=== 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: 225305aa1575fcf7b2a0a553af71d6a327fd5eb1 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/01 job id: 50071905 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/02 job id: 50071906 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/03 job id: 50071908 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/04 job id: 50071909 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/05 job id: 50071910 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/06 job id: 50071912 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/07 job id: 50071913 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/08 job id: 50071915 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/09 job id: 50071916 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/10 job id: 50071918 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/11 job id: 50071919 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/12 job id: 50071921 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/13 job id: 50071922 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/14 job id: 50071923 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/15 job id: 50071924 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/16 job id: 50071926 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/17 job id: 50071927 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/18 job id: 50071928 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/19 job id: 50071929 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/20 job id: 50071931 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/21 job id: 50071932 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/22 job id: 50071933 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/23 job id: 50071934 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/24 job id: 50071935 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/25 job id: 50071937 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/26 job id: 50071938 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/27 job id: 50071939 --- 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/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 0.018 0.034 133.096 133.097 farming_run 1 2.0 132.546 132.548 133.068 133.072 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.491730E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 114.032 114.033 qs_energies 1 2.0 0.000 0.000 113.828 113.830 mp2_main 1 3.0 0.000 0.000 111.773 111.776 mp2_gpw_main 1 4.0 0.027 0.033 110.795 110.798 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.049 92.207 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.565 54.724 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.164 40.967 45.778 get_2c_integrals 1 6.0 0.008 0.009 36.493 37.311 integrate_v_rspace 273 8.0 0.438 0.454 24.565 29.024 pw_transfer 6555 10.6 0.380 0.394 26.754 27.036 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.378 25.585 grid_integrate_task_list 273 9.0 20.482 25.366 20.482 25.366 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.029 22.907 23.097 compute_2c_integrals 1 7.0 0.002 0.002 18.992 18.993 rpa_ri_compute_en 1 5.0 0.019 0.023 18.642 18.780 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.555 18.729 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.417 18.553 18.728 cp_fm_cholesky_decompose 12 8.2 17.427 18.209 17.427 18.209 cholesky_decomp 1 7.0 0.000 0.000 16.345 17.126 fft3d_s 5443 13.4 16.083 16.288 16.104 16.309 ao_to_mo_and_store_B_mult_1 272 7.0 10.775 15.325 10.775 15.325 calculate_wavefunction 272 8.0 5.399 5.488 12.307 13.048 rpa_num_int 1 6.0 0.001 0.019 10.596 10.606 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.580 10.584 calc_mat_Q 8 8.0 0.000 0.000 9.451 9.534 contract_S_to_Q 8 9.0 0.000 0.000 8.876 8.960 calc_potential_gpw 544 9.5 0.004 0.004 8.225 8.678 parallel_gemm_fm 14 9.1 0.000 0.000 8.469 8.587 parallel_gemm_fm_cosma 14 10.1 8.469 8.587 8.469 8.587 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.053 8.312 potential_pw2rs 545 10.0 0.106 0.108 7.464 8.199 create_integ_mat 1 6.0 0.016 0.032 7.794 7.804 collocate_single_gaussian 272 10.0 0.039 0.041 7.314 7.600 array2fm 1 7.0 0.000 0.000 6.685 7.215 pw_scatter_s 2720 13.7 4.347 4.533 4.347 4.533 pw_gather_s 2722 13.2 3.439 3.845 3.439 3.845 array2fm_buffer_send 1 8.0 2.984 3.171 2.984 3.171 pw_poisson_solve 545 10.5 1.118 1.169 2.198 2.335 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.797845, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2810.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 0.024 0.035 417.919 417.921 farming_run 1 2.0 417.141 417.154 417.884 417.886 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.239192E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 215.897 215.897 qs_energies 1 2.0 0.000 0.000 215.671 215.689 scf_env_do_scf 1 3.0 0.000 0.000 113.949 113.949 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.064 113.071 rebuild_ks_matrix 4 6.0 0.000 0.000 113.063 113.070 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.063 113.070 hfx_ks_matrix 4 8.0 0.001 0.001 112.718 112.723 integrate_four_center 4 9.0 0.153 0.465 112.718 112.723 integrate_four_center_main 4 10.0 0.100 0.694 101.586 104.305 integrate_four_center_bin 265 11.0 101.486 104.246 101.486 104.246 mp2_main 1 3.0 0.000 0.000 101.433 101.452 mp2_gpw_main 1 4.0 0.047 0.066 100.582 100.601 init_scf_loop 1 4.0 0.000 0.000 95.836 95.837 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.977 74.974 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.826 54.823 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 41.496 46.530 integrate_v_rspace 95 8.0 0.397 0.562 27.960 32.822 pw_transfer 2240 10.6 0.145 0.168 29.541 29.949 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.548 28.928 mp2_ri_gpw_compute_en 1 5.0 0.059 0.077 26.461 28.513 grid_integrate_task_list 95 9.0 23.377 28.469 23.377 28.469 ao_to_mo_and_store_B_mult_1 91 7.0 10.656 28.176 10.656 28.176 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.295 26.662 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.898 24.805 24.814 get_2c_integrals 1 6.0 0.000 0.000 20.056 20.087 compute_2c_integrals 1 7.0 0.002 0.003 19.038 19.040 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.666 18.914 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.847 18.665 18.913 fft3d_s 1823 13.4 18.485 18.824 18.499 18.837 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.111 18.111 calculate_wavefunction 91 8.0 2.010 2.035 9.623 9.832 mp2_ri_gpw_compute_en_expansio 172 7.0 0.560 0.584 8.697 9.204 potential_pw2rs 186 10.0 0.033 0.035 8.463 8.965 local_gemm 172 8.0 8.137 8.639 8.137 8.639 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.184 8.587 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.517 7.875 8.419 calc_potential_gpw 182 9.5 0.002 0.002 7.894 8.051 collocate_single_gaussian 91 10.0 0.017 0.021 7.770 8.014 integrate_four_center_load 4 10.0 0.000 0.000 6.758 6.762 hfx_load_balance 1 11.0 0.000 0.000 6.757 6.762 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.446 6.346 6.446 mp_sendrecv_dm3 2068 8.0 5.908 6.422 5.908 6.422 mp_sync 37 10.5 3.122 6.086 3.122 6.086 pw_gather_s 912 13.2 4.495 4.982 4.495 4.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.578517, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.115904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 54.263 54.264 qs_mol_dyn_low 1 2.0 0.003 0.003 54.008 54.017 qs_forces 11 3.9 0.002 0.002 53.949 53.950 qs_energies 11 4.9 0.001 0.001 52.480 52.503 scf_env_do_scf 11 5.9 0.000 0.001 46.310 46.311 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.308 44.308 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.418 34.935 qs_scf_new_mos 108 7.5 0.000 0.001 34.480 34.760 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.479 34.759 ot_scf_mini 108 9.5 0.002 0.002 32.787 32.965 multiply_cannon 2286 13.5 0.178 0.186 26.819 28.329 multiply_cannon_loop 2286 14.5 1.832 1.934 26.187 27.733 velocity_verlet 10 3.0 0.002 0.005 26.380 26.381 ot_mini 108 10.5 0.001 0.001 19.949 20.209 qs_ot_get_derivative 108 11.5 0.001 0.001 16.790 16.987 mp_waitall_1 245248 16.5 8.694 15.245 8.694 15.245 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 6.240 13.630 multiply_cannon_multrec 54864 15.5 3.655 5.693 7.890 11.378 qs_ot_get_p 119 10.4 0.001 0.001 8.064 8.338 rebuild_ks_matrix 119 8.3 0.000 0.000 7.721 7.875 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.720 7.874 mp_sum_l 7287 12.8 5.704 7.489 5.704 7.489 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.796 6.940 multiply_cannon_sync_h2d 54864 15.5 5.213 6.469 5.213 6.469 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.725 6.183 dbcsr_mm_accdrv_process 76910 16.1 1.835 2.899 4.149 5.915 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.579 5.691 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.207 5.256 init_scf_run 11 5.9 0.000 0.001 4.922 4.922 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.922 4.922 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.392 4.393 sum_up_and_integrate 119 10.3 0.001 0.002 4.342 4.348 integrate_v_rspace 119 11.3 0.002 0.002 4.331 4.338 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.206 4.207 cp_fm_redistribute_end 50 14.0 2.147 4.184 2.152 4.186 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.026 4.150 calculate_rho_elec 119 8.7 0.011 0.017 4.026 4.149 cp_fm_diag_elpa_base 50 14.0 2.029 4.080 2.033 4.087 calculate_dm_sparse 119 9.5 0.000 0.000 3.099 3.237 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.935 3.136 apply_single 119 13.6 0.000 0.000 2.935 3.136 jit_kernel_multiply 13 15.8 2.250 2.956 2.250 2.956 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.848 2.826 acc_transpose_blocks 54864 15.5 0.225 0.249 2.246 2.808 ot_diis_step 108 11.5 0.006 0.006 2.806 2.807 calculate_first_density_matrix 1 7.0 0.000 0.000 2.755 2.762 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.566 2.651 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.369 2.370 density_rs2pw 119 9.7 0.004 0.004 2.047 2.194 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.105 2.155 grid_integrate_task_list 119 12.3 2.016 2.108 2.016 2.108 wfi_extrapolate 11 7.9 0.001 0.001 2.096 2.096 mp_sum_d 4137 12.0 1.488 2.078 1.488 2.078 init_scf_loop 11 6.9 0.000 0.000 1.987 1.987 potential_pw2rs 119 12.3 0.004 0.004 1.737 1.748 pw_transfer 1439 11.6 0.052 0.056 1.581 1.646 make_m2s 4572 13.5 0.053 0.055 1.575 1.619 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.571 1.606 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.506 1.574 make_images 4572 14.5 0.133 0.139 1.494 1.538 transfer_rs2pw 487 10.6 0.005 0.006 1.294 1.446 acc_transpose_blocks_sync 164592 16.5 1.192 1.431 1.192 1.431 mp_alltoall_d11v 2130 13.8 1.261 1.422 1.261 1.422 grid_collocate_task_list 119 9.7 1.350 1.410 1.350 1.410 mp_waitany 12084 13.8 1.232 1.377 1.232 1.377 transfer_pw2rs 487 13.2 0.005 0.006 1.274 1.280 fft3d_ps 1201 14.6 0.369 0.473 1.219 1.279 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.889 1.277 fft_wrap_pw1pw2_140 487 13.2 0.107 0.114 1.165 1.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.264000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.882176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 762437. 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 0.007 0.030 38.312 38.313 qs_mol_dyn_low 1 2.0 0.003 0.003 38.118 38.146 qs_forces 11 3.9 0.014 0.021 37.891 37.891 qs_energies 11 4.9 0.001 0.002 36.227 36.242 scf_env_do_scf 11 5.9 0.001 0.001 31.253 31.253 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.872 28.872 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.813 22.169 qs_scf_new_mos 108 7.5 0.001 0.001 20.389 20.618 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.389 20.618 ot_scf_mini 108 9.5 0.002 0.003 19.458 19.623 velocity_verlet 10 3.0 0.001 0.002 18.250 18.254 multiply_cannon 2286 13.5 0.206 0.214 16.735 18.215 multiply_cannon_loop 2286 14.5 1.187 1.248 15.597 17.135 ot_mini 108 10.5 0.001 0.001 12.011 12.239 mp_waitall_1 200699 16.5 5.405 10.507 5.405 10.507 qs_ot_get_derivative 108 11.5 0.001 0.001 9.659 9.827 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 3.955 9.282 multiply_cannon_multrec 27432 15.5 1.833 4.313 6.123 9.006 rebuild_ks_matrix 119 8.3 0.000 0.000 6.669 6.821 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.669 6.820 dbcsr_mm_accdrv_process 47894 16.0 3.576 5.808 4.212 6.263 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.865 6.003 qs_ot_get_p 119 10.4 0.001 0.001 4.544 4.764 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.601 4.513 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.060 4.056 apply_single 119 13.6 0.000 0.000 3.060 4.056 mp_sum_l 7287 12.8 2.052 4.017 2.052 4.017 init_scf_run 11 5.9 0.000 0.001 3.762 3.762 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.762 3.762 sum_up_and_integrate 119 10.3 0.001 0.001 3.613 3.620 integrate_v_rspace 119 11.3 0.002 0.003 3.599 3.607 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.545 3.581 calculate_rho_elec 119 8.7 0.021 0.024 3.545 3.580 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.932 2.950 make_m2s 4572 13.5 0.051 0.053 2.570 2.895 make_images 4572 14.5 0.205 0.243 2.480 2.808 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.541 2.542 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.285 2.372 init_scf_loop 11 6.9 0.000 0.000 2.361 2.362 ot_diis_step 108 11.5 0.011 0.011 2.303 2.304 calculate_first_density_matrix 1 7.0 0.000 0.001 2.277 2.280 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.234 2.234 cp_fm_redistribute_end 50 14.0 1.133 2.211 1.136 2.213 multiply_cannon_sync_h2d 27432 15.5 1.690 2.187 1.690 2.187 cp_fm_diag_elpa_base 50 14.0 1.046 2.125 1.074 2.160 calculate_dm_sparse 119 9.5 0.000 0.001 2.067 2.148 acc_transpose_blocks 27432 15.5 0.114 0.120 1.612 1.952 grid_integrate_task_list 119 12.3 1.827 1.930 1.827 1.930 density_rs2pw 119 9.7 0.004 0.004 1.841 1.928 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.855 1.857 pw_transfer 1439 11.6 0.065 0.069 1.807 1.838 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.715 1.748 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.662 1.702 jit_kernel_multiply 9 16.3 0.578 1.699 0.578 1.699 make_images_data 4572 15.5 0.047 0.053 1.230 1.586 prepare_preconditioner 11 7.9 0.000 0.000 1.492 1.518 make_preconditioner 11 8.9 0.000 0.000 1.492 1.