=== 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: 7e92e2ab9bf27c2a5bb388384f6c7748634383a1 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, 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 (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/01 job id: 47240653 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/02 job id: 47240655 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/03 job id: 47240658 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/04 job id: 47240659 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/05 job id: 47240660 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/06 job id: 47240661 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/07 job id: 47240662 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/08 job id: 47240663 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/09 job id: 47240664 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/10 job id: 47240665 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/11 job id: 47240666 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/12 job id: 47240667 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/13 job id: 47240668 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/14 job id: 47240669 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/15 job id: 47240670 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/16 job id: 47240671 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/17 job id: 47240672 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/18 job id: 47240673 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/19 job id: 47240675 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/20 job id: 47240676 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/21 job id: 47240677 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/22 job id: 47240678 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/23 job id: 47240679 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/24 job id: 47240680 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/25 job id: 47240681 --- 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/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/26 job id: 47240682 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/27 job id: 47240683 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 134.673 134.674 farming_run 1 2.0 134.199 134.201 134.641 134.644 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465844E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.025 115.310 115.310 qs_energies 1 2.0 0.000 0.000 115.092 115.095 mp2_main 1 3.0 0.000 0.000 112.964 112.967 mp2_gpw_main 1 4.0 0.019 0.025 111.902 111.905 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.814 93.271 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.887 55.343 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.332 46.583 get_2c_integrals 1 6.0 0.008 0.009 37.261 37.756 integrate_v_rspace 273 8.0 0.432 0.449 24.717 29.734 pw_transfer 6555 10.6 0.373 0.390 27.424 27.906 fft_wrap_pw1pw2 5465 11.4 0.045 0.046 26.100 26.601 grid_integrate_task_list 273 9.0 20.529 26.031 20.529 26.031 fft_wrap_pw1pw2_100 2178 12.4 1.192 1.430 23.623 24.132 compute_2c_integrals 1 7.0 0.002 0.002 19.703 19.705 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.880 19.418 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.440 18.877 19.414 rpa_ri_compute_en 1 5.0 0.019 0.022 18.979 19.121 cp_fm_cholesky_decompose 12 8.2 17.536 18.035 17.536 18.035 cholesky_decomp 1 7.0 0.000 0.000 16.398 16.903 fft3d_s 5443 13.4 16.151 16.623 16.174 16.645 ao_to_mo_and_store_B_mult_1 272 7.0 10.721 15.310 10.721 15.310 calculate_wavefunction 272 8.0 5.409 5.563 12.541 13.151 rpa_num_int 1 6.0 0.000 0.000 10.703 10.704 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.671 10.702 calc_mat_Q 8 8.0 0.000 0.000 9.484 9.586 contract_S_to_Q 8 9.0 0.000 0.000 8.905 9.005 calc_potential_gpw 544 9.5 0.005 0.006 8.261 8.683 parallel_gemm_fm 14 9.1 0.000 0.000 8.499 8.597 parallel_gemm_fm_cosma 14 10.1 8.499 8.597 8.499 8.597 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.208 8.551 potential_pw2rs 545 10.0 0.107 0.110 7.675 8.306 collocate_single_gaussian 272 10.0 0.039 0.041 7.469 7.770 create_integ_mat 1 6.0 0.021 0.028 7.726 7.726 array2fm 1 7.0 0.000 0.000 6.701 7.135 pw_scatter_s 2720 13.7 4.446 4.662 4.446 4.662 pw_gather_s 2722 13.2 3.903 4.291 3.903 4.291 array2fm_buffer_send 1 8.0 2.932 3.122 2.932 3.122 pw_poisson_solve 545 10.5 1.105 1.143 2.135 2.359 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.900691, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.039 397.223 397.226 farming_run 1 2.0 396.350 396.363 397.183 397.188 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.224716E+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 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 26992263. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 209.437 209.438 qs_energies 1 2.0 0.003 0.021 209.179 209.190 scf_env_do_scf 1 3.0 0.000 0.000 106.279 106.280 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.338 105.348 rebuild_ks_matrix 4 6.0 0.000 0.000 105.337 105.346 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.066 105.337 105.346 hfx_ks_matrix 4 8.0 0.001 0.001 104.944 104.949 integrate_four_center 4 9.0 0.145 0.462 104.943 104.948 mp2_main 1 3.0 0.003 0.029 102.581 102.587 mp2_gpw_main 1 4.0 0.038 0.090 101.659 101.670 integrate_four_center_main 4 10.0 0.085 0.501 96.531 98.933 integrate_four_center_bin 264 11.0 96.446 98.907 96.446 98.907 init_scf_loop 1 4.0 0.000 0.000 92.111 92.111 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 74.424 75.569 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.994 55.134 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.732 46.765 integrate_v_rspace 95 8.0 0.396 0.566 28.124 33.113 pw_transfer 2240 10.6 0.144 0.165 29.938 30.348 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.962 29.366 mp2_ri_gpw_compute_en 1 5.0 0.055 0.065 27.060 28.849 grid_integrate_task_list 95 9.0 23.410 28.567 23.410 28.567 ao_to_mo_and_store_B_mult_1 91 7.0 10.583 28.547 10.583 28.547 fft_wrap_pw1pw2_100 730 12.4 1.261 1.446 26.660 27.061 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.896 25.220 25.230 get_2c_integrals 1 6.0 0.001 0.009 20.309 20.370 compute_2c_integrals 1 7.0 0.002 0.003 19.282 19.296 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.832 19.155 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.842 18.830 19.153 fft3d_s 1823 13.4 18.448 18.788 18.462 18.801 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.167 14.167 calculate_wavefunction 91 8.0 2.012 2.042 9.732 9.950 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.585 8.760 9.329 potential_pw2rs 186 10.0 0.033 0.035 8.635 9.180 local_gemm 172 8.0 8.204 8.770 8.204 8.770 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.007 8.222 8.674 mp2_ri_gpw_compute_en_comm 22 7.0 0.491 0.514 8.234 8.640 collocate_single_gaussian 91 10.0 0.017 0.029 7.875 8.176 calc_potential_gpw 182 9.5 0.002 0.002 7.899 8.151 mp_sendrecv_dm3 2068 8.0 6.290 6.707 6.290 6.707 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.425 6.348 6.425 mp_sync 37 10.5 3.112 6.001 3.112 6.001 pw_gather_s 912 13.2 4.919 5.383 4.919 5.383 pw_scatter_s 910 13.7 3.931 4.271 3.931 4.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.653425, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 451.502080E+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 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 55.525 55.526 qs_mol_dyn_low 1 2.0 0.003 0.003 55.291 55.298 qs_forces 11 3.9 0.002 0.003 55.229 55.231 qs_energies 11 4.9 0.001 0.001 53.672 53.697 scf_env_do_scf 11 5.9 0.000 0.001 47.544 47.544 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.369 45.369 qs_scf_new_mos 108 7.5 0.000 0.001 34.451 34.718 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.450 34.718 dbcsr_multiply_generic 2286 12.5 0.094 0.099 33.711 34.215 ot_scf_mini 108 9.5 0.002 0.002 32.697 32.890 multiply_cannon 2286 13.5 0.188 0.194 26.002 27.824 velocity_verlet 10 3.0 0.001 0.001 27.492 27.493 multiply_cannon_loop 2286 14.5 1.480 1.580 25.236 27.140 ot_mini 108 10.5 0.001 0.001 19.564 19.815 qs_ot_get_derivative 108 11.5 0.001 0.001 16.512 16.681 mp_waitall_1 245248 16.5 8.648 14.636 8.648 14.636 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.974 12.927 multiply_cannon_multrec 54864 15.5 4.253 6.590 7.585 11.001 qs_ot_get_p 119 10.4 0.001 0.001 8.400 8.676 rebuild_ks_matrix 119 8.3 0.000 0.000 8.287 8.426 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.287 8.426 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.302 7.426 mp_sum_l 7287 12.8 5.438 7.276 5.438 7.276 multiply_cannon_sync_h2d 54864 15.5 5.852 6.869 5.852 6.869 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.957 6.422 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.583 5.627 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.080 5.191 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.711 4.848 calculate_rho_elec 119 8.7 0.012 0.017 4.710 4.848 sum_up_and_integrate 119 10.3 0.012 0.014 4.831 4.839 integrate_v_rspace 119 11.3 0.002 0.003 4.819 4.829 init_scf_run 11 5.9 0.000 0.001 4.819 4.819 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.819 4.819 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.744 4.744 dbcsr_mm_accdrv_process 76910 16.1 1.128 1.800 3.253 4.582 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.487 4.488 cp_fm_redistribute_end 50 14.0 2.289 4.452 2.299 4.457 cp_fm_diag_elpa_base 50 14.0 2.149 4.325 2.155 4.336 rs_pw_transfer 974 11.9 0.012 0.013 3.105 3.207 calculate_dm_sparse 119 9.5 0.000 0.001 3.030 3.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.943 3.136 apply_single 119 13.6 0.000 0.000 2.943 3.135 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.811 3.019 ot_diis_step 108 11.5 0.006 0.006 2.753 2.754 jit_kernel_multiply 13 15.8 2.066 2.734 2.066 2.734 density_rs2pw 119 9.7 0.004 0.005 2.588 2.713 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.486 2.548 calculate_first_density_matrix 1 7.0 0.000 0.000 2.499 2.503 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.306 2.309 acc_transpose_blocks 54864 15.5 0.226 0.248 1.792 2.261 wfi_extrapolate 11 7.9 0.001 0.001 2.230 2.230 init_scf_loop 11 6.9 0.000 0.000 2.151 2.151 potential_pw2rs 119 12.3 0.004 0.004 2.131 2.150 pw_transfer 1439 11.6 0.052 0.059 2.059 2.139 grid_integrate_task_list 119 12.3 2.005 2.115 2.005 2.115 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.982 2.065 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.993 2.061 make_m2s 4572 13.5 0.054 0.056 1.955 2.023 mp_sum_d 4135 12.0 1.372 1.976 1.372 1.976 make_images 4572 14.5 0.132 0.137 1.873 1.939 fft3d_ps 1201 14.6 0.373 0.480 1.622 1.701 mp_alltoall_d11v 2130 13.8 1.485 1.701 1.485 1.701 fft_wrap_pw1pw2_140 487 13.2 0.186 0.205 1.513 1.593 mp_waitany 12084 13.8 1.332 1.498 1.332 1.498 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.453 1.487 grid_collocate_task_list 119 9.7 1.361 1.424 1.361 1.424 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.780 1.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.526000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 487.104512E+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 10306 303. MP_Sync 54 MP_Alltoall 2060 1614464. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 53.364 53.365 qs_mol_dyn_low 1 2.0 0.003 0.003 53.106 53.113 qs_forces 11 3.9 0.002 0.002 53.012 53.012 qs_energies 11 4.9 0.001 0.001 50.942 50.952 scf_env_do_scf 11 5.9 0.000 0.001 44.471 44.472 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 39.721 39.722 velocity_verlet 10 3.0 0.001 0.001 28.200 28.202 dbcsr_multiply_generic 2286 12.5 0.105 0.109 26.304 27.192 qs_scf_new_mos 108 7.5 0.001 0.001 26.486 26.743 qs_scf_loop_do_ot 108 8.5 0.001 0.001 26.486 26.743 ot_scf_mini 108 9.5 0.002 0.003 25.252 25.429 multiply_cannon 2286 13.5 0.208 0.217 18.557 20.986 multiply_cannon_loop 2286 14.5 0.912 0.982 16.754 19.615 ot_mini 108 10.5 0.001 0.001 15.248 15.531 mp_waitall_1 200699 16.5 9.418 14.989 9.418 14.989 qs_ot_get_derivative 108 11.5 0.001 0.001 12.290 12.465 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 5.721 11.422 rebuild_ks_matrix 119 8.3 0.000 0.000 9.620 9.894 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 9.619 9.893 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.479 8.718 multiply_cannon_multrec 27432 15.5 1.875 4.398 5.856 8.611 qs_ot_get_p 119 10.4 0.001 0.001 6.424 6.697 sum_up_and_integrate 119 10.3 0.025 0.028 6.281 6.287 integrate_v_rspace 119 11.3 0.002 0.003 6.256 6.266 qs_rho_update_rho_low 119 7.7 0.001 0.001 6.011 6.143 calculate_rho_elec 119 8.7 0.021 0.024 6.011 6.142 rs_pw_transfer 974 11.9 0.010 0.011 5.511 5.977 dbcsr_mm_accdrv_process 47894 16.0 3.074 5.072 3.912 5.780 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.949 5.738 mp_sum_l 7287 12.8 3.197 5.558 3.197 5.558 init_scf_run 11 5.9 0.000 0.001 5.007 5.007 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.006 5.007 density_rs2pw 119 9.7 0.004 0.004 4.175 4.699 init_scf_loop 11 6.9 0.000 0.000 4.697 4.697 make_m2s 4572 13.5 0.052 0.054 4.100 4.540 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.331 4.513 apply_single 119 13.6 0.000 0.000 3.331 4.513 make_images 4572 14.5 0.201 0.238 4.010 4.448 qs_ot_p2m_diag 50 11.0 0.009 0.013 4.077 4.107 potential_pw2rs 119 12.3 0.006 0.006 3.856 3.903 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.554 3.555 prepare_preconditioner 11 7.9 0.000 0.000 3.399 3.430 make_preconditioner 11 8.9 0.000 0.000 3.399 3.430 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.233 3.305 pw_transfer 1439 11.6 0.066 0.070 3.098 3.163 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 3.005 3.073 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.929 2.930 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.815 2.897 cp_fm_redistribute_end 50 14.0 1.474 2.855 1.485 2.862 ot_diis_step 108 11.5 0.011 0.011 2.819 2.821 cp_fm_diag_elpa_base 50 14.0 1.334 2.698 1.366 2.749 make_images_data 4572 15.5 0.045 0.051 1.790 2.687 fft3d_ps 1201 14.6 0.529 0.586 2.559 2.619 calculate_dm_sparse 119 9.5 0.000 0.001 2.487 2.594 multiply_cannon_sync_h2d 27432 15.5 1.923 2.591 1.923 2.591 calculate_first_density_matrix 1 7.0 0.000 0.000 2.537 2.