=== 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: 9ee8bdf25fd706f00eae9f0d758bddb493a004a9 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/01 job id: 45062283 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/02 job id: 45062284 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/03 job id: 45062285 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/04 job id: 45062286 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/05 job id: 45062287 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/06 job id: 45062288 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/07 job id: 45062289 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/08 job id: 45062290 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/09 job id: 45062291 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/10 job id: 45062293 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/11 job id: 45062294 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/12 job id: 45062295 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/13 job id: 45062297 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/14 job id: 45062298 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/15 job id: 45062300 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/16 job id: 45062301 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/17 job id: 45062304 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/18 job id: 45062306 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/19 job id: 45062307 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/20 job id: 45062308 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/21 job id: 45062309 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/22 job id: 45062310 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/23 job id: 45062312 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/24 job id: 45062313 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/25 job id: 45062314 --- 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/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/26 job id: 45062315 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 133.876 133.876 farming_run 1 2.0 133.408 133.409 133.848 133.852 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460158E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 115.560 115.561 qs_energies 1 2.0 0.000 0.000 115.357 115.357 mp2_main 1 3.0 0.000 0.000 113.481 113.481 mp2_gpw_main 1 4.0 0.019 0.026 112.629 112.629 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 93.670 94.072 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.472 55.874 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.792 47.108 get_2c_integrals 1 6.0 0.000 0.001 37.289 38.024 integrate_v_rspace 273 8.0 0.439 0.449 25.205 30.323 pw_transfer 6555 10.6 0.372 0.395 27.421 27.970 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.080 26.651 grid_integrate_task_list 273 9.0 21.005 26.637 21.005 26.637 fft_wrap_pw1pw2_100 2178 12.4 1.172 1.385 23.599 24.169 compute_2c_integrals 1 7.0 0.002 0.003 19.658 19.658 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.878 19.393 mp2_eri_2c_integrate_gpw 1 9.0 2.372 2.425 18.874 19.389 rpa_ri_compute_en 1 5.0 0.006 0.021 18.853 19.103 cp_fm_cholesky_decompose 12 8.2 17.553 18.295 17.553 18.295 cholesky_decomp 1 7.0 0.000 0.000 16.474 17.224 fft3d_s 5443 13.4 16.196 16.679 16.218 16.701 ao_to_mo_and_store_B_mult_1 272 7.0 10.844 15.562 10.844 15.562 calculate_wavefunction 272 8.0 5.377 5.529 12.502 13.113 rpa_num_int 1 6.0 0.000 0.001 10.517 10.517 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.479 10.494 calc_mat_Q 8 8.0 0.000 0.000 9.350 9.442 contract_S_to_Q 8 9.0 0.000 0.000 8.771 8.859 calc_potential_gpw 544 9.5 0.005 0.006 8.293 8.654 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.224 8.528 parallel_gemm_fm 14 9.1 0.000 0.000 8.355 8.427 parallel_gemm_fm_cosma 14 10.1 8.355 8.427 8.355 8.427 potential_pw2rs 545 10.0 0.108 0.110 7.674 8.274 create_integ_mat 1 6.0 0.006 0.008 7.819 7.819 collocate_single_gaussian 272 10.0 0.039 0.042 7.457 7.778 array2fm 1 7.0 0.000 0.000 6.719 7.214 pw_scatter_s 2720 13.7 4.437 4.674 4.437 4.674 pw_gather_s 2722 13.2 3.863 4.155 3.863 4.155 array2fm_buffer_send 1 8.0 2.988 3.128 2.988 3.128 pw_poisson_solve 545 10.5 1.123 1.181 2.174 2.362 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.629364, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.045 396.068 396.070 farming_run 1 2.0 395.367 395.371 396.018 396.033 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.224815E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2369671. 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 0.012 0.043 209.706 209.706 qs_energies 1 2.0 0.000 0.000 209.458 209.467 scf_env_do_scf 1 3.0 0.000 0.000 106.073 106.073 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.133 105.141 rebuild_ks_matrix 4 6.0 0.000 0.000 105.132 105.139 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.132 105.139 hfx_ks_matrix 4 8.0 0.001 0.001 104.732 104.736 integrate_four_center 4 9.0 0.144 0.456 104.732 104.735 mp2_main 1 3.0 0.000 0.000 103.086 103.095 mp2_gpw_main 1 4.0 0.037 0.059 102.036 102.045 integrate_four_center_main 4 10.0 0.072 0.360 96.574 99.358 integrate_four_center_bin 264 11.0 96.502 99.356 96.502 99.356 init_scf_loop 1 4.0 0.000 0.000 91.713 91.713 mp2_ri_gpw_compute_in 1 5.0 0.067 0.078 74.961 76.012 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.493 55.541 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.288 47.314 integrate_v_rspace 95 8.0 0.399 0.569 28.683 33.513 pw_transfer 2240 10.6 0.144 0.165 29.950 30.430 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.962 29.464 grid_integrate_task_list 95 9.0 23.953 29.006 23.953 29.006 ao_to_mo_and_store_B_mult_1 91 7.0 10.521 28.820 10.521 28.820 mp2_ri_gpw_compute_en 1 5.0 0.056 0.067 26.922 28.695 fft_wrap_pw1pw2_100 730 12.4 1.271 1.461 26.656 27.175 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.835 1.904 25.220 25.232 get_2c_integrals 1 6.0 0.000 0.000 20.374 20.406 compute_2c_integrals 1 7.0 0.004 0.015 19.350 19.365 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.879 19.188 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.834 18.876 19.187 fft3d_s 1823 13.4 18.423 18.789 18.437 18.802 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.358 14.358 calculate_wavefunction 91 8.0 2.012 2.041 9.735 9.974 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.601 8.801 9.641 potential_pw2rs 186 10.0 0.034 0.035 8.617 9.231 local_gemm 172 8.0 8.243 9.075 8.243 9.075 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.515 8.199 8.622 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.253 8.597 calc_potential_gpw 182 9.5 0.002 0.002 7.958 8.231 collocate_single_gaussian 91 10.0 0.017 0.024 7.881 8.126 mp_sendrecv_dm3 2068 8.0 6.246 6.670 6.246 6.670 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.418 6.340 6.418 mp_sync 38 10.4 2.830 5.696 2.830 5.696 pw_gather_s 912 13.2 4.905 5.360 4.905 5.360 pw_scatter_s 910 13.7 3.957 4.357 3.957 4.357 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.038784, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.960832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578578. 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 0.016 0.032 54.252 54.255 qs_mol_dyn_low 1 2.0 0.003 0.003 53.910 53.918 qs_forces 11 3.9 0.009 0.011 53.844 53.846 qs_energies 11 4.9 0.001 0.002 52.350 52.360 scf_env_do_scf 11 5.9 0.000 0.001 45.877 45.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.687 43.687 dbcsr_multiply_generic 2286 12.5 0.095 0.097 34.340 34.877 qs_scf_new_mos 108 7.5 0.000 0.001 33.034 33.317 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.034 33.317 ot_scf_mini 108 9.5 0.002 0.002 31.301 31.483 multiply_cannon 2286 13.5 0.180 0.188 26.484 28.220 multiply_cannon_loop 2286 14.5 1.492 1.560 25.713 27.412 velocity_verlet 10 3.0 0.001 0.001 26.745 26.745 ot_mini 108 10.5 0.001 0.001 19.953 20.193 qs_ot_get_derivative 108 11.5 0.001 0.001 16.921 17.089 mp_waitall_1 245248 16.5 8.907 15.264 8.907 15.264 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.181 13.724 multiply_cannon_multrec 54864 15.5 4.181 6.464 7.788 11.294 rebuild_ks_matrix 119 8.3 0.000 0.000 8.171 8.313 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 8.171 8.313 multiply_cannon_sync_h2d 54864 15.5 5.842 7.990 5.842 7.990 mp_sum_l 7207 12.9 5.691 7.483 5.691 7.483 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.224 7.356 qs_ot_get_p 119 10.4 0.001 0.001 6.715 7.032 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.758 6.206 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.596 5.713 init_scf_run 11 5.9 0.000 0.001 5.164 5.165 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.164 5.164 dbcsr_mm_accdrv_process 76910 16.1 1.190 1.829 3.529 4.927 sum_up_and_integrate 119 10.3 0.012 0.014 4.712 4.719 integrate_v_rspace 119 11.3 0.002 0.002 4.700 4.708 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.469 4.583 calculate_rho_elec 119 8.7 0.011 0.017 4.469 4.582 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.925 4.082 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.772 3.835 calculate_dm_sparse 119 9.5 0.000 0.000 3.125 3.256 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.894 3.083 apply_single 119 13.6 0.000 0.000 2.893 3.083 rs_pw_transfer 974 11.9 0.011 0.012 2.919 3.033 jit_kernel_multiply 13 15.8 2.275 3.008 2.275 3.008 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.956 2.957 calculate_first_density_matrix 1 7.0 0.001 0.001 2.894 2.898 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.699 2.699 ot_diis_step 108 11.5 0.006 0.006 2.697 2.698 cp_fm_redistribute_end 50 14.0 2.444 2.663 2.450 2.665 cp_fm_diag_elpa_base 50 14.0 0.213 2.566 0.214 2.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.565 2.567 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.478 2.555 density_rs2pw 119 9.7 0.004 0.005 2.433 2.539 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.215 2.273 acc_transpose_blocks 54864 15.5 0.234 0.255 1.783 2.213 wfi_extrapolate 11 7.9 0.001 0.001 2.203 2.204 init_scf_loop 11 6.9 0.000 0.000 2.169 2.169 grid_integrate_task_list 119 12.3 2.024 2.116 2.024 2.116 pw_transfer 1439 11.6 0.053 0.059 2.024 2.106 mp_sum_d 4127 12.0 1.415 2.073 1.415 2.073 potential_pw2rs 119 12.3 0.004 0.004 2.038 2.054 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.947 2.033 make_m2s 4572 13.5 0.053 0.056 1.850 1.904 make_images 4572 14.5 0.132 0.137 1.767 1.822 fft3d_ps 1201 14.6 0.369 0.477 1.718 1.802 fft_wrap_pw1pw2_140 487 13.2 0.082 0.097 1.510 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.557 1.579 mp_alltoall_d11v 2130 13.8 1.402 1.571 1.402 1.571 mp_waitany 12084 13.8 1.254 1.375 1.254 1.375 grid_collocate_task_list 119 9.7 1.292 1.358 1.292 1.358 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.794 1.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.255000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 488.763392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1031904. 