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.397 1.453 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.095 1.435 wfi_extrapolate 11 7.9 0.001 0.001 1.427 1.427 fft_wrap_pw1pw2_140 487 13.2 0.127 0.132 1.377 1.410 mp_alltoall_d11v 2130 13.8 1.188 1.394 1.188 1.394 potential_pw2rs 119 12.3 0.006 0.006 1.374 1.380 fft3d_ps 1201 14.6 0.523 0.579 1.347 1.373 grid_collocate_task_list 119 9.7 1.289 1.345 1.289 1.345 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.249 1.260 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.190 1.236 transfer_rs2pw 487 10.6 0.005 0.006 1.032 1.152 mp_sum_d 4137 12.0 0.528 1.001 0.528 1.001 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.937 0.949 mp_allgather_i34 2286 14.5 0.564 0.947 0.564 0.947 acc_transpose_blocks_sync 82296 16.5 0.811 0.935 0.811 0.935 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.663 0.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.860 0.874 transfer_pw2rs 487 13.2 0.004 0.005 0.854 0.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.313000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.636364, yerr=1.610913 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 520.589312E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 31.900 31.900 qs_mol_dyn_low 1 2.0 0.003 0.003 31.701 31.710 qs_forces 11 3.9 0.002 0.002 31.645 31.646 qs_energies 11 4.9 0.001 0.001 30.111 30.113 scf_env_do_scf 11 5.9 0.000 0.001 25.404 25.404 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.975 22.975 dbcsr_multiply_generic 2286 12.5 0.096 0.100 16.777 16.878 qs_scf_new_mos 108 7.5 0.001 0.001 15.325 15.341 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.324 15.340 velocity_verlet 10 3.0 0.001 0.002 15.192 15.193 ot_scf_mini 108 9.5 0.002 0.002 14.569 14.586 multiply_cannon 2286 13.5 0.192 0.196 13.583 14.231 multiply_cannon_loop 2286 14.5 0.857 0.882 12.838 13.519 ot_mini 108 10.5 0.001 0.001 9.012 9.032 qs_ot_get_derivative 108 11.5 0.001 0.001 7.527 7.545 multiply_cannon_multrec 18288 15.5 1.867 2.907 6.932 7.268 rebuild_ks_matrix 119 8.3 0.000 0.001 5.957 5.979 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.957 5.978 dbcsr_mm_accdrv_process 38222 16.0 4.899 5.668 4.973 5.768 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.253 5.274 sum_up_and_integrate 119 10.3 0.001 0.001 3.545 3.550 integrate_v_rspace 119 11.3 0.002 0.003 3.533 3.539 init_scf_run 11 5.9 0.000 0.001 3.458 3.458 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.458 3.458 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.869 3.439 qs_ot_get_p 119 10.4 0.001 0.001 3.328 3.351 mp_waitall_1 158411 16.6 2.640 3.325 2.640 3.325 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.235 3.239 calculate_rho_elec 119 8.7 0.031 0.031 3.235 3.238 init_scf_loop 11 6.9 0.000 0.000 2.414 2.415 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.965 2.341 apply_single 119 13.6 0.000 0.000 1.965 2.341 calculate_first_density_matrix 1 7.0 0.000 0.000 2.250 2.251 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.229 2.235 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.478 2.219 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.931 1.931 density_rs2pw 119 9.7 0.004 0.004 1.808 1.897 make_m2s 4572 13.5 0.044 0.045 1.770 1.897 grid_integrate_task_list 119 12.3 1.803 1.895 1.803 1.895 pw_transfer 1439 11.6 0.065 0.069 1.839 1.848 calculate_dm_sparse 119 9.5 0.000 0.001 1.836 1.846 make_images 4572 14.5 0.191 0.203 1.684 1.811 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.746 1.756 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.703 1.710 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.687 1.697 cp_fm_diag_elpa_base 50 14.0 1.664 1.675 1.685 1.695 acc_transpose_blocks 18288 15.5 0.078 0.081 1.604 1.674 prepare_preconditioner 11 7.9 0.000 0.000 1.660 1.663 make_preconditioner 11 8.9 0.000 0.000 1.660 1.663 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.656 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.516 1.600 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.493 1.500 ot_diis_step 108 11.5 0.011 0.011 1.469 1.469 fft_wrap_pw1pw2_140 487 13.2 0.179 0.182 1.412 1.423 mp_sum_l 7287 12.8 1.028 1.414 1.028 1.414 potential_pw2rs 119 12.3 0.007 0.008 1.355 1.360 grid_collocate_task_list 119 9.7 1.235 1.321 1.235 1.321 fft3d_ps 1201 14.6 0.544 0.567 1.301 1.312 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.176 1.181 multiply_cannon_sync_h2d 18288 15.5 1.036 1.171 1.036 1.171 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 transfer_rs2pw 487 10.6 0.005 0.005 0.983 1.090 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.033 1.034 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.951 0.968 make_images_data 4572 15.5 0.047 0.051 0.749 0.890 mp_alltoall_d11v 2130 13.8 0.654 0.848 0.654 0.848 transfer_pw2rs 487 13.2 0.004 0.004 0.834 0.837 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.650 0.810 acc_transpose_blocks_kernels 18288 16.5 0.218 0.224 0.779 0.795 acc_transpose_blocks_sync 54864 16.5 0.728 0.793 0.728 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.774 0.777 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.674 0.742 cp_fm_cholesky_invert 11 10.9 0.698 0.702 0.698 0.702 mp_alltoall_z22v 1201 16.6 0.608 0.674 0.608 0.674 mp_waitany 9880 13.7 0.548 0.669 0.548 0.669 transfer_rs2pw_140 130 11.5 0.115 0.121 0.541 0.648 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.900000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 554.856448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.025 36.272 36.272 qs_mol_dyn_low 1 2.0 0.003 0.003 36.048 36.056 qs_forces 11 3.9 0.002 0.003 35.707 35.708 qs_energies 11 4.9 0.001 0.001 34.009 34.014 scf_env_do_scf 11 5.9 0.000 0.001 29.078 29.079 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.718 25.718 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.799 19.931 velocity_verlet 10 3.0 0.010 0.022 18.591 18.604 qs_scf_new_mos 108 7.5 0.001 0.001 17.723 17.775 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.722 17.774 ot_scf_mini 108 9.5 0.002 0.003 16.690 16.751 multiply_cannon 2286 13.5 0.216 0.224 16.044 16.486 multiply_cannon_loop 2286 14.5 1.564 1.627 15.103 15.430 ot_mini 108 10.5 0.001 0.001 10.302 10.369 multiply_cannon_multrec 27432 15.5 2.446 3.151 9.016 9.328 qs_ot_get_derivative 108 11.5 0.001 0.001 8.432 8.494 dbcsr_mm_accdrv_process 47916 15.9 6.090 7.771 6.469 7.833 rebuild_ks_matrix 119 8.3 0.000 0.000 6.258 6.310 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.257 6.310 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.541 5.587 init_scf_run 11 5.9 0.000 0.001 3.610 3.610 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.610 3.610 qs_ot_get_p 119 10.4 0.001 0.001 3.529 3.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.038 3.517 sum_up_and_integrate 119 10.3 0.001 0.001 3.430 3.438 integrate_v_rspace 119 11.3 0.002 0.003 3.419 3.427 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.331 3.361 calculate_rho_elec 119 8.7 0.040 0.046 3.330 3.361 init_scf_loop 11 6.9 0.000 0.000 3.341 3.342 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.199 2.665 apply_single 119 13.6 0.000 0.000 2.198 2.664 acc_transpose_blocks 27432 15.5 0.119 0.124 2.436 2.530 prepare_preconditioner 11 7.9 0.000 0.000 2.522 2.530 make_preconditioner 11 8.9 0.000 0.000 2.522 2.529 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.126 2.455 make_m2s 4572 13.5 0.054 0.055 2.262 2.373 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.338 2.368 make_images 4572 14.5 0.273 0.334 2.155 2.265 calculate_first_density_matrix 1 7.0 0.000 0.000 2.244 2.246 calculate_dm_sparse 119 9.5 0.000 0.000 2.151 2.212 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.135 2.144 mp_waitall_1 137007 16.6 1.596 2.077 1.596 2.077 pw_transfer 1439 11.6 0.064 0.068 1.958 1.995 grid_integrate_task_list 119 12.3 1.826 1.909 1.826 1.909 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.866 1.906 density_rs2pw 119 9.7 0.004 0.004 1.773 1.863 ot_diis_step 108 11.5 0.012 0.012 1.829 1.829 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.796 1.796 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.732 1.732 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.635 mp_sum_l 7287 12.8 1.057 1.614 1.057 1.614 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.552 1.593 acc_transpose_blocks_sync 82296 16.5 1.474 1.557 1.474 1.557 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.500 1.510 cp_fm_diag_elpa_base 50 14.0 1.467 1.485 1.498 1.508 fft3d_ps 1201 14.6 0.590 0.650 1.355 1.380 grid_collocate_task_list 119 9.7 1.246 1.331 1.246 1.331 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.319 wfi_extrapolate 11 7.9 0.001 0.001 1.313 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.254 potential_pw2rs 119 12.3 0.008 0.009 1.240 1.242 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.684 1.171 cp_fm_upper_to_full 72 14.2 0.821 1.170 0.821 1.170 jit_kernel_multiply 6 16.2 0.314 1.082 0.314 1.082 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.067 1.067 dbcsr_complete_redistribute 329 12.2 0.132 0.158 0.770 1.047 transfer_rs2pw 487 10.6 0.004 0.005 0.864 0.959 make_images_data 4572 15.5 0.048 0.053 0.841 0.959 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.726 0.899 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.876 mp_alltoall_d11v 2130 13.8 0.740 0.849 0.740 0.849 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.569 0.839 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.815 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.272000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.909091, yerr=2.391445 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 630.190080E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 30.974 30.975 qs_mol_dyn_low 1 2.0 0.003 0.003 30.737 30.756 qs_forces 11 3.9 0.002 0.002 30.644 30.645 qs_energies 11 4.9 0.001 0.001 28.914 28.917 scf_env_do_scf 11 5.9 0.008 0.051 23.967 23.967 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 21.253 21.254 velocity_verlet 10 3.0 0.002 0.002 15.603 15.608 dbcsr_multiply_generic 2286 12.5 0.095 0.098 14.194 14.356 qs_scf_new_mos 108 7.5 0.001 0.001 13.077 13.115 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.076 13.114 ot_scf_mini 108 9.5 0.002 0.002 12.298 12.334 multiply_cannon 2286 13.5 0.225 0.232 11.101 11.475 multiply_cannon_loop 2286 14.5 0.643 0.674 10.072 10.351 ot_mini 108 10.5 0.001 0.001 7.052 7.094 multiply_cannon_multrec 9144 15.5 1.729 1.939 6.156 6.415 rebuild_ks_matrix 119 8.3 0.000 0.000 6.067 6.100 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.067 6.099 qs_ot_get_derivative 108 11.5 0.001 0.001 5.685 5.718 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.405 5.434 dbcsr_mm_accdrv_process 12550 15.8 3.610 4.331 4.319 4.411 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.645 3.654 calculate_rho_elec 119 8.7 0.060 0.061 3.645 3.653 sum_up_and_integrate 119 10.3 0.001 0.001 3.633 3.638 integrate_v_rspace 119 11.3 0.003 0.003 3.623 3.627 init_scf_run 11 5.9 0.000 0.001 3.483 3.483 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.482 3.483 qs_ot_get_p 119 10.4 0.001 0.001 3.196 3.234 init_scf_loop 11 6.9 0.000 0.000 2.685 2.694 mp_waitall_1 115863 16.7 1.984 2.674 1.984 2.674 make_m2s 4572 13.5 0.033 0.035 2.111 2.265 pw_transfer 1439 11.6 0.066 0.067 2.241 2.252 calculate_first_density_matrix 1 7.0 0.000 0.001 2.251 2.252 make_images 4572 14.5 0.269 0.300 2.022 2.175 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.148 2.159 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.082 2.086 density_rs2pw 119 9.7 0.003 0.003 1.990 2.057 grid_integrate_task_list 119 12.3 1.877 1.957 1.877 1.957 prepare_preconditioner 11 7.9 0.000 0.000 1.887 1.891 make_preconditioner 11 8.9 0.000 0.000 1.887 1.891 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.853 1.877 calculate_dm_sparse 119 9.5 0.000 0.000 1.853 1.872 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.822 1.824 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.765 1.798 fft_wrap_pw1pw2_140 487 13.2 0.324 0.331 1.745 1.