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.432 2.435 hybrid_alltoall_any 4725 16.4 0.051 0.114 1.549 2.430 wfi_extrapolate 11 7.9 0.001 0.001 2.378 2.378 mp_allgather_i34 2286 14.5 1.238 2.351 1.238 2.351 mp_alltoall_d11v 2130 13.8 2.025 2.220 2.025 2.220 jit_kernel_multiply 11 16.1 0.784 2.209 0.784 2.209 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.918 2.023 make_images_sizes 4572 15.5 0.005 0.005 1.499 2.006 mp_alltoall_i44 4572 16.5 1.494 2.001 1.494 2.001 grid_integrate_task_list 119 12.3 1.826 1.972 1.826 1.972 fft_wrap_pw1pw2_140 487 13.2 0.200 0.212 1.783 1.849 mp_sendrecv_dv 16779 12.7 1.659 1.706 1.659 1.706 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.640 1.695 cp_fm_cholesky_invert 11 10.9 1.642 1.650 1.642 1.650 acc_transpose_blocks 27432 15.5 0.110 0.114 1.297 1.625 mp_alltoall_z22v 1201 16.6 1.467 1.591 1.467 1.591 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.603 1.424 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.437 1.459 grid_collocate_task_list 119 9.7 1.278 1.425 1.278 1.425 mp_sum_d 4135 12.0 0.910 1.418 0.910 1.418 dbcsr_complete_redistribute 329 12.2 0.078 0.108 1.294 1.406 mp_waitany 5720 13.7 0.713 1.253 0.713 1.253 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.204 1.239 rs_pw_transfer_RS2PW_50 119 11.7 0.271 0.284 1.172 1.233 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.146 0.675 1.186 copy_dbcsr_to_fm 153 11.3 0.002 0.002 1.051 1.181 cp_fm_cholesky_decompose 22 10.9 1.176 1.180 1.176 1.180 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.617 1.080 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=53.365000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.727273, yerr=1.482682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 521.920512E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.076 35.489 35.490 qs_mol_dyn_low 1 2.0 0.004 0.010 35.088 35.147 qs_forces 11 3.9 0.002 0.003 35.014 35.017 qs_energies 11 4.9 0.003 0.019 33.331 33.338 scf_env_do_scf 11 5.9 0.001 0.002 28.270 28.270 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 25.374 25.375 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.291 17.507 velocity_verlet 10 3.0 0.001 0.005 16.896 16.898 qs_scf_new_mos 108 7.5 0.001 0.001 16.384 16.408 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.383 16.407 ot_scf_mini 108 9.5 0.002 0.003 15.604 15.619 multiply_cannon 2286 13.5 0.196 0.199 13.339 14.334 multiply_cannon_loop 2286 14.5 0.640 0.669 12.465 13.545 ot_mini 108 10.5 0.001 0.001 9.409 9.432 qs_ot_get_derivative 108 11.5 0.001 0.001 7.814 7.833 multiply_cannon_multrec 18288 15.5 1.963 2.953 6.755 7.118 rebuild_ks_matrix 119 8.3 0.000 0.000 6.878 6.914 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.029 6.877 6.914 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.041 6.073 dbcsr_mm_accdrv_process 38222 16.0 3.927 5.197 4.708 5.671 mp_waitall_1 158411 16.6 3.234 4.613 3.234 4.613 sum_up_and_integrate 119 10.3 0.031 0.031 4.371 4.376 integrate_v_rspace 119 11.3 0.003 0.010 4.340 4.348 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.908 3.925 qs_ot_get_p 119 10.4 0.001 0.001 3.895 3.925 calculate_rho_elec 119 8.7 0.031 0.031 3.908 3.925 init_scf_run 11 5.9 0.000 0.001 3.724 3.724 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.724 3.724 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.909 3.552 rs_pw_transfer 974 11.9 0.009 0.010 2.813 3.059 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.671 2.985 init_scf_loop 11 6.9 0.001 0.005 2.875 2.876 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.627 2.637 density_rs2pw 119 9.7 0.004 0.004 2.349 2.605 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.093 2.416 apply_single 119 13.6 0.000 0.000 2.093 2.416 calculate_first_density_matrix 1 7.0 0.001 0.009 2.355 2.357 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.327 2.327 make_m2s 4572 13.5 0.046 0.059 2.098 2.204 pw_transfer 1439 11.6 0.066 0.070 2.176 2.195 make_images 4572 14.5 0.191 0.203 2.010 2.118 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.083 2.106 potential_pw2rs 119 12.3 0.007 0.009 2.084 2.092 prepare_preconditioner 11 7.9 0.000 0.001 2.032 2.034 make_preconditioner 11 8.9 0.000 0.002 2.032 2.034 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.020 2.031 cp_fm_diag_elpa_base 50 14.0 1.993 2.009 2.018 2.029 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.870 1.955 grid_integrate_task_list 119 12.3 1.800 1.894 1.800 1.894 calculate_dm_sparse 119 9.5 0.000 0.001 1.876 1.894 mp_sum_l 7287 12.8 1.504 1.872 1.504 1.872 jit_kernel_multiply 10 15.9 0.729 1.859 0.729 1.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.813 1.819 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.763 1.771 fft_wrap_pw1pw2_140 487 13.2 0.252 0.259 1.588 1.607 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.588 1.606 fft3d_ps 1201 14.6 0.535 0.556 1.572 1.596 multiply_cannon_sync_h2d 18288 15.5 1.325 1.552 1.325 1.552 ot_diis_step 108 11.5 0.011 0.011 1.549 1.549 grid_collocate_task_list 119 9.7 1.246 1.398 1.246 1.398 wfi_extrapolate 11 7.9 0.001 0.001 1.309 1.309 acc_transpose_blocks 18288 15.5 0.076 0.079 1.264 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.203 1.213 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.390 1.104 make_images_data 4572 15.5 0.045 0.048 0.950 1.103 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.978 1.012 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.821 1.003 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.988 0.990 mp_alltoall_d11v 2130 13.8 0.861 0.953 0.861 0.953 mp_alltoall_z22v 1201 16.6 0.862 0.947 0.862 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.919 cp_fm_cholesky_invert 11 10.9 0.891 0.897 0.891 0.897 mp_waitany 9880 13.7 0.600 0.873 0.600 0.873 acc_transpose_blocks_kernels 18288 16.5 0.211 0.221 0.814 0.836 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.563 0.812 mp_allgather_i34 2286 14.5 0.335 0.755 0.335 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.640 0.730 dbcsr_complete_redistribute 329 12.2 0.105 0.137 0.610 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.490000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=2.104698 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 556.638208E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.081 44.938 44.939 qs_mol_dyn_low 1 2.0 0.003 0.004 44.533 44.541 qs_forces 11 3.9 0.009 0.055 44.472 44.473 qs_energies 11 4.9 0.004 0.021 42.528 42.535 scf_env_do_scf 11 5.9 0.001 0.001 36.549 36.551 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 31.829 31.830 velocity_verlet 10 3.0 0.001 0.001 21.814 21.816 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.067 21.307 qs_scf_new_mos 108 7.5 0.001 0.001 20.832 20.893 qs_scf_loop_do_ot 108 8.5 0.001 0.002 20.832 20.893 ot_scf_mini 108 9.5 0.003 0.003 19.751 19.818 multiply_cannon 2286 13.5 0.225 0.258 15.503 16.097 multiply_cannon_loop 2286 14.5 0.949 0.983 14.132 14.789 ot_mini 108 10.5 0.001 0.001 11.397 11.467 qs_ot_get_derivative 108 11.5 0.001 0.001 9.417 9.478 multiply_cannon_multrec 27432 15.5 2.395 3.089 8.834 9.221 rebuild_ks_matrix 119 8.3 0.000 0.000 8.034 8.101 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 8.034 8.101 dbcsr_mm_accdrv_process 47916 15.9 5.316 7.369 6.348 7.612 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.183 7.241 qs_ot_get_p 119 10.4 0.001 0.002 5.357 5.430 sum_up_and_integrate 119 10.3 0.035 0.037 5.048 5.057 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.987 5.034 calculate_rho_elec 119 8.7 0.040 0.046 4.986 5.034 integrate_v_rspace 119 11.3 0.003 0.003 5.013 5.022 mp_waitall_1 137007 16.6 3.672 4.756 3.672 4.756 init_scf_loop 11 6.9 0.001 0.006 4.686 4.686 init_scf_run 11 5.9 0.000 0.001 4.399 4.399 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.398 4.399 rs_pw_transfer 974 11.9 0.009 0.009 3.915 4.151 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.259 3.747 prepare_preconditioner 11 7.9 0.000 0.002 3.667 3.676 make_preconditioner 11 8.9 0.000 0.001 3.666 3.676 make_m2s 4572 13.5 0.054 0.057 3.438 3.660 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.229 3.579 make_images 4572 14.5 0.271 0.330 3.329 3.550 qs_ot_p2m_diag 50 11.0 0.016 0.023 3.518 3.541 density_rs2pw 119 9.7 0.004 0.004 3.294 3.485 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.072 3.072 pw_transfer 1439 11.6 0.066 0.071 2.716 2.763 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.723 2.757 potential_pw2rs 119 12.3 0.009 0.010 2.708 2.726 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.241 2.698 apply_single 119 13.6 0.000 0.000 2.240 2.698 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.623 2.673 calculate_first_density_matrix 1 7.0 0.001 0.008 2.615 2.616 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.545 2.556 cp_fm_diag_elpa_base 50 14.0 2.496 2.519 2.542 2.552 mp_sum_l 7287 12.8 1.721 2.382 1.721 2.382 calculate_dm_sparse 119 9.5 0.000 0.000 2.292 2.348 fft3d_ps 1201 14.6 0.565 0.613 2.063 2.106 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 1.056 2.062 jit_kernel_multiply 10 15.8 0.971 2.012 0.971 2.012 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.005 2.006 fft_wrap_pw1pw2_140 487 13.2 0.287 0.298 1.917 1.970 ot_diis_step 108 11.5 0.012 0.012 1.918 1.918 grid_integrate_task_list 119 12.3 1.798 1.888 1.798 1.888 make_images_data 4572 15.5 0.045 0.048 1.507 1.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.735 1.751 wfi_extrapolate 11 7.9 0.001 0.001 1.692 1.692 hybrid_alltoall_any 4725 16.4 0.062 0.151 1.262 1.543 acc_transpose_blocks 27432 15.5 0.111 0.113 1.508 1.542 mp_alltoall_z22v 1201 16.6 1.381 1.420 1.381 1.420 dbcsr_complete_redistribute 329 12.2 0.121 0.147 1.118 1.413 mp_alltoall_d11v 2130 13.8 1.214 1.405 1.214 1.405 cp_fm_cholesky_invert 11 10.9 1.371 1.377 1.371 1.377 grid_collocate_task_list 119 9.7 1.251 1.370 1.251 1.370 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.356 1.370 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.357 1.369 cp_fm_upper_to_full 72 14.2 0.888 1.272 0.888 1.272 make_images_sizes 4572 15.5 0.005 0.005 0.907 1.252 mp_alltoall_i44 4572 16.5 0.903 1.247 0.903 1.247 mp_sendrecv_dv 8211 12.7 1.157 1.188 1.157 1.188 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.168 1.169 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.743 1.039 multiply_cannon_sync_h2d 27432 15.5 0.950 1.002 0.950 1.002 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.955 0.961 mp_allgather_i34 2286 14.5 0.669 0.918 0.669 0.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=44.939000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.272727, yerr=6.607034 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 615.350272E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.066 31.601 31.602 qs_mol_dyn_low 1 2.0 0.003 0.003 31.210 31.218 qs_forces 11 3.9 0.002 0.003 31.154 31.155 qs_energies 11 4.9 0.001 0.001 29.388 29.391 scf_env_do_scf 11 5.9 0.000 0.001 24.135 24.135 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 21.257 21.257 velocity_verlet 10 3.0 0.001 0.001 15.832 15.834 dbcsr_multiply_generic 2286 12.5 0.091 0.097 13.453 13.582 qs_scf_new_mos 108 7.5 0.001 0.001 12.504 12.529 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.504 12.528 ot_scf_mini 108 9.5 0.002 0.002 11.773 11.796 multiply_cannon 2286 13.5 0.231 0.238 10.395 10.818 multiply_cannon_loop 2286 14.5 0.332 0.344 9.383 9.608 rebuild_ks_matrix 119 8.3 0.000 0.000 6.505 6.533 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.505 6.532 ot_mini 108 10.5 0.001 0.001 6.503 6.528 multiply_cannon_multrec 9144 15.5 1.697 2.017 6.081 6.417 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.813 5.839 qs_ot_get_derivative 108 11.5 0.001 0.001 5.111 5.134 dbcsr_mm_accdrv_process 12550 15.8 3.216 4.316 4.282 4.375 sum_up_and_integrate 119 10.3 0.038 0.042 4.015 4.021 integrate_v_rspace 119 11.3 0.003 0.003 3.977 3.983 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.935 3.943 calculate_rho_elec 119 8.7 0.060 0.061 3.935 3.943 init_scf_run 11 5.9 0.000 0.001 3.737 3.737 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.737 3.737 qs_ot_get_p 119 10.4 0.001 0.001 3.344 3.374 init_scf_loop 11 6.9 0.000 0.000 2.852 2.853 mp_waitall_1 115863 16.7 2.135 2.754 2.135 2.754 calculate_first_density_matrix 1 7.0 0.000 0.000 2.423 2.424 density_rs2pw 119 9.7 0.004 0.004 2.190 2.332 pw_transfer 1439 11.6 0.066 0.070 2.301 2.311 rs_pw_transfer 974 11.9 0.008 0.008 2.112 2.244 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.229 2.232 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.207 2.216 make_m2s 4572 13.5 0.034 0.035 2.087 2.211 make_images 4572 14.5 0.267 0.302 1.997 2.120 prepare_preconditioner 11 7.9 0.000 0.000 2.010 2.014 make_preconditioner 11 8.9 0.000 0.000 2.010 2.014 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.002 2.003 jit_kernel_multiply 10 15.7 1.028 1.951 1.028 1.951 grid_integrate_task_list 119 12.3 1.866 1.945 1.866 1.