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 0.013 0.031 39.871 39.871 qs_mol_dyn_low 1 2.0 0.003 0.003 39.684 39.692 qs_forces 11 3.9 0.002 0.002 39.621 39.622 qs_energies 11 4.9 0.001 0.002 37.901 37.905 scf_env_do_scf 11 5.9 0.000 0.001 32.792 32.793 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.129 30.129 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.581 21.946 qs_scf_new_mos 108 7.5 0.001 0.001 20.474 20.708 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.473 20.708 ot_scf_mini 108 9.5 0.002 0.003 19.534 19.708 velocity_verlet 10 3.0 0.001 0.001 19.007 19.008 multiply_cannon 2286 13.5 0.207 0.213 16.386 18.098 multiply_cannon_loop 2286 14.5 0.896 0.973 15.223 16.791 ot_mini 108 10.5 0.001 0.001 12.021 12.265 mp_waitall_1 200699 16.5 5.968 11.167 5.968 11.167 qs_ot_get_derivative 108 11.5 0.001 0.001 9.542 9.724 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.239 9.384 multiply_cannon_multrec 27432 15.5 1.971 4.321 5.849 8.795 rebuild_ks_matrix 119 8.3 0.000 0.000 7.497 7.638 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.496 7.637 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.639 6.767 dbcsr_mm_accdrv_process 47894 16.0 3.081 5.128 3.810 5.659 qs_ot_get_p 119 10.4 0.001 0.001 4.593 4.820 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.685 4.529 sum_up_and_integrate 119 10.3 0.024 0.026 4.443 4.448 integrate_v_rspace 119 11.3 0.002 0.002 4.419 4.426 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.060 4.136 apply_single 119 13.6 0.000 0.000 3.060 4.135 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.040 4.075 calculate_rho_elec 119 8.7 0.021 0.024 4.040 4.074 mp_sum_l 7207 12.9 2.147 4.066 2.147 4.066 init_scf_run 11 5.9 0.000 0.001 3.875 3.875 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.875 3.875 rs_pw_transfer 974 11.9 0.010 0.011 2.924 3.324 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.063 3.085 make_m2s 4572 13.5 0.052 0.053 2.600 2.840 multiply_cannon_sync_h2d 27432 15.5 2.177 2.778 2.177 2.778 density_rs2pw 119 9.7 0.004 0.004 2.348 2.762 make_images 4572 14.5 0.198 0.235 2.512 2.748 init_scf_loop 11 6.9 0.000 0.000 2.636 2.636 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.620 2.620 ot_diis_step 108 11.5 0.011 0.011 2.424 2.425 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.221 2.311 calculate_first_density_matrix 1 7.0 0.001 0.001 2.297 2.300 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.241 2.242 cp_fm_redistribute_end 50 14.0 1.850 2.211 1.855 2.212 potential_pw2rs 119 12.3 0.006 0.006 2.149 2.164 calculate_dm_sparse 119 9.5 0.000 0.000 2.090 2.163 cp_fm_diag_elpa_base 50 14.0 0.341 2.089 0.354 2.141 pw_transfer 1439 11.6 0.065 0.071 2.079 2.117 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.987 2.028 grid_integrate_task_list 119 12.3 1.842 1.919 1.842 1.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.898 fft3d_ps 1201 14.6 0.514 0.569 1.690 1.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.660 1.700 prepare_preconditioner 11 7.9 0.000 0.000 1.670 1.698 make_preconditioner 11 8.9 0.000 0.000 1.670 1.698 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.566 1.625 make_images_data 4572 15.5 0.044 0.051 1.188 1.616 jit_kernel_multiply 8 16.1 0.676 1.600 0.676 1.600 fft_wrap_pw1pw2_140 487 13.2 0.078 0.085 1.539 1.581 wfi_extrapolate 11 7.9 0.001 0.001 1.519 1.519 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.044 1.517 acc_transpose_blocks 27432 15.5 0.108 0.113 1.176 1.467 mp_alltoall_d11v 2130 13.8 1.291 1.386 1.291 1.386 mp_allgather_i34 2286 14.5 0.609 1.376 0.609 1.376 grid_collocate_task_list 119 9.7 1.231 1.357 1.231 1.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.280 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.193 1.244 mp_sum_d 4127 12.0 0.593 1.052 0.593 1.052 mp_waitany 5720 13.7 0.527 0.967 0.527 0.967 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.942 0.943 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.150 0.543 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.916 0.929 mp_alltoall_z22v 1201 16.6 0.764 0.873 0.764 0.873 make_images_sizes 4572 15.5 0.005 0.005 0.602 0.867 mp_alltoall_i44 4572 16.5 0.598 0.862 0.598 0.862 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.609 0.803 0.862 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.646 0.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.871000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.482682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 524.627968E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 32.271 32.272 qs_mol_dyn_low 1 2.0 0.003 0.003 32.059 32.067 qs_forces 11 3.9 0.002 0.002 31.999 32.000 qs_energies 11 4.9 0.001 0.001 30.396 30.398 scf_env_do_scf 11 5.9 0.000 0.001 25.855 25.855 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 23.227 23.228 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.274 16.376 velocity_verlet 10 3.0 0.001 0.001 15.494 15.496 qs_scf_new_mos 108 7.5 0.001 0.001 14.885 14.909 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.884 14.908 ot_scf_mini 108 9.5 0.002 0.002 14.147 14.161 multiply_cannon 2286 13.5 0.195 0.204 12.985 13.718 multiply_cannon_loop 2286 14.5 0.629 0.658 12.213 13.060 ot_mini 108 10.5 0.001 0.001 8.732 8.747 qs_ot_get_derivative 108 11.5 0.001 0.001 7.208 7.223 multiply_cannon_multrec 18288 15.5 1.932 2.899 6.729 7.062 rebuild_ks_matrix 119 8.3 0.000 0.000 6.487 6.511 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.487 6.510 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.738 5.757 dbcsr_mm_accdrv_process 38222 16.0 4.055 5.246 4.713 5.426 sum_up_and_integrate 119 10.3 0.030 0.031 4.039 4.044 integrate_v_rspace 119 11.3 0.002 0.003 4.009 4.018 mp_waitall_1 158411 16.6 2.664 3.957 2.664 3.957 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.510 3.520 calculate_rho_elec 119 8.7 0.030 0.031 3.509 3.520 init_scf_run 11 5.9 0.000 0.001 3.379 3.379 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.379 3.379 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.696 3.309 qs_ot_get_p 119 10.4 0.001 0.001 3.236 3.250 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.453 2.632 init_scf_loop 11 6.9 0.000 0.000 2.610 2.611 rs_pw_transfer 974 11.9 0.009 0.010 2.336 2.585 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.042 2.366 apply_single 119 13.6 0.000 0.000 2.042 2.366 density_rs2pw 119 9.7 0.004 0.004 2.030 2.286 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.184 2.190 calculate_first_density_matrix 1 7.0 0.000 0.001 2.179 2.180 pw_transfer 1439 11.6 0.065 0.070 1.985 1.996 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.921 1.922 make_m2s 4572 13.5 0.044 0.045 1.775 1.918 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.892 1.907 grid_integrate_task_list 119 12.3 1.808 1.864 1.808 1.864 calculate_dm_sparse 119 9.5 0.000 0.000 1.822 1.838 make_images 4572 14.5 0.188 0.199 1.690 1.832 prepare_preconditioner 11 7.9 0.000 0.000 1.805 1.807 make_preconditioner 11 8.9 0.000 0.000 1.805 1.807 potential_pw2rs 119 12.3 0.007 0.008 1.798 1.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.659 1.744 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.682 1.684 cp_fm_redistribute_end 50 14.0 1.256 1.663 1.258 1.664 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.635 1.636 cp_fm_diag_elpa_base 50 14.0 0.390 1.580 0.404 1.624 multiply_cannon_sync_h2d 18288 15.5 1.400 1.603 1.400 1.603 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.577 1.589 fft3d_ps 1201 14.6 0.521 0.540 1.572 1.587 fft_wrap_pw1pw2_140 487 13.2 0.090 0.093 1.515 1.527 ot_diis_step 108 11.5 0.011 0.011 1.499 1.500 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.491 1.497 mp_sum_l 7207 12.9 1.127 1.487 1.127 1.487 grid_collocate_task_list 119 9.7 1.211 1.382 1.211 1.382 jit_kernel_multiply 7 16.2 0.608 1.278 0.608 1.278 acc_transpose_blocks 18288 15.5 0.076 0.077 1.215 1.241 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.179 1.184 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 make_images_data 4572 15.5 0.044 0.048 0.787 0.949 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 0.941 0.943 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.896 0.919 hybrid_alltoall_any 4725 16.4 0.054 0.111 0.672 0.849 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.823 acc_transpose_blocks_kernels 18288 16.5 0.207 0.217 0.779 0.796 mp_alltoall_d11v 2130 13.8 0.659 0.784 0.659 0.784 mp_waitany 9880 13.7 0.517 0.771 0.517 0.771 mp_alltoall_z22v 1201 16.6 0.657 0.755 0.657 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.657 0.754 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.124 0.504 0.750 cp_fm_cholesky_invert 11 10.9 0.727 0.730 0.727 0.730 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.272000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.909091, yerr=3.232071 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.736512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 35.176 35.177 qs_mol_dyn_low 1 2.0 0.003 0.003 34.980 34.989 qs_forces 11 3.9 0.002 0.002 34.766 34.767 qs_energies 11 4.9 0.001 0.002 33.031 33.037 scf_env_do_scf 11 5.9 0.000 0.001 28.069 28.070 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 24.656 24.657 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.915 18.019 velocity_verlet 10 3.0 0.001 0.001 17.975 17.983 qs_scf_new_mos 108 7.5 0.001 0.001 16.058 16.114 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.057 16.113 ot_scf_mini 108 9.5 0.002 0.003 15.123 15.178 multiply_cannon 2286 13.5 0.227 0.268 14.268 14.763 multiply_cannon_loop 2286 14.5 0.930 0.961 13.332 13.765 ot_mini 108 10.5 0.001 0.001 9.183 9.248 multiply_cannon_multrec 27432 15.5 2.332 3.023 8.573 8.955 qs_ot_get_derivative 108 11.5 0.001 0.001 7.369 7.424 dbcsr_mm_accdrv_process 47916 15.9 5.382 7.098 6.149 7.352 rebuild_ks_matrix 119 8.3 0.000 0.000 6.663 6.715 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.662 6.715 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.908 5.954 sum_up_and_integrate 119 10.3 0.035 0.038 3.876 3.884 integrate_v_rspace 119 11.3 0.002 0.003 3.841 3.849 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.611 3.641 calculate_rho_elec 119 8.7 0.040 0.046 3.610 3.641 init_scf_run 11 5.9 0.000 0.001 3.630 3.630 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.630 3.630 qs_ot_get_p 119 10.4 0.001 0.001 3.342 3.415 init_scf_loop 11 6.9 0.000 0.000 3.392 3.393 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.657 3.122 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.104 2.575 apply_single 119 13.6 0.000 0.000 2.104 2.575 prepare_preconditioner 11 7.9 0.000 0.000 2.554 2.561 make_preconditioner 11 8.9 0.000 0.000 2.554 2.561 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.164 2.486 mp_waitall_1 137007 16.6 1.756 2.344 1.756 2.344 make_m2s 4572 13.5 0.054 0.055 2.194 2.311 calculate_first_density_matrix 1 7.0 0.001 0.001 2.250 2.252 rs_pw_transfer 974 11.9 0.009 0.009 2.040 2.233 make_images 4572 14.5 0.268 0.327 2.088 2.205 density_rs2pw 119 9.7 0.004 0.004 2.015 2.204 pw_transfer 1439 11.6 0.065 0.071 2.124 2.165 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.077 2.087 calculate_dm_sparse 119 9.5 0.000 0.000 2.024 2.085 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.031 2.078 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.962 1.991 grid_integrate_task_list 119 12.3 1.820 1.920 1.820 1.920 jit_kernel_multiply 10 15.8 0.707 1.821 0.707 1.