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.694 1.696 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.547 1.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.520 1.531 acc_transpose_blocks 9144 15.5 0.042 0.044 1.472 1.518 fft3d_ps 1201 14.6 0.651 0.663 1.499 1.511 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.486 1.493 cp_fm_diag_elpa_base 50 14.0 1.458 1.476 1.485 1.491 potential_pw2rs 119 12.3 0.010 0.011 1.412 1.416 grid_collocate_task_list 119 9.7 1.308 1.380 1.308 1.380 ot_diis_step 108 11.5 0.012 0.013 1.338 1.338 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.271 1.301 apply_single 119 13.6 0.000 0.000 1.271 1.301 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.228 1.228 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.907 1.198 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.180 1.187 jit_kernel_multiply 6 15.3 0.669 1.181 0.669 1.181 wfi_extrapolate 11 7.9 0.001 0.001 1.174 1.174 make_images_data 4572 15.5 0.042 0.046 0.944 1.168 transfer_rs2pw 487 10.6 0.004 0.004 0.938 1.015 cp_fm_cholesky_invert 11 10.9 0.988 0.992 0.988 0.992 mp_alltoall_d11v 2130 13.8 0.885 0.959 0.885 0.959 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.485 0.959 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.872 0.921 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.906 0.916 mp_allgather_i34 2286 14.5 0.344 0.858 0.344 0.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.786 transfer_pw2rs 487 13.2 0.003 0.003 0.763 0.766 acc_transpose_blocks_sync 27432 16.5 0.718 0.763 0.718 0.763 mp_alltoall_z22v 1201 16.6 0.722 0.761 0.722 0.761 mp_sum_l 7287 12.8 0.526 0.750 0.526 0.750 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.676 0.725 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.696 0.698 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.465 0.645 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.975000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=597.272727, yerr=6.467976 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 777.318400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 42.476 42.478 qs_mol_dyn_low 1 2.0 0.003 0.003 42.274 42.281 qs_forces 11 3.9 0.002 0.002 42.207 42.208 qs_energies 11 4.9 0.001 0.001 40.223 40.226 scf_env_do_scf 11 5.9 0.001 0.001 34.585 34.585 scf_env_do_scf_inner_loop 108 6.5 0.006 0.014 26.828 26.829 velocity_verlet 10 3.0 0.015 0.016 23.957 23.962 dbcsr_multiply_generic 2286 12.5 0.105 0.106 18.854 19.101 qs_scf_new_mos 108 7.5 0.001 0.001 17.174 17.272 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.173 17.271 ot_scf_mini 108 9.5 0.002 0.002 16.012 16.114 multiply_cannon 2286 13.5 0.295 0.303 14.972 15.947 multiply_cannon_loop 2286 14.5 0.872 0.884 13.721 14.735 ot_mini 108 10.5 0.001 0.001 9.749 9.870 multiply_cannon_multrec 9144 15.5 3.361 4.638 8.791 8.884 qs_ot_get_derivative 108 11.5 0.001 0.001 7.715 7.817 init_scf_loop 11 6.9 0.000 0.000 7.730 7.732 rebuild_ks_matrix 119 8.3 0.000 0.000 7.220 7.360 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.220 7.360 prepare_preconditioner 11 7.9 0.000 0.000 6.776 6.789 make_preconditioner 11 8.9 0.000 0.000 6.776 6.788 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.317 6.671 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.514 6.641 dbcsr_mm_accdrv_process 12550 15.8 4.487 5.845 5.297 6.565 cp_fm_upper_to_full 72 14.2 3.169 4.575 3.169 4.575 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.167 4.174 calculate_rho_elec 119 8.7 0.117 0.120 4.166 4.174 sum_up_and_integrate 119 10.3 0.001 0.001 3.817 3.824 integrate_v_rspace 119 11.3 0.003 0.003 3.807 3.814 init_scf_run 11 5.9 0.000 0.001 3.585 3.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.585 3.585 qs_ot_get_p 119 10.4 0.001 0.001 3.345 3.481 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.800 3.222 mp_waitall_1 94719 16.7 2.136 3.138 2.136 3.138 pw_transfer 1439 11.6 0.068 0.069 3.013 3.019 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.915 2.920 dbcsr_complete_redistribute 329 12.2 0.281 0.286 1.982 2.802 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.686 2.507 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.212 2.498 apply_single 119 13.6 0.000 0.000 2.211 2.498 fft_wrap_pw1pw2_140 487 13.2 0.648 0.653 2.483 2.491 make_m2s 4572 13.5 0.037 0.037 2.308 2.450 make_images 4572 14.5 0.356 0.391 2.186 2.328 mp_alltoall_i22 627 13.8 1.434 2.294 1.434 2.294 calculate_dm_sparse 119 9.5 0.000 0.000 2.253 2.272 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.453 2.267 density_rs2pw 119 9.7 0.003 0.004 2.214 2.225 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.250 2.175 calculate_first_density_matrix 1 7.0 0.000 0.000 2.130 2.131 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.063 2.113 grid_integrate_task_list 119 12.3 2.065 2.073 2.065 2.073 ot_diis_step 108 11.5 0.014 0.014 2.010 2.010 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.973 1.975 acc_transpose_blocks 9144 15.5 0.044 0.045 1.838 1.854 fft3d_ps 1201 14.6 0.868 0.889 1.837 1.840 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.811 1.811 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.773 1.774 mp_sum_l 7287 12.8 1.022 1.728 1.022 1.728 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.620 1.621 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.564 1.606 grid_collocate_task_list 119 9.7 1.511 1.529 1.511 1.529 wfi_extrapolate 11 7.9 0.001 0.001 1.392 1.392 cp_fm_cholesky_invert 11 10.9 1.383 1.386 1.383 1.386 potential_pw2rs 119 12.3 0.014 0.014 1.363 1.365 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.335 1.335 cp_fm_diag_elpa_base 50 14.0 1.190 1.243 1.334 1.334 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.330 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.256 1.264 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.046 1.253 mp_alltoall_d11v 2130 13.8 1.132 1.151 1.132 1.151 make_images_data 4572 15.5 0.046 0.049 0.956 1.138 acc_transpose_blocks_sync 27432 16.5 1.112 1.128 1.112 1.128 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.102 1.121 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.984 1.038 jit_kernel_multiply 6 15.5 0.781 1.013 0.781 1.013 qs_create_task_list 11 7.9 0.001 0.002 0.929 0.940 generate_qs_task_list 11 8.9 0.368 0.387 0.928 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.916 0.928 mp_alltoall_z22v 1201 16.6 0.833 0.863 0.833 0.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.478000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.000000, yerr=15.231546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 500.506624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 83.932 83.934 qs_mol_dyn_low 1 2.0 0.003 0.003 83.683 83.694 qs_forces 11 3.9 0.003 0.003 83.588 83.589 qs_energies 11 4.9 0.001 0.002 80.702 80.715 scf_env_do_scf 11 5.9 0.001 0.001 71.769 71.772 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.825 65.825 dbcsr_multiply_generic 2055 12.4 0.107 0.111 51.604 52.001 qs_scf_new_mos 99 7.5 0.000 0.001 48.272 48.386 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.271 48.386 ot_scf_mini 99 9.5 0.002 0.002 45.907 46.005 velocity_verlet 10 3.0 0.001 0.002 44.465 44.466 multiply_cannon 2055 13.4 0.175 0.183 42.872 44.098 multiply_cannon_loop 2055 14.4 1.799 1.851 41.798 43.052 ot_mini 99 10.5 0.001 0.001 27.144 27.267 qs_ot_get_derivative 99 11.5 0.001 0.001 20.268 20.393 multiply_cannon_multrec 49320 15.4 11.367 12.054 17.438 18.222 rebuild_ks_matrix 110 8.3 0.000 0.001 14.765 15.002 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.764 15.002 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.943 13.170 mp_waitall_1 220248 16.4 11.211 12.115 11.211 12.115 qs_ot_get_p 110 10.4 0.001 0.001 10.282 10.489 multiply_cannon_sync_h2d 49320 15.4 9.611 10.333 9.611 10.333 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.949 8.487 multiply_cannon_metrocomm3 49320 15.4 0.082 0.086 6.580 7.891 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.258 7.727 apply_single 110 13.6 0.000 0.000 7.258 7.726 sum_up_and_integrate 110 10.3 0.001 0.003 7.227 7.240 integrate_v_rspace 110 11.3 0.002 0.003 7.202 7.222 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.861 6.881 init_scf_run 11 5.9 0.000 0.001 6.879 6.880 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.879 6.879 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.679 6.833 calculate_rho_elec 110 8.6 0.020 0.025 6.678 6.832 ot_diis_step 99 11.5 0.005 0.006 6.636 6.637 dbcsr_mm_accdrv_process 87628 16.1 3.070 3.177 5.941 6.204 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.990 5.990 init_scf_loop 11 6.9 0.000 0.000 5.914 5.914 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.400 5.461 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.364 5.385 cp_fm_diag_elpa_base 48 14.0 5.349 5.370 5.362 5.383 mp_sum_l 6594 12.7 4.063 4.994 4.063 4.994 make_m2s 4110 13.4 0.061 0.065 4.205 4.307 make_images 4110 14.4 0.178 0.191 4.109 4.212 wfi_extrapolate 11 7.9 0.001 0.001 4.125 4.125 density_rs2pw 110 9.6 0.004 0.004 3.669 3.926 calculate_dm_sparse 110 9.5 0.001 0.001 3.766 3.844 prepare_preconditioner 11 7.9 0.000 0.000 3.693 3.709 make_preconditioner 11 8.9 0.000 0.000 3.693 3.709 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.506 3.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.542 3.547 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.466 3.510 grid_integrate_task_list 110 12.3 3.247 3.404 3.247 3.404 pw_transfer 1331 11.6 0.055 0.069 3.196 3.304 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.203 3.246 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.177 3.237 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.107 3.219 potential_pw2rs 110 12.3 0.005 0.006 2.815 2.856 mp_alltoall_d11v 2046 13.8 2.170 2.745 2.170 2.745 fft_wrap_pw1pw2_140 451 13.1 0.309 0.330 2.607 2.720 calculate_first_density_matrix 1 7.0 0.000 0.001 2.649 2.653 jit_kernel_multiply 13 15.9 2.589 2.642 2.589 2.642 acc_transpose_blocks 49320 15.4 0.213 0.220 2.562 2.629 fft3d_ps 1111 14.6 0.798 0.887 2.403 2.489 transfer_rs2pw 451 10.6 0.005 0.006 2.138 2.391 grid_collocate_task_list 110 9.6 2.156 2.263 2.156 2.263 mp_waitany 14300 13.8 1.839 2.165 1.839 2.165 cp_fm_cholesky_invert 11 10.9 2.048 2.053 2.048 2.053 make_images_data 4110 15.4 0.043 0.047 1.845 2.026 transfer_pw2rs 451 13.1 0.006 0.006 1.948 1.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.952 1.973 mp_sum_d 3891 11.9 1.405 1.968 1.405 1.968 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.588 1.888 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.680 1.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.934000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.391445 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 593.379328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2010410. 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 0.008 0.029 68.318 68.321 qs_mol_dyn_low 1 2.0 0.003 0.003 68.109 68.120 qs_forces 11 3.9 0.003 0.003 68.043 68.043 qs_energies 11 4.9 0.001 0.001 64.739 64.743 scf_env_do_scf 11 5.9 0.003 0.005 56.280 56.283 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 48.837 48.837 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.895 38.086 velocity_verlet 10 3.0 0.003 0.004 35.756 35.757 qs_scf_new_mos 99 7.5 0.001 0.001 33.116 33.222 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.116 33.222 multiply_cannon 2055 13.4 0.223 0.244 31.313 32.281 ot_scf_mini 99 9.5 0.003 0.003 31.445 31.550 multiply_cannon_loop 2055 14.4 1.159 1.186 30.037 30.990 ot_mini 99 10.5 0.001 0.001 18.394 18.501 multiply_cannon_multrec 24660 15.4 6.984 8.770 14.014 15.924 rebuild_ks_matrix 110 8.3 0.000 0.000 13.467 13.595 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.467 13.595 qs_ot_get_derivative 99 11.5 0.001 0.001 12.562 12.675 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.848 11.963 mp_waitall_1 176588 16.5 7.482 9.910 7.482 9.910 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.141 8.027 init_scf_loop 11 6.9 0.000 0.000 7.407 7.410 multiply_cannon_sync_h2d 24660 15.4 6.407 7.351 6.407 7.351 dbcsr_mm_accdrv_process 52282 16.1 5.596 6.414 6.861 7.182 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.540 7.181 apply_single 110 13.6 0.000 0.001 6.540 7.180 qs_ot_get_p 110 10.4 0.001 0.001 6.398 6.547 sum_up_and_integrate 110 10.3 0.001 0.002 6.303 6.314 integrate_v_rspace 110 11.3 0.002 0.003 6.276 6.289 init_scf_run 11 5.9 0.000 0.001 6.091 6.092 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.091 6.092 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.804 5.816 calculate_rho_elec 110 8.6 0.039 0.047 5.804 5.815 ot_diis_step 99 11.5 0.010 0.010 5.781 5.781 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.710 5.475 prepare_preconditioner 11 7.9 0.000 0.000 5.398 5.421 make_preconditioner 11 8.9 0.000 0.000 5.397 5.421 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.982 5.132 make_m2s 4110 13.4 0.057 0.060 4.179 4.648 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.521 4.540 make_images 4110 14.4 0.405 0.467 4.067 4.533 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.057 4.057 pw_transfer 1331 11.6 0.066 0.074 3.497 3.633 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.391 3.530 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.484 3.500 cp_fm_diag_elpa_base 48 14.0 3.438 3.456 3.481 3.497 wfi_extrapolate 11 7.9 0.001 0.001 3.483 3.483 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.273 3.343 grid_integrate_task_list 110 12.3 3.156 3.323 3.156 3.323 density_rs2pw 110 9.6 0.004 0.004 3.056 3.245 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.228 3.230 fft_wrap_pw1pw2_140 451 13.1 0.357 0.376 2.922 3.061 calculate_dm_sparse 110 9.5 0.001 0.001 2.980 3.