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.879 1.909 calculate_dm_sparse 119 9.5 0.000 0.000 1.870 1.888 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.774 1.775 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.756 1.767 fft_wrap_pw1pw2_140 487 13.2 0.366 0.375 1.754 1.765 potential_pw2rs 119 12.3 0.010 0.011 1.677 1.683 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.661 1.667 cp_fm_diag_elpa_base 50 14.0 1.629 1.647 1.658 1.665 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.588 1.596 fft3d_ps 1201 14.6 0.565 0.578 1.537 1.549 grid_collocate_task_list 119 9.7 1.302 1.428 1.302 1.428 ot_diis_step 108 11.5 0.012 0.013 1.370 1.370 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.320 1.333 wfi_extrapolate 11 7.9 0.001 0.001 1.255 1.255 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.219 1.244 apply_single 119 13.6 0.000 0.000 1.219 1.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.242 1.243 make_images_data 4572 15.5 0.039 0.043 1.013 1.225 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.960 1.199 acc_transpose_blocks 9144 15.5 0.038 0.039 1.033 1.057 mp_alltoall_d11v 2130 13.8 0.946 1.051 0.946 1.051 cp_fm_cholesky_invert 11 10.9 1.041 1.046 1.041 1.046 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.411 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.867 0.915 mp_alltoall_z22v 1201 16.6 0.846 0.875 0.846 0.875 mp_allgather_i34 2286 14.5 0.333 0.859 0.333 0.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.824 0.826 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.805 0.814 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.788 0.811 multiply_cannon_sync_h2d 9144 15.5 0.718 0.788 0.718 0.788 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.535 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.695 0.751 mp_sum_l 7287 12.8 0.545 0.708 0.545 0.708 jit_kernel_transpose 5 15.6 0.671 0.695 0.671 0.695 dbcsr_complete_redistribute 329 12.2 0.161 0.170 0.612 0.650 mp_waitany 5200 13.7 0.502 0.633 0.502 0.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.602000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.272727, yerr=4.453618 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 739.663872E+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 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.058 41.992 41.992 qs_mol_dyn_low 1 2.0 0.003 0.003 41.746 41.752 qs_forces 11 3.9 0.002 0.003 41.647 41.648 qs_energies 11 4.9 0.001 0.001 39.648 39.651 scf_env_do_scf 11 5.9 0.001 0.001 33.667 33.667 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.713 25.715 velocity_verlet 10 3.0 0.001 0.001 23.608 23.619 dbcsr_multiply_generic 2286 12.5 0.100 0.103 17.739 17.865 qs_scf_new_mos 108 7.5 0.001 0.001 15.951 16.044 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.950 16.043 ot_scf_mini 108 9.5 0.002 0.002 14.856 14.955 multiply_cannon 2286 13.5 0.303 0.309 13.860 14.754 multiply_cannon_loop 2286 14.5 0.347 0.352 12.582 13.494 ot_mini 108 10.5 0.001 0.001 8.886 9.001 multiply_cannon_multrec 9144 15.5 3.550 5.046 8.766 8.844 init_scf_loop 11 6.9 0.000 0.000 7.925 7.928 rebuild_ks_matrix 119 8.3 0.000 0.000 7.272 7.416 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.272 7.416 prepare_preconditioner 11 7.9 0.000 0.000 6.962 6.976 make_preconditioner 11 8.9 0.000 0.000 6.962 6.976 qs_ot_get_derivative 108 11.5 0.001 0.001 6.846 6.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.480 6.852 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.562 6.692 dbcsr_mm_accdrv_process 12550 15.8 4.251 5.670 5.092 6.493 cp_fm_upper_to_full 72 14.2 3.200 4.623 3.200 4.623 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.329 4.394 calculate_rho_elec 119 8.7 0.119 0.121 4.328 4.393 sum_up_and_integrate 119 10.3 0.064 0.065 3.974 3.981 init_scf_run 11 5.9 0.000 0.001 3.965 3.965 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.965 3.965 integrate_v_rspace 119 11.3 0.004 0.004 3.910 3.917 mp_waitall_1 94719 16.7 2.390 3.402 2.390 3.402 qs_ot_get_p 119 10.4 0.001 0.001 3.212 3.342 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.478 2.915 dbcsr_complete_redistribute 329 12.2 0.286 0.292 2.024 2.876 pw_transfer 1439 11.6 0.069 0.069 2.686 2.692 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.588 2.594 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.712 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.422 2.490 make_m2s 4572 13.5 0.038 0.038 2.297 2.483 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.214 2.464 apply_single 119 13.6 0.000 0.000 2.214 2.464 make_images 4572 14.5 0.350 0.380 2.175 2.360 calculate_dm_sparse 119 9.5 0.000 0.001 2.272 2.332 mp_alltoall_i22 627 13.8 1.497 2.331 1.497 2.331 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.477 2.296 density_rs2pw 119 9.7 0.004 0.004 2.219 2.242 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.321 2.227 fft_wrap_pw1pw2_140 487 13.2 0.616 0.622 2.180 2.187 grid_integrate_task_list 119 12.3 1.995 2.021 1.995 2.021 ot_diis_step 108 11.5 0.014 0.015 2.013 2.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.994 1.996 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.959 1.960 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.705 1.755 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.745 1.746 mp_sum_l 7287 12.8 1.039 1.734 1.039 1.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.697 1.706 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.684 1.684 fft3d_ps 1201 14.6 0.596 0.606 1.605 1.610 rs_pw_transfer 974 11.9 0.009 0.009 1.504 1.538 grid_collocate_task_list 119 9.7 1.455 1.496 1.455 1.496 cp_fm_cholesky_invert 11 10.9 1.492 1.496 1.492 1.496 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.146 1.414 wfi_extrapolate 11 7.9 0.001 0.001 1.412 1.412 potential_pw2rs 119 12.3 0.014 0.014 1.404 1.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.388 1.388 cp_fm_diag_elpa_base 50 14.0 1.241 1.294 1.386 1.387 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.324 1.386 make_images_data 4572 15.5 0.042 0.045 1.091 1.325 mp_alltoall_d11v 2130 13.8 1.182 1.273 1.182 1.273 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.150 1.172 jit_kernel_multiply 6 15.4 0.814 1.133 0.814 1.133 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.081 1.094 multiply_cannon_sync_h2d 9144 15.5 1.037 1.042 1.037 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.986 1.041 acc_transpose_blocks 9144 15.5 0.039 0.039 0.947 0.955 qs_create_task_list 11 7.9 0.001 0.001 0.929 0.939 generate_qs_task_list 11 8.9 0.367 0.388 0.928 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.935 mp_alltoall_z22v 1201 16.6 0.874 0.893 0.874 0.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.992000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=695.000000, yerr=11.224972 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.235712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2431597. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 98.485 98.486 qs_mol_dyn_low 1 2.0 0.003 0.003 98.127 98.136 qs_forces 11 3.9 0.003 0.004 97.743 97.744 qs_energies 11 4.9 0.001 0.001 94.608 94.623 scf_env_do_scf 11 5.9 0.000 0.001 84.997 84.999 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 77.211 77.212 qs_scf_new_mos 99 7.5 0.001 0.001 55.519 55.636 qs_scf_loop_do_ot 99 8.5 0.001 0.001 55.518 55.635 dbcsr_multiply_generic 2055 12.4 0.109 0.111 54.852 55.128 ot_scf_mini 99 9.5 0.002 0.002 52.932 52.993 velocity_verlet 10 3.0 0.001 0.001 48.210 48.214 multiply_cannon 2055 13.4 0.180 0.186 43.433 44.613 multiply_cannon_loop 2055 14.4 1.520 1.552 41.944 43.155 ot_mini 99 10.5 0.001 0.001 29.114 29.179 qs_ot_get_derivative 99 11.5 0.001 0.001 22.262 22.322 multiply_cannon_multrec 49320 15.4 12.114 12.799 17.004 17.818 rebuild_ks_matrix 110 8.3 0.000 0.000 17.034 17.117 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 17.033 17.117 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.025 15.117 mp_waitall_1 220248 16.4 13.763 14.868 13.763 14.868 qs_ot_get_p 110 10.4 0.001 0.001 14.690 14.772 multiply_cannon_sync_h2d 49320 15.4 9.846 10.544 9.846 10.544 qs_ot_p2m_diag 48 11.0 0.013 0.019 10.456 10.513 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.003 9.169 calculate_rho_elec 110 8.6 0.021 0.026 9.002 9.168 cp_dbcsr_syevd 48 12.0 0.003 0.003 9.119 9.121 sum_up_and_integrate 110 10.3 0.036 0.043 8.958 8.973 integrate_v_rspace 110 11.3 0.003 0.004 8.922 8.945 cp_fm_diag_elpa 48 13.0 0.000 0.000 8.080 8.105 cp_fm_diag_elpa_base 48 14.0 8.044 8.068 8.076 8.102 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.467 8.016 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.946 7.902 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.242 7.830 apply_single 110 13.6 0.000 0.001 7.242 7.829 init_scf_loop 11 6.9 0.000 0.000 7.746 7.747 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.481 7.527 init_scf_run 11 5.9 0.000 0.001 7.352 7.352 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.351 7.352 ot_diis_step 99 11.5 0.005 0.005 6.625 6.625 mp_sum_l 6594 12.7 5.223 6.445 5.223 6.445 rs_pw_transfer 902 11.9 0.011 0.013 5.817 6.387 density_rs2pw 110 9.6 0.004 0.005 5.652 6.211 make_m2s 4110 13.4 0.062 0.065 5.731 5.882 make_images 4110 14.4 0.178 0.192 5.631 5.784 pw_transfer 1331 11.6 0.057 0.067 5.498 5.646 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.408 5.551 prepare_preconditioner 11 7.9 0.000 0.000 5.345 5.358 make_preconditioner 11 8.9 0.000 0.000 5.345 5.358 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.098 5.153 dbcsr_mm_accdrv_process 87628 16.1 1.926 1.991 4.767 5.028 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 3.394 4.877 fft3d_ps 1111 14.6 0.785 0.875 4.603 4.727 wfi_extrapolate 11 7.9 0.001 0.001 4.502 4.502 potential_pw2rs 110 12.3 0.006 0.007 4.358 4.435 fft_wrap_pw1pw2_140 451 13.1 0.448 0.495 4.263 4.408 calculate_dm_sparse 110 9.5 0.001 0.001 3.977 4.096 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.739 3.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.670 3.675 grid_integrate_task_list 110 12.3 3.236 3.503 3.236 3.503 mp_alltoall_z22v 1111 16.6 3.274 3.487 3.274 3.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.258 3.296 mp_alltoall_d11v 2046 13.8 2.567 3.080 2.567 3.080 cp_fm_cholesky_invert 11 10.9 2.870 2.877 2.870 2.877 calculate_first_density_matrix 1 7.0 0.000 0.000 2.713 2.716 make_images_data 4110 15.4 0.044 0.047 2.434 2.710 jit_kernel_multiply 13 15.9 2.565 2.596 2.565 2.596 mp_waitany 14300 13.8 1.997 2.508 1.997 2.508 grid_collocate_task_list 110 9.6 2.158 2.414 2.158 2.414 hybrid_alltoall_any 4261 16.3 0.083 0.476 2.030 2.397 make_images_sizes 4110 15.4 0.004 0.005 1.759 2.354 mp_alltoall_i44 4110 16.4 1.754 2.350 1.754 2.350 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.150 2.169 acc_transpose_blocks 49320 15.4 0.211 0.219 2.100 2.150 mp_sum_d 3889 11.9 1.506 2.002 1.506 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=98.486000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.909091, yerr=2.502891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 588.488704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2530369. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.111 70.786 70.790 qs_mol_dyn_low 1 2.0 0.005 0.014 70.327 70.337 qs_forces 11 3.9 0.013 0.062 70.236 70.238 qs_energies 11 4.9 0.003 0.011 66.789 66.798 scf_env_do_scf 11 5.9 0.001 0.001 58.093 58.096 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.243 50.244 dbcsr_multiply_generic 2055 12.4 0.117 0.121 38.060 38.344 velocity_verlet 10 3.0 0.002 0.007 37.063 37.064 qs_scf_new_mos 99 7.5 0.001 0.001 33.548 33.710 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.547 33.709 multiply_cannon 2055 13.4 0.221 0.243 31.165 32.232 ot_scf_mini 99 9.5 0.003 0.004 31.872 32.024 multiply_cannon_loop 2055 14.4 0.927 0.947 29.840 31.044 ot_mini 99 10.5 0.001 0.002 18.566 18.725 multiply_cannon_multrec 24660 15.4 7.652 9.582 13.824 15.747 rebuild_ks_matrix 110 8.3 0.000 0.000 14.049 14.211 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.022 14.049 14.211 qs_ot_get_derivative 99 11.5 0.001 0.002 12.685 12.831 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.347 12.491 mp_waitall_1 176588 16.5 7.839 10.309 7.839 10.309 multiply_cannon_sync_h2d 24660 15.4 7.042 8.219 7.042 8.219 init_scf_loop 11 6.9 0.001 0.004 7.813 7.814 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.257 7.597 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.632 7.354 apply_single 110 13.6 0.000 0.001 6.632 7.354 sum_up_and_integrate 110 10.3 0.053 0.059 6.808 6.818 qs_ot_get_p 110 10.4 0.001 0.001 6.612 6.791 integrate_v_rspace 110 11.3 0.003 0.003 6.755 6.767 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.379 6.393 calculate_rho_elec 110 8.6 0.040 0.048 6.379 6.391 dbcsr_mm_accdrv_process 52282 16.1 4.551 5.452 6.012 6.315 init_scf_run 11 5.9 0.000 0.001 6.231 6.232 scf_env_initial_rho_setup 11 6.9 0.003 0.005 6.231 6.232 ot_diis_step 99 11.5 0.010 0.013 5.819 5.819 prepare_preconditioner 11 7.9 0.000 0.000 5.746 5.769 make_preconditioner 11 8.9 0.000 0.001 5.746 5.769 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.805 5.580 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.307 5.469 make_m2s 4110 13.4 0.057 0.059 4.296 4.742 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.700 4.723 make_images 4110 14.4 0.401 0.447 4.187 4.630 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.253 4.253 density_rs2pw 110 9.6 0.004 0.005 3.434 3.944 pw_transfer 1331 11.6 0.067 0.074 3.775 3.921 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.667 3.816 rs_pw_transfer 902 11.9 0.012 0.014 3.159 3.737 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.669 3.683 cp_fm_diag_elpa_base 48 14.0 3.618 3.640 3.666 3.680 wfi_extrapolate 11 7.9 0.001 0.001 3.599 3.599 grid_integrate_task_list 110 12.3 3.148 3.372 3.148 3.372 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.219 3.