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.798 1.799 ot_diis_step 108 11.5 0.012 0.012 1.772 1.772 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.760 1.761 fft3d_ps 1201 14.6 0.559 0.612 1.706 1.744 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.680 1.732 potential_pw2rs 119 12.3 0.008 0.010 1.625 1.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.618 1.630 mp_sum_l 7207 12.9 1.025 1.566 1.025 1.566 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.519 1.520 cp_fm_redistribute_end 50 14.0 1.006 1.496 1.007 1.497 cp_fm_diag_elpa_base 50 14.0 0.467 1.434 0.487 1.473 acc_transpose_blocks 27432 15.5 0.111 0.115 1.437 1.456 grid_collocate_task_list 119 9.7 1.222 1.357 1.222 1.357 wfi_extrapolate 11 7.9 0.001 0.001 1.332 1.332 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.700 1.219 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.217 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.116 cp_fm_upper_to_full 72 13.5 0.795 1.101 0.795 1.101 multiply_cannon_sync_h2d 27432 15.5 1.005 1.083 1.005 1.083 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.076 1.076 dbcsr_complete_redistribute 329 12.2 0.121 0.143 0.763 1.035 make_images_data 4572 15.5 0.045 0.049 0.862 0.994 hybrid_alltoall_any 4725 16.4 0.061 0.149 0.748 0.908 mp_alltoall_d11v 2130 13.8 0.739 0.908 0.739 0.908 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.850 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.564 0.831 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.816 0.824 mp_alltoall_z22v 1201 16.6 0.732 0.770 0.732 0.770 cp_fm_cholesky_invert 11 10.9 0.764 0.767 0.764 0.767 mp_alltoall_i22 627 13.8 0.426 0.723 0.426 0.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.177000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.000000, yerr=2.763397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 614.227968E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 28.723 28.725 qs_mol_dyn_low 1 2.0 0.003 0.003 28.523 28.532 qs_forces 11 3.9 0.002 0.002 27.992 27.993 qs_energies 11 4.9 0.001 0.001 26.264 26.267 scf_env_do_scf 11 5.9 0.000 0.001 21.634 21.635 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.095 19.095 velocity_verlet 10 3.0 0.001 0.001 14.953 14.985 dbcsr_multiply_generic 2286 12.5 0.092 0.096 12.016 12.089 qs_scf_new_mos 108 7.5 0.001 0.001 10.898 10.925 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.897 10.925 ot_scf_mini 108 9.5 0.002 0.002 10.217 10.244 multiply_cannon 2286 13.5 0.229 0.239 9.526 9.962 multiply_cannon_loop 2286 14.5 0.328 0.339 8.625 8.813 rebuild_ks_matrix 119 8.3 0.000 0.000 6.083 6.102 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.015 6.082 6.102 multiply_cannon_multrec 9144 15.5 1.592 1.856 5.768 6.087 ot_mini 108 10.5 0.001 0.001 5.660 5.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.416 5.434 qs_ot_get_derivative 108 11.5 0.001 0.001 4.356 4.383 dbcsr_mm_accdrv_process 12550 15.8 3.235 4.095 4.075 4.138 sum_up_and_integrate 119 10.3 0.038 0.041 3.720 3.724 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.688 3.697 calculate_rho_elec 119 8.7 0.060 0.061 3.688 3.696 integrate_v_rspace 119 11.3 0.003 0.003 3.682 3.686 init_scf_run 11 5.9 0.000 0.001 3.204 3.204 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.204 3.204 qs_ot_get_p 119 10.4 0.001 0.001 2.792 2.838 init_scf_loop 11 6.9 0.000 0.000 2.519 2.520 pw_transfer 1439 11.6 0.066 0.069 2.205 2.217 density_rs2pw 119 9.7 0.004 0.004 2.003 2.141 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.111 2.125 calculate_first_density_matrix 1 7.0 0.001 0.001 2.077 2.078 mp_waitall_1 115863 16.7 1.460 1.946 1.460 1.946 rs_pw_transfer 974 11.9 0.008 0.008 1.768 1.926 grid_integrate_task_list 119 12.3 1.845 1.916 1.845 1.916 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.880 1.883 make_m2s 4572 13.5 0.034 0.035 1.724 1.877 make_images 4572 14.5 0.266 0.302 1.634 1.787 fft3d_ps 1201 14.6 0.562 0.573 1.771 1.782 fft_wrap_pw1pw2_140 487 13.2 0.087 0.092 1.746 1.763 prepare_preconditioner 11 7.9 0.000 0.000 1.744 1.748 make_preconditioner 11 8.9 0.000 0.000 1.744 1.748 calculate_dm_sparse 119 9.5 0.000 0.000 1.704 1.723 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.634 1.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.528 1.529 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.487 1.499 potential_pw2rs 119 12.3 0.010 0.010 1.452 1.456 jit_kernel_multiply 8 15.8 0.802 1.449 0.802 1.449 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.400 1.401 grid_collocate_task_list 119 9.7 1.270 1.392 1.270 1.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.375 1.382 cp_fm_redistribute_end 50 14.0 0.698 1.379 0.699 1.379 cp_fm_diag_elpa_base 50 14.0 0.635 1.308 0.679 1.364 ot_diis_step 108 11.5 0.012 0.013 1.292 1.292 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.222 1.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.169 1.189 apply_single 119 13.6 0.000 0.000 1.169 1.188 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.149 1.162 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.116 1.120 wfi_extrapolate 11 7.9 0.001 0.001 1.077 1.077 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.799 1.024 make_images_data 4572 15.5 0.039 0.042 0.809 1.000 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 acc_transpose_blocks 9144 15.5 0.038 0.039 0.897 0.907 cp_fm_cholesky_invert 11 10.9 0.863 0.866 0.863 0.866 mp_alltoall_d11v 2130 13.8 0.790 0.858 0.790 0.858 multiply_cannon_sync_h2d 9144 15.5 0.715 0.797 0.715 0.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.789 0.791 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.721 0.728 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.676 0.726 mp_alltoall_z22v 1201 16.6 0.657 0.709 0.657 0.709 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.350 0.690 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.657 0.662 mp_allgather_i34 2286 14.5 0.237 0.651 0.237 0.651 yz_to_x 606 15.1 0.265 0.276 0.610 0.629 x_to_yz 595 16.2 0.276 0.288 0.589 0.597 mp_waitany 5200 13.7 0.457 0.596 0.457 0.596 rs_pw_transfer_RS2PW_140 130 11.5 0.101 0.104 0.426 0.583 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.725000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.181818, yerr=10.734974 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 750.587904E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_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 0.025 0.043 43.184 43.185 qs_mol_dyn_low 1 2.0 0.003 0.003 42.878 42.885 qs_forces 11 3.9 0.002 0.002 42.816 42.818 qs_energies 11 4.9 0.001 0.001 40.748 40.752 scf_env_do_scf 11 5.9 0.001 0.001 34.774 34.774 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.744 26.745 velocity_verlet 10 3.0 0.001 0.001 24.168 24.174 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.809 17.978 qs_scf_new_mos 108 7.5 0.001 0.001 16.445 16.541 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.445 16.541 ot_scf_mini 108 9.5 0.002 0.002 15.342 15.449 multiply_cannon 2286 13.5 0.302 0.309 13.754 14.630 multiply_cannon_loop 2286 14.5 0.342 0.348 12.475 13.380 ot_mini 108 10.5 0.001 0.001 9.067 9.191 multiply_cannon_multrec 9144 15.5 3.369 4.743 8.553 8.638 init_scf_loop 11 6.9 0.000 0.000 8.000 8.001 rebuild_ks_matrix 119 8.3 0.000 0.000 7.619 7.771 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.015 7.619 7.770 qs_ot_get_derivative 108 11.5 0.001 0.001 7.007 7.112 prepare_preconditioner 11 7.9 0.000 0.000 7.002 7.015 make_preconditioner 11 8.9 0.000 0.000 7.002 7.015 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.870 7.008 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.567 6.890 dbcsr_mm_accdrv_process 12550 15.8 4.159 5.607 5.061 6.376 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.609 4.670 calculate_rho_elec 119 8.7 0.118 0.121 4.608 4.669 cp_fm_upper_to_full 72 14.2 3.229 4.602 3.229 4.602 sum_up_and_integrate 119 10.3 0.065 0.066 4.222 4.229 integrate_v_rspace 119 11.3 0.003 0.003 4.157 4.166 init_scf_run 11 5.9 0.000 0.001 3.918 3.918 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.918 3.918 mp_waitall_1 94719 16.7 2.679 3.674 2.679 3.674 qs_ot_get_p 119 10.4 0.001 0.001 3.490 3.641 pw_transfer 1439 11.6 0.068 0.069 3.016 3.024 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.620 3.010 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.917 2.926 dbcsr_complete_redistribute 329 12.2 0.287 0.292 2.006 2.839 make_m2s 4572 13.5 0.038 0.038 2.412 2.582 fft3d_ps 1201 14.6 0.596 0.605 2.542 2.550 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.681 2.517 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.171 2.486 apply_single 119 13.6 0.000 0.000 2.171 2.486 density_rs2pw 119 9.7 0.004 0.004 2.444 2.463 make_images 4572 14.5 0.351 0.380 2.290 2.460 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.429 2.440 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.478 2.373 calculate_first_density_matrix 1 7.0 0.001 0.001 2.283 2.347 calculate_dm_sparse 119 9.5 0.000 0.000 2.252 2.306 mp_alltoall_i22 627 13.8 1.435 2.290 1.435 2.290 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.429 2.257 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.133 2.135 grid_integrate_task_list 119 12.3 2.059 2.078 2.059 2.078 ot_diis_step 108 11.5 0.014 0.015 2.029 2.029 mp_sum_l 7207 12.9 1.109 1.901 1.109 1.901 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.886 1.887 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.849 1.850 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.784 1.840 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.777 1.778 rs_pw_transfer 974 11.9 0.009 0.009 1.690 1.717 potential_pw2rs 119 12.3 0.014 0.015 1.608 1.611 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.588 1.609 cp_fm_cholesky_invert 11 10.9 1.562 1.565 1.562 1.565 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.529 1.530 cp_fm_diag_elpa_base 50 14.0 1.381 1.432 1.528 1.528 grid_collocate_task_list 119 9.7 1.491 1.510 1.491 1.510 wfi_extrapolate 11 7.9 0.001 0.001 1.505 1.505 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.185 1.403 jit_kernel_multiply 8 15.3 0.875 1.398 0.875 1.398 make_images_data 4572 15.5 0.043 0.046 1.142 1.379 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.303 1.361 mp_alltoall_d11v 2130 13.8 1.219 1.288 1.219 1.288 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.171 1.196 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.108 1.125 yz_to_x 606 15.1 0.459 0.471 1.055 1.068 multiply_cannon_sync_h2d 9144 15.5 1.046 1.048 1.046 1.048 mp_alltoall_z22v 1201 16.6 0.995 1.039 0.995 1.039 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.979 1.