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.893 2.945 hybrid_alltoall_any 4261 16.3 0.105 0.456 2.006 2.814 make_images_data 4110 15.4 0.048 0.052 2.277 2.774 fft3d_ps 1111 14.6 1.115 1.352 2.484 2.633 cp_fm_cholesky_invert 11 10.9 2.526 2.533 2.526 2.533 calculate_first_density_matrix 1 7.0 0.000 0.001 2.514 2.515 mp_sum_l 6594 12.7 1.767 2.448 1.767 2.448 grid_collocate_task_list 110 9.6 2.169 2.312 2.169 2.312 potential_pw2rs 110 12.3 0.008 0.009 2.186 2.194 acc_transpose_blocks 24660 15.4 0.115 0.118 1.961 1.991 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.971 1.990 mp_alltoall_d11v 2046 13.8 1.742 1.958 1.742 1.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.862 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.775 1.791 jit_kernel_multiply 9 16.3 0.898 1.760 0.898 1.760 mp_allgather_i34 2055 14.4 0.584 1.591 0.584 1.591 transfer_rs2pw 451 10.6 0.006 0.007 1.393 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.553 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.771 1.518 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.384 1.488 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.321000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.818182, yerr=7.333834 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 667.176960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.030 59.293 59.294 qs_mol_dyn_low 1 2.0 0.003 0.003 59.002 59.013 qs_forces 11 3.9 0.003 0.003 58.929 58.931 qs_energies 11 4.9 0.001 0.001 55.776 55.780 scf_env_do_scf 11 5.9 0.001 0.001 47.754 47.754 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.359 39.359 velocity_verlet 10 3.0 0.001 0.001 32.145 32.147 dbcsr_multiply_generic 2055 12.4 0.109 0.114 28.598 28.851 qs_scf_new_mos 99 7.5 0.001 0.001 24.830 24.926 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.830 24.926 multiply_cannon 2055 13.4 0.212 0.243 22.457 23.702 ot_scf_mini 99 9.5 0.002 0.003 23.585 23.694 multiply_cannon_loop 2055 14.4 0.810 0.833 21.285 22.196 ot_mini 99 10.5 0.001 0.001 13.559 13.664 rebuild_ks_matrix 110 8.3 0.000 0.000 12.067 12.190 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.067 12.189 multiply_cannon_multrec 16440 15.4 3.759 5.435 9.996 11.585 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.616 10.728 mp_waitall_1 139946 16.5 6.643 9.404 6.643 9.404 qs_ot_get_derivative 99 11.5 0.001 0.001 9.167 9.276 init_scf_loop 11 6.9 0.000 0.000 8.363 8.364 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.132 6.800 prepare_preconditioner 11 7.9 0.000 0.000 6.640 6.658 make_preconditioner 11 8.9 0.000 0.000 6.640 6.658 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.952 6.317 sum_up_and_integrate 110 10.3 0.001 0.002 6.218 6.232 dbcsr_mm_accdrv_process 34862 16.1 5.346 5.713 6.082 6.227 integrate_v_rspace 110 11.3 0.003 0.003 6.192 6.206 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.683 5.696 calculate_rho_elec 110 8.6 0.059 0.059 5.682 5.695 init_scf_run 11 5.9 0.000 0.001 5.647 5.647 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.647 5.647 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.938 5.352 apply_single 110 13.6 0.000 0.000 4.937 5.352 qs_ot_get_p 110 10.4 0.001 0.001 5.188 5.318 make_m2s 4110 13.4 0.049 0.051 4.111 4.503 make_images 4110 14.4 0.398 0.518 3.995 4.389 ot_diis_step 99 11.5 0.010 0.011 4.363 4.364 multiply_cannon_sync_h2d 16440 15.4 3.299 4.088 3.299 4.088 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.105 3.754 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.607 3.610 pw_transfer 1331 11.6 0.065 0.071 3.576 3.586 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.469 3.482 grid_integrate_task_list 110 12.3 3.191 3.373 3.191 3.373 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.242 3.242 density_rs2pw 110 9.6 0.004 0.004 2.926 3.167 fft_wrap_pw1pw2_140 451 13.1 0.460 0.472 3.004 3.030 wfi_extrapolate 11 7.9 0.001 0.001 2.933 2.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.809 2.809 make_images_data 4110 15.4 0.046 0.050 2.316 2.776 hybrid_alltoall_any 4261 16.3 0.108 0.375 2.059 2.770 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.722 2.731 cp_fm_diag_elpa_base 48 14.0 2.658 2.688 2.721 2.729 calculate_first_density_matrix 1 7.0 0.000 0.000 2.624 2.626 calculate_dm_sparse 110 9.5 0.001 0.001 2.533 2.575 cp_fm_cholesky_invert 11 10.9 2.565 2.571 2.565 2.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.481 2.532 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.400 2.463 grid_collocate_task_list 110 9.6 2.212 2.414 2.212 2.414 fft3d_ps 1111 14.6 1.119 1.134 2.385 2.395 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.858 2.368 mp_irecv_dv 48980 15.7 0.783 2.232 0.783 2.232 mp_alltoall_d11v 2046 13.8 1.743 2.088 1.743 2.088 potential_pw2rs 110 12.3 0.011 0.011 2.071 2.075 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.948 1.949 mp_sum_l 6594 12.7 1.401 1.946 1.401 1.946 dbcsr_complete_redistribute 325 12.2 0.321 0.345 1.438 1.893 cp_fm_upper_to_full 70 14.2 1.392 1.780 1.392 1.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.657 acc_transpose_blocks 16440 15.4 0.078 0.081 1.578 1.632 cp_fm_cholesky_decompose 22 10.9 1.544 1.565 1.544 1.565 mp_allgather_i34 2055 14.4 0.486 1.502 0.486 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.472 1.488 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.466 transfer_rs2pw 451 10.6 0.005 0.006 1.245 1.453 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.966 1.408 mp_waitany 17072 13.8 1.118 1.378 1.118 1.378 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.260 1.267 rs_gather_matrices 110 12.3 0.137 0.149 0.880 1.208 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.294000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.363636, yerr=9.422147 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 735.514624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 65.290 65.291 qs_mol_dyn_low 1 2.0 0.003 0.003 65.054 65.065 qs_forces 11 3.9 0.003 0.003 64.990 64.990 qs_energies 11 4.9 0.001 0.001 61.601 61.605 scf_env_do_scf 11 5.9 0.001 0.001 53.269 53.272 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.784 41.785 velocity_verlet 10 3.0 0.002 0.002 36.942 36.944 dbcsr_multiply_generic 2055 12.4 0.118 0.122 30.684 30.856 qs_scf_new_mos 99 7.5 0.001 0.001 27.224 27.327 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.223 27.326 ot_scf_mini 99 9.5 0.002 0.003 25.545 25.647 multiply_cannon 2055 13.4 0.236 0.258 23.781 24.863 multiply_cannon_loop 2055 14.4 1.439 1.490 22.204 22.704 ot_mini 99 10.5 0.001 0.001 14.748 14.881 multiply_cannon_multrec 24660 15.4 4.101 6.661 13.040 14.299 rebuild_ks_matrix 110 8.3 0.000 0.000 11.953 12.062 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.030 11.952 12.061 init_scf_loop 11 6.9 0.000 0.000 11.447 11.447 qs_ot_get_derivative 99 11.5 0.001 0.001 10.567 10.675 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.545 10.641 dbcsr_mm_accdrv_process 52304 16.0 7.716 9.049 8.782 9.747 prepare_preconditioner 11 7.9 0.000 0.000 9.720 9.738 make_preconditioner 11 8.9 0.000 0.000 9.720 9.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.960 9.420 sum_up_and_integrate 110 10.3 0.001 0.002 6.177 6.190 integrate_v_rspace 110 11.3 0.003 0.003 6.152 6.163 mp_waitall_1 121746 16.5 4.039 6.002 4.039 6.002 make_m2s 4110 13.4 0.059 0.061 5.316 5.825 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.810 5.818 calculate_rho_elec 110 8.6 0.077 0.081 5.809 5.818 qs_ot_get_p 110 10.4 0.001 0.001 5.653 5.814 make_images 4110 14.4 0.575 0.699 5.178 5.681 init_scf_run 11 5.9 0.000 0.001 5.619 5.619 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.618 5.619 cp_fm_upper_to_full 70 14.2 3.350 4.857 3.350 4.857 ot_diis_step 99 11.5 0.011 0.011 4.146 4.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.077 4.140 apply_single 110 13.6 0.000 0.000 4.077 4.140 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.819 3.835 dbcsr_complete_redistribute 325 12.2 0.416 0.457 2.623 3.782 pw_transfer 1331 11.6 0.065 0.074 3.725 3.757 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.619 3.653 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.517 3.579 grid_integrate_task_list 110 12.3 3.304 3.420 3.304 3.420 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.331 3.388 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.254 3.255 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.113 3.254 fft_wrap_pw1pw2_140 451 13.1 0.529 0.543 3.130 3.167 make_images_data 4110 15.4 0.049 0.053 2.622 3.134 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.248 3.133 density_rs2pw 110 9.6 0.004 0.004 2.926 3.099 calculate_dm_sparse 110 9.5 0.001 0.001 3.029 3.069 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.315 3.009 wfi_extrapolate 11 7.9 0.001 0.001 2.962 2.962 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.751 2.872 mp_alltoall_i22 605 13.7 1.661 2.823 1.661 2.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.806 2.807 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.730 2.740 cp_fm_diag_elpa_base 48 14.0 2.586 2.641 2.728 2.738 acc_transpose_blocks 24660 15.4 0.112 0.114 2.476 2.601 cp_fm_cholesky_invert 11 10.9 2.591 2.599 2.591 2.599 calculate_first_density_matrix 1 7.0 0.006 0.025 2.561 2.564 multiply_cannon_sync_h2d 24660 15.4 2.401 2.509 2.401 2.509 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.461 2.495 fft3d_ps 1111 14.6 1.150 1.186 2.446 2.463 grid_collocate_task_list 110 9.6 2.262 2.370 2.262 2.370 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.226 2.227 potential_pw2rs 110 12.3 0.012 0.013 1.984 1.988 mp_alltoall_d11v 2046 13.8 1.711 1.955 1.711 1.955 jit_kernel_multiply 10 15.9 0.721 1.869 0.721 1.869 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.797 1.829 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.703 cp_fm_cholesky_decompose 22 10.9 1.614 1.674 1.614 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.658 mp_allgather_i34 2055 14.4 0.664 1.647 0.664 1.647 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.591 1.616 acc_transpose_blocks_sync 73980 16.4 1.436 1.552 1.436 1.552 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.838 1.539 mp_sum_l 6594 12.7 0.892 1.498 0.892 1.498 mp_irecv_dv 62702 16.1 0.735 1.458 0.735 1.458 transfer_rs2pw 451 10.6 0.005 0.006 1.139 1.319 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.291000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.181818, yerr=8.547389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 858.894336E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.058 0.074 55.264 55.265 qs_mol_dyn_low 1 2.0 0.003 0.003 54.822 54.831 qs_forces 11 3.9 0.012 0.013 54.728 54.730 qs_energies 11 4.9 0.001 0.001 51.062 51.067 scf_env_do_scf 11 5.9 0.000 0.001 42.806 42.806 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.180 35.180 velocity_verlet 10 3.0 0.001 0.002 31.014 31.018 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.192 23.368 qs_scf_new_mos 99 7.5 0.001 0.001 20.682 20.744 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.682 20.743 ot_scf_mini 99 9.5 0.002 0.002 19.430 19.474 multiply_cannon 2055 13.4 0.236 0.255 17.692 18.846 multiply_cannon_loop 2055 14.4 0.606 0.628 16.475 16.854 rebuild_ks_matrix 110 8.3 0.000 0.000 11.600 11.672 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.599 11.672 ot_mini 99 10.5 0.001 0.001 10.849 10.896 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.316 10.380 multiply_cannon_multrec 8220 15.4 3.150 4.878 7.543 8.933 mp_waitall_1 103326 16.6 5.903 7.665 5.903 7.665 init_scf_loop 11 6.9 0.000 0.000 7.581 7.581 qs_ot_get_derivative 99 11.5 0.001 0.001 7.073 7.116 sum_up_and_integrate 110 10.3 0.001 0.002 6.188 6.202 integrate_v_rspace 110 11.3 0.003 0.003 6.161 6.175 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.954 5.972 calculate_rho_elec 110 8.6 0.113 0.113 5.954 5.972 prepare_preconditioner 11 7.9 0.000 0.000 5.942 5.950 make_preconditioner 11 8.9 0.000 0.000 5.942 5.950 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.533 5.603 dbcsr_mm_accdrv_process 17442 15.9 3.170 4.121 4.253 5.164 init_scf_run 11 5.9 0.000 0.001 5.126 5.126 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.126 5.126 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.183 4.785 qs_ot_get_p 110 10.4 0.001 0.001 4.663 4.720 make_m2s 4110 13.4 0.037 0.039 4.129 4.315 make_images 4110 14.4 0.645 0.707 3.999 4.180 pw_transfer 1331 11.6 0.065 0.070 4.065 4.077 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.958 3.973 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.803 3.871 apply_single 110 13.6 0.000 0.000 3.802 3.870 ot_diis_step 99 11.5 0.012 0.012 3.751 3.752 grid_integrate_task_list 110 12.3 3.371 3.541 3.371 3.541 fft_wrap_pw1pw2_140 451 13.1 0.717 0.732 3.439 3.453 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.299 3.303 density_rs2pw 110 9.6 0.004 0.004 2.992 3.115 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.963 2.964 cp_fm_cholesky_invert 11 10.9 2.854 2.859 2.854 2.859 wfi_extrapolate 11 7.9 0.001 0.001 2.691 2.691 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.675 2.