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.288 3.290 fft_wrap_pw1pw2_140 451 13.1 0.521 0.538 3.086 3.236 calculate_dm_sparse 110 9.5 0.001 0.001 2.981 3.008 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.933 2.999 make_images_data 4110 15.4 0.047 0.051 2.420 2.874 hybrid_alltoall_any 4261 16.3 0.103 0.446 2.108 2.837 fft3d_ps 1111 14.6 1.110 1.333 2.639 2.788 cp_fm_cholesky_invert 11 10.9 2.752 2.760 2.752 2.760 mp_sum_l 6594 12.7 2.024 2.742 2.024 2.742 potential_pw2rs 110 12.3 0.008 0.009 2.552 2.569 calculate_first_density_matrix 1 7.0 0.001 0.006 2.526 2.532 grid_collocate_task_list 110 9.6 2.103 2.530 2.103 2.530 mp_alltoall_d11v 2046 13.8 1.836 2.137 1.836 2.137 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 1.998 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.902 1.906 mp_waitany 10164 13.8 1.290 1.885 1.290 1.885 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.861 1.877 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.785 1.733 mp_allgather_i34 2055 14.4 0.655 1.721 0.655 1.721 jit_kernel_multiply 9 16.3 1.111 1.673 1.111 1.673 acc_transpose_blocks 24660 15.4 0.112 0.117 1.559 1.655 mp_irecv_dv 57340 16.2 0.656 1.608 0.656 1.608 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 1.012 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.475 dbcsr_complete_redistribute 325 12.2 0.238 0.300 1.156 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.790000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.000000, yerr=6.687845 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 661.942272E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 63.587 63.587 qs_mol_dyn_low 1 2.0 0.003 0.003 63.287 63.297 qs_forces 11 3.9 0.003 0.004 62.779 62.780 qs_energies 11 4.9 0.001 0.001 59.486 59.491 scf_env_do_scf 11 5.9 0.000 0.001 51.405 51.405 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.065 42.066 velocity_verlet 10 3.0 0.001 0.001 34.900 34.916 dbcsr_multiply_generic 2055 12.4 0.108 0.113 29.332 29.705 qs_scf_new_mos 99 7.5 0.001 0.001 26.386 26.506 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.386 26.505 ot_scf_mini 99 9.5 0.002 0.003 25.125 25.256 multiply_cannon 2055 13.4 0.211 0.222 22.507 23.690 multiply_cannon_loop 2055 14.4 0.620 0.637 21.222 22.551 ot_mini 99 10.5 0.001 0.001 14.214 14.351 rebuild_ks_matrix 110 8.3 0.000 0.000 12.801 12.985 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.801 12.985 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.256 11.419 mp_waitall_1 139946 16.5 7.796 10.832 7.796 10.832 multiply_cannon_multrec 16440 15.4 3.785 4.876 9.490 10.541 qs_ot_get_derivative 99 11.5 0.001 0.001 9.596 9.725 init_scf_loop 11 6.9 0.000 0.000 9.298 9.300 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.816 7.935 prepare_preconditioner 11 7.9 0.000 0.000 7.474 7.495 make_preconditioner 11 8.9 0.000 0.000 7.474 7.495 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.777 7.140 sum_up_and_integrate 110 10.3 0.060 0.061 6.834 6.849 integrate_v_rspace 110 11.3 0.003 0.004 6.773 6.789 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.333 6.348 calculate_rho_elec 110 8.6 0.059 0.059 6.333 6.348 qs_ot_get_p 110 10.4 0.001 0.001 5.955 6.127 dbcsr_mm_accdrv_process 34862 16.1 4.495 5.100 5.562 5.703 init_scf_run 11 5.9 0.000 0.001 5.643 5.643 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.643 5.643 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.100 5.607 apply_single 110 13.6 0.000 0.000 5.099 5.607 make_m2s 4110 13.4 0.050 0.051 4.475 4.824 make_images 4110 14.4 0.394 0.511 4.359 4.708 density_rs2pw 110 9.6 0.004 0.005 3.355 4.596 ot_diis_step 99 11.5 0.010 0.011 4.571 4.571 multiply_cannon_sync_h2d 16440 15.4 3.626 4.242 3.626 4.242 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.204 4.211 rs_pw_transfer 902 11.9 0.010 0.011 2.959 4.178 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.300 3.946 pw_transfer 1331 11.6 0.066 0.075 3.862 3.873 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.838 3.838 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.755 3.767 grid_integrate_task_list 110 12.3 3.163 3.377 3.163 3.377 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.255 3.266 cp_fm_diag_elpa_base 48 14.0 3.182 3.213 3.253 3.263 fft_wrap_pw1pw2_140 451 13.1 0.636 0.645 3.189 3.204 cp_fm_cholesky_invert 11 10.9 3.178 3.185 3.178 3.185 wfi_extrapolate 11 7.9 0.001 0.001 3.151 3.151 make_images_data 4110 15.4 0.043 0.047 2.612 3.117 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.342 3.018 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.834 2.836 fft3d_ps 1111 14.6 1.093 1.101 2.578 2.586 calculate_dm_sparse 110 9.5 0.001 0.001 2.543 2.578 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.497 2.568 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.913 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.478 2.539 mp_waitany 17072 13.8 1.232 2.516 1.232 2.516 potential_pw2rs 110 12.3 0.011 0.012 2.501 2.512 grid_collocate_task_list 110 9.6 2.120 2.509 2.120 2.509 mp_sum_l 6594 12.7 1.731 2.491 1.731 2.491 mp_irecv_dv 48980 15.7 0.843 2.429 0.843 2.429 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.394 mp_alltoall_d11v 2046 13.8 1.971 2.274 1.971 2.274 jit_kernel_multiply 12 16.5 0.679 2.244 0.679 2.244 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.181 0.921 2.148 dbcsr_complete_redistribute 325 12.2 0.322 0.343 1.531 1.982 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.966 1.967 cp_fm_cholesky_decompose 22 10.9 1.786 1.809 1.786 1.809 mp_allgather_i34 2055 14.4 0.611 1.793 0.611 1.793 cp_fm_upper_to_full 70 14.2 1.388 1.785 1.388 1.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.743 1.764 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.035 1.493 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.476 1.489 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.358 1.479 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.344 1.354 mp_alltoall_z22v 1111 16.6 1.274 1.298 1.274 1.298 rs_gather_matrices 110 12.3 0.232 0.256 1.024 1.275 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.587000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.272727, yerr=7.793831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 725.569536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.037 78.537 78.538 qs_mol_dyn_low 1 2.0 0.003 0.003 78.212 78.222 qs_forces 11 3.9 0.003 0.003 78.141 78.141 qs_energies 11 4.9 0.002 0.009 74.547 74.551 scf_env_do_scf 11 5.9 0.001 0.001 65.371 65.374 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.788 50.789 velocity_verlet 10 3.0 0.001 0.001 41.801 41.803 qs_scf_new_mos 99 7.5 0.001 0.001 33.191 33.300 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.190 33.299 dbcsr_multiply_generic 2055 12.4 0.115 0.120 32.635 32.870 ot_scf_mini 99 9.5 0.003 0.005 31.436 31.540 multiply_cannon 2055 13.4 0.244 0.269 23.563 24.571 multiply_cannon_loop 2055 14.4 0.893 0.921 21.743 22.361 ot_mini 99 10.5 0.001 0.001 16.282 16.413 init_scf_loop 11 6.9 0.001 0.005 14.530 14.531 multiply_cannon_multrec 24660 15.4 4.179 6.594 12.829 13.926 rebuild_ks_matrix 110 8.3 0.000 0.000 13.609 13.741 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.609 13.741 prepare_preconditioner 11 7.9 0.000 0.001 12.624 12.639 make_preconditioner 11 8.9 0.000 0.001 12.624 12.639 make_full_inverse_cholesky 11 9.9 0.000 0.001 10.739 12.300 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.068 12.184 qs_ot_get_derivative 99 11.5 0.001 0.001 11.924 12.036 qs_ot_get_p 110 10.4 0.001 0.002 9.748 9.891 dbcsr_mm_accdrv_process 52304 16.0 7.072 8.653 8.503 9.435 mp_waitall_1 121746 16.5 6.312 8.610 6.312 8.610 sum_up_and_integrate 110 10.3 0.067 0.071 7.637 7.649 integrate_v_rspace 110 11.3 0.003 0.003 7.570 7.583 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.555 7.567 calculate_rho_elec 110 8.6 0.078 0.082 7.554 7.566 make_m2s 4110 13.4 0.060 0.061 7.019 7.507 make_images 4110 14.4 0.577 0.701 6.875 7.359 qs_ot_p2m_diag 48 11.0 0.055 0.064 7.270 7.287 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.621 6.623 init_scf_run 11 5.9 0.000 0.001 6.265 6.265 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.264 6.265 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.562 5.571 cp_fm_diag_elpa_base 48 14.0 5.347 5.426 5.557 5.567 cp_fm_upper_to_full 70 14.2 3.569 5.190 3.569 5.190 density_rs2pw 110 9.6 0.004 0.004 4.285 4.804 pw_transfer 1331 11.6 0.066 0.076 4.588 4.628 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.481 4.526 rs_pw_transfer 902 11.9 0.010 0.011 3.975 4.522 ot_diis_step 99 11.5 0.011 0.011 4.293 4.295 make_images_data 4110 15.4 0.046 0.049 3.651 4.289 dbcsr_complete_redistribute 325 12.2 0.409 0.461 3.072 4.287 cp_fm_cholesky_invert 11 10.9 4.262 4.273 4.262 4.273 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.171 4.261 apply_single 110 13.6 0.000 0.000 4.171 4.261 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 4.075 4.137 hybrid_alltoall_any 4261 16.3 0.121 0.459 3.062 4.121 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.668 3.731 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.377 3.586 fft_wrap_pw1pw2_140 451 13.1 0.664 0.688 3.513 3.563 grid_integrate_task_list 110 12.3 3.244 3.470 3.244 3.470 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.542 3.429 wfi_extrapolate 11 7.9 0.001 0.001 3.411 3.412 multiply_cannon_sync_h2d 24660 15.4 3.126 3.319 3.126 3.319 fft3d_ps 1111 14.6 1.089 1.119 3.247 3.269 potential_pw2rs 110 12.3 0.013 0.014 3.203 3.219 calculate_dm_sparse 110 9.5 0.001 0.001 3.186 3.204 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.070 3.073 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.875 3.060 mp_alltoall_i22 605 13.7 1.813 3.058 1.813 3.058 calculate_first_density_matrix 1 7.0 0.000 0.003 2.736 2.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.608 2.648 cp_fm_cholesky_decompose 22 10.9 2.583 2.618 2.583 2.618 mp_alltoall_d11v 2046 13.8 2.288 2.534 2.288 2.534 grid_collocate_task_list 110 9.6 2.224 2.479 2.224 2.479 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.304 2.305 jit_kernel_multiply 10 15.8 1.098 2.066 1.098 2.066 mp_alltoall_z22v 1111 16.6 1.943 2.015 1.943 2.015 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.965 1.992 mp_sum_l 6594 12.7 1.375 1.915 1.375 1.915 make_images_sizes 4110 15.4 0.005 0.005 1.127 1.885 mp_alltoall_i44 4110 16.4 1.123 1.881 1.123 1.881 mp_waitany 13376 13.8 1.388 1.869 1.388 1.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.804 mp_allgather_i34 2055 14.4 0.919 1.747 0.919 1.747 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.689 1.706 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.701 acc_transpose_blocks 24660 15.4 0.104 0.106 1.595 1.