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.963 0.977 qs_create_task_list 11 7.9 0.000 0.000 0.936 0.952 generate_qs_task_list 11 8.9 0.371 0.390 0.936 0.952 acc_transpose_blocks 9144 15.5 0.038 0.039 0.903 0.918 x_to_yz 595 16.2 0.480 0.490 0.879 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.185000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.636364, yerr=11.267594 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.547008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.055 85.962 85.963 qs_mol_dyn_low 1 2.0 0.003 0.008 85.485 85.495 qs_forces 11 3.9 0.017 0.020 85.411 85.414 qs_energies 11 4.9 0.003 0.010 82.427 82.455 scf_env_do_scf 11 5.9 0.000 0.001 73.309 73.311 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.310 67.311 dbcsr_multiply_generic 2055 12.4 0.108 0.111 52.298 52.628 qs_scf_new_mos 99 7.5 0.000 0.001 48.694 48.834 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.694 48.833 ot_scf_mini 99 9.5 0.002 0.002 46.253 46.363 velocity_verlet 10 3.0 0.001 0.002 44.188 44.189 multiply_cannon 2055 13.4 0.175 0.181 42.526 43.304 multiply_cannon_loop 2055 14.4 1.545 1.587 41.451 42.265 ot_mini 99 10.5 0.001 0.001 27.894 28.001 qs_ot_get_derivative 99 11.5 0.001 0.002 20.998 21.101 multiply_cannon_multrec 49320 15.4 12.115 12.979 17.147 17.871 rebuild_ks_matrix 110 8.3 0.000 0.000 15.419 15.545 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 15.419 15.544 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.578 13.702 mp_waitall_1 220248 16.4 11.925 12.933 11.925 12.933 multiply_cannon_sync_h2d 49320 15.4 9.930 10.542 9.930 10.542 qs_ot_get_p 110 10.4 0.001 0.001 9.774 9.902 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.887 8.257 sum_up_and_integrate 110 10.3 0.037 0.045 7.814 7.826 integrate_v_rspace 110 11.3 0.003 0.004 7.777 7.798 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.272 7.797 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.184 7.768 apply_single 110 13.6 0.000 0.001 7.183 7.768 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.278 7.421 calculate_rho_elec 110 8.6 0.021 0.026 7.278 7.420 init_scf_run 11 5.9 0.000 0.001 6.972 6.972 scf_env_initial_rho_setup 11 6.9 0.000 0.003 6.972 6.972 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.578 6.640 ot_diis_step 99 11.5 0.010 0.037 6.609 6.610 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.396 6.438 mp_sum_l 6514 12.8 5.142 5.974 5.142 5.974 init_scf_loop 11 6.9 0.002 0.017 5.963 5.964 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.419 5.420 dbcsr_mm_accdrv_process 87628 16.1 2.090 2.218 4.910 5.192 rs_pw_transfer 902 11.9 0.012 0.013 4.467 5.080 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.920 4.922 cp_fm_redistribute_end 48 14.0 4.294 4.895 4.298 4.897 cp_fm_diag_elpa_base 48 14.0 0.591 4.745 0.595 4.778 density_rs2pw 110 9.6 0.004 0.005 4.033 4.670 make_m2s 4110 13.4 0.060 0.065 4.171 4.314 wfi_extrapolate 11 7.9 0.001 0.003 4.255 4.255 make_images 4110 14.4 0.177 0.191 4.076 4.222 calculate_dm_sparse 110 9.5 0.000 0.001 3.807 3.913 pw_transfer 1331 11.6 0.054 0.068 3.778 3.859 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.690 3.775 multiply_cannon_metrocomm1 49320 15.4 0.058 0.063 2.691 3.743 prepare_preconditioner 11 7.9 0.000 0.000 3.605 3.624 make_preconditioner 11 8.9 0.000 0.000 3.605 3.624 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.562 3.566 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.379 3.429 grid_integrate_task_list 110 12.3 3.242 3.414 3.242 3.414 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.313 3.373 fft3d_ps 1111 14.6 0.785 0.884 3.218 3.294 potential_pw2rs 110 12.3 0.006 0.007 3.249 3.291 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.186 3.247 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 3.038 3.126 mp_alltoall_d11v 2046 13.8 2.288 2.781 2.288 2.781 mp_waitany 14300 13.8 1.901 2.689 1.901 2.689 calculate_first_density_matrix 1 7.0 0.001 0.003 2.615 2.628 jit_kernel_multiply 13 15.9 2.536 2.551 2.536 2.551 grid_collocate_task_list 110 9.6 2.083 2.296 2.083 2.296 acc_transpose_blocks 49320 15.4 0.227 0.240 2.183 2.269 make_images_data 4110 15.4 0.042 0.045 1.886 2.061 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.999 2.015 mp_alltoall_z22v 1111 16.6 1.729 2.011 1.729 2.011 cp_fm_cholesky_invert 11 10.9 2.004 2.008 2.004 2.008 mp_sum_d 3881 11.9 1.567 2.005 1.567 2.005 hybrid_alltoall_any 4261 16.3 0.081 0.481 1.635 1.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.963000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.443648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2463693. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 70.389 70.390 qs_mol_dyn_low 1 2.0 0.003 0.003 70.126 70.136 qs_forces 11 3.9 0.002 0.002 70.018 70.019 qs_energies 11 4.9 0.001 0.002 66.592 66.597 scf_env_do_scf 11 5.9 0.000 0.001 58.005 58.008 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 50.056 50.056 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.802 37.977 velocity_verlet 10 3.0 0.001 0.001 37.611 37.613 qs_scf_new_mos 99 7.5 0.001 0.001 32.977 33.129 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.977 33.128 multiply_cannon 2055 13.4 0.221 0.241 30.966 32.125 ot_scf_mini 99 9.5 0.003 0.003 31.303 31.459 multiply_cannon_loop 2055 14.4 0.924 0.957 29.649 30.621 ot_mini 99 10.5 0.001 0.001 18.430 18.589 multiply_cannon_multrec 24660 15.4 7.660 8.824 13.832 15.149 rebuild_ks_matrix 110 8.3 0.000 0.000 14.380 14.500 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.379 14.499 qs_ot_get_derivative 99 11.5 0.001 0.001 12.617 12.771 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.656 12.764 mp_waitall_1 176588 16.5 7.928 10.552 7.928 10.552 multiply_cannon_sync_h2d 24660 15.4 7.031 8.470 7.031 8.470 multiply_cannon_metrocomm3 24660 15.4 0.071 0.074 5.124 8.266 init_scf_loop 11 6.9 0.000 0.000 7.902 7.903 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.516 7.310 apply_single 110 13.6 0.000 0.001 6.516 7.309 sum_up_and_integrate 110 10.3 0.052 0.058 7.146 7.157 integrate_v_rspace 110 11.3 0.002 0.003 7.095 7.109 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.475 6.487 calculate_rho_elec 110 8.6 0.040 0.047 6.475 6.487 qs_ot_get_p 110 10.4 0.001 0.001 6.212 6.385 dbcsr_mm_accdrv_process 52282 16.1 4.709 5.756 6.011 6.282 init_scf_run 11 5.9 0.000 0.001 6.197 6.197 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.196 6.197 prepare_preconditioner 11 7.9 0.000 0.000 5.841 5.870 make_preconditioner 11 8.9 0.000 0.000 5.841 5.870 ot_diis_step 99 11.5 0.010 0.010 5.764 5.764 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.831 5.589 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.411 5.571 make_m2s 4110 13.4 0.056 0.059 4.375 4.810 make_images 4110 14.4 0.396 0.438 4.266 4.697 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.282 4.301 pw_transfer 1331 11.6 0.066 0.072 4.021 4.162 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.915 4.059 density_rs2pw 110 9.6 0.004 0.005 3.517 4.048 rs_pw_transfer 902 11.9 0.012 0.013 3.327 3.873 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.857 3.858 wfi_extrapolate 11 7.9 0.001 0.001 3.630 3.630 fft3d_ps 1111 14.6 1.102 1.316 3.267 3.397 fft_wrap_pw1pw2_140 451 13.1 0.202 0.220 3.225 3.373 grid_integrate_task_list 110 12.3 3.144 3.364 3.144 3.364 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.287 3.289 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.199 3.281 cp_fm_redistribute_end 48 14.0 2.459 3.262 2.461 3.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.248 3.250 cp_fm_diag_elpa_base 48 14.0 0.766 3.112 0.797 3.198 calculate_dm_sparse 110 9.5 0.001 0.001 2.978 3.010 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.893 2.943 potential_pw2rs 110 12.3 0.008 0.009 2.879 2.901 make_images_data 4110 15.4 0.046 0.049 2.443 2.890 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.155 2.885 cp_fm_cholesky_invert 11 10.9 2.681 2.688 2.681 2.688 grid_collocate_task_list 110 9.6 2.055 2.483 2.055 2.483 calculate_first_density_matrix 1 7.0 0.000 0.001 2.472 2.475 mp_sum_l 6514 12.8 1.834 2.434 1.834 2.434 mp_alltoall_d11v 2046 13.8 1.983 2.296 1.983 2.296 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.967 1.988 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.889 1.903 mp_waitany 10164 13.8 1.276 1.884 1.276 1.884 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.862 1.863 mp_allgather_i34 2055 14.4 0.651 1.680 0.651 1.680 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.771 1.672 jit_kernel_multiply 8 16.3 0.954 1.595 0.954 1.595 mp_alltoall_z22v 1111 16.6 1.471 1.589 1.471 1.589 mp_irecv_dv 57340 16.2 0.649 1.557 0.649 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.551 acc_transpose_blocks 24660 15.4 0.112 0.115 1.512 1.532 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.973 1.506 dbcsr_complete_redistribute 325 12.2 0.240 0.298 1.209 1.479 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.474 cp_fm_cholesky_decompose 22 10.9 1.455 1.463 1.455 1.463 mp_sum_d 3881 11.9 0.933 1.444 0.933 1.444 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.390000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.363636, yerr=7.138535 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 659.529728E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 61.097 61.098 qs_mol_dyn_low 1 2.0 0.003 0.003 60.783 60.793 qs_forces 11 3.9 0.002 0.005 60.716 60.717 qs_energies 11 4.9 0.003 0.010 57.465 57.468 scf_env_do_scf 11 5.9 0.001 0.003 49.258 49.258 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.435 40.435 velocity_verlet 10 3.0 0.001 0.002 33.327 33.328 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.654 28.969 qs_scf_new_mos 99 7.5 0.001 0.001 25.048 25.139 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.048 25.138 ot_scf_mini 99 9.5 0.003 0.003 23.826 23.931 multiply_cannon 2055 13.4 0.210 0.225 22.319 23.467 multiply_cannon_loop 2055 14.4 0.611 0.627 21.114 22.318 ot_mini 99 10.5 0.001 0.001 13.655 13.761 rebuild_ks_matrix 110 8.3 0.000 0.000 12.610 12.741 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.609 12.740 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.110 11.232 multiply_cannon_multrec 16440 15.4 3.775 4.757 9.511 10.463 mp_waitall_1 139946 16.5 7.403 10.293 7.403 10.293 qs_ot_get_derivative 99 11.5 0.001 0.001 9.121 9.225 init_scf_loop 11 6.9 0.001 0.005 8.785 8.786 multiply_cannon_metrocomm3 16440 15.4 0.043 0.046 4.684 7.401 prepare_preconditioner 11 7.9 0.000 0.000 6.964 6.980 make_preconditioner 11 8.9 0.000 0.001 6.964 6.980 sum_up_and_integrate 110 10.3 0.059 0.060 6.692 6.708 integrate_v_rspace 110 11.3 0.003 0.003 6.632 6.649 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.259 6.616 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.174 6.185 calculate_rho_elec 110 8.6 0.058 0.059 6.174 6.184 dbcsr_mm_accdrv_process 34862 16.1 4.717 5.308 5.592 5.773 init_scf_run 11 5.9 0.000 0.001 5.714 5.715 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.714 5.715 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.058 5.536 apply_single 110 13.6 0.000 0.000 5.058 5.536 qs_ot_get_p 110 10.4 0.001 0.001 5.317 5.470 make_m2s 4110 13.4 0.049 0.051 4.162 4.524 density_rs2pw 110 9.6 0.004 0.005 3.258 4.516 ot_diis_step 99 11.5 0.011 0.011 4.501 4.501 make_images 4110 14.4 0.390 0.508 4.048 4.412 multiply_cannon_sync_h2d 16440 15.4 3.663 4.251 3.663 4.251 rs_pw_transfer 902 11.9 0.010 0.012 2.793 4.052 pw_transfer 1331 11.6 0.065 0.072 3.851 3.864 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.106 3.767 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.744 3.760 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.684 3.689 grid_integrate_task_list 110 12.3 3.161 3.406 3.161 3.406 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.334 3.335 fft_wrap_pw1pw2_140 451 13.1 0.213 0.219 3.204 3.218 wfi_extrapolate 11 7.9 0.001 0.001 3.064 3.065 fft3d_ps 1111 14.6 1.087 1.101 3.043 3.059 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.909 2.911 make_images_data 4110 15.4 0.043 0.047 2.415 2.899 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.795 2.797 cp_fm_redistribute_end 48 14.0 1.745 2.770 1.747 2.771 cp_fm_cholesky_invert 11 10.9 2.726 2.732 2.726 2.732 hybrid_alltoall_any 4261 16.3 0.105 0.378 2.137 2.730 cp_fm_diag_elpa_base 48 14.0 0.961 2.616 1.019 2.