 3.669 3.680 make_images_sizes 222 9.7 0.000 0.000 1.428 3.529 mp_alltoall_i44 222 10.7 1.428 3.529 1.428 3.529 ls_scf_store_result 1 5.0 0.000 0.000 3.394 3.405 calculate_norms 792 9.8 3.242 3.277 3.242 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.084 3.139 qs_energies_init_hamiltonians 1 3.0 0.060 0.073 2.931 2.931 dbcsr_merge_all 275 8.9 0.891 0.915 2.871 2.920 dbcsr_complete_redistribute 5 7.6 1.434 1.469 2.773 2.894 dbcsr_data_release 12724 10.6 2.331 2.894 2.331 2.894 mp_allgather_i34 111 8.7 0.956 2.880 0.956 2.880 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.742 2.754 matrix_ls_to_qs 2 6.0 0.000 0.000 2.404 2.527 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.759 2.517 dbcsr_sort_data 325 11.1 2.445 2.515 2.445 2.515 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.339 2.342 dbcsr_new_transposed 4 7.5 0.243 0.253 2.292 2.316 rebuild_ks_matrix 3 7.3 0.000 0.000 2.273 2.276 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.273 2.276 dbcsr_frobenius_norm 74 6.6 2.054 2.140 2.197 2.242 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.206 dbcsr_add_anytype 103 7.2 0.860 0.894 2.132 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.782000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6903.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a85683755d81adf0cf830f47158f5f9130b730a2 Summary: empty Status: OK