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.553 2.555 grid_collocate_task_list 110 9.6 2.369 2.543 2.369 2.543 hybrid_alltoall_any 4261 16.3 0.198 0.843 2.148 2.524 fft3d_ps 1111 14.6 1.289 1.297 2.504 2.519 calculate_dm_sparse 110 9.5 0.001 0.001 2.487 2.518 multiply_cannon_sync_h2d 8220 15.4 2.344 2.507 2.344 2.507 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.458 2.465 cp_fm_diag_elpa_base 48 14.0 2.403 2.422 2.456 2.464 make_images_data 4110 15.4 0.040 0.046 2.179 2.453 calculate_first_density_matrix 1 7.0 0.000 0.000 2.340 2.341 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.126 2.156 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.026 2.060 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 2.013 potential_pw2rs 110 12.3 0.015 0.015 1.992 1.994 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.915 1.942 mp_alltoall_d11v 2046 13.8 1.605 1.797 1.605 1.797 cp_fm_cholesky_decompose 22 10.9 1.668 1.703 1.668 1.703 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.505 1.624 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.613 1.621 dbcsr_complete_redistribute 325 12.2 0.578 0.629 1.455 1.547 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.483 1.491 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.438 1.449 mp_irecv_dv 24056 15.7 0.455 1.445 0.455 1.445 acc_transpose_blocks 8220 15.4 0.040 0.042 1.377 1.411 mp_allgather_i34 2055 14.4 0.387 1.411 0.387 1.411 qs_create_task_list 11 7.9 0.001 0.001 1.211 1.313 generate_qs_task_list 11 8.9 0.376 0.444 1.210 1.312 transfer_rs2pw 451 10.6 0.005 0.005 1.092 1.240 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.725 1.212 mp_waitany 9240 13.8 1.040 1.182 1.040 1.182 jit_kernel_multiply 7 15.3 0.762 1.167 0.762 1.167 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.115 1.149 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.265000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=809.545455, yerr=11.765617 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 1.409221E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.080 87.236 87.236 qs_mol_dyn_low 1 2.0 0.003 0.003 86.801 86.813 qs_forces 11 3.9 0.003 0.003 86.613 86.614 qs_energies 11 4.9 0.001 0.002 82.472 82.474 scf_env_do_scf 11 5.9 0.000 0.001 72.430 72.431 velocity_verlet 10 3.0 0.002 0.003 55.631 55.653 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.457 44.459 dbcsr_multiply_generic 2055 12.4 0.129 0.141 29.358 29.452 init_scf_loop 11 6.9 0.000 0.000 27.904 27.907 qs_scf_new_mos 99 7.5 0.001 0.001 26.583 26.616 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.583 26.615 prepare_preconditioner 11 7.9 0.000 0.000 25.893 25.901 make_preconditioner 11 8.9 0.000 0.000 25.893 25.901 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.321 25.380 ot_scf_mini 99 9.5 0.002 0.002 24.782 24.808 multiply_cannon 2055 13.4 0.335 0.355 22.342 23.082 multiply_cannon_loop 2055 14.4 0.830 0.844 20.560 20.831 cp_fm_upper_to_full 70 14.2 12.681 17.981 12.681 17.981 ot_mini 99 10.5 0.001 0.001 14.003 14.031 rebuild_ks_matrix 110 8.3 0.000 0.000 13.927 13.961 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.927 13.960 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.635 12.667 dbcsr_complete_redistribute 325 12.2 1.007 1.035 7.123 10.295 multiply_cannon_multrec 8220 15.4 4.056 4.208 9.756 9.906 qs_ot_get_derivative 99 11.5 0.001 0.001 9.451 9.477 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.108 9.283 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.557 8.700 mp_alltoall_i22 605 13.7 5.182 8.379 5.182 8.379 mp_waitall_1 84994 16.7 7.338 8.195 7.338 8.195 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.447 7.483 calculate_rho_elec 110 8.6 0.222 0.223 7.447 7.482 sum_up_and_integrate 110 10.3 0.002 0.002 6.887 6.901 integrate_v_rspace 110 11.3 0.004 0.004 6.860 6.873 pw_transfer 1331 11.6 0.075 0.075 5.786 5.793 init_scf_run 11 5.9 0.000 0.001 5.779 5.779 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.779 5.779 make_m2s 4110 13.4 0.042 0.043 5.295 5.777 dbcsr_mm_accdrv_process 11614 15.7 3.925 4.104 5.552 5.744 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.669 5.676 make_images 4110 14.4 0.881 0.937 5.108 5.589 qs_ot_get_p 110 10.4 0.001 0.001 5.512 5.546 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.703 5.168 apply_single 110 13.6 0.000 0.000 4.703 5.168 cp_fm_cholesky_invert 11 10.9 5.141 5.146 5.141 5.146 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.754 5.026 fft_wrap_pw1pw2_140 451 13.1 1.365 1.371 4.984 4.989 ot_diis_step 99 11.5 0.015 0.015 4.538 4.538 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.955 3.961 density_rs2pw 110 9.6 0.004 0.004 3.918 3.928 grid_integrate_task_list 110 12.3 3.678 3.743 3.678 3.743 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.664 3.665 hybrid_alltoall_any 4261 16.3 0.265 0.575 2.783 3.511 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.984 3.441 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.439 3.440 fft3d_ps 1111 14.6 1.880 1.889 3.377 3.382 make_images_data 4110 15.4 0.045 0.048 2.765 3.378 wfi_extrapolate 11 7.9 0.001 0.001 3.312 3.312 calculate_dm_sparse 110 9.5 0.001 0.001 3.193 3.219 multiply_cannon_sync_h2d 8220 15.4 3.118 3.147 3.118 3.147 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.879 2.882 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.862 2.862 cp_fm_diag_elpa_base 48 14.0 2.328 2.525 2.860 2.861 grid_collocate_task_list 110 9.6 2.670 2.686 2.670 2.686 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.628 2.642 potential_pw2rs 110 12.3 0.021 0.021 2.425 2.428 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.390 2.402 calculate_first_density_matrix 1 7.0 0.000 0.000 2.349 2.349 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.188 2.250 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.095 2.210 mp_alltoall_d11v 2046 13.8 2.002 2.071 2.002 2.071 cp_fm_cholesky_decompose 22 10.9 1.964 1.977 1.964 1.977 qs_create_task_list 11 7.9 0.001 0.001 1.878 1.925 generate_qs_task_list 11 8.9 0.732 0.786 1.877 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.889 1.892 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.739 1.778 acc_transpose_blocks 8220 15.4 0.041 0.041 1.744 1.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.236000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1251.727273, yerr=69.106099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 632.209408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 3613077. 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 0.008 0.028 207.708 207.709 qs_mol_dyn_low 1 2.0 0.003 0.003 207.283 207.298 qs_forces 11 3.9 0.005 0.005 207.202 207.202 qs_energies 11 4.9 0.002 0.003 201.630 201.648 scf_env_do_scf 11 5.9 0.001 0.001 185.110 185.114 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 164.887 164.889 dbcsr_multiply_generic 2527 12.6 0.179 0.188 127.320 128.115 qs_scf_new_mos 118 7.6 0.001 0.001 125.537 125.723 qs_scf_loop_do_ot 118 8.6 0.001 0.001 125.537 125.723 velocity_verlet 10 3.0 0.001 0.002 124.230 124.231 ot_scf_mini 118 9.6 0.003 0.003 118.735 118.968 multiply_cannon 2527 13.6 0.237 0.244 103.072 104.684 multiply_cannon_loop 2527 14.6 2.422 2.473 100.865 102.425 ot_mini 118 10.6 0.001 0.001 67.428 67.627 multiply_cannon_multrec 60648 15.6 31.985 34.279 42.006 43.808 qs_ot_get_derivative 118 11.6 0.001 0.001 42.430 42.642 rebuild_ks_matrix 129 8.3 0.001 0.001 33.580 33.754 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 33.579 33.753 mp_waitall_1 269254 16.5 29.541 32.463 29.541 32.463 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.166 30.328 qs_ot_get_p 129 10.4 0.001 0.001 28.995 29.228 multiply_cannon_sync_h2d 60648 15.6 26.519 28.182 26.519 28.182 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.537 25.135 apply_single 129 13.6 0.001 0.001 24.537 25.135 ot_diis_step 118 11.6 0.008 0.008 24.731 24.732 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.409 22.463 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.127 20.239 init_scf_loop 11 6.9 0.000 0.001 20.150 20.151 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.466 19.467 multiply_cannon_metrocomm3 60648 15.6 0.117 0.121 16.534 18.578 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.501 16.528 cp_fm_diag_elpa_base 83 14.4 16.433 16.463 16.497 16.525 prepare_preconditioner 11 7.9 0.000 0.000 15.604 15.646 make_preconditioner 11 8.9 0.000 0.000 15.604 15.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.859 15.045 make_m2s 5054 13.6 0.105 0.113 13.795 14.198 make_images 5054 14.6 0.401 0.421 13.608 14.024 sum_up_and_integrate 129 10.3 0.002 0.004 13.914 13.929 integrate_v_rspace 129 11.3 0.003 0.004 13.856 13.875 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.029 13.240 calculate_rho_elec 129 8.7 0.045 0.064 13.028 13.239 init_scf_run 11 5.9 0.000 0.001 12.397 12.397 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.396 12.397 mp_sum_l 8010 12.9 9.184 10.173 9.184 10.173 dbcsr_mm_accdrv_process 125468 16.2 4.761 4.897 9.584 10.137 wfi_extrapolate 11 7.9 0.001 0.001 9.072 9.072 cp_fm_cholesky_invert 11 10.9 8.884 8.892 8.884 8.892 calculate_dm_sparse 129 9.5 0.001 0.001 8.609 8.732 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.421 8.583 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.132 8.204 multiply_cannon_metrocomm1 60648 15.6 0.092 0.095 6.595 7.886 make_images_data 5054 15.6 0.066 0.072 6.773 7.759 pw_transfer 1559 11.6 0.073 0.097 7.538 7.718 grid_integrate_task_list 129 12.3 7.124 7.587 7.124 7.587 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 7.336 7.516 hybrid_alltoall_any 5240 16.5 0.295 2.271 5.931 7.318 density_rs2pw 129 9.7 0.006 0.007 6.635 7.153 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.786 6.795 fft_wrap_pw1pw2_140 527 13.2 0.863 0.910 6.474 6.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.895 6.028 fft3d_ps 1301 14.7 2.214 2.871 5.331 5.664 mp_alltoall_d11v 2423 14.1 4.231 5.428 4.231 5.428 grid_collocate_task_list 129 9.7 4.879 5.265 4.879 5.265 cp_fm_cholesky_decompose 22 10.9 4.576 4.590 4.576 4.590 potential_pw2rs 129 12.3 0.009 0.010 4.387 4.402 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.709000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.181818, yerr=6.379202 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/16/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 2.183246E+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 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 848.732160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463835059104 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5244621. 