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.562 2.614 calculate_first_density_matrix 1 7.0 0.001 0.003 2.549 2.551 mp_waitany 17072 13.8 1.202 2.532 1.202 2.532 calculate_dm_sparse 110 9.5 0.001 0.001 2.493 2.525 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.920 2.516 grid_collocate_task_list 110 9.6 2.086 2.465 2.086 2.465 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.367 2.425 potential_pw2rs 110 12.3 0.010 0.010 2.394 2.402 mp_irecv_dv 48980 15.7 0.850 2.389 0.850 2.389 mp_alltoall_d11v 2046 13.8 1.883 2.320 1.883 2.320 rs_pw_transfer_RS2PW_140 121 11.5 0.183 0.188 0.905 2.165 mp_sum_l 6514 12.8 1.560 2.162 1.560 2.162 qs_energies_init_hamiltonians 11 5.9 0.025 0.035 2.043 2.044 dbcsr_complete_redistribute 325 12.2 0.329 0.394 1.569 2.037 cp_fm_upper_to_full 70 13.6 1.393 1.862 1.393 1.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.718 cp_fm_cholesky_decompose 22 10.9 1.674 1.689 1.674 1.689 mp_allgather_i34 2055 14.4 0.538 1.632 0.538 1.632 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.501 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.034 1.491 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.456 1.470 rs_gather_matrices 110 12.3 0.234 0.267 0.998 1.357 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.282 1.291 mp_alltoall_z22v 1111 16.6 1.191 1.234 1.191 1.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.098000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.909091, yerr=8.447211 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 734.027776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.095 66.098 66.099 qs_mol_dyn_low 1 2.0 0.003 0.003 65.702 65.711 qs_forces 11 3.9 0.008 0.010 65.635 65.635 qs_energies 11 4.9 0.002 0.008 62.185 62.194 scf_env_do_scf 11 5.9 0.001 0.002 53.843 53.846 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 41.993 41.994 velocity_verlet 10 3.0 0.001 0.001 37.500 37.502 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.541 29.857 qs_scf_new_mos 99 7.5 0.001 0.001 26.619 26.732 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.618 26.732 ot_scf_mini 99 9.5 0.003 0.003 24.994 25.114 multiply_cannon 2055 13.4 0.243 0.265 22.492 23.605 multiply_cannon_loop 2055 14.4 0.881 0.902 21.151 21.895 ot_mini 99 10.5 0.001 0.001 14.370 14.510 multiply_cannon_multrec 24660 15.4 4.204 6.646 12.596 13.705 rebuild_ks_matrix 110 8.3 0.000 0.000 12.360 12.485 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.016 12.359 12.484 init_scf_loop 11 6.9 0.001 0.006 11.809 11.810 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.906 11.014 qs_ot_get_derivative 99 11.5 0.001 0.001 10.188 10.312 prepare_preconditioner 11 7.9 0.000 0.000 10.039 10.052 make_preconditioner 11 8.9 0.000 0.000 10.039 10.052 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.255 9.719 dbcsr_mm_accdrv_process 52304 16.0 6.701 8.168 8.246 9.184 sum_up_and_integrate 110 10.3 0.068 0.071 6.636 6.652 integrate_v_rspace 110 11.3 0.003 0.003 6.568 6.583 mp_waitall_1 121746 16.5 4.545 6.437 4.545 6.437 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.345 6.355 calculate_rho_elec 110 8.6 0.077 0.081 6.344 6.354 qs_ot_get_p 110 10.4 0.001 0.001 5.655 5.824 make_m2s 4110 13.4 0.059 0.061 5.422 5.651 init_scf_run 11 5.9 0.000 0.001 5.562 5.562 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.562 5.562 make_images 4110 14.4 0.573 0.691 5.283 5.510 cp_fm_upper_to_full 70 13.8 3.277 4.664 3.277 4.664 ot_diis_step 99 11.5 0.011 0.012 4.146 4.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.044 4.145 apply_single 110 13.6 0.000 0.000 4.044 4.145 pw_transfer 1331 11.6 0.065 0.074 4.005 4.036 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.900 3.934 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.835 3.851 dbcsr_complete_redistribute 325 12.2 0.416 0.459 2.695 3.838 density_rs2pw 110 9.6 0.004 0.005 3.237 3.743 grid_integrate_task_list 110 12.3 3.272 3.495 3.272 3.495 multiply_cannon_sync_h2d 24660 15.4 3.184 3.464 3.184 3.464 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.574 3.414 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.345 3.391 fft_wrap_pw1pw2_140 451 13.1 0.202 0.215 3.330 3.366 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.357 3.358 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.134 3.253 fft3d_ps 1111 14.6 1.091 1.130 3.192 3.217 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.055 3.126 make_images_data 4110 15.4 0.046 0.050 2.755 3.051 wfi_extrapolate 11 7.9 0.001 0.001 3.019 3.019 rs_pw_transfer 902 11.9 0.010 0.011 2.480 3.003 hybrid_alltoall_any 4261 16.3 0.119 0.456 2.426 3.003 calculate_dm_sparse 110 9.5 0.001 0.001 2.907 2.939 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.775 2.879 mp_alltoall_i22 605 13.7 1.657 2.821 1.657 2.821 cp_fm_cholesky_invert 11 10.9 2.812 2.820 2.812 2.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.772 2.775 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.765 2.767 cp_fm_redistribute_end 48 14.0 1.382 2.736 1.384 2.738 cp_fm_diag_elpa_base 48 14.0 1.273 2.599 1.350 2.711 calculate_first_density_matrix 1 7.0 0.001 0.002 2.450 2.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.405 2.450 grid_collocate_task_list 110 9.6 2.181 2.440 2.181 2.440 jit_kernel_multiply 13 15.8 1.217 2.349 1.217 2.349 potential_pw2rs 110 12.3 0.013 0.013 2.270 2.283 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.258 2.265 mp_alltoall_d11v 2046 13.8 1.883 2.109 1.883 2.109 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.730 cp_fm_cholesky_decompose 22 10.9 1.677 1.727 1.677 1.727 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.702 1.717 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.702 mp_waitany 13376 13.8 1.071 1.580 1.071 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.537 acc_transpose_blocks 24660 15.4 0.106 0.107 1.498 1.523 mp_sum_l 6514 12.8 0.946 1.518 0.946 1.518 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.862 1.402 mp_alltoall_z22v 1111 16.6 1.241 1.341 1.241 1.341 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.178 0.810 1.332 mp_irecv_dv 62702 16.1 0.763 1.329 0.763 1.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.099000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.272727, yerr=10.349272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 829.980672E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.038 55.986 55.987 qs_mol_dyn_low 1 2.0 0.003 0.003 55.752 55.761 qs_forces 11 3.9 0.002 0.002 55.673 55.675 qs_energies 11 4.9 0.001 0.001 51.966 51.970 scf_env_do_scf 11 5.9 0.000 0.001 43.574 43.574 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 35.736 35.736 velocity_verlet 10 3.0 0.001 0.001 31.535 31.539 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.765 22.867 qs_scf_new_mos 99 7.5 0.001 0.001 20.333 20.388 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.332 20.388 ot_scf_mini 99 9.5 0.002 0.002 19.093 19.138 multiply_cannon 2055 13.4 0.247 0.258 17.360 18.640 multiply_cannon_loop 2055 14.4 0.318 0.330 16.040 16.251 rebuild_ks_matrix 110 8.3 0.000 0.000 12.194 12.228 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 12.194 12.228 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.838 10.870 ot_mini 99 10.5 0.001 0.001 10.304 10.344 multiply_cannon_multrec 8220 15.4 3.216 4.508 7.447 8.567 mp_waitall_1 103326 16.6 6.177 7.953 6.177 7.953 init_scf_loop 11 6.9 0.000 0.000 7.788 7.789 sum_up_and_integrate 110 10.3 0.079 0.080 6.673 6.686 integrate_v_rspace 110 11.3 0.003 0.003 6.594 6.608 qs_ot_get_derivative 99 11.5 0.001 0.001 6.500 6.546 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.443 6.453 calculate_rho_elec 110 8.6 0.115 0.116 6.442 6.453 prepare_preconditioner 11 7.9 0.000 0.000 6.091 6.099 make_preconditioner 11 8.9 0.000 0.000 6.091 6.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.672 5.745 init_scf_run 11 5.9 0.000 0.001 5.226 5.226 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.225 5.226 dbcsr_mm_accdrv_process 17442 15.9 2.730 3.946 4.100 5.007 qs_ot_get_p 110 10.4 0.001 0.001 4.850 4.903 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.129 4.603 make_m2s 4110 13.4 0.038 0.040 4.166 4.405 make_images 4110 14.4 0.634 0.686 4.036 4.274 pw_transfer 1331 11.6 0.065 0.071 4.223 4.235 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.116 4.132 ot_diis_step 99 11.5 0.012 0.012 3.783 3.783 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.699 3.747 apply_single 110 13.6 0.000 0.000 3.699 3.746 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.562 3.582 grid_integrate_task_list 110 12.3 3.357 3.571 3.357 3.571 density_rs2pw 110 9.6 0.004 0.004 3.183 3.557 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.509 3.513 fft3d_ps 1111 14.6 1.151 1.195 3.359 3.373 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.201 3.201 multiply_cannon_sync_h2d 8220 15.4 2.902 3.129 2.902 3.129 cp_fm_cholesky_invert 11 10.9 2.986 2.990 2.986 2.990 wfi_extrapolate 11 7.9 0.001 0.001 2.812 2.812 make_images_data 4110 15.4 0.038 0.043 2.382 2.771 qs_energies_init_hamiltonians 11 5.9 0.010 0.020 2.699 2.710 hybrid_alltoall_any 4261 16.3 0.199 0.865 2.310 2.692 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.663 2.663 rs_pw_transfer 902 11.9 0.010 0.011 2.243 2.641 cp_fm_redistribute_end 48 14.0 0.682 2.636 0.686 2.637 cp_fm_diag_elpa_base 48 14.0 1.770 2.440 1.943 2.600 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.537 2.538 grid_collocate_task_list 110 9.6 2.277 2.537 2.277 2.537 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.511 calculate_first_density_matrix 1 7.0 0.001 0.001 2.317 2.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.185 2.202 potential_pw2rs 110 12.3 0.015 0.016 2.194 2.199 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 1.978 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.879 1.909 mp_alltoall_d11v 2046 13.8 1.664 1.899 1.664 1.899 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.789 1.810 jit_kernel_multiply 10 15.8 1.064 1.774 1.064 1.774 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.692 1.698 cp_fm_cholesky_decompose 22 10.9 1.676 1.688 1.676 1.688 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.526 1.651 mp_allgather_i34 2055 14.4 0.511 1.603 0.511 1.603 dbcsr_complete_redistribute 325 12.2 0.556 0.582 1.440 1.533 mp_waitany 9240 13.8 1.075 1.503 1.075 1.503 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.450 qs_create_task_list 11 7.9 0.000 0.000 1.236 1.340 generate_qs_task_list 11 8.9 0.380 0.447 1.236 1.339 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.479 1.302 mp_irecv_dv 24056 15.7 0.455 1.258 0.455 1.258 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.166 0.788 1.195 rs_gather_matrices 110 12.3 0.327 0.367 0.954 1.154 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.124 1.153 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.799 1.146 mp_alltoall_z22v 1111 16.6 1.098 1.125 1.098 1.