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=78.538000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.636364, yerr=7.607087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 826.028032E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 55.957 55.957 qs_mol_dyn_low 1 2.0 0.003 0.003 55.708 55.717 qs_forces 11 3.9 0.003 0.003 55.645 55.646 qs_energies 11 4.9 0.001 0.001 51.916 51.922 scf_env_do_scf 11 5.9 0.000 0.001 43.625 43.625 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.720 35.721 velocity_verlet 10 3.0 0.001 0.001 31.571 31.574 dbcsr_multiply_generic 2055 12.4 0.103 0.108 22.810 22.981 qs_scf_new_mos 99 7.5 0.001 0.001 20.496 20.550 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.496 20.549 ot_scf_mini 99 9.5 0.002 0.002 19.256 19.301 multiply_cannon 2055 13.4 0.244 0.253 17.362 18.433 multiply_cannon_loop 2055 14.4 0.323 0.336 16.026 16.388 rebuild_ks_matrix 110 8.3 0.000 0.000 12.011 12.044 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.010 12.043 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.641 10.671 ot_mini 99 10.5 0.001 0.001 10.465 10.510 multiply_cannon_multrec 8220 15.4 3.206 4.883 7.406 8.644 mp_waitall_1 103326 16.6 6.170 7.862 6.170 7.862 init_scf_loop 11 6.9 0.000 0.000 7.856 7.858 qs_ot_get_derivative 99 11.5 0.001 0.001 6.599 6.643 sum_up_and_integrate 110 10.3 0.080 0.081 6.613 6.625 integrate_v_rspace 110 11.3 0.003 0.003 6.532 6.544 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.424 6.440 calculate_rho_elec 110 8.6 0.115 0.116 6.424 6.439 prepare_preconditioner 11 7.9 0.000 0.000 6.189 6.193 make_preconditioner 11 8.9 0.000 0.000 6.189 6.193 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.772 5.844 init_scf_run 11 5.9 0.000 0.001 5.157 5.157 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.156 5.157 dbcsr_mm_accdrv_process 17442 15.9 2.824 3.936 4.070 4.964 qs_ot_get_p 110 10.4 0.001 0.001 4.864 4.919 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.136 4.656 make_m2s 4110 13.4 0.039 0.040 4.175 4.418 make_images 4110 14.4 0.637 0.696 4.044 4.286 pw_transfer 1331 11.6 0.067 0.073 3.989 4.015 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.881 3.911 ot_diis_step 99 11.5 0.012 0.014 3.843 3.843 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.751 3.802 apply_single 110 13.6 0.000 0.000 3.751 3.802 grid_integrate_task_list 110 12.3 3.370 3.552 3.370 3.552 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.511 3.516 density_rs2pw 110 9.6 0.004 0.004 3.123 3.496 fft_wrap_pw1pw2_140 451 13.1 0.831 0.842 3.346 3.385 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.200 3.202 cp_fm_cholesky_invert 11 10.9 3.060 3.065 3.060 3.065 multiply_cannon_sync_h2d 8220 15.4 2.898 3.047 2.898 3.047 wfi_extrapolate 11 7.9 0.001 0.001 2.803 2.803 make_images_data 4110 15.4 0.038 0.043 2.378 2.792 hybrid_alltoall_any 4261 16.3 0.201 0.867 2.290 2.701 rs_pw_transfer 902 11.9 0.010 0.010 2.294 2.691 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.668 2.676 cp_fm_diag_elpa_base 48 14.0 2.612 2.637 2.666 2.674 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.657 2.658 grid_collocate_task_list 110 9.6 2.314 2.572 2.314 2.572 calculate_dm_sparse 110 9.5 0.001 0.001 2.460 2.500 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.495 2.496 fft3d_ps 1111 14.6 1.139 1.182 2.428 2.448 calculate_first_density_matrix 1 7.0 0.000 0.000 2.251 2.252 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.138 2.161 potential_pw2rs 110 12.3 0.015 0.016 2.135 2.139 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.987 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.914 1.939 mp_alltoall_d11v 2046 13.8 1.636 1.901 1.636 1.901 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.820 1.843 cp_fm_cholesky_decompose 22 10.9 1.717 1.740 1.717 1.740 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.709 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.497 1.625 mp_allgather_i34 2055 14.4 0.523 1.589 0.523 1.589 dbcsr_complete_redistribute 325 12.2 0.556 0.574 1.437 1.522 mp_waitany 9240 13.8 1.106 1.513 1.106 1.513 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.430 1.441 jit_kernel_multiply 8 15.8 0.935 1.361 0.935 1.361 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.484 1.346 qs_create_task_list 11 7.9 0.001 0.001 1.215 1.320 generate_qs_task_list 11 8.9 0.374 0.441 1.214 1.319 mp_irecv_dv 24056 15.7 0.460 1.307 0.460 1.307 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.796 1.214 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.167 0.811 1.207 rs_gather_matrices 110 12.3 0.325 0.369 0.938 1.152 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.109 1.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.957000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.181818, yerr=9.916178 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 1.357816E+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 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.056 88.814 88.815 qs_mol_dyn_low 1 2.0 0.003 0.003 88.379 88.387 qs_forces 11 3.9 0.003 0.004 87.664 87.665 qs_energies 11 4.9 0.001 0.001 83.472 83.474 scf_env_do_scf 11 5.9 0.001 0.001 73.033 73.033 velocity_verlet 10 3.0 0.001 0.001 56.639 56.726 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.329 44.332 init_scf_loop 11 6.9 0.000 0.000 28.631 28.636 dbcsr_multiply_generic 2055 12.4 0.118 0.123 28.561 28.630 prepare_preconditioner 11 7.9 0.000 0.000 26.613 26.620 make_preconditioner 11 8.9 0.000 0.000 26.613 26.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.716 26.088 qs_scf_new_mos 99 7.5 0.001 0.001 25.993 26.025 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.992 26.024 ot_scf_mini 99 9.5 0.002 0.002 24.245 24.265 multiply_cannon 2055 13.4 0.352 0.384 21.622 22.439 multiply_cannon_loop 2055 14.4 0.342 0.346 19.854 20.174 cp_fm_upper_to_full 70 14.2 12.883 18.504 12.883 18.504 rebuild_ks_matrix 110 8.3 0.000 0.001 14.138 14.166 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.137 14.165 ot_mini 99 10.5 0.001 0.001 13.423 13.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.780 12.805 dbcsr_complete_redistribute 325 12.2 1.016 1.062 7.475 10.701 multiply_cannon_multrec 8220 15.4 4.435 4.599 9.870 10.003 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.449 9.693 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.882 9.097 qs_ot_get_derivative 99 11.5 0.001 0.001 8.952 8.974 mp_alltoall_i22 605 13.7 5.498 8.760 5.498 8.760 mp_waitall_1 84994 16.7 7.328 8.169 7.328 8.169 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.810 7.840 calculate_rho_elec 110 8.6 0.227 0.227 7.809 7.840 sum_up_and_integrate 110 10.3 0.150 0.151 7.381 7.394 integrate_v_rspace 110 11.3 0.004 0.004 7.231 7.245 init_scf_run 11 5.9 0.000 0.001 6.153 6.154 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.153 6.153 make_m2s 4110 13.4 0.042 0.043 5.259 5.742 qs_ot_get_p 110 10.4 0.001 0.001 5.683 5.708 dbcsr_mm_accdrv_process 11614 15.7 3.265 3.687 5.294 5.554 make_images 4110 14.4 0.882 0.933 5.070 5.552 cp_fm_cholesky_invert 11 10.9 5.293 5.298 5.293 5.298 pw_transfer 1331 11.6 0.075 0.075 5.096 5.101 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.582 5.052 apply_single 110 13.6 0.000 0.000 4.582 5.051 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.978 4.983 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.578 4.964 ot_diis_step 99 11.5 0.015 0.016 4.454 4.454 fft_wrap_pw1pw2_140 451 13.1 1.335 1.343 4.339 4.348 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.150 4.157 multiply_cannon_sync_h2d 8220 15.4 3.949 3.952 3.949 3.952 density_rs2pw 110 9.6 0.004 0.004 3.794 3.848 grid_integrate_task_list 110 12.3 3.685 3.751 3.685 3.751 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.704 3.705 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.677 3.678 hybrid_alltoall_any 4261 16.3 0.258 0.555 2.876 3.604 make_images_data 4110 15.4 0.041 0.044 2.851 3.457 wfi_extrapolate 11 7.9 0.001 0.001 3.330 3.330 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.863 3.302 calculate_dm_sparse 110 9.5 0.001 0.001 3.237 3.258 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.106 3.106 cp_fm_diag_elpa_base 48 14.0 2.556 2.759 3.104 3.104 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.008 3.012 fft3d_ps 1111 14.6 1.305 1.310 2.910 2.917 grid_collocate_task_list 110 9.6 2.677 2.701 2.677 2.701 calculate_first_density_matrix 1 7.0 0.000 0.000 2.697 2.700 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.518 2.530 potential_pw2rs 110 12.3 0.021 0.022 2.447 2.450 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.376 2.393 rs_pw_transfer 902 11.9 0.011 0.011 2.332 2.366 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.198 2.260 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.180 mp_alltoall_d11v 2046 13.8 1.985 2.068 1.985 2.068 cp_fm_cholesky_decompose 22 10.9 2.004 2.028 2.004 2.028 jit_kernel_multiply 10 15.3 1.825 2.009 1.825 2.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.959 1.961 qs_create_task_list 11 7.9 0.023 0.023 1.903 1.952 generate_qs_task_list 11 8.9 0.732 0.783 1.880 1.929 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.767 1.805 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.789 1.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.815000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1220.272727, yerr=46.622451 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 632.885248E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57456. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 2468810. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.028 217.918 217.919 qs_mol_dyn_low 1 2.0 0.003 0.003 217.441 217.453 qs_forces 11 3.9 0.005 0.005 217.356 217.357 qs_energies 11 4.9 0.001 0.001 211.671 211.686 scf_env_do_scf 11 5.9 0.001 0.001 187.958 187.962 scf_env_do_scf_inner_loop 118 6.6 0.006 0.025 166.910 166.912 dbcsr_multiply_generic 2527 12.6 0.179 0.183 132.810 133.336 velocity_verlet 10 3.0 0.001 0.002 126.470 126.471 qs_scf_new_mos 118 7.6 0.001 0.001 125.767 125.905 qs_scf_loop_do_ot 118 8.6 0.001 0.001 125.766 125.904 ot_scf_mini 118 9.6 0.003 0.003 119.068 119.207 multiply_cannon 2527 13.6 0.240 0.246 109.137 110.586 multiply_cannon_loop 2527 14.6 2.107 2.162 106.886 108.306 ot_mini 118 10.6 0.001 0.001 66.519 66.691 multiply_cannon_multrec 60648 15.6 33.673 35.752 41.967 44.235 qs_ot_get_derivative 118 11.6 0.001 0.001 41.450 41.615 rebuild_ks_matrix 129 8.3 0.001 0.001 34.375 34.619 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.018 34.374 34.618 mp_waitall_1 269254 16.5 28.836 32.055 28.836 32.055 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.859 31.093 qs_ot_get_p 129 10.4 0.001 0.001 30.421 30.676 multiply_cannon_sync_h2d 60648 15.6 27.856 29.443 27.856 29.443 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.602 25.362 apply_single 129 13.6 0.001 0.001 24.601 25.361 ot_diis_step 118 11.6 0.008 0.008 24.807 24.808 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.741 23.799 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.097 21.098 init_scf_loop 11 6.9 0.000 0.001 20.968 20.969 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.512 19.648 init_scf_run 11 5.9 0.000 0.001 19.550 19.551 scf_env_initial_rho_setup 11 6.9 0.001 0.001 19.550 19.550 multiply_cannon_metrocomm3 60648 15.6 0.114 0.121 15.976 18.410 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.034 18.066 cp_fm_diag_elpa_base 83 14.4 17.957 17.987 18.030 18.063 prepare_preconditioner 11 7.9 0.000 0.000 16.300 16.351 make_preconditioner 11 8.9 0.000 0.000 16.300 16.351 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.512 15.727 sum_up_and_integrate 129 10.3 0.090 0.109 14.650 14.668 integrate_v_rspace 129 11.3 0.004 0.005 14.560 14.580 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.327 14.468 calculate_rho_elec 129 8.7 0.047 0.065 14.326 14.467 make_m2s 5054 13.6 0.106 0.114 14.158 14.461 make_images 5054 14.6 0.406 0.427 13.978 14.288 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 13.704 13.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 12.842 12.942 density_rs2pw 129 9.7 0.006 0.007 7.497 10.904 calculate_first_density_matrix 1 7.0 0.000 0.002 10.109 10.114 acc_transpose_blocks 60648 15.6 0.296 0.307 9.921 10.019 rs_pw_transfer 1054 12.0 0.016 0.019 6.146 9.548 cp_fm_cholesky_invert 11 10.9 9.280 9.288 9.280 9.288 wfi_extrapolate 11 7.9 0.001 0.001 9.253 9.253 mp_sum_l 8010 12.9 8.226 9.085 8.226 9.085 calculate_dm_sparse 129 9.5 0.001 0.001 8.531 8.657 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.271 8.399 dbcsr_mm_accdrv_process 125468 16.2 3.152 3.297 7.843 8.398 acc_transpose_blocks_kernels 60648 16.6 0.694 0.730 8.322 8.367 pw_transfer 1559 11.6 0.075 0.092 8.131 8.328 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 7.926 8.119 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.870 7.979 make_images_data 5054 15.6 0.068 0.073 6.983 7.968 grid_integrate_task_list 129 12.3 7.105 7.654 7.105 7.654 jit_kernel_transpose 5 15.6 7.627 7.651 7.627 7.651 multiply_cannon_metrocomm1 60648 15.6 0.091 0.094 6.028 7.459 hybrid_alltoall_any 5240 16.5 0.295 2.298 6.102 7.323 fft_wrap_pw1pw2_140 527 13.2 1.290 1.334 6.866 7.073 mp_waitany 16140 13.9 2.808 6.203 2.808 6.203 fft3d_ps 1301 14.7 2.154 2.813 5.663 6.009 grid_collocate_task_list 129 9.7 4.701 5.987 4.701 5.987 mp_alltoall_d11v 2423 14.1 4.455 5.725 4.455 5.725 rs_pw_transfer_RS2PW_140 140 11.5 0.282 0.300 2.204 5.617 potential_pw2rs 129 12.3 0.009 0.011 4.880 4.910 cp_fm_cholesky_decompose 22 10.9 4.788 4.810 4.788 4.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=217.919000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.636364, yerr=6.342823 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.926272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5620351. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.094 195.103 195.104 qs_mol_dyn_low 1 2.0 0.003 0.004 194.489 194.511 qs_forces 11 3.9 0.005 0.016 194.383 194.385 qs_energies 11 4.9 0.004 0.013 187.454 187.467 scf_env_do_scf 11 5.9 0.001 0.002 170.464 170.474 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 136.770 136.773 velocity_verlet 10 3.0 0.006 0.015 121.744 121.746 dbcsr_multiply_generic 2507 12.6 0.189 0.194 97.722 99.040 qs_scf_new_mos 117 7.6 0.001 0.001 95.836 96.198 qs_scf_loop_do_ot 117 8.6 0.001 0.002 95.835 96.197 ot_scf_mini 117 9.6 0.004 0.005 91.011 91.408 multiply_cannon 2507 13.6 0.477 0.524 77.623 81.541 multiply_cannon_loop 2507 14.6 1.251 1.282 74.344 76.878 ot_mini 117 10.6 0.001 0.001 49.956 50.367 mp_waitall_1 214728 16.6 24.439 37.707 24.439 37.707 multiply_cannon_multrec 30084 15.6 22.246 26.356 31.891 36.295 rebuild_ks_matrix 128 8.3 0.001 0.001 33.930 34.407 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 33.929 34.406 init_scf_loop 11 6.9 0.002 0.006 33.602 33.603 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.464 30.910 prepare_preconditioner 11 7.9 0.000 0.000 29.110 29.168 make_preconditioner 11 8.9 0.001 0.006 29.109 29.168 qs_ot_get_derivative 117 11.6 0.001 0.002 28.011 28.419 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.785 28.331 multiply_cannon_metrocomm3 30084 15.6 0.091 0.097 15.387 28.169 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.020 23.149 apply_single 128 13.6 0.001 0.001 22.020 23.149 qs_ot_get_p 128 10.4 0.001 0.002 22.704 23.145 multiply_cannon_sync_h2d 30084 15.6 19.529 21.868 19.529 21.868 ot_diis_step 117 11.6 0.014 0.015 21.780 21.782 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.893 17.930 cp_fm_cholesky_invert 11 10.9 17.033 17.045 17.033 17.045 cp_dbcsr_syevd 83 12.4 0.006 0.006 16.719 16.720 make_m2s 5014 13.6 0.093 0.098 14.447 15.953 make_images 5014 14.6 1.163 1.347 14.234 15.737 sum_up_and_integrate 128 10.3 0.116 0.134 15.128 15.162 integrate_v_rspace 128 11.3 0.004 0.004 15.012 15.051 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.563 14.599 calculate_rho_elec 128 8.7 0.089 0.106 14.563 14.598 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.415 13.441 cp_fm_diag_elpa_base 83 14.4 13.138 13.266 13.408 13.437 init_scf_run 11 5.9 0.001 0.008 11.826 11.829 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.825 11.829 qs_ot_get_derivative_diag 77 12.4 0.003 0.008 11.037 11.307 make_images_data 5014 15.6 0.064 0.073 8.733 10.691 multiply_cannon_metrocomm4 27577 15.6 0.105 0.119 3.739 10.436 density_rs2pw 128 9.7 0.007 0.008 7.781 10.274 mp_irecv_dv 69486 16.3 3.537 10.050 3.537 10.050 hybrid_alltoall_any 5200 16.5 0.344 1.509 7.430 9.798 dbcsr_mm_accdrv_process 62242 16.2 4.552 5.257 9.100 9.685 pw_transfer 1547 11.6 0.087 0.095 9.123 9.181 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 8.896 8.958 rs_pw_transfer 1046 11.9 0.014 0.016 6.330 8.909 wfi_extrapolate 11 7.9 0.001 0.001 8.575 8.575 fft_wrap_pw1pw2_140 523 13.2 1.341 1.393 7.786 7.862 grid_integrate_task_list 128 12.3 7.139 7.547 7.139 7.547 cp_fm_cholesky_decompose 22 10.9 7.177 7.257 7.177 7.257 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.180 6.922 calculate_dm_sparse 128 9.5 0.001 0.001 6.484 6.614 fft3d_ps 1291 14.7 2.812 2.978 6.195 6.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.212 6.220 grid_collocate_task_list 128 9.7 4.783 5.966 4.783 5.966 mp_sum_l 7950 12.9 3.877 5.750 3.877 5.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.328 5.498 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.338 5.406 mp_waitany 11748 13.9 2.746 5.376 2.746 5.376 mp_alltoall_d11v 2415 14.1 4.390 5.373 4.390 5.373 potential_pw2rs 128 12.3 0.016 0.019 5.310 5.332 mp_allgather_i34 2507 14.6 1.851 5.152 1.851 5.152 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.378 2.295 4.855 dbcsr_complete_redistribute 395 12.7 0.775 0.861 3.210 4.083 mp_sum_d 4465 12.1 2.664 4.068 2.664 4.068 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.104000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.272727, yerr=1.285649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 941.334528E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58141. MP_Allreduce 11095 999. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.037 180.314 180.315 qs_mol_dyn_low 1 2.0 0.003 0.003 179.578 179.591 qs_forces 11 3.9 0.004 0.005 179.477 179.480 qs_energies 11 4.9 0.001 0.001 172.746 172.754 scf_env_do_scf 11 5.9 0.001 0.001 157.093 157.093 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 121.129 121.129 velocity_verlet 10 3.0 0.001 0.001 114.425 114.427 dbcsr_multiply_generic 2485 12.5 0.177 0.184 81.994 83.103 qs_scf_new_mos 116 7.6 0.001 0.001 82.259 82.571 qs_scf_loop_do_ot 116 8.6 0.001 0.001 82.258 82.570 ot_scf_mini 116 9.6 0.003 0.004 78.134 78.502 multiply_cannon 2485 13.5 0.495 0.521 61.750 66.375 multiply_cannon_loop 2485 14.5 0.856 0.890 58.610 61.339 ot_mini 116 10.6 0.001 0.001 42.912 43.285 init_scf_loop 11 6.9 0.000 0.000 35.856 35.857 mp_waitall_1 169034 16.6 25.995 35.430 25.995 35.430 prepare_preconditioner 11 7.9 0.000 0.000 31.762 31.811 make_preconditioner 11 8.9 0.000 0.000 31.762 31.811 rebuild_ks_matrix 127 8.3 0.001 0.001 31.177 31.652 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 31.177 31.652 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.284 30.776 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.998 28.430 multiply_cannon_metrocomm3 19880 15.5 0.058 0.061 15.766 25.136 multiply_cannon_multrec 19880 15.5 13.192 16.305 21.785 24.903 qs_ot_get_derivative 116 11.6 0.001 0.002 23.020 23.392 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.009 21.174 apply_single 127 13.6 0.001 0.001 20.008 21.174 qs_ot_get_p 127 10.4 0.001 0.001 20.505 20.942 ot_diis_step 116 11.6 0.017 0.020 19.787 19.788 qs_ot_p2m_diag 82 11.4 0.263 0.269 16.174 16.188 make_m2s 4970 13.5 0.082 0.087 15.128 16.024 make_images 4970 14.5 1.170 1.260 14.895 15.788 multiply_cannon_sync_h2d 19880 15.5 14.089 15.389 14.089 15.389 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.109 15.110 cp_fm_cholesky_invert 11 10.9 14.924 14.934 14.924 14.934 sum_up_and_integrate 127 10.3 0.130 0.145 14.893 14.920 integrate_v_rspace 127 11.3 0.004 0.004 14.763 14.794 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.734 14.761 calculate_rho_elec 127 8.7 0.131 0.146 14.733 14.760 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.776 11.801 cp_fm_diag_elpa_base 82 14.4 11.349 11.517 11.771 11.797 make_images_data 4970 15.5 0.058 0.067 9.436 10.754 init_scf_run 11 5.9 0.000 0.001 10.571 10.572 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.571 10.571 hybrid_alltoall_any 5155 16.4 0.431 1.967 8.151 9.833 density_rs2pw 127 9.7 0.006 0.007 7.525 9.591 multiply_cannon_metrocomm4 17395 15.5 0.062 0.072 3.470 9.471 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.071 9.345 pw_transfer 1535 11.6 0.086 0.104 9.166 9.279 mp_irecv_dv 49801 16.2 3.350 9.228 3.350 9.228 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.942 9.061 dbcsr_mm_accdrv_process 41158 16.2 4.359 4.879 8.059 8.170 fft_wrap_pw1pw2_140 519 13.2 1.404 1.438 7.851 7.986 grid_integrate_task_list 127 12.3 7.241 7.704 7.241 7.704 cp_fm_cholesky_decompose 22 10.9 7.646 7.686 7.646 7.686 rs_pw_transfer 1038 11.9 0.013 0.014 5.528 7.585 wfi_extrapolate 11 7.9 0.001 0.001 7.576 7.577 cp_fm_upper_to_full 104 14.8 5.675 7.309 5.675 7.309 dbcsr_complete_redistribute 393 12.7 1.168 1.190 4.820 6.586 fft3d_ps 1281 14.7 2.719 2.943 6.081 6.165 grid_collocate_task_list 127 9.7 4.938 5.837 4.938 5.837 calculate_dm_sparse 127 9.5 0.001 0.001 5.724 5.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.446 5.452 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.582 5.350 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.578 5.258 mp_alltoall_d11v 2401 14.1 4.570 5.244 4.570 5.244 potential_pw2rs 127 12.3 0.021 0.024 4.977 5.007 mp_allgather_i34 2485 14.5 1.673 4.873 1.673 4.873 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.554 4.692 mp_sum_l 7884 12.9 3.318 4.656 3.318 4.656 mp_waitany 11660 13.9 2.403 4.502 2.403 4.502 transfer_fm_to_dbcsr 11 9.9 0.020 0.027 2.459 4.205 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.084 4.126 mp_alltoall_i22 712 14.1 2.061 4.043 2.061 4.043 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.350 1.937 4.008 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.797 3.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.676 3.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.315000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.818182, yerr=9.589940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 1.174491E+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 4020 57954. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.117 194.336 194.337 qs_mol_dyn_low 1 2.0 0.003 0.004 193.514 193.597 qs_forces 11 3.9 0.004 0.004 193.396 193.401 qs_energies 11 4.9 0.005 0.031 186.130 186.141 scf_env_do_scf 11 5.9 0.001 0.003 169.016 169.023 velocity_verlet 10 3.0 0.001 0.002 127.542 127.545 scf_env_do_scf_inner_loop 117 6.6 0.003 0.011 121.769 121.771 qs_scf_new_mos 117 7.6 0.001 0.001 83.189 83.532 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.189 83.531 dbcsr_multiply_generic 2507 12.6 0.186 0.194 80.413 81.258 ot_scf_mini 117 9.6 0.003 0.004 78.490 78.804 multiply_cannon 2507 13.6 0.555 0.600 54.608 57.270 multiply_cannon_loop 2507 14.6 1.193 1.220 50.824 52.922 init_scf_loop 11 6.9 0.003 0.020 47.121 47.122 ot_mini 117 10.6 0.001 0.001 42.972 43.262 prepare_preconditioner 11 7.9 0.000 0.000 42.929 42.947 make_preconditioner 11 8.9 0.000 0.001 42.929 42.947 make_full_inverse_cholesky 11 9.9 0.012 0.024 36.563 41.547 multiply_cannon_multrec 30084 15.6 14.279 19.492 26.186 30.846 rebuild_ks_matrix 128 8.3 0.001 0.001 30.447 30.827 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.034 30.446 30.827 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.422 27.767 mp_waitall_1 147882 16.7 17.733 27.185 17.733 27.185 qs_ot_get_derivative 117 11.6 0.001 0.002 23.262 23.572 make_m2s 5014 13.6 0.098 0.101 21.316 22.260 make_images 5014 14.6 1.958 2.290 21.009 21.955 qs_ot_get_p 128 10.4 0.001 0.002 21.627 21.947 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.043 19.688 apply_single 128 13.6 0.001 0.001 19.042 19.688 ot_diis_step 117 11.6 0.017 0.018 19.566 19.569 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.247 17.304 cp_fm_upper_to_full 105 14.8 11.497 16.906 11.497 16.906 cp_fm_cholesky_invert 11 10.9 16.565 16.575 16.565 16.575 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.941 15.943 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.286 15.329 calculate_rho_elec 128 8.7 0.176 0.192 15.286 15.328 sum_up_and_integrate 128 10.3 0.140 0.152 15.038 15.065 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.143 14.996 integrate_v_rspace 128 11.3 0.004 0.004 14.898 14.929 make_images_data 5014 15.6 0.062 0.068 11.519 13.260 multiply_cannon_sync_h2d 30084 15.6 11.693 13.075 11.693 13.075 dbcsr_complete_redistribute 395 12.7 1.506 1.644 9.048 12.842 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.693 12.703 cp_fm_diag_elpa_base 83 14.4 11.688 12.007 12.685 12.696 hybrid_alltoall_any 5200 16.5 0.526 2.221 10.109 12.057 dbcsr_mm_accdrv_process 62264 16.2 7.266 8.073 11.480 12.038 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.633 11.417 init_scf_run 11 5.9 0.000 0.001 11.131 11.132 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.131 11.132 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.762 9.978 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.343 9.962 pw_transfer 1547 11.6 0.087 0.100 9.440 9.527 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.214 9.304 mp_alltoall_i22 716 14.1 5.586 9.249 5.586 9.249 density_rs2pw 128 9.7 0.006 0.006 7.270 8.528 fft_wrap_pw1pw2_140 523 13.2 1.558 1.594 8.139 8.239 grid_integrate_task_list 128 12.3 7.480 8.001 7.480 8.001 cp_fm_cholesky_decompose 22 10.9 7.848 7.942 7.848 7.942 wfi_extrapolate 11 7.9 0.001 0.001 7.877 7.877 multiply_cannon_metrocomm4 25070 15.6 0.080 0.092 2.744 6.992 mp_irecv_dv 76098 16.2 2.597 6.730 2.597 6.730 calculate_dm_sparse 128 9.5 0.001 0.001 6.389 6.472 mp_alltoall_d11v 2415 14.1 5.495 6.331 5.495 6.331 rs_pw_transfer 1046 11.9 0.013 0.015 4.927 6.275 fft3d_ps 1291 14.7 2.840 2.933 6.137 6.204 grid_collocate_task_list 128 9.7 5.164 6.001 5.164 6.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.496 5.565 potential_pw2rs 128 12.3 0.023 0.023 4.757 4.776 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.499 4.594 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.364 4.440 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.411 4.411 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.216 4.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=194.337000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1100.636364, yerr=28.362179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.470214E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 180.706 180.707 qs_mol_dyn_low 1 2.0 0.003 0.003 180.263 180.275 qs_forces 11 3.9 0.004 0.004 180.122 180.130 qs_energies 11 4.9 0.001 0.001 172.440 172.447 scf_env_do_scf 11 5.9 0.001 0.001 154.731 154.732 velocity_verlet 10 3.0 0.001 0.001 117.722 117.727 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.592 117.593 qs_scf_new_mos 117 7.6 0.001 0.001 78.088 78.221 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.088 78.220 dbcsr_multiply_generic 2507 12.6 0.180 0.193 75.058 75.418 ot_scf_mini 117 9.6 0.003 0.004 73.572 73.642 multiply_cannon 2507 13.6 0.583 0.624 54.441 59.190 multiply_cannon_loop 2507 14.6 0.447 0.457 49.798 50.569 ot_mini 117 10.6 0.001 0.001 40.136 40.204 init_scf_loop 11 6.9 0.000 0.000 36.985 36.986 mp_waitall_1 125778 16.7 26.555 33.701 26.555 33.701 prepare_preconditioner 11 7.9 0.000 0.000 32.968 32.999 make_preconditioner 11 8.9 0.000 0.000 32.968 32.999 make_full_inverse_cholesky 11 9.9 0.018 0.027 30.877 31.148 rebuild_ks_matrix 128 8.3 0.001 0.001 30.442 30.574 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.441 30.574 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.630 27.751 multiply_cannon_multrec 10028 15.6 10.383 14.561 17.682 20.626 qs_ot_get_p 128 10.4 0.001 0.001 20.407 20.470 qs_ot_get_derivative 117 11.6 0.001 0.002 20.116 20.191 ot_diis_step 117 11.6 0.020 0.020 19.956 19.956 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.531 19.741 apply_single 128 13.6 0.001 0.001 19.531 19.740 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.230 19.402 make_m2s 5014 13.6 0.068 0.071 16.879 19.285 make_images 5014 14.6 2.320 2.785 16.570 18.978 cp_fm_cholesky_invert 11 10.9 18.916 18.922 18.916 18.922 qs_ot_p2m_diag 83 11.4 0.496 0.501 16.442 16.460 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.873 15.898 calculate_rho_elec 128 8.7 0.259 0.269 15.872 15.897 sum_up_and_integrate 128 10.3 0.181 0.191 15.393 15.444 