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 0.021 0.036 189.256 189.260 qs_mol_dyn_low 1 2.0 0.003 0.005 188.790 188.838 qs_forces 11 3.9 0.007 0.025 188.687 188.701 qs_energies 11 4.9 0.007 0.037 181.940 181.963 scf_env_do_scf 11 5.9 0.001 0.001 165.696 165.707 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.083 133.087 velocity_verlet 10 3.0 0.001 0.002 118.642 118.645 dbcsr_multiply_generic 2507 12.6 0.187 0.194 97.398 98.668 qs_scf_new_mos 117 7.6 0.001 0.001 94.734 95.321 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.733 95.320 ot_scf_mini 117 9.6 0.004 0.004 89.930 90.594 multiply_cannon 2507 13.6 0.502 0.558 77.130 81.326 multiply_cannon_loop 2507 14.6 1.562 1.629 74.193 76.793 ot_mini 117 10.6 0.001 0.001 50.670 51.338 mp_waitall_1 214728 16.6 24.480 39.046 24.480 39.046 multiply_cannon_multrec 30084 15.6 20.846 25.116 31.456 36.179 rebuild_ks_matrix 128 8.3 0.001 0.001 32.122 32.834 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.034 32.121 32.833 init_scf_loop 11 6.9 0.000 0.000 32.525 32.526 multiply_cannon_metrocomm3 30084 15.6 0.098 0.104 16.105 30.530 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.874 29.524 qs_ot_get_derivative 117 11.6 0.001 0.002 28.609 29.259 prepare_preconditioner 11 7.9 0.000 0.000 28.243 28.304 make_preconditioner 11 8.9 0.000 0.000 28.243 28.304 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.942 27.481 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.087 23.249 apply_single 128 13.6 0.001 0.001 22.086 23.248 qs_ot_get_p 128 10.4 0.001 0.002 21.389 22.035 ot_diis_step 117 11.6 0.014 0.015 21.884 21.886 multiply_cannon_sync_h2d 30084 15.6 17.870 19.902 17.870 19.902 cp_fm_cholesky_invert 11 10.9 16.554 16.566 16.554 16.566 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.480 16.515 make_m2s 5014 13.6 0.087 0.092 14.257 15.488 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.271 15.271 make_images 5014 14.6 1.168 1.369 14.039 15.269 sum_up_and_integrate 128 10.3 0.002 0.004 13.979 14.010 integrate_v_rspace 128 11.3 0.003 0.003 13.919 13.954 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.321 13.347 calculate_rho_elec 128 8.7 0.087 0.104 13.320 13.346 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.101 12.138 cp_fm_diag_elpa_base 83 14.4 11.847 11.947 12.095 12.128 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.189 11.685 init_scf_run 11 5.9 0.000 0.001 11.417 11.419 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.417 11.419 multiply_cannon_metrocomm4 27577 15.6 0.099 0.112 3.920 10.912 dbcsr_mm_accdrv_process 62242 16.2 5.392 6.147 10.062 10.660 mp_irecv_dv 69486 16.3 3.717 10.511 3.717 10.511 make_images_data 5014 15.6 0.066 0.076 8.360 9.870 hybrid_alltoall_any 5200 16.5 0.348 1.534 7.027 9.292 pw_transfer 1547 11.6 0.084 0.096 8.587 8.658 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.363 8.435 wfi_extrapolate 11 7.9 0.001 0.001 8.295 8.295 grid_integrate_task_list 128 12.3 7.181 7.562 7.181 7.562 fft_wrap_pw1pw2_140 523 13.2 0.937 0.971 7.395 7.476 density_rs2pw 128 9.7 0.006 0.006 7.028 7.450 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.666 7.431 cp_fm_cholesky_decompose 22 10.9 6.850 6.949 6.850 6.949 calculate_dm_sparse 128 9.5 0.001 0.001 6.489 6.619 mp_sum_l 7950 12.9 4.212 6.427 4.212 6.427 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.025 6.035 fft3d_ps 1291 14.7 2.829 3.021 5.916 5.983 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.391 5.464 grid_collocate_task_list 128 9.7 5.003 5.366 5.003 5.366 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.157 5.319 mp_alltoall_d11v 2415 14.1 4.126 4.592 4.126 4.592 mp_allgather_i34 2507 14.6 1.454 4.544 1.454 4.544 potential_pw2rs 128 12.3 0.015 0.017 4.471 4.486 mp_sum_d 4481 12.1 2.675 4.081 2.675 4.081 dbcsr_complete_redistribute 395 12.7 0.762 0.836 3.090 3.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.260000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=806.454545, yerr=6.344126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+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 6705500928 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 950.599680E+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 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58221. MP_Allreduce 11080 1000. 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 0.051 0.065 173.016 173.018 qs_mol_dyn_low 1 2.0 0.003 0.003 172.429 172.462 qs_forces 11 3.9 0.004 0.006 172.317 172.319 qs_energies 11 4.9 0.001 0.002 165.775 165.785 scf_env_do_scf 11 5.9 0.001 0.002 150.241 150.243 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.716 115.717 velocity_verlet 10 3.0 0.001 0.002 110.184 110.187 dbcsr_multiply_generic 2485 12.5 0.182 0.186 81.342 82.522 qs_scf_new_mos 116 7.6 0.001 0.001 79.183 79.535 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.182 79.534 ot_scf_mini 116 9.6 0.003 0.004 74.981 75.372 multiply_cannon 2485 13.5 0.503 0.520 62.156 66.781 multiply_cannon_loop 2485 14.5 1.114 1.173 59.151 62.411 ot_mini 116 10.6 0.001 0.001 42.285 42.674 init_scf_loop 11 6.9 0.000 0.000 34.429 34.430 mp_waitall_1 169034 16.6 24.073 33.532 24.073 33.532 prepare_preconditioner 11 7.9 0.000 0.000 30.475 30.518 make_preconditioner 11 8.9 0.000 0.000 30.475 30.518 rebuild_ks_matrix 127 8.3 0.001 0.001 29.848 30.346 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.033 29.847 30.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.129 29.528 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.842 27.312 multiply_cannon_multrec 19880 15.5 13.045 15.969 22.820 25.705 multiply_cannon_metrocomm3 19880 15.5 0.061 0.064 15.010 24.747 qs_ot_get_derivative 116 11.6 0.001 0.002 22.958 23.354 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.370 20.433 apply_single 127 13.6 0.001 0.001 19.370 20.433 ot_diis_step 116 11.6 0.017 0.018 19.210 19.210 qs_ot_get_p 127 10.4 0.001 0.001 18.235 18.701 make_m2s 4970 13.5 0.077 0.083 14.289 15.538 make_images 4970 14.5 1.138 1.212 14.053 15.296 multiply_cannon_sync_h2d 19880 15.5 13.596 15.161 13.596 15.161 cp_fm_cholesky_invert 11 10.9 14.124 14.132 14.124 14.132 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.054 14.061 sum_up_and_integrate 127 10.3 0.002 0.002 13.710 13.733 integrate_v_rspace 127 11.3 0.003 0.003 13.651 13.674 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.358 13.396 calculate_rho_elec 127 8.7 0.127 0.142 13.358 13.396 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.016 13.016 init_scf_run 11 5.9 0.000 0.001 10.520 10.520 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.519 10.520 make_images_data 4970 15.5 0.063 0.073 8.555 10.207 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.928 9.951 cp_fm_diag_elpa_base 82 14.4 9.530 9.674 9.924 9.948 hybrid_alltoall_any 5155 16.4 0.443 2.006 7.449 9.499 multiply_cannon_metrocomm4 17395 15.5 0.064 0.074 3.414 9.456 dbcsr_mm_accdrv_process 41158 16.2 5.862 6.423 9.239 9.364 mp_irecv_dv 49801 16.2 3.286 9.200 3.286 9.200 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.843 9.130 pw_transfer 1535 11.6 0.084 0.099 8.628 8.726 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.406 8.511 grid_integrate_task_list 127 12.3 7.298 7.717 7.298 7.717 fft_wrap_pw1pw2_140 519 13.2 1.049 1.079 7.494 7.602 wfi_extrapolate 11 7.9 0.001 0.001 7.402 7.402 density_rs2pw 127 9.7 0.005 0.006 6.788 7.300 cp_fm_upper_to_full 104 14.8 5.725 7.235 5.725 7.235 cp_fm_cholesky_decompose 22 10.9 7.210 7.226 7.210 7.226 dbcsr_complete_redistribute 393 12.7 1.159 1.189 4.497 6.239 calculate_dm_sparse 127 9.5 0.001 0.001 5.905 6.022 fft3d_ps 1281 14.7 2.798 3.033 5.746 5.847 grid_collocate_task_list 127 9.7 5.111 5.528 5.111 5.528 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.452 5.457 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.616 5.213 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.369 5.112 mp_allgather_i34 2485 14.5 1.508 4.752 1.508 4.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.554 4.716 mp_alltoall_d11v 2401 14.1 4.126 4.606 4.126 4.606 mp_sum_l 7884 12.9 3.075 4.583 3.075 4.583 potential_pw2rs 127 12.3 0.019 0.021 4.248 4.256 transfer_fm_to_dbcsr 11 9.9 0.020 0.024 2.326 4.059 mp_alltoall_i22 712 14.1 1.963 3.908 1.963 3.908 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.862 3.892 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.732 3.732 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.499 3.538 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.018000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=903.545455, yerr=10.924232 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/18/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 4.353791E+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 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 1.155416E+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 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58091. MP_Allreduce 11102 1082. 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 0.011 0.032 189.982 189.983 qs_mol_dyn_low 1 2.0 0.003 0.003 189.467 189.481 qs_forces 11 3.9 0.004 0.004 189.355 189.361 qs_energies 11 4.9 0.001 0.002 182.226 182.238 scf_env_do_scf 11 5.9 0.001 0.001 165.552 165.564 velocity_verlet 10 3.0 0.002 0.002 125.462 125.465 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 119.197 119.198 qs_scf_new_mos 117 7.6 0.001 0.001 82.599 82.917 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.598 82.916 dbcsr_multiply_generic 2507 12.6 0.192 0.198 81.439 82.224 ot_scf_mini 117 9.6 0.003 0.003 78.018 78.356 multiply_cannon 2507 13.6 0.549 0.599 56.298 58.741 multiply_cannon_loop 2507 14.6 1.868 1.916 52.531 54.365 init_scf_loop 11 6.9 0.000 0.000 46.234 46.235 ot_mini 117 10.6 0.001 0.001 43.633 43.962 prepare_preconditioner 11 7.9 0.000 0.000 42.088 42.114 make_preconditioner 11 8.9 0.000 0.000 42.088 42.114 make_full_inverse_cholesky 11 9.9 0.013 0.023 35.741 40.839 multiply_cannon_multrec 30084 15.6 13.656 19.192 26.841 32.142 rebuild_ks_matrix 128 8.3 0.001 0.001 29.571 29.867 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 29.570 29.866 mp_waitall_1 147882 16.7 17.499 27.492 17.499 27.492 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.695 26.961 qs_ot_get_derivative 117 11.6 0.001 0.002 23.943 24.283 make_m2s 5014 13.6 0.091 0.096 20.645 21.797 make_images 5014 14.6 1.949 2.289 20.348 21.497 qs_ot_get_p 128 10.4 0.001 0.002 20.470 20.821 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.153 19.725 apply_single 128 13.6 0.001 0.001 19.152 19.725 ot_diis_step 117 11.6 0.017 0.018 19.563 19.564 cp_fm_upper_to_full 105 14.8 11.496 17.012 11.496 17.012 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.182 16.232 cp_fm_cholesky_invert 11 10.9 16.038 16.046 16.038 16.046 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.439 15.388 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.760 14.762 sum_up_and_integrate 128 10.3 0.002 0.003 14.028 14.061 integrate_v_rspace 128 11.3 0.003 0.004 13.968 14.002 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.873 13.916 calculate_rho_elec 128 8.7 0.169 0.185 13.872 13.915 dbcsr_mm_accdrv_process 62264 16.2 8.553 9.677 12.755 13.432 make_images_data 5014 15.6 0.065 0.074 11.133 13.141 dbcsr_complete_redistribute 395 12.7 1.480 1.592 9.044 12.923 hybrid_alltoall_any 5200 16.5 0.531 2.212 9.808 12.037 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.689 11.538 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.506 11.519 cp_fm_diag_elpa_base 83 14.4 10.526 10.833 11.500 11.511 multiply_cannon_sync_h2d 30084 15.6 10.456 11.293 10.456 11.293 init_scf_run 11 5.9 0.000 0.001 10.745 10.747 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.745 10.746 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.888 10.131 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.325 10.054 pw_transfer 1547 11.6 0.085 0.102 9.270 9.334 mp_alltoall_i22 716 14.1 5.579 9.321 5.579 9.321 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.045 9.114 fft_wrap_pw1pw2_140 523 13.2 1.247 1.280 8.031 8.119 grid_integrate_task_list 128 12.3 7.536 7.965 7.536 7.965 wfi_extrapolate 11 7.9 0.001 0.001 7.678 7.678 cp_fm_cholesky_decompose 22 10.9 7.488 7.603 7.488 7.603 density_rs2pw 128 9.7 0.005 0.006 6.955 7.404 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.769 7.389 mp_irecv_dv 76098 16.2 2.616 7.114 2.616 7.114 calculate_dm_sparse 128 9.5 0.001 0.001 6.261 6.358 fft3d_ps 1291 14.7 2.996 3.062 6.093 6.148 grid_collocate_task_list 128 9.7 5.285 5.648 5.285 5.648 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.482 5.547 mp_alltoall_d11v 2415 14.1 4.820 5.376 4.820 5.376 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.934 5.107 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.426 4.508 qs_energies_init_hamiltonians 11 5.9 0.004 0.008 4.442 4.443 potential_pw2rs 128 12.3 0.023 0.024 4.349 4.361 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.245 4.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.983000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1093.545455, yerr=18.593076 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.535119E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.075 170.552 170.565 qs_mol_dyn_low 1 2.0 0.003 0.003 170.073 170.147 qs_forces 11 3.9 0.005 0.007 169.866 169.874 qs_energies 11 4.9 0.001 0.002 162.457 162.468 scf_env_do_scf 11 5.9 0.001 0.001 145.125 145.137 velocity_verlet 10 3.0 0.002 0.002 112.451 112.462 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 109.760 109.761 dbcsr_multiply_generic 2527 12.6 0.184 0.191 72.572 73.056 qs_scf_new_mos 118 7.6 0.001 0.001 72.436 72.565 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.435 72.565 ot_scf_mini 118 9.6 0.003 0.004 67.991 68.114 multiply_cannon 2527 13.6 0.566 0.585 53.956 56.916 multiply_cannon_loop 2527 14.6 0.817 0.851 50.804 51.974 ot_mini 118 10.6 0.001 0.001 37.988 38.099 init_scf_loop 11 6.9 0.000 0.000 35.210 35.211 prepare_preconditioner 11 7.9 0.000 0.000 31.368 31.400 make_preconditioner 11 8.9 0.000 0.000 31.368 31.399 mp_waitall_1 126780 16.7 24.903 31.397 24.903 31.397 rebuild_ks_matrix 129 8.3 0.001 0.001 29.375 29.512 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.374 29.512 make_full_inverse_cholesky 11 9.9 0.016 0.028 29.199 29.469 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.798 26.923 multiply_cannon_multrec 10108 15.6 10.381 16.162 18.019 22.629 qs_ot_get_derivative 118 11.6 0.002 0.002 20.601 20.716 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 12.972 20.706 cp_fm_cholesky_invert 11 10.9 17.806 17.812 17.806 17.812 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.427 17.610 apply_single 129 13.6 0.001 0.001 17.427 17.610 qs_ot_get_p 129 10.4 0.001 0.001 17.210 17.364 ot_diis_step 118 11.6 0.020 0.022 17.318 17.318 make_m2s 5054 13.6 0.064 0.067 14.664 15.626 make_images 5054 14.6 2.177 2.594 14.353 15.307 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.310 14.348 calculate_rho_elec 129 8.7 0.255 0.266 14.309 14.348 sum_up_and_integrate 129 10.3 0.002 0.002 14.268 14.322 integrate_v_rspace 129 11.3 0.004 0.004 14.207 14.263 qs_ot_p2m_diag 83 11.4 0.496 0.502 13.463 13.478 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.298 12.298 multiply_cannon_sync_h2d 10108 15.6 10.784 11.236 10.784 11.236 init_scf_run 11 5.9 0.000 0.001 10.483 10.483 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.483 10.483 pw_transfer 1559 11.6 0.085 0.093 9.971 10.003 make_images_data 5054 15.6 0.055 0.065 8.395 9.819 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.747 9.782 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.437 9.449 hybrid_alltoall_any 5240 16.5 0.844 3.799 8.238 9.447 cp_fm_diag_elpa_base 83 14.4 9.202 9.285 9.433 9.445 fft_wrap_pw1pw2_140 527 13.2 1.692 1.736 8.645 8.679 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.015 8.092 grid_integrate_task_list 129 12.3 7.815 8.067 7.815 8.067 dbcsr_mm_accdrv_process 20926 16.1 3.303 4.523 7.285 8.022 cp_fm_cholesky_decompose 22 10.9 7.804 7.905 7.804 7.905 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.321 7.859 wfi_extrapolate 11 7.9 0.001 0.001 7.467 7.467 density_rs2pw 129 9.7 0.005 0.005 7.001 7.177 calculate_dm_sparse 129 9.5 0.001 0.001 6.132 6.207 fft3d_ps 1301 14.7 3.158 3.256 6.168 6.200 multiply_cannon_metrocomm4 7581 15.6 0.026 0.029 1.893 5.820 grid_collocate_task_list 129 9.7 5.605 5.768 5.605 5.768 dbcsr_complete_redistribute 395 12.7 2.089 2.176 5.345 5.736 mp_irecv_dv 29086 15.9 1.852 5.731 1.852 5.731 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 5.279 5.283 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.211 5.249 mp_alltoall_d11v 2423 14.1 4.701 5.021 4.701 5.021 mp_allgather_i34 2527 14.6 1.323 4.570 1.323 4.570 potential_pw2rs 129 12.3 0.026 0.026 4.383 4.