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.987000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.818182, yerr=10.844506 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 1.383752E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_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 0.022 0.041 89.524 89.525 qs_mol_dyn_low 1 2.0 0.003 0.003 89.138 89.148 qs_forces 11 3.9 0.012 0.013 89.069 89.070 qs_energies 11 4.9 0.001 0.002 84.821 84.823 scf_env_do_scf 11 5.9 0.001 0.001 74.786 74.786 velocity_verlet 10 3.0 0.001 0.001 57.337 57.343 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 45.409 45.410 init_scf_loop 11 6.9 0.000 0.000 29.303 29.305 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.675 28.801 prepare_preconditioner 11 7.9 0.000 0.000 27.215 27.226 make_preconditioner 11 8.9 0.000 0.000 27.215 27.226 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.320 26.678 qs_scf_new_mos 99 7.5 0.001 0.001 26.484 26.569 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.483 26.568 ot_scf_mini 99 9.5 0.002 0.002 24.731 24.797 multiply_cannon 2055 13.4 0.340 0.356 21.561 22.235 multiply_cannon_loop 2055 14.4 0.339 0.341 19.729 20.106 cp_fm_upper_to_full 70 14.2 13.091 18.700 13.091 18.700 rebuild_ks_matrix 110 8.3 0.001 0.001 14.589 14.673 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.015 14.589 14.672 ot_mini 99 10.5 0.001 0.001 13.716 13.786 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.225 13.301 dbcsr_complete_redistribute 325 12.2 1.023 1.042 7.500 10.836 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.446 9.779 multiply_cannon_multrec 8220 15.4 4.372 4.567 9.534 9.669 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.880 9.178 qs_ot_get_derivative 99 11.5 0.001 0.001 9.058 9.124 mp_alltoall_i22 605 13.7 5.514 8.874 5.514 8.874 mp_waitall_1 84994 16.7 7.655 8.720 7.655 8.720 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.020 8.057 calculate_rho_elec 110 8.6 0.227 0.227 8.019 8.056 sum_up_and_integrate 110 10.3 0.150 0.152 7.678 7.691 integrate_v_rspace 110 11.3 0.004 0.006 7.528 7.542 qs_ot_get_p 110 10.4 0.001 0.001 5.803 5.881 make_m2s 4110 13.4 0.043 0.044 5.279 5.804 init_scf_run 11 5.9 0.000 0.001 5.743 5.743 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.743 5.743 make_images 4110 14.4 0.873 0.920 5.091 5.616 cp_fm_cholesky_invert 11 10.9 5.593 5.598 5.593 5.598 pw_transfer 1331 11.6 0.075 0.076 5.444 5.451 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.328 5.334 dbcsr_mm_accdrv_process 11614 15.7 3.252 3.656 5.020 5.272 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.733 5.248 apply_single 110 13.6 0.000 0.000 4.733 5.248 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.841 5.110 ot_diis_step 99 11.5 0.016 0.016 4.638 4.638 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.606 4.614 fft3d_ps 1111 14.6 1.290 1.299 4.522 4.527 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.207 4.216 density_rs2pw 110 9.6 0.004 0.004 3.965 4.033 multiply_cannon_sync_h2d 8220 15.4 3.957 3.961 3.957 3.961 grid_integrate_task_list 110 12.3 3.755 3.877 3.755 3.877 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.759 3.759 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.687 3.689 hybrid_alltoall_any 4261 16.3 0.255 0.550 2.900 3.637 make_images_data 4110 15.4 0.042 0.044 2.893 3.541 wfi_extrapolate 11 7.9 0.001 0.001 3.358 3.359 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.881 3.336 calculate_dm_sparse 110 9.5 0.001 0.001 3.136 3.161 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.149 3.149 cp_fm_diag_elpa_base 48 14.0 2.599 2.802 3.146 3.146 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.866 2.867 grid_collocate_task_list 110 9.6 2.704 2.772 2.704 2.772 potential_pw2rs 110 12.3 0.021 0.022 2.602 2.610 rs_pw_transfer 902 11.9 0.011 0.011 2.385 2.478 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.426 2.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.363 2.389 qs_env_update_s_mstruct 11 6.9 0.002 0.005 2.206 2.267 calculate_first_density_matrix 1 7.0 0.001 0.001 2.261 2.262 mp_alltoall_d11v 2046 13.8 2.095 2.218 2.095 2.218 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.189 cp_fm_cholesky_decompose 22 10.9 2.084 2.116 2.084 2.116 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.974 1.981 qs_create_task_list 11 7.9 0.000 0.000 1.897 1.946 generate_qs_task_list 11 8.9 0.735 0.789 1.897 1.946 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.823 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.525000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1233.363636, yerr=72.150440 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.043776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.057 205.676 205.678 qs_mol_dyn_low 1 2.0 0.004 0.008 205.235 205.248 qs_forces 11 3.9 0.005 0.019 205.141 205.142 qs_energies 11 4.9 0.002 0.007 199.456 199.476 scf_env_do_scf 11 5.9 0.001 0.002 182.694 182.698 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 161.685 161.687 dbcsr_multiply_generic 2507 12.6 0.178 0.182 123.934 124.734 velocity_verlet 10 3.0 0.002 0.005 124.268 124.269 qs_scf_new_mos 117 7.6 0.001 0.001 120.758 121.007 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.757 121.006 ot_scf_mini 117 9.6 0.003 0.004 114.173 114.385 multiply_cannon 2507 13.6 0.237 0.246 100.717 102.945 multiply_cannon_loop 2507 14.6 2.105 2.170 98.496 100.520 ot_mini 117 10.6 0.001 0.002 65.541 65.765 multiply_cannon_multrec 60168 15.6 33.260 35.287 41.601 43.709 qs_ot_get_derivative 117 11.6 0.001 0.002 40.703 40.930 rebuild_ks_matrix 128 8.3 0.001 0.001 34.255 34.599 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.022 34.254 34.598 mp_waitall_1 267128 16.5 28.300 31.426 28.300 31.426 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.798 31.101 multiply_cannon_sync_h2d 60168 15.6 27.339 29.188 27.339 29.188 qs_ot_get_p 128 10.4 0.001 0.001 26.756 27.002 apply_preconditioner_dbcsr 128 12.6 0.000 0.002 24.354 24.943 apply_single 128 13.6 0.001 0.001 24.354 24.943 ot_diis_step 117 11.6 0.008 0.014 24.486 24.487 init_scf_loop 11 6.9 0.001 0.004 20.932 20.933 qs_ot_p2m_diag 83 11.4 0.078 0.091 19.972 20.054 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.772 18.931 multiply_cannon_metrocomm3 60168 15.6 0.116 0.124 15.681 17.870 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.394 17.396 prepare_preconditioner 11 7.9 0.000 0.001 16.238 16.284 make_preconditioner 11 8.9 0.000 0.001 16.238 16.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.439 15.613 sum_up_and_integrate 128 10.3 0.090 0.108 14.669 14.687 integrate_v_rspace 128 11.3 0.004 0.005 14.579 14.602 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.428 14.434 cp_fm_redistribute_end 83 14.4 11.403 14.347 11.414 14.350 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.253 14.350 calculate_rho_elec 128 8.7 0.046 0.065 14.252 14.349 make_m2s 5014 13.6 0.104 0.114 13.913 14.207 cp_fm_diag_elpa_base 83 14.4 2.894 13.964 2.925 14.093 make_images 5014 14.6 0.400 0.419 13.731 14.034 init_scf_run 11 5.9 0.000 0.001 12.528 12.528 scf_env_initial_rho_setup 11 6.9 0.003 0.028 12.527 12.528 density_rs2pw 128 9.7 0.006 0.007 7.577 10.834 mp_sum_l 7870 13.0 8.042 9.540 8.042 9.540 rs_pw_transfer 1046 11.9 0.016 0.019 6.202 9.428 cp_fm_cholesky_invert 11 10.9 9.303 9.311 9.303 9.311 wfi_extrapolate 11 7.9 0.005 0.030 9.253 9.253 pw_transfer 1547 11.6 0.076 0.107 8.332 8.604 calculate_dm_sparse 128 9.5 0.001 0.001 8.422 8.510 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 8.127 8.405 dbcsr_mm_accdrv_process 124484 16.2 3.306 3.448 7.904 8.391 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.134 8.256 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.815 7.894 multiply_cannon_metrocomm1 60168 15.6 0.085 0.089 5.832 7.831 make_images_data 5014 15.6 0.065 0.071 6.829 7.790 grid_integrate_task_list 128 12.3 7.039 7.531 7.039 7.531 hybrid_alltoall_any 5200 16.5 0.290 2.264 5.986 7.364 fft_wrap_pw1pw2_140 523 13.2 0.444 0.518 6.978 7.300 fft3d_ps 1291 14.7 2.108 2.870 6.914 7.101 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.670 6.679 mp_waitany 16020 13.9 2.738 5.987 2.738 5.987 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.796 5.946 mp_alltoall_d11v 2415 14.1 4.367 5.884 4.367 5.884 grid_collocate_task_list 128 9.7 4.554 5.776 4.554 5.776 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.289 2.153 5.415 potential_pw2rs 128 12.3 0.009 0.011 5.030 5.072 cp_fm_cholesky_decompose 22 10.9 4.701 4.714 4.701 4.714 mp_sum_d 4462 12.2 3.591 4.296 3.591 4.296 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.678000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.272727, yerr=6.523955 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 828.248064E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57283. MP_Allreduce 11237 947. MP_Sync 168 MP_Alltoall 1955 5616512. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 191.960 191.962 qs_mol_dyn_low 1 2.0 0.003 0.005 191.589 191.603 qs_forces 11 3.9 0.003 0.004 191.500 191.501 qs_energies 11 4.9 0.002 0.009 184.620 184.633 scf_env_do_scf 11 5.9 0.001 0.003 168.213 168.223 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 134.874 134.876 velocity_verlet 10 3.0 0.002 0.006 120.575 120.577 dbcsr_multiply_generic 2485 12.5 0.189 0.197 96.894 98.110 qs_scf_new_mos 116 7.6 0.001 0.001 94.374 94.856 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.373 94.855 ot_scf_mini 116 9.6 0.004 0.005 89.574 90.141 multiply_cannon 2485 13.5 0.477 0.533 76.588 81.148 multiply_cannon_loop 2485 14.5 1.238 1.285 73.341 75.888 ot_mini 116 10.6 0.001 0.001 49.760 50.262 mp_waitall_1 212858 16.6 25.055 38.431 25.055 38.431 multiply_cannon_multrec 29820 15.5 21.543 25.964 31.051 35.987 rebuild_ks_matrix 127 8.3 0.001 0.001 33.561 34.173 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.025 33.560 34.173 init_scf_loop 11 6.9 0.001 0.005 33.247 33.248 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.165 30.738 multiply_cannon_metrocomm3 29820 15.5 0.096 0.104 16.057 28.801 prepare_preconditioner 11 7.9 0.000 0.001 28.727 28.798 make_preconditioner 11 8.9 0.000 0.002 28.727 28.798 qs_ot_get_derivative 116 11.6 0.001 0.002 27.830 28.374 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.381 27.911 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.008 23.170 apply_single 127 13.6 0.001 0.001 22.008 23.170 qs_ot_get_p 127 10.4 0.001 0.001 21.754 22.398 ot_diis_step 116 11.6 0.014 0.015 21.748 21.750 multiply_cannon_sync_h2d 29820 15.5 18.806 21.387 18.806 21.387 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.941 16.980 cp_fm_cholesky_invert 11 10.9 16.747 16.760 16.747 16.760 make_m2s 4970 13.5 0.088 0.093 14.448 15.921 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.780 15.781 make_images 4970 14.5 1.139 1.327 14.239 15.707 sum_up_and_integrate 127 10.3 0.116 0.131 15.059 15.087 integrate_v_rspace 127 11.3 0.003 0.004 14.943 14.978 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.490 14.533 calculate_rho_elec 127 8.7 0.087 0.105 14.490 14.532 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.636 12.650 cp_fm_redistribute_end 82 14.4 7.378 12.569 7.389 12.571 cp_fm_diag_elpa_base 82 14.4 4.945 12.056 5.158 12.412 init_scf_run 11 5.9 0.000 0.001 11.601 