integrate_v_rspace 128 11.3 0.004 0.004 15.211 15.270 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.259 15.262 make_images_data 5014 15.6 0.051 0.060 10.281 12.880 hybrid_alltoall_any 5200 16.5 0.814 3.815 10.074 12.495 multiply_cannon_sync_h2d 10028 15.6 11.613 12.260 11.613 12.260 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.050 12.059 cp_fm_diag_elpa_base 83 14.4 11.793 11.885 12.045 12.054 init_scf_run 11 5.9 0.000 0.001 10.835 10.835 scf_env_initial_rho_setup 11 6.9 0.018 0.023 10.835 10.835 pw_transfer 1547 11.6 0.086 0.096 9.895 9.919 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.669 9.697 fft_wrap_pw1pw2_140 523 13.2 1.911 1.954 8.462 8.499 cp_fm_cholesky_decompose 22 10.9 8.338 8.477 8.338 8.477 grid_integrate_task_list 128 12.3 7.741 8.267 7.741 8.267 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.143 8.196 density_rs2pw 128 9.7 0.006 0.006 7.279 8.173 wfi_extrapolate 11 7.9 0.001 0.001 7.769 7.770 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.643 7.734 dbcsr_mm_accdrv_process 20762 16.1 2.599 3.456 6.923 7.590 mp_allgather_i34 2507 14.6 2.854 7.371 2.854 7.371 mp_alltoall_d11v 2415 14.1 5.329 6.639 5.329 6.639 calculate_dm_sparse 128 9.5 0.001 0.001 6.176 6.248 fft3d_ps 1291 14.7 2.764 2.829 6.147 6.185 grid_collocate_task_list 128 9.7 5.470 6.147 5.470 6.147 dbcsr_complete_redistribute 395 12.7 2.118 2.152 5.291 5.713 rs_pw_transfer 1046 11.9 0.012 0.013 4.480 5.412 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.275 5.310 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.291 5.292 potential_pw2rs 128 12.3 0.027 0.027 4.709 4.727 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.146 4.168 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.850 4.010 mp_irecv_dv 28860 15.9 1.815 3.943 1.815 3.943 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.555 3.877 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.696 3.798 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.740 3.769 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.401 3.755 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.723 3.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=180.707000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1375.727273, yerr=37.251213 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.172282E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11179 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.045 299.669 299.670 qs_mol_dyn_low 1 2.0 0.003 0.003 299.113 299.125 qs_forces 11 3.9 0.005 0.005 298.989 298.993 qs_energies 11 4.9 0.001 0.001 289.564 289.575 scf_env_do_scf 11 5.9 0.001 0.001 267.176 267.183 velocity_verlet 10 3.0 0.001 0.001 216.297 216.309 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 141.837 141.839 init_scf_loop 11 6.9 0.000 0.000 125.083 125.086 prepare_preconditioner 11 7.9 0.000 0.000 120.028 120.054 make_preconditioner 11 8.9 0.000 0.000 120.028 120.054 make_full_inverse_cholesky 11 9.9 0.037 0.039 96.033 117.152 qs_scf_new_mos 118 7.6 0.001 0.001 90.498 90.630 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.497 90.629 ot_scf_mini 118 9.6 0.004 0.004 85.684 85.716 dbcsr_multiply_generic 2535 12.6 0.212 0.224 83.399 83.943 cp_fm_upper_to_full 106 14.8 53.709 76.687 53.709 76.687 multiply_cannon 2535 13.6 0.704 0.759 58.736 59.217 multiply_cannon_loop 2535 14.6 0.478 0.491 55.181 56.437 ot_mini 118 10.6 0.001 0.001 44.293 44.322 dbcsr_complete_redistribute 397 12.7 3.982 4.016 29.658 42.845 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.309 39.506 rebuild_ks_matrix 129 8.3 0.001 0.001 38.665 38.705 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 38.664 38.705 transfer_fm_to_dbcsr 11 9.9 0.029 0.034 23.949 36.961 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.622 35.661 mp_alltoall_i22 720 14.1 21.753 35.072 21.753 35.072 cp_fm_cholesky_invert 11 10.9 33.237 33.243 33.237 33.243 mp_waitall_1 104820 16.8 28.236 31.955 28.236 31.955 qs_ot_get_p 129 10.4 0.001 0.001 25.832 25.845 qs_ot_get_derivative 118 11.6 0.002 0.002 24.183 24.216 qs_ot_p2m_diag 84 11.4 0.889 0.894 21.785 21.814 make_m2s 5070 13.6 0.077 0.080 20.087 20.913 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.816 20.850 calculate_rho_elec 129 8.7 0.485 0.486 20.816 20.850 make_images 5070 14.6 3.805 3.950 19.603 20.433 ot_diis_step 118 11.6 0.022 0.023 20.084 20.084 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.052 20.053 sum_up_and_integrate 129 10.3 0.323 0.325 19.850 19.926 multiply_cannon_metrocomm3 10140 15.6 0.023 0.025 18.774 19.908 integrate_v_rspace 129 11.3 0.004 0.004 19.527 19.603 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.262 19.456 apply_single 129 13.6 0.001 0.001 19.262 19.456 multiply_cannon_multrec 10140 15.6 10.592 12.381 18.032 18.096 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.817 16.818 cp_fm_diag_elpa_base 84 14.4 12.366 14.026 16.813 16.813 multiply_cannon_sync_h2d 10140 15.6 15.783 15.789 15.783 15.789 hybrid_alltoall_any 5257 16.5 1.309 3.063 10.856 12.691 init_scf_run 11 5.9 0.000 0.001 12.357 12.358 scf_env_initial_rho_setup 11 6.9 0.031 0.045 12.357 12.357 make_images_data 5070 15.6 0.059 0.064 10.698 12.317 pw_transfer 1559 11.6 0.093 0.096 11.557 11.563 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.320 11.327 mp_alltoall_d11v 2429 14.1 8.569 10.231 8.569 10.231 fft_wrap_pw1pw2_140 527 13.2 3.096 3.132 10.046 10.053 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.501 9.532 wfi_extrapolate 11 7.9 0.001 0.001 9.201 9.201 dbcsr_mm_accdrv_process 20958 16.1 3.828 5.697 7.200 9.027 cp_fm_cholesky_decompose 22 10.9 8.925 8.944 8.925 8.944 grid_integrate_task_list 129 12.3 8.656 8.845 8.656 8.845 density_rs2pw 129 9.7 0.005 0.006 8.108 8.266 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.980 7.981 calculate_dm_sparse 129 9.5 0.001 0.001 6.666 6.788 grid_collocate_task_list 129 9.7 6.491 6.566 6.491 6.566 rs_scatter_matrices 140 9.7 3.694 4.803 6.192 6.421 fft3d_ps 1301 14.7 2.868 2.875 6.360 6.384 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.303 6.349 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.020 6.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.670000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2884.181818, yerr=160.627643 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 1.259663E+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 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.118 85.551 85.562 qs_energies 1 2.0 0.000 0.000 84.919 84.929 ls_scf 1 3.0 0.000 0.000 83.980 83.991 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.780 72.946 multiply_cannon 111 7.7 0.017 0.020 55.967 57.352 multiply_cannon_loop 111 8.7 0.211 0.226 52.546 54.158 ls_scf_main 1 4.0 0.000 0.000 52.531 52.539 density_matrix_trs4 2 5.0 0.002 0.003 46.930 46.997 ls_scf_init_scf 1 4.0 0.000 0.000 28.397 28.401 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.215 27.263 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.055 25.071 mp_waitall_1 11031 10.9 22.587 24.988 22.587 24.988 multiply_cannon_multrec 2664 9.7 8.194 8.949 15.591 17.364 multiply_cannon_sync_h2d 2664 9.7 13.542 14.787 13.542 14.787 make_m2s 222 7.7 0.009 0.012 13.188 13.705 make_images 222 8.7 0.101 0.112 13.166 13.684 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.690 12.172 make_images_data 222 9.7 0.005 0.006 7.719 8.338 hybrid_alltoall_any 227 10.6 0.218 1.836 6.655 8.272 dbcsr_mm_accdrv_process 4760 10.4 0.501 0.639 7.014 8.039 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.515 8.027 dbcsr_mm_accdrv_process_sort 4760 11.4 6.309 7.218 6.309 7.218 calculate_norms 4752 9.8 5.536 6.169 5.536 6.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.045 5.221 mp_sum_l 887 5.1 3.101 4.887 3.101 4.887 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.060 4.373 mp_irecv_dv 6231 10.9 2.043 4.338 2.043 4.338 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.305 3.665 make_images_sizes 222 9.7 0.000 0.000 0.793 3.616 mp_alltoall_i44 222 10.7 0.793 3.616 0.793 3.616 arnoldi_extremal 4 6.8 0.000 0.001 3.261 3.286 arnoldi_normal_ev 4 7.8 0.001 0.003 3.261 3.285 build_subspace 16 8.4 0.009 0.012 3.172 3.175 ls_scf_post 1 4.0 0.000 0.000 3.052 3.063 ls_scf_store_result 1 5.0 0.000 0.000 2.858 2.906 dbcsr_special_finalize 555 9.7 0.005 0.006 2.314 2.812 dbcsr_merge_single_wm 555 10.7 0.450 0.591 2.306 2.802 make_images_pack 222 9.7 2.213 2.637 2.215 2.639 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.325 2.565 dbcsr_sort_data 658 11.4 2.113 2.524 2.113 2.524 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.455 2.066 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.358 2.426 buffer_matrices_ensure_size 222 8.7 1.759 2.118 1.759 2.118 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.851 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.841 1.846 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.017 1.841 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.562000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 2.112573E+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 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.045 91.410 91.412 qs_energies 1 2.0 0.000 0.000 90.934 90.938 ls_scf 1 3.0 0.000 0.000 89.623 89.627 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.555 75.856 multiply_cannon 111 7.7 0.028 0.038 53.386 56.945 ls_scf_main 1 4.0 0.000 0.000 55.150 55.156 multiply_cannon_loop 111 8.7 0.117 0.125 50.087 52.826 density_matrix_trs4 2 5.0 0.002 0.003 49.372 49.549 ls_scf_init_scf 1 4.0 0.000 0.000 30.986 30.987 mp_waitall_1 9105 10.9 21.386 30.101 21.386 30.101 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.756 29.856 multiply_cannon_multrec 1332 9.7 13.118 16.978 22.420 27.454 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.326 27.342 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.884 20.729 make_m2s 222 7.7 0.008 0.009 15.536 16.255 make_images 222 8.7 1.588 1.979 15.505 16.226 dbcsr_mm_accdrv_process 4041 10.4 0.270 0.451 8.903 10.471 dbcsr_mm_accdrv_process_sort 4041 11.4 8.498 10.024 8.498 10.024 make_images_data 222 9.7 0.004 0.005 8.946 10.004 hybrid_alltoall_any 227 10.6 0.521 2.466 8.389 9.323 mp_sum_l 887 5.1 5.322 8.004 5.322 8.004 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.231 7.699 mp_irecv_dv 3311 11.0 3.211 7.640 3.211 7.640 calculate_norms 2376 9.8 6.006 6.841 6.006 6.841 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.127 6.535 multiply_cannon_sync_h2d 1332 9.7 4.767 5.888 4.767 5.888 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.025 5.190 arnoldi_extremal 4 6.8 0.000 0.001 4.671 4.692 arnoldi_normal_ev 4 7.8 0.001 0.004 4.671 4.692 build_subspace 16 8.4 0.014 0.021 4.416 4.419 ls_scf_post 1 4.0 0.000 0.000 3.487 3.491 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.148 3.371 ls_scf_store_result 1 5.0 0.000 0.000 3.247 3.311 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.226 2.745 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.671 2.761 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.193 2.676 mp_allgather_i34 111 8.7 0.989 2.456 0.989 2.456 make_images_pack 222 9.7 2.033 2.414 2.036 2.417 dbcsr_sort_data 436 11.2 1.826 2.100 1.826 2.100 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.888 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.875 1.878 dbcsr_data_new 4174 10.1 1.620 1.838 1.620 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.412000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1714.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 2.735616E+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 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.052 101.800 101.802 qs_energies 1 2.0 0.000 0.000 101.157 101.162 ls_scf 1 3.0 0.000 0.000 99.736 99.741 dbcsr_multiply_generic 111 6.7 0.016 0.017 83.924 84.286 multiply_cannon 111 7.7 0.042 0.155 59.887 63.942 multiply_cannon_loop 111 8.7 0.101 0.112 56.249 60.184 ls_scf_main 1 4.0 0.000 0.000 57.983 57.986 density_matrix_trs4 2 5.0 0.002 0.003 51.890 52.022 ls_scf_init_scf 1 4.0 0.000 0.000 38.147 38.149 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.883 36.984 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 34.483 34.500 mp_waitall_1 7281 11.0 24.730 34.377 24.730 34.377 multiply_cannon_multrec 888 9.7 12.634 15.175 21.236 24.387 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.663 23.011 make_m2s 222 7.7 0.007 0.009 17.395 18.659 make_images 222 8.7 1.984 2.325 17.356 18.618 make_images_data 222 9.7 0.004 0.005 10.036 11.176 hybrid_alltoall_any 227 10.6 0.622 2.884 9.654 10.958 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.419 8.140 9.378 mp_sum_l 887 5.1 5.093 9.072 5.093 9.072 dbcsr_mm_accdrv_process_sort 3754 11.4 7.771 8.959 7.771 8.959 compute_matrix_preconditioner 1 6.0 0.000 0.001 7.828 7.835 multiply_cannon_sync_h2d 888 9.7 5.934 7.350 5.934 7.350 acc_transpose_blocks 888 9.7 0.004 0.005 6.575 7.242 acc_transpose_blocks_kernels 888 10.7 0.011 0.013 6.533 7.199 jit_kernel_transpose 1 13.0 6.522 7.187 6.522 7.187 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.931 7.180 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.460 6.837 mp_irecv_dv 2335 11.1 2.445 6.774 2.445 6.774 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.639 6.640 arnoldi_extremal 4 6.8 0.000 0.001 5.361 5.383 arnoldi_normal_ev 4 7.8 0.001 0.005 5.361 5.382 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.929 5.150 build_subspace 16 8.4 0.014 0.020 5.056 5.064 calculate_norms 1584 9.8 4.306 4.726 4.306 4.726 mp_allgather_i34 111 8.7 1.465 3.900 1.465 3.900 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.562 3.868 ls_scf_post 1 4.0 0.000 0.000 3.606 3.613 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.598 3.017 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.341 3.434 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.964 3.070 make_images_sizes 222 9.7 0.000 0.000 1.025 2.383 mp_alltoall_i44 222 10.7 1.025 2.382 1.025 2.382 dbcsr_sort_data 325 11.1 1.864 2.156 1.864 2.156 make_images_pack 222 9.7 1.820 2.121 1.823 2.124 dbcsr_data_release 9322 10.9 1.314 2.052 1.314 2.052 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=101.802000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2152.