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.080 4.141 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.582 3.869 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.486 3.803 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.691 3.787 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.666 3.711 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.450 3.535 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.483 3.498 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.565000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1429.454545, yerr=38.031957 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.168756E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 59236. MP_Allreduce 11065 1504. MP_Sync 86 MP_Alltoall 1700 36954382. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.177 288.461 288.502 qs_mol_dyn_low 1 2.0 0.003 0.003 287.705 287.726 qs_forces 11 3.9 0.005 0.005 287.587 287.589 qs_energies 11 4.9 0.002 0.002 278.743 278.751 scf_env_do_scf 11 5.9 0.001 0.001 256.040 256.048 velocity_verlet 10 3.0 0.002 0.002 208.444 208.455 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 132.275 132.277 init_scf_loop 11 6.9 0.000 0.000 123.493 123.496 prepare_preconditioner 11 7.9 0.000 0.000 118.640 118.658 make_preconditioner 11 8.9 0.000 0.000 118.640 118.658 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.745 115.816 qs_scf_new_mos 116 7.6 0.001 0.001 88.000 88.037 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.999 88.036 ot_scf_mini 116 9.6 0.004 0.004 83.212 83.226 dbcsr_multiply_generic 2485 12.5 0.214 0.225 80.911 81.304 cp_fm_upper_to_full 104 14.8 52.685 75.575 52.685 75.575 multiply_cannon 2485 13.5 0.673 0.722 58.256 59.860 multiply_cannon_loop 2485 14.5 1.044 1.055 54.275 55.400 ot_mini 116 10.6 0.001 0.001 43.465 43.479 dbcsr_complete_redistribute 393 12.7 3.984 4.040 29.887 42.791 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.389 39.246 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.852 36.614 rebuild_ks_matrix 127 8.3 0.001 0.001 34.362 34.380 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.362 34.380 mp_alltoall_i22 712 14.1 21.639 34.355 21.639 34.355 cp_fm_cholesky_invert 11 10.9 32.997 33.003 32.997 33.003 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.836 31.865 mp_waitall_1 102768 16.8 27.388 31.737 27.388 31.737 qs_ot_get_p 127 10.4 0.001 0.001 24.781 24.824 qs_ot_get_derivative 116 11.6 0.002 0.002 24.323 24.333 qs_ot_p2m_diag 82 11.4 0.869 0.875 20.794 20.822 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.508 19.644 make_m2s 4970 13.5 0.072 0.075 18.167 19.644 make_images 4970 14.5 3.036 3.229 17.689 19.168 ot_diis_step 116 11.6 0.021 0.022 19.116 19.117 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.029 19.033 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.479 18.561 apply_single 127 13.6 0.001 0.001 18.478 18.561 multiply_cannon_multrec 9940 15.5 10.204 11.997 17.990 18.083 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.158 17.172 calculate_rho_elec 127 8.7 0.474 0.475 17.158 17.172 sum_up_and_integrate 127 10.3 0.002 0.002 16.255 16.345 integrate_v_rspace 127 11.3 0.004 0.005 16.192 16.282 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.914 15.914 cp_fm_diag_elpa_base 82 14.4 11.539 13.122 15.910 15.911 multiply_cannon_sync_h2d 9940 15.5 14.187 14.209 14.187 14.209 pw_transfer 1535 11.6 0.095 0.096 13.016 13.025 hybrid_alltoall_any 5155 16.4 1.304 3.045 10.554 12.828 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 12.780 12.787 make_images_data 4970 15.5 0.063 0.069 10.211 12.346 init_scf_run 11 5.9 0.000 0.001 12.261 12.264 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.260 12.264 fft_wrap_pw1pw2_140 519 13.2 3.193 3.277 11.396 11.405 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.520 9.529 dbcsr_mm_accdrv_process 20590 16.0 4.125 5.767 7.538 9.394 wfi_extrapolate 11 7.9 0.001 0.001 9.096 9.096 cp_fm_cholesky_decompose 22 10.9 8.868 8.891 8.868 8.891 grid_integrate_task_list 127 12.3 8.492 8.686 8.492 8.686 density_rs2pw 127 9.7 0.005 0.005 8.245 8.289 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.229 8.230 fft3d_ps 1281 14.7 3.947 3.960 7.285 7.337 mp_alltoall_d11v 2401 14.1 6.878 7.023 6.878 7.023 calculate_dm_sparse 127 9.5 0.001 0.001 6.663 6.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.479 6.524 grid_collocate_task_list 127 9.7 6.364 6.440 6.364 6.440 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.189 6.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.502000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2858.000000, yerr=156.441682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 1.260642E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.013 0.036 85.077 85.078 qs_energies 1 2.0 0.000 0.000 84.585 84.597 ls_scf 1 3.0 0.000 0.000 83.645 83.657 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.475 72.629 multiply_cannon 111 7.7 0.017 0.020 55.778 56.927 multiply_cannon_loop 111 8.7 0.230 0.245 52.383 53.730 ls_scf_main 1 4.0 0.000 0.000 52.210 52.221 density_matrix_trs4 2 5.0 0.002 0.003 46.679 46.748 ls_scf_init_scf 1 4.0 0.000 0.000 28.401 28.403 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.286 27.332 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.050 25.076 mp_waitall_1 11031 10.9 22.208 24.945 22.208 24.945 multiply_cannon_multrec 2664 9.7 8.140 8.955 15.632 17.455 multiply_cannon_sync_h2d 2664 9.7 13.554 15.537 13.554 15.537 make_m2s 222 7.7 0.009 0.011 13.076 13.595 make_images 222 8.7 0.098 0.108 13.054 13.576 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.494 12.752 make_images_data 222 9.7 0.004 0.005 7.602 8.208 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.724 7.109 8.091 hybrid_alltoall_any 227 10.6 0.216 1.829 6.511 8.021 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.443 7.501 dbcsr_mm_accdrv_process_sort 4760 11.4 6.296 7.221 6.296 7.221 calculate_norms 4752 9.8 5.539 6.244 5.539 6.244 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.063 5.191 mp_sum_l 887 5.1 3.086 4.467 3.086 4.467 make_images_sizes 222 9.7 0.000 0.000 0.807 3.752 mp_alltoall_i44 222 10.7 0.807 3.752 0.807 3.752 arnoldi_extremal 4 6.8 0.000 0.000 3.320 3.344 arnoldi_normal_ev 4 7.8 0.009 0.019 3.320 3.344 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.287 3.344 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.059 3.300 mp_irecv_dv 6231 10.9 2.043 3.286 2.043 3.286 build_subspace 16 8.4 0.009 0.012 3.208 3.210 ls_scf_post 1 4.0 0.000 0.000 3.034 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.849 2.892 dbcsr_special_finalize 555 9.7 0.005 0.006 2.313 2.737 dbcsr_merge_single_wm 555 10.7 0.455 0.622 2.305 2.728 make_images_pack 222 9.7 2.207 2.666 2.209 2.668 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.315 2.560 dbcsr_sort_data 658 11.4 2.108 2.488 2.108 2.488 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.461 2.067 2.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.314 2.407 buffer_matrices_ensure_size 222 8.7 1.748 2.102 1.748 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.765 1.766 rebuild_ks_matrix 3 7.3 0.000 0.000 1.756 1.757 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.756 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.078000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 2.138763E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.038 89.570 89.571 qs_energies 1 2.0 0.000 0.000 89.128 89.132 ls_scf 1 3.0 0.000 0.000 87.748 87.752 dbcsr_multiply_generic 111 6.7 0.015 0.015 73.716 74.001 multiply_cannon 111 7.7 0.027 0.039 52.491 56.694 ls_scf_main 1 4.0 0.000 0.000 54.201 54.202 multiply_cannon_loop 111 8.7 0.136 0.148 49.749 53.021 density_matrix_trs4 2 5.0 0.002 0.003 48.557 48.720 ls_scf_init_scf 1 4.0 0.000 0.000 29.910 29.912 mp_waitall_1 9105 10.9 20.546 29.343 20.546 29.343 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.740 28.828 multiply_cannon_multrec 1332 9.7 13.226 17.061 22.623 27.611 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.099 26.108 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.328 20.269 make_m2s 222 7.7 0.006 0.008 15.038 15.625 make_images 222 8.7 1.368 1.693 15.007 15.595 dbcsr_mm_accdrv_process 4041 10.4 0.320 0.558 8.994 10.591 dbcsr_mm_accdrv_process_sort 4041 11.4 8.496 10.033 8.496 10.033 make_images_data 222 9.7 0.004 0.005 8.705 9.636 hybrid_alltoall_any 227 10.6 0.543 2.556 8.185 9.506 mp_sum_l 887 5.1 4.706 7.677 4.706 7.677 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.231 7.673 mp_irecv_dv 3311 11.0 3.210 7.616 3.210 7.616 calculate_norms 2376 9.8 6.058 6.782 6.058 6.782 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.629 6.196 multiply_cannon_sync_h2d 1332 9.7 4.857 6.141 4.857 6.141 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.906 5.148 arnoldi_extremal 4 6.8 0.000 0.000 4.667 4.679 arnoldi_normal_ev 4 7.8 0.001 0.005 4.666 4.678 build_subspace 16 8.4 0.014 0.021 4.416 4.419 ls_scf_post 1 4.0 0.000 0.000 3.637 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.371 3.459 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.182 3.425 dbcsr_matrix_vector_mult_local 304 10.0 2.774 3.266 2.776 3.268 mp_allgather_i34 111 8.7 0.723 2.923 0.723 2.923 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.620 2.736 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.160 2.563 dbcsr_data_new 4174 10.1 2.127 2.427 2.127 2.427 make_images_pack 222 9.7 1.835 2.135 1.837 2.137 dbcsr_sort_data 436 11.2 1.839 2.077 1.839 2.077 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.842 1.845 rebuild_ks_matrix 3 7.3 0.000 0.000 1.830 1.832 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.830 1.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.571000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1866.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 2.832773E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.045 91.107 91.107 qs_energies 1 2.0 0.000 0.000 90.579 90.582 ls_scf 1 3.0 0.000 0.000 89.165 89.169 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.950 74.330 ls_scf_main 1 4.0 0.000 0.000 55.603 55.607 multiply_cannon 111 7.7 0.029 0.057 51.499 55.152 multiply_cannon_loop 111 8.7 0.116 0.129 48.750 52.751 density_matrix_trs4 2 5.0 0.002 0.003 49.743 49.929 mp_waitall_1 7281 11.0 22.602 32.550 22.602 32.550 ls_scf_init_scf 1 4.0 0.000 0.000 29.915 29.918 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.731 28.804 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.404 26.423 multiply_cannon_multrec 888 9.7 12.689 15.438 21.369 24.748 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.425 22.215 make_m2s 222 7.7 0.006 0.007 15.864 16.636 make_images 222 8.7 1.581 1.852 15.826 16.595 make_images_data 222 9.7 0.004 0.004 9.186 10.220 hybrid_alltoall_any 227 10.6 0.641 2.950 8.847 9.817 dbcsr_mm_accdrv_process 3754 10.4 0.299 0.497 8.203 9.488 dbcsr_mm_accdrv_process_sort 3754 11.4 7.769 8.991 7.769 8.991 mp_sum_l 887 5.1 4.888 8.222 4.888 8.222 multiply_cannon_sync_h2d 888 9.7 6.040 7.257 6.040 7.257 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.439 7.130 mp_irecv_dv 2335 11.1 2.424 7.074 2.424 7.074 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.763 6.556 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.635 6.555 arnoldi_extremal 4 6.8 0.000 0.000 5.071 5.094 arnoldi_normal_ev 4 7.8 0.001 0.005 5.071 5.094 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.690 4.915 build_subspace 16 8.4 0.014 0.020 4.772 4.779 calculate_norms 1584 9.8 4.382 4.703 4.382 4.703 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.464 3.789 ls_scf_post 1 4.0 0.000 0.000 3.647 3.652 dbcsr_matrix_vector_mult_local 304 10.0 3.053 3.628 3.055 3.630 mp_allgather_i34 111 8.7 0.868 3.560 0.868 3.560 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.844 2.955 dbcsr_data_new 4116 9.9 2.111 2.437 2.111 2.437 dbcsr_sort_data 325 11.1 1.867 2.141 1.867 2.141 make_images_sizes 222 9.7 0.000 0.000 0.855 1.939 mp_alltoall_i44 222 10.7 0.854 1.938 0.854 1.938 make_images_pack 222 9.7 1.626 1.889 1.629 1.891 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 dbcsr_finalize 304 7.8 0.026 0.032 1.604 1.872 rebuild_ks_matrix 3 7.3 0.000 0.000 1.869 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.869 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.107000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2252.