11.602 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.600 11.602 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.944 11.371 multiply_cannon_metrocomm4 27335 15.5 0.094 0.110 3.793 10.623 make_images_data 4970 15.5 0.062 0.071 8.770 10.493 density_rs2pw 127 9.7 0.006 0.007 7.900 10.281 mp_irecv_dv 68888 16.3 3.601 10.240 3.601 10.240 hybrid_alltoall_any 5155 16.4 0.338 1.490 7.531 9.958 dbcsr_mm_accdrv_process 61726 16.2 4.591 5.482 8.977 9.568 pw_transfer 1535 11.6 0.086 0.109 9.235 9.323 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 9.010 9.106 rs_pw_transfer 1038 11.9 0.014 0.017 6.165 8.551 wfi_extrapolate 11 7.9 0.001 0.001 8.493 8.493 fft_wrap_pw1pw2_140 519 13.2 0.469 0.524 7.866 7.979 grid_integrate_task_list 127 12.3 7.099 7.539 7.099 7.539 fft3d_ps 1281 14.7 2.729 2.902 7.388 7.428 cp_fm_cholesky_decompose 22 10.9 7.026 7.094 7.026 7.094 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.232 6.996 calculate_dm_sparse 127 9.5 0.001 0.001 6.458 6.595 mp_sum_l 7804 13.0 4.089 6.158 4.089 6.158 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.107 6.115 grid_collocate_task_list 127 9.7 4.662 5.825 4.662 5.825 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.340 5.420 potential_pw2rs 127 12.3 0.014 0.016 5.369 5.398 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.229 5.381 mp_allgather_i34 2485 14.5 1.842 5.109 1.842 5.109 mp_waitany 11660 13.9 2.540 4.984 2.540 4.984 mp_alltoall_d11v 2401 14.1 4.244 4.875 4.244 4.875 rs_pw_transfer_RS2PW_140 138 11.5 0.351 0.376 2.070 4.457 dbcsr_complete_redistribute 393 12.7 0.771 0.876 3.222 4.072 mp_sum_d 4441 12.1 2.691 4.013 2.691 4.013 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.804 3.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.962000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.636364, yerr=4.006194 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 931.012608E+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 4083 57222. MP_Allreduce 11246 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 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 0.015 0.030 176.820 176.820 qs_mol_dyn_low 1 2.0 0.003 0.003 176.449 176.461 qs_forces 11 3.9 0.003 0.003 176.159 176.162 qs_energies 11 4.9 0.005 0.006 169.484 169.493 scf_env_do_scf 11 5.9 0.001 0.001 153.830 153.831 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 118.651 118.651 velocity_verlet 10 3.0 0.001 0.001 113.190 113.198 dbcsr_multiply_generic 2485 12.5 0.181 0.185 80.778 81.851 qs_scf_new_mos 116 7.6 0.001 0.001 80.196 80.532 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.196 80.531 ot_scf_mini 116 9.6 0.003 0.004 76.078 76.449 multiply_cannon 2485 13.5 0.498 0.522 61.273 65.638 multiply_cannon_loop 2485 14.5 0.843 0.870 58.257 60.857 ot_mini 116 10.6 0.001 0.001 42.074 42.443 init_scf_loop 11 6.9 0.000 0.000 35.074 35.075 mp_waitall_1 169034 16.6 24.755 34.016 24.755 34.016 rebuild_ks_matrix 127 8.3 0.001 0.001 30.863 31.408 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.027 30.863 31.407 prepare_preconditioner 11 7.9 0.000 0.000 31.005 31.059 make_preconditioner 11 8.9 0.000 0.000 31.005 31.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.656 30.043 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.767 28.262 multiply_cannon_multrec 19880 15.5 13.358 16.314 22.066 24.976 multiply_cannon_metrocomm3 19880 15.5 0.059 0.064 15.195 24.830 qs_ot_get_derivative 116 11.6 0.001 0.002 22.442 22.811 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.755 20.629 apply_single 127 13.6 0.001 0.001 19.754 20.628 qs_ot_get_p 127 10.4 0.001 0.001 19.576 20.026 ot_diis_step 116 11.6 0.017 0.018 19.530 19.530 multiply_cannon_sync_h2d 19880 15.5 14.160 15.767 14.160 15.767 qs_ot_p2m_diag 82 11.4 0.262 0.268 15.345 15.354 make_m2s 4970 13.5 0.080 0.085 14.524 15.216 make_images 4970 14.5 1.155 1.231 14.291 14.981 sum_up_and_integrate 127 10.3 0.133 0.144 14.806 14.833 integrate_v_rspace 127 11.3 0.004 0.004 14.673 14.704 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.508 14.558 calculate_rho_elec 127 8.7 0.131 0.146 14.507 14.558 cp_fm_cholesky_invert 11 10.9 14.491 14.500 14.491 14.500 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.346 14.348 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.178 11.179 cp_fm_redistribute_end 82 14.4 4.224 11.107 4.238 11.110 cp_fm_diag_elpa_base 82 14.4 6.447 10.517 6.850 10.974 init_scf_run 11 5.9 0.000 0.001 10.653 10.653 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.653 10.653 make_images_data 4970 15.5 0.060 0.067 8.931 10.049 density_rs2pw 127 9.7 0.006 0.006 7.382 9.418 multiply_cannon_metrocomm4 17395 15.5 0.060 0.073 3.440 9.410 hybrid_alltoall_any 5155 16.4 0.431 1.987 7.893 9.362 pw_transfer 1535 11.6 0.085 0.106 9.227 9.346 mp_irecv_dv 49801 16.2 3.320 9.154 3.320 9.154 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.005 9.129 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.780 9.043 dbcsr_mm_accdrv_process 41158 16.2 4.349 5.305 8.172 8.305 fft_wrap_pw1pw2_140 519 13.2 0.474 0.524 7.941 8.077 grid_integrate_task_list 127 12.3 7.216 7.817 7.216 7.817 wfi_extrapolate 11 7.9 0.001 0.001 7.520 7.520 cp_fm_upper_to_full 104 14.5 5.739 7.465 5.739 7.465 cp_fm_cholesky_decompose 22 10.9 7.369 7.404 7.369 7.404 fft3d_ps 1281 14.7 2.647 2.880 7.279 7.366 rs_pw_transfer 1038 11.9 0.013 0.015 5.250 7.318 dbcsr_complete_redistribute 393 12.7 1.182 1.215 4.635 6.418 calculate_dm_sparse 127 9.5 0.001 0.001 5.755 5.837 grid_collocate_task_list 127 9.7 4.860 5.709 4.860 5.709 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.575 5.585 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.511 5.187 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.399 5.178 mp_alltoall_d11v 2401 14.1 4.462 5.113 4.462 5.113 potential_pw2rs 127 12.3 0.020 0.022 4.937 4.953 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.627 4.801 mp_sum_l 7804 13.0 3.174 4.505 3.174 4.505 mp_waitany 11660 13.9 2.345 4.383 2.345 4.383 mp_allgather_i34 2485 14.5 1.564 4.349 1.564 4.349 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.329 4.051 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.000 4.040 rs_pw_transfer_RS2PW_140 138 11.5 0.333 0.352 1.904 3.935 mp_alltoall_i22 712 14.1 1.923 3.787 1.923 3.787 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.747 3.748 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.606 3.658 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.820000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=882.909091, yerr=10.281975 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.167303E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.046 187.964 187.966 qs_mol_dyn_low 1 2.0 0.003 0.003 187.544 187.559 qs_forces 11 3.9 0.003 0.003 187.430 187.439 qs_energies 11 4.9 0.002 0.008 180.167 180.174 scf_env_do_scf 11 5.9 0.001 0.006 163.374 163.381 velocity_verlet 10 3.0 0.001 0.001 124.848 124.851 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 116.342 116.344 qs_scf_new_mos 116 7.6 0.001 0.001 78.656 78.917 qs_scf_loop_do_ot 116 8.6 0.001 0.003 78.655 78.916 dbcsr_multiply_generic 2485 12.5 0.188 0.192 78.272 78.870 ot_scf_mini 116 9.6 0.003 0.005 74.205 74.450 multiply_cannon 2485 13.5 0.550 0.583 54.094 56.796 multiply_cannon_loop 2485 14.5 1.164 1.195 50.423 51.904 init_scf_loop 11 6.9 0.001 0.005 46.910 46.913 prepare_preconditioner 11 7.9 0.000 0.001 42.730 42.756 make_preconditioner 11 8.9 0.000 0.002 42.730 42.756 ot_mini 116 10.6 0.001 0.001 41.763 41.993 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.411 41.348 multiply_cannon_multrec 29820 15.5 14.121 19.173 25.892 30.488 rebuild_ks_matrix 127 8.3 0.001 0.001 30.082 30.270 qs_ks_build_kohn_sham_matrix 127 9.3 0.021 0.023 30.082 30.269 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.052 27.221 mp_waitall_1 146592 16.7 16.996 27.054 16.996 27.054 qs_ot_get_derivative 116 11.6 0.001 0.002 22.382 22.631 make_m2s 4970 13.5 0.094 0.100 19.848 20.853 make_images 4970 14.5 1.914 2.283 19.553 20.558 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.871 19.403 apply_single 127 13.6 0.001 0.001 18.871 19.403 ot_diis_step 116 11.6 0.017 0.018 19.253 19.255 qs_ot_get_p 127 10.4 0.001 0.002 18.880 19.136 cp_fm_cholesky_invert 11 10.9 16.538 16.546 16.538 16.546 cp_fm_upper_to_full 104 14.7 10.976 16.224 10.976 16.224 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.355 15.115 sum_up_and_integrate 127 10.3 0.140 0.152 14.942 14.970 qs_ot_p2m_diag 82 11.4 0.338 0.386 14.891 14.944 integrate_v_rspace 127 11.3 0.004 0.004 14.801 14.835 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.748 14.791 calculate_rho_elec 127 8.7 0.174 0.189 14.747 14.790 cp_dbcsr_syevd 82 12.4 0.005 0.007 13.627 13.629 dbcsr_complete_redistribute 393 12.7 1.501 1.636 8.925 12.646 make_images_data 4970 15.5 0.063 0.068 10.632 12.302 multiply_cannon_sync_h2d 29820 15.5 11.589 12.145 11.589 12.145 dbcsr_mm_accdrv_process 61748 16.2 7.187 8.333 11.347 11.893 hybrid_alltoall_any 5155 16.4 0.507 2.053 9.657 11.189 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.508 11.182 init_scf_run 11 5.9 0.000 0.001 10.901 10.902 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.900 10.902 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.554 10.556 cp_fm_redistribute_end 82 14.4 1.827 10.484 1.842 10.488 cp_fm_diag_elpa_base 82 14.4 8.047 9.872 8.614 10.355 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.296 9.909 pw_transfer 1535 11.6 0.087 0.105 9.615 9.696 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.389 9.475 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.222 9.406 mp_alltoall_i22 712 14.1 5.498 9.138 5.498 9.138 density_rs2pw 127 9.7 0.006 0.006 7.236 8.659 fft_wrap_pw1pw2_140 519 13.2 0.480 0.493 8.311 8.385 grid_integrate_task_list 127 12.3 7.421 7.914 7.421 7.914 cp_fm_cholesky_decompose 22 10.9 7.809 7.905 7.809 7.905 wfi_extrapolate 11 7.9 0.001 0.001 7.793 7.793 fft3d_ps 1281 14.7 2.757 2.835 7.623 7.697 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.723 6.899 mp_irecv_dv 75445 16.2 2.579 6.646 2.579 6.646 rs_pw_transfer 1038 11.9 0.013 0.015 4.794 6.285 calculate_dm_sparse 127 9.5 0.001 0.001 6.128 6.219 mp_alltoall_d11v 2401 14.1 4.950 5.819 4.950 5.819 grid_collocate_task_list 127 9.7 5.040 5.703 5.040 5.