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 3.370869E+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 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.076 97.949 97.951 qs_energies 1 2.0 0.000 0.000 97.154 97.165 ls_scf 1 3.0 0.002 0.006 95.458 95.470 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.743 79.053 ls_scf_main 1 4.0 0.001 0.011 59.289 59.292 multiply_cannon 111 7.7 0.054 0.099 51.916 56.620 density_matrix_trs4 2 5.0 0.003 0.006 53.063 53.172 multiply_cannon_loop 111 8.7 0.116 0.128 46.855 49.748 ls_scf_init_scf 1 4.0 0.000 0.001 32.891 32.892 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.628 31.692 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.104 29.130 mp_waitall_1 6369 11.0 22.766 28.511 22.766 28.511 multiply_cannon_multrec 1332 9.7 14.345 17.085 22.311 25.599 make_m2s 222 7.7 0.008 0.010 21.132 22.623 make_images 222 8.7 3.151 3.611 21.080 22.574 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.117 16.913 make_images_data 222 9.7 0.004 0.005 11.797 13.343 hybrid_alltoall_any 227 10.6 0.798 3.772 11.274 12.896 dbcsr_mm_accdrv_process 3641 10.4 0.219 0.407 7.595 9.124 dbcsr_mm_accdrv_process_sort 3641 11.4 7.222 8.709 7.222 8.709 mp_sum_l 887 5.1 4.121 8.083 4.121 8.083 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.114 6.360 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.083 6.187 mp_irecv_dv 3229 10.9 2.060 6.110 2.060 6.110 multiply_cannon_sync_h2d 1332 9.7 5.512 5.981 5.512 5.981 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.678 5.672 arnoldi_extremal 4 6.8 0.007 0.032 5.454 5.481 arnoldi_normal_ev 4 7.8 0.008 0.037 5.447 5.481 build_subspace 16 8.4 0.014 0.021 5.072 5.084 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.544 4.712 mp_allgather_i34 111 8.7 2.180 4.663 2.180 4.663 calculate_norms 2376 9.8 4.207 4.500 4.207 4.500 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.635 3.961 dbcsr_matrix_vector_mult_local 304 10.0 3.157 3.682 3.159 3.683 dbcsr_sort_data 658 11.4 3.104 3.399 3.104 3.399 ls_scf_post 1 4.0 0.002 0.009 3.277 3.288 dbcsr_special_finalize 555 9.7 0.006 0.012 2.853 3.160 dbcsr_merge_single_wm 555 10.7 0.539 0.648 2.844 3.152 ls_scf_dm_to_ks 2 5.0 0.001 0.008 3.094 3.140 ls_scf_store_result 1 5.0 0.000 0.000 2.992 3.077 dbcsr_data_release 10477 10.7 1.590 2.411 1.590 2.411 dbcsr_finalize 304 7.8 0.050 0.061 1.820 2.000 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.991 1.992 rebuild_ks_matrix 3 7.3 0.000 0.000 1.967 1.968 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.967 1.968 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.951000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2767.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 4.688957E+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 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.060 100.613 100.614 qs_energies 1 2.0 0.000 0.000 99.801 99.806 ls_scf 1 3.0 0.000 0.000 97.868 97.877 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.986 79.242 ls_scf_main 1 4.0 0.000 0.000 62.721 62.722 multiply_cannon 111 7.7 0.096 0.227 55.738 60.632 density_matrix_trs4 2 5.0 0.002 0.003 55.389 55.476 multiply_cannon_loop 111 8.7 0.069 0.077 51.298 53.715 mp_waitall_1 5436 11.0 27.017 33.084 27.017 33.084 ls_scf_init_scf 1 4.0 0.000 0.000 31.456 31.461 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.150 30.187 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.988 28.000 multiply_cannon_multrec 444 9.7 14.025 16.448 21.123 23.537 make_m2s 222 7.7 0.006 0.007 18.159 20.580 make_images 222 8.7 3.733 4.447 18.095 20.519 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.395 16.184 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.033 14.812 make_images_data 222 9.7 0.003 0.004 10.384 12.688 hybrid_alltoall_any 227 10.6 0.788 3.769 10.106 12.527 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.349 6.806 7.951 multiply_cannon_sync_h2d 444 9.7 6.498 7.848 6.498 7.848 dbcsr_mm_accdrv_process_sort 3003 11.4 6.492 7.606 6.492 7.606 mp_allgather_i34 111 8.7 2.666 7.037 2.666 7.037 arnoldi_extremal 4 6.8 0.000 0.001 5.931 5.944 arnoldi_normal_ev 4 7.8 0.002 0.005 5.930 5.944 build_subspace 16 8.4 0.015 0.020 5.536 5.546 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.616 4.788 mp_sum_l 887 5.1 3.138 4.664 3.138 4.664 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.709 4.517 mp_irecv_dv 1241 11.2 1.691 4.490 1.691 4.490 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.203 4.358 dbcsr_matrix_vector_mult_local 304 10.0 3.605 4.060 3.607 4.062 make_images_sizes 222 9.7 0.000 0.000 1.107 3.835 mp_alltoall_i44 222 10.7 1.107 3.835 1.107 3.835 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.699 3.811 ls_scf_post 1 4.0 0.000 0.000 3.690 3.694 calculate_norms 792 9.8 3.541 3.683 3.541 3.683 ls_scf_store_result 1 5.0 0.000 0.000 3.462 3.510 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.209 3.481 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.300 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.127 2.127 dbcsr_merge_all 275 8.9 0.473 0.530 2.051 2.126 rebuild_ks_matrix 3 7.3 0.000 0.000 2.094 2.095 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.094 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.614000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_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 8.761221E+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 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.092 0.112 112.789 112.790 qs_energies 1 2.0 0.000 0.000 111.322 111.334 ls_scf 1 3.0 0.000 0.000 108.402 108.415 dbcsr_multiply_generic 111 6.7 0.023 0.028 81.687 81.832 ls_scf_main 1 4.0 0.000 0.000 66.320 66.322 density_matrix_trs4 2 5.0 0.002 0.003 56.985 57.040 multiply_cannon 111 7.7 0.137 0.268 54.056 56.433 multiply_cannon_loop 111 8.7 0.068 0.071 50.534 51.282 ls_scf_init_scf 1 4.0 0.000 0.000 38.380 38.381 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.780 36.797 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 33.921 33.929 mp_waitall_1 4527 11.1 22.885 26.145 22.885 26.145 make_m2s 222 7.7 0.007 0.007 23.988 25.001 make_images 222 8.7 4.586 4.988 23.878 24.890 multiply_cannon_multrec 444 9.7 17.967 18.583 22.619 23.220 hybrid_alltoall_any 227 10.6 1.661 3.624 12.948 16.027 make_images_data 222 9.7 0.003 0.004 13.227 15.680 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.064 14.766 multiply_cannon_sync_h2d 444 9.7 8.848 8.890 8.848 8.890 arnoldi_extremal 4 6.8 0.001 0.001 7.481 7.495 arnoldi_normal_ev 4 7.8 0.003 0.009 7.481 7.494 build_subspace 16 8.4 0.026 0.036 6.922 6.928 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.507 5.557 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.254 5.350 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.072 5.326 dbcsr_matrix_vector_mult_local 304 10.0 4.862 5.141 4.864 5.144 compute_matrix_preconditioner 1 6.0 0.002 0.002 4.607 4.614 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.315 4.483 4.610 dbcsr_mm_accdrv_process_sort 1814 11.4 4.183 4.318 4.183 4.318 acc_transpose_blocks 444 9.7 0.002 0.002 3.568 4.067 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 3.532 4.033 jit_kernel_transpose 1 13.0 3.527 4.028 3.527 4.028 ls_scf_post 1 4.0 0.000 0.000 3.702 3.714 make_images_sizes 222 9.7 0.000 0.000 1.412 3.574 mp_alltoall_i44 222 10.7 1.411 3.573 1.411 3.573 mp_allgather_i34 111 8.7 1.127 3.504 1.127 3.504 ls_scf_store_result 1 5.0 0.000 0.000 3.439 3.459 calculate_norms 792 9.8 3.239 3.283 3.239 3.283 dbcsr_finalize 304 7.8 0.082 0.090 3.092 3.172 dbcsr_merge_all 275 8.9 0.893 0.918 2.877 2.950 dbcsr_complete_redistribute 5 7.6 1.430 1.460 2.796 2.916 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.889 2.889 dbcsr_data_release 12724 10.6 2.337 2.864 2.337 2.864 matrix_ls_to_qs 2 6.0 0.000 0.000 2.446 2.585 dbcsr_sort_data 325 11.1 2.438 2.496 2.438 2.496 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.494 2.496 rebuild_ks_matrix 3 7.3 0.000 0.000 2.427 2.428 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.427 2.428 dbcsr_new_transposed 4 7.5 0.243 0.250 2.326 2.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=112.790000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6852.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e92e2ab9bf27c2a5bb388384f6c7748634383a1_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 585.625600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 945678128. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.085 312.956 312.958 qs_mol_dyn_low 1 2.0 0.003 0.004 311.864 311.886 qs_forces 5 3.8 0.004 0.005 311.789 311.791 qs_energies 5 4.8 0.001 0.053 308.180 308.213 scf_env_do_scf 5 5.8 0.001 0.012 288.718 288.722 scf_env_do_scf_inner_loop 105 6.6 0.003 0.025 249.005 249.008 qs_scf_new_mos 105 7.6 0.001 0.001 194.514 194.910 qs_scf_loop_do_ot 105 8.6 0.001 0.001 194.513 194.909 ot_scf_mini 105 9.6 0.003 0.004 184.400 184.706 dbcsr_multiply_generic 1445 12.2 0.126 0.134 141.455 142.714 velocity_verlet 4 3.0 0.001 0.006 126.413 126.414 multiply_cannon 1445 13.2 0.279 0.291 119.304 122.467 multiply_cannon_loop 1445 14.2 2.399 2.540 116.402 119.622 qs_ot_get_p 112 10.4 0.001 0.001 90.350 90.690 qs_ot_p2m_diag 40 11.0 0.020 0.030 77.580 77.667 cp_dbcsr_syevd 40 12.0 0.002 0.003 74.062 74.065 cp_fm_syevd 40 13.0 0.000 0.001 67.328 67.488 ot_mini 105 10.6 0.001 0.001 63.170 63.382 cp_fm_redistribute_end 40 14.0 30.726 61.298 30.737 61.305 cp_fm_syevd_base 40 14.0 30.544 61.117 30.544 61.117 mp_waitall_1 488190 16.1 40.828 51.298 40.828 51.298 multiply_cannon_multrec 69360 15.2 31.104 36.283 38.739 44.229 qs_ot_get_derivative 55 11.6 0.001 0.001 41.384 41.682 multiply_cannon_metrocomm3 69360 15.2 0.194 0.206 27.029 40.145 init_scf_loop 7 6.6 0.000 0.009 39.644 39.649 rebuild_ks_matrix 110 8.4 0.000 0.000 34.788 34.991 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.034 34.787 34.991 multiply_cannon_sync_h2d 69360 15.2 29.737 34.233 29.737 34.233 prepare_preconditioner 7 7.6 0.000 0.000 33.755 33.792 make_preconditioner 7 8.6 0.000 0.013 33.755 33.792 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.929 32.112 qs_rho_update_rho_low 110 7.6 0.001 0.001 27.015 27.383 calculate_rho_elec 110 8.6 0.030 0.033 27.014 27.382 make_full_inverse_cholesky 7 9.6 0.000 0.008 24.795 24.874 rs_pw_transfer 690 11.5 0.011 0.012 23.348 24.763 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.149 23.361 apply_single 62 13.6 0.000 0.000 23.149 23.361 density_rs2pw 110 9.6 0.006 0.008 20.272 21.637 ot_new_cg_direction 55 11.6 0.000 0.001 21.128 21.130 cp_fm_cholesky_invert 7 10.6 16.689 16.705 16.689 16.705 init_scf_run 5 5.8 0.000 0.001 16.593 16.593 scf_env_initial_rho_setup 5 6.8 0.012 0.028 16.593 16.593 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.503 15.815 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.934 15.090 pw_transfer 1645 12.4 0.085 0.107 13.963 14.359 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.819 14.221 mp_sum_l 4764 12.2 12.650 13.529 12.650 13.529 sum_up_and_integrate 60 10.3 0.028 0.030 13.287 13.303 integrate_v_rspace 60 11.3 0.001 0.002 13.259 13.275 check_diag 80 13.5 8.589 8.860 11.753 11.915 calculate_dm_sparse 110 9.5 0.000 0.001 11.667 11.883 fft_wrap_pw1pw2_240 915 15.0 1.213 1.310 11.341 11.690 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 11.379 11.396 qs_vxc_create 110 10.4 0.002 0.004 11.281 11.342 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 11.099 11.194 acc_transpose_blocks 69360 15.2 0.363 0.383 10.366 10.997 calculate_first_density_matrix 1 7.0 0.000 0.002 10.876 10.896 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 10.818 10.890 multiply_cannon_metrocomm1 69360 15.2 0.092 0.098 5.246 10.456 rs_pw_transfer_RS2PW_30 110 11.6 1.588 1.662 9.298 10.251 fft3d_pb 915 16.0 2.384 2.670 9.370 9.841 make_m2s 2890 13.2 0.081 0.089 9.077 9.770 mp_sendrecv_dv 168740 12.6 9.518 9.698 9.518 9.698 make_images 2890 14.2 0.244 0.263 8.967 9.662 acc_transpose_blocks_kernels 69360 16.2 0.862 0.914 8.420 9.007 potential_pw2rs 60 12.3 0.003 0.003 8.934 8.991 make_full_single_inverse 7 9.6 0.001 0.002 8.511 8.549 dbcsr_mm_accdrv_process 154766 15.8 4.035 4.200 7.509 8.323 jit_kernel_transpose 5 15.0 7.559 8.130 7.559 8.130 xc_rho_set_and_dset_create 110 12.4 0.077 0.119 7.740 8.005 xc_vxc_pw_create 60 11.3 0.039 0.056 7.639 7.698 cp_fm_cholesky_decompose 14 10.2 7.498 7.509 7.498 7.509 mp_alltoall_z22v 2340 17.7 6.573 7.334 6.573 7.334 xc_pw_derive 510 13.4 0.006 0.007 6.888 6.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=312.958000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7e92e2ab9bf27c2a5bb388384f6c7748634383a1 Summary: empty Status: OK