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 3.380916E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.044 96.985 96.985 qs_energies 1 2.0 0.000 0.000 96.314 96.318 ls_scf 1 3.0 0.000 0.000 94.618 94.623 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.395 78.606 ls_scf_main 1 4.0 0.000 0.000 58.771 58.771 multiply_cannon 111 7.7 0.037 0.082 51.731 55.938 density_matrix_trs4 2 5.0 0.002 0.003 52.661 52.783 multiply_cannon_loop 111 8.7 0.154 0.166 46.738 49.387 ls_scf_init_scf 1 4.0 0.000 0.000 32.678 32.680 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.471 31.529 mp_waitall_1 6369 11.0 22.736 30.247 22.736 30.247 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.010 29.021 multiply_cannon_multrec 1332 9.7 14.146 17.313 22.140 25.450 make_m2s 222 7.7 0.007 0.007 21.205 22.496 make_images 222 8.7 3.146 3.610 21.155 22.448 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.199 16.625 make_images_data 222 9.7 0.004 0.004 11.816 13.541 hybrid_alltoall_any 227 10.6 0.800 3.786 11.208 13.003 dbcsr_mm_accdrv_process 3641 10.4 0.285 0.483 7.623 9.144 dbcsr_mm_accdrv_process_sort 3641 11.4 7.182 8.647 7.182 8.647 mp_sum_l 887 5.1 3.873 6.933 3.873 6.933 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.097 6.061 multiply_cannon_sync_h2d 1332 9.7 5.462 6.051 5.462 6.051 mp_irecv_dv 3229 10.9 2.072 5.984 2.072 5.984 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.006 5.358 arnoldi_extremal 4 6.8 0.000 0.000 5.243 5.254 arnoldi_normal_ev 4 7.8 0.016 0.027 5.243 5.254 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.555 5.170 build_subspace 16 8.4 0.014 0.021 4.893 4.903 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.522 4.692 calculate_norms 2376 9.8 4.176 4.496 4.176 4.496 mp_allgather_i34 111 8.7 2.098 4.281 2.098 4.281 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.616 3.919 dbcsr_matrix_vector_mult_local 304 10.0 3.229 3.729 3.231 3.731 dbcsr_sort_data 658 11.4 3.077 3.350 3.077 3.350 ls_scf_post 1 4.0 0.000 0.000 3.170 3.174 dbcsr_special_finalize 555 9.7 0.006 0.007 2.828 3.160 dbcsr_merge_single_wm 555 10.7 0.540 0.663 2.819 3.152 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.006 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.923 2.973 dbcsr_data_release 10477 10.7 1.604 2.414 1.604 2.414 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.985000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2741.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 4.730360E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.055 91.867 91.867 qs_energies 1 2.0 0.000 0.000 91.129 91.134 ls_scf 1 3.0 0.000 0.000 89.157 89.162 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.834 71.084 ls_scf_main 1 4.0 0.000 0.000 56.632 56.632 multiply_cannon 111 7.7 0.078 0.129 52.955 56.420 multiply_cannon_loop 111 8.7 0.088 0.094 50.397 51.672 density_matrix_trs4 2 5.0 0.002 0.003 49.693 49.769 ls_scf_init_scf 1 4.0 0.000 0.000 29.104 29.105 mp_waitall_1 5436 11.0 24.722 28.850 24.722 28.850 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.851 27.885 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.780 25.790 multiply_cannon_multrec 444 9.7 13.549 16.595 20.708 22.250 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.628 15.813 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.753 14.741 make_m2s 222 7.7 0.005 0.005 13.392 14.328 make_images 222 8.7 2.038 2.474 13.325 14.259 hybrid_alltoall_any 227 10.6 0.801 3.793 8.069 9.623 make_images_data 222 9.7 0.003 0.004 8.158 9.601 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.406 6.852 7.991 multiply_cannon_sync_h2d 444 9.7 6.798 7.947 6.798 7.947 dbcsr_mm_accdrv_process_sort 3003 11.4 6.490 7.590 6.490 7.590 arnoldi_extremal 4 6.8 0.000 0.000 5.776 5.786 arnoldi_normal_ev 4 7.8 0.002 0.005 5.776 5.786 build_subspace 16 8.4 0.015 0.020 5.391 5.401 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.373 4.554 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.174 4.367 dbcsr_matrix_vector_mult_local 304 10.0 3.735 4.182 3.737 4.184 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.560 4.064 mp_irecv_dv 1241 11.2 1.547 4.032 1.547 4.032 mp_sum_l 887 5.1 2.480 3.948 2.480 3.948 mp_allgather_i34 111 8.7 1.138 3.699 1.138 3.699 calculate_norms 792 9.8 3.613 3.678 3.613 3.678 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.607 ls_scf_post 1 4.0 0.000 0.000 3.421 3.426 ls_scf_store_result 1 5.0 0.000 0.000 3.208 3.247 make_images_sizes 222 9.7 0.000 0.000 0.851 3.122 mp_alltoall_i44 222 10.7 0.851 3.122 0.851 3.122 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.711 2.894 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.337 dbcsr_data_new 4608 9.7 1.790 2.237 1.790 2.237 dbcsr_merge_all 275 8.9 0.479 0.523 2.053 2.174 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.029 2.030 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.996 1.997 qs_energies_init_hamiltonians 1 3.0 0.010 0.015 1.955 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.867000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_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 8.884130E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.089 105.287 105.288 qs_energies 1 2.0 0.000 0.000 103.958 103.966 ls_scf 1 3.0 0.000 0.000 100.980 100.987 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.592 74.733 ls_scf_main 1 4.0 0.000 0.000 63.608 63.609 density_matrix_trs4 2 5.0 0.002 0.003 54.585 54.635 multiply_cannon 111 7.7 0.166 0.226 48.241 50.151 multiply_cannon_loop 111 8.7 0.098 0.101 45.258 45.958 ls_scf_init_scf 1 4.0 0.000 0.000 33.596 33.597 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.027 32.054 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.342 29.350 mp_waitall_1 4527 11.1 21.814 25.341 21.814 25.341 make_m2s 222 7.7 0.005 0.005 22.687 23.753 make_images 222 8.7 3.571 3.869 22.579 23.645 multiply_cannon_multrec 444 9.7 17.723 18.414 22.416 23.123 hybrid_alltoall_any 227 10.6 1.653 3.612 12.840 15.810 make_images_data 222 9.7 0.003 0.004 13.086 15.198 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.134 10.654 multiply_cannon_sync_h2d 444 9.7 8.792 8.852 8.792 8.852 arnoldi_extremal 4 6.8 0.000 0.000 7.369 7.378 arnoldi_normal_ev 4 7.8 0.003 0.009 7.369 7.378 build_subspace 16 8.4 0.026 0.036 6.829 6.842 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.494 5.642 dbcsr_matrix_vector_mult_local 304 10.0 5.091 5.376 5.094 5.378 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.169 5.265 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.828 5.078 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.355 4.510 4.640 dbcsr_mm_accdrv_process_sort 1814 11.4 4.175 4.308 4.175 4.308 ls_scf_post 1 4.0 0.000 0.000 3.776 3.784 make_images_sizes 222 9.7 0.000 0.000 1.480 3.600 mp_alltoall_i44 222 10.7 1.479 3.599 1.479 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.500 3.536 mp_allgather_i34 111 8.7 1.074 3.441 1.074 3.441 calculate_norms 792 9.8 3.228 3.266 3.228 3.266 dbcsr_finalize 304 7.8 0.082 0.089 3.073 3.166 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.947 2.947 dbcsr_merge_all 275 8.9 0.882 0.912 2.858 2.945 dbcsr_complete_redistribute 5 7.6 1.434 1.469 2.746 2.866 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.520 dbcsr_sort_data 325 11.1 2.437 2.502 2.437 2.502 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.461 2.464 rebuild_ks_matrix 3 7.3 0.000 0.000 2.395 2.397 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.395 2.397 dbcsr_new_transposed 4 7.5 0.240 0.250 2.288 2.307 dbcsr_data_new 6591 9.6 1.835 2.259 1.835 2.259 dbcsr_frobenius_norm 74 6.6 2.058 2.138 2.215 2.255 dbcsr_add_d 103 6.2 0.000 0.000 2.123 2.202 dbcsr_add_anytype 103 7.2 0.858 0.890 2.123 2.202 dbcsr_data_release 12724 10.6 1.975 2.179 1.975 2.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.288000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6991.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/225305aa1575fcf7b2a0a553af71d6a327fd5eb1_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.981440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1500530417. 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 0.019 0.047 223.690 223.692 qs_mol_dyn_low 1 2.0 0.004 0.017 222.513 222.624 qs_forces 5 3.8 0.005 0.017 222.366 222.367 qs_energies 5 4.8 0.001 0.027 219.450 219.467 scf_env_do_scf 5 5.8 0.000 0.001 205.856 205.858 scf_env_do_scf_inner_loop 105 6.6 0.002 0.012 179.810 179.819 qs_scf_new_mos 105 7.6 0.000 0.001 140.999 141.171 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.999 141.170 dbcsr_multiply_generic 1445 12.2 0.129 0.137 131.822 132.432 ot_scf_mini 105 9.6 0.003 0.004 131.123 131.240 multiply_cannon 1445 13.2 0.272 0.283 113.256 115.294 multiply_cannon_loop 1445 14.2 2.837 2.985 111.572 112.945 velocity_verlet 4 3.0 0.002 0.016 105.054 105.055 ot_mini 105 10.6 0.001 0.003 59.578 59.712 multiply_cannon_multrec 69360 15.2 29.685 34.846 39.720 44.686 qs_ot_get_p 112 10.4 0.001 0.001 41.489 41.839 mp_waitall_1 488190 16.1 34.394 39.859 34.394 39.859 qs_ot_get_derivative 55 11.6 0.001 0.001 37.866 38.006 multiply_cannon_sync_h2d 69360 15.2 29.161 33.086 29.161 33.086 multiply_cannon_metrocomm3 69360 15.2 0.202 0.214 25.630 31.545 qs_ot_p2m_diag 40 11.0 0.020 0.031 30.479 30.641 rebuild_ks_matrix 110 8.4 0.000 0.000 28.367 28.541 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.034 28.367 28.541 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.322 27.322 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.049 26.208 init_scf_loop 7 6.6 0.000 0.000 26.017 26.018 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.979 23.213 apply_single 62 13.6 0.000 0.000 22.979 23.213 cp_fm_syevd 40 13.0 0.000 0.000 22.331 22.481 prepare_preconditioner 7 7.6 0.000 0.000 21.229 21.257 make_preconditioner 7 8.6 0.000 0.000 21.229 21.257 ot_new_cg_direction 55 11.6 0.001 0.002 20.983 20.983 cp_fm_redistribute_end 40 14.0 8.851 17.665 8.856 17.666 cp_fm_syevd_base 40 14.0 8.803 17.617 8.803 17.617 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.077 16.458 calculate_rho_elec 110 8.6 0.030 0.032 16.076 16.457 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.879 15.142 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.071 14.131 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.622 13.718 mp_sum_l 4764 12.2 11.896 12.770 11.896 12.770 calculate_dm_sparse 110 9.5 0.000 0.001 11.534 11.706 pw_transfer 1645 12.4 0.080 0.097 11.395 11.577 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.257 11.440 init_scf_run 5 5.8 0.000 0.001 10.990 10.991 scf_env_initial_rho_setup 5 6.8 0.002 0.005 10.990 10.991 density_rs2pw 110 9.6 0.005 0.006 10.388 10.884 dbcsr_mm_accdrv_process 154766 15.8 6.213 6.433 9.903 10.713 qs_vxc_create 110 10.4 0.002 0.004 10.309 10.357 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.116 10.179 fft_wrap_pw1pw2_240 915 15.0 0.835 0.916 9.886 10.086 cp_fm_cholesky_invert 7 10.6 9.837 9.845 9.837 9.845 check_diag 80 13.5 8.595 8.887 9.322 9.473 fft3d_pb 915 16.0 2.412 2.664 8.055 8.280 sum_up_and_integrate 60 10.3 0.001 0.004 8.092 8.113 integrate_v_rspace 60 11.3 0.002 0.034 8.075 8.097 transfer_rs2pw 445 10.6 0.007 0.008 7.164 7.630 xc_rho_set_and_dset_create 110 12.4 0.075 0.095 7.241 7.515 acc_transpose_blocks 69360 15.2 0.360 0.378 6.570 7.031 make_full_single_inverse 7 9.6 0.001 0.001 6.901 6.928 xc_vxc_pw_create 60 11.3 0.039 0.049 6.862 6.909 multiply_cannon_metrocomm1 69360 15.2 0.094 0.100 4.558 6.908 make_m2s 2890 13.2 0.078 0.090 6.165 6.739 make_images 2890 14.2 0.236 0.258 6.057 6.628 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.359 6.375 calculate_first_density_matrix 1 7.0 0.000 0.001 6.222 6.238 xc_pw_derive 510 13.4 0.006 0.007 6.065 6.131 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.952 6.012 mp_alltoall_z22v 2340 17.7 5.042 5.326 5.042 5.326 mp_waitany 7680 13.5 4.291 4.824 4.291 4.824 multiply_cannon_metrocomm4 67915 15.2 0.184 0.202 2.017 4.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=223.692000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 225305aa1575fcf7b2a0a553af71d6a327fd5eb1 Summary: empty Status: OK