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.429 5.486 potential_pw2rs 127 12.3 0.022 0.022 4.796 4.814 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.465 4.468 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.359 4.427 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.347 4.420 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.182 4.241 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.966000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.181818, yerr=25.394409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.496351E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 176.089 176.090 qs_mol_dyn_low 1 2.0 0.003 0.003 175.713 175.726 qs_forces 11 3.9 0.003 0.003 175.599 175.603 qs_energies 11 4.9 0.002 0.002 168.050 168.058 scf_env_do_scf 11 5.9 0.001 0.001 150.838 150.844 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 115.035 115.036 velocity_verlet 10 3.0 0.001 0.001 114.977 114.981 qs_scf_new_mos 117 7.6 0.001 0.001 75.798 75.877 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.797 75.876 dbcsr_multiply_generic 2507 12.6 0.183 0.187 74.620 75.256 ot_scf_mini 117 9.6 0.003 0.004 71.347 71.429 multiply_cannon 2507 13.6 0.583 0.607 54.845 59.376 multiply_cannon_loop 2507 14.6 0.442 0.459 50.173 51.322 ot_mini 117 10.6 0.001 0.001 39.999 40.092 init_scf_loop 11 6.9 0.000 0.000 35.654 35.655 mp_waitall_1 125778 16.7 26.426 34.194 26.426 34.194 prepare_preconditioner 11 7.9 0.000 0.000 31.671 31.702 make_preconditioner 11 8.9 0.000 0.000 31.671 31.702 rebuild_ks_matrix 128 8.3 0.001 0.001 30.323 30.456 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.020 30.323 30.456 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.527 29.783 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.574 27.698 multiply_cannon_multrec 10028 15.6 10.391 15.584 17.648 21.884 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.922 20.101 ot_diis_step 117 11.6 0.020 0.022 20.047 20.047 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.601 19.964 apply_single 128 13.6 0.001 0.001 19.601 19.963 qs_ot_get_derivative 117 11.6 0.001 0.002 19.873 19.956 qs_ot_get_p 128 10.4 0.001 0.001 18.383 18.520 make_m2s 5014 13.6 0.065 0.070 16.079 18.382 cp_fm_cholesky_invert 11 10.9 18.135 18.142 18.135 18.142 make_images 5014 14.6 2.257 2.652 15.772 18.079 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.598 15.636 calculate_rho_elec 128 8.7 0.259 0.268 15.597 15.636 sum_up_and_integrate 128 10.3 0.180 0.190 15.290 15.343 integrate_v_rspace 128 11.3 0.004 0.004 15.110 15.173 qs_ot_p2m_diag 83 11.4 0.495 0.500 14.526 14.540 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.382 13.383 make_images_data 5014 15.6 0.051 0.059 9.809 12.260 multiply_cannon_sync_h2d 10028 15.6 11.552 12.255 11.552 12.255 hybrid_alltoall_any 5200 16.5 0.845 3.832 9.640 12.179 init_scf_run 11 5.9 0.000 0.001 10.371 10.371 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.370 10.371 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.294 10.303 cp_fm_diag_elpa_base 83 14.4 10.048 10.125 10.287 10.298 pw_transfer 1547 11.6 0.085 0.093 10.079 10.121 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.856 9.898 fft_wrap_pw1pw2_140 523 13.2 0.496 0.519 8.689 8.732 grid_integrate_task_list 128 12.3 7.750 8.158 7.750 8.158 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.000 8.061 density_rs2pw 128 9.7 0.005 0.006 7.181 8.061 fft3d_ps 1291 14.7 2.704 2.783 8.016 8.040 cp_fm_cholesky_decompose 22 10.9 7.860 7.954 7.860 7.954 dbcsr_mm_accdrv_process 20762 16.1 2.592 3.484 6.890 7.528 wfi_extrapolate 11 7.9 0.001 0.001 7.437 7.437 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.433 7.428 mp_allgather_i34 2507 14.6 2.898 7.323 2.898 7.323 calculate_dm_sparse 128 9.5 0.001 0.001 6.104 6.190 grid_collocate_task_list 128 9.7 5.391 6.062 5.391 6.062 mp_alltoall_d11v 2415 14.1 5.171 6.019 5.171 6.019 dbcsr_complete_redistribute 395 12.7 2.117 2.201 5.229 5.591 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.885 5.335 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 5.329 5.331 mp_irecv_dv 28860 15.9 1.849 5.251 1.849 5.251 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.132 5.141 rs_pw_transfer 1046 11.9 0.012 0.013 4.213 5.139 potential_pw2rs 128 12.3 0.027 0.028 4.680 4.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.058 4.099 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.553 3.834 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.777 3.824 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.394 3.711 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.658 3.669 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.593 3.653 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.508 3.534 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.090000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1395.363636, yerr=44.208746 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.021566E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_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 0.080 0.105 305.118 305.120 qs_mol_dyn_low 1 2.0 0.003 0.003 304.506 304.519 qs_forces 11 3.9 0.003 0.003 304.411 304.414 qs_energies 11 4.9 0.002 0.002 294.906 294.913 scf_env_do_scf 11 5.9 0.001 0.001 272.198 272.214 velocity_verlet 10 3.0 0.001 0.001 219.732 219.739 scf_env_do_scf_inner_loop 116 6.6 0.004 0.010 143.062 143.064 init_scf_loop 11 6.9 0.000 0.000 128.876 128.878 prepare_preconditioner 11 7.9 0.000 0.000 123.685 123.716 make_preconditioner 11 8.9 0.000 0.000 123.685 123.716 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.203 120.830 qs_scf_new_mos 116 7.6 0.001 0.001 91.984 92.090 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.983 92.089 ot_scf_mini 116 9.6 0.003 0.004 87.219 87.287 dbcsr_multiply_generic 2485 12.5 0.212 0.218 84.021 84.700 cp_fm_upper_to_full 104 14.8 54.647 79.207 54.647 79.207 multiply_cannon 2485 13.5 0.703 0.753 58.818 59.602 multiply_cannon_loop 2485 14.5 0.467 0.477 55.034 56.626 dbcsr_complete_redistribute 393 12.7 3.973 4.029 31.148 45.189 ot_mini 116 10.6 0.001 0.001 44.893 44.952 copy_fm_to_dbcsr 208 11.6 0.002 0.002 27.740 41.828 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.436 39.362 rebuild_ks_matrix 127 8.3 0.001 0.001 38.651 38.727 qs_ks_build_kohn_sham_matrix 127 9.3 0.041 0.041 38.650 38.726 mp_alltoall_i22 712 14.1 23.250 37.455 23.250 37.455 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.498 35.579 cp_fm_cholesky_invert 11 10.9 34.301 34.307 34.301 34.307 mp_waitall_1 102768 16.8 28.972 32.490 28.972 32.490 qs_ot_get_p 127 10.4 0.001 0.001 26.800 26.861 qs_ot_get_derivative 116 11.6 0.002 0.002 24.466 24.532 qs_ot_p2m_diag 82 11.4 0.867 0.871 22.709 22.742 make_m2s 4970 13.5 0.075 0.076 20.457 21.567 make_images 4970 14.5 3.731 3.904 19.981 21.092 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.985 20.987 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.851 20.869 calculate_rho_elec 127 8.7 0.479 0.480 20.851 20.869 ot_diis_step 116 11.6 0.022 0.022 20.391 20.392 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.950 20.286 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.655 19.896 apply_single 127 13.6 0.001 0.001 19.654 19.895 sum_up_and_integrate 127 10.3 0.319 0.320 19.785 19.863 integrate_v_rspace 127 11.3 0.004 0.004 19.466 19.547 multiply_cannon_multrec 9940 15.5 10.378 12.112 17.978 18.146 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.753 17.754 cp_fm_diag_elpa_base 82 14.4 13.264 14.968 17.748 17.749 multiply_cannon_sync_h2d 9940 15.5 15.546 15.572 15.546 15.572 hybrid_alltoall_any 5155 16.4 1.289 3.003 11.267 13.207 make_images_data 4970 15.5 0.059 0.064 11.184 13.135 init_scf_run 11 5.9 0.000 0.001 12.673 12.673 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.673 12.673 pw_transfer 1535 11.6 0.092 0.092 12.533 12.539 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.300 12.306 fft_wrap_pw1pw2_140 519 13.2 0.536 0.540 10.917 10.934 fft3d_ps 1281 14.7 2.721 2.728 10.355 10.363 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.497 9.555 mp_alltoall_d11v 2401 14.1 8.262 9.213 8.262 9.213 wfi_extrapolate 11 7.9 0.001 0.001 9.203 9.203 cp_fm_cholesky_decompose 22 10.9 9.103 9.200 9.103 9.200 dbcsr_mm_accdrv_process 20590 16.0 3.691 5.598 7.362 9.168 grid_integrate_task_list 127 12.3 8.525 8.686 8.525 8.686 density_rs2pw 127 9.7 0.005 0.006 8.505 8.647 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.028 8.028 calculate_dm_sparse 127 9.5 0.001 0.001 6.741 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.579 6.658 grid_collocate_task_list 127 9.7 6.317 6.337 6.317 6.337 rs_scatter_matrices 138 9.7 3.616 4.451 5.995 6.214 copy_dbcsr_to_fm 185 11.7 0.004 0.005 6.081 6.194 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.120000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2704.272727, yerr=167.450992 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 1.261900E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.018 0.037 85.120 85.121 qs_energies 1 2.0 0.000 0.000 84.653 84.679 ls_scf 1 3.0 0.000 0.000 83.733 83.763 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.533 72.671 multiply_cannon 111 7.7 0.018 0.021 55.851 56.912 multiply_cannon_loop 111 8.7 0.210 0.226 52.457 53.869 ls_scf_main 1 4.0 0.000 0.000 52.299 52.299 density_matrix_trs4 2 5.0 0.002 0.003 46.753 46.823 ls_scf_init_scf 1 4.0 0.000 0.000 28.384 28.392 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.272 27.330 mp_waitall_1 11031 10.9 22.306 25.327 22.306 25.327 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.963 24.981 multiply_cannon_multrec 2664 9.7 8.211 8.974 15.517 17.275 multiply_cannon_sync_h2d 2664 9.7 13.795 15.313 13.795 15.313 make_m2s 222 7.7 0.008 0.011 13.081 13.534 make_images 222 8.7 0.099 0.110 13.059 13.514 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.566 12.452 make_images_data 222 9.7 0.004 0.005 7.660 8.277 hybrid_alltoall_any 227 10.6 0.215 1.833 6.604 8.012 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.406 7.933 dbcsr_mm_accdrv_process 4760 10.4 0.504 0.605 6.925 7.914 dbcsr_mm_accdrv_process_sort 4760 11.4 6.218 7.122 6.218 7.122 calculate_norms 4752 9.8 5.529 6.255 5.529 6.255 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.162 mp_sum_l 807 5.4 3.072 4.767 3.072 4.767 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.052 3.695 mp_irecv_dv 6231 10.9 2.035 3.660 2.035 3.660 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.296 3.530 make_images_sizes 222 9.7 0.000 0.000 0.729 3.509 mp_alltoall_i44 222 10.7 0.729 3.508 0.729 3.508 arnoldi_extremal 4 6.8 0.000 0.000 3.252 3.271 arnoldi_normal_ev 4 7.8 0.001 0.004 3.251 3.271 build_subspace 16 8.4 0.009 0.012 3.152 3.155 ls_scf_post 1 4.0 0.000 0.000 3.050 3.076 ls_scf_store_result 1 5.0 0.000 0.000 2.845 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.339 2.797 dbcsr_merge_single_wm 555 10.7 0.460 0.593 2.331 2.789 make_images_pack 222 9.7 2.207 2.621 2.209 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.322 2.561 dbcsr_sort_data 658 11.4 2.128 2.544 2.128 2.544 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.454 2.066 2.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.308 2.381 buffer_matrices_ensure_size 222 8.7 1.757 2.063 1.757 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.792 1.793 rebuild_ks_matrix 3 7.3 0.000 0.000 1.782 1.783 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.782 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.121000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 2.104209E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_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 0.027 0.044 90.975 90.976 qs_energies 1 2.0 0.000 0.000 90.513 90.518 ls_scf 1 3.0 0.000 0.000 89.195 89.200 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.052 75.376 multiply_cannon 111 7.7 0.030 0.043 53.372 56.769 ls_scf_main 1 4.0 0.000 0.000 54.787 54.791 multiply_cannon_loop 111 8.7 0.116 0.124 50.076 53.019 density_matrix_trs4 2 5.0 0.002 0.003 49.019 49.209 ls_scf_init_scf 1 4.0 0.000 0.000 30.832 30.833 mp_waitall_1 9105 10.9 21.090 30.183 21.090 30.183 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.596 29.703 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.200 27.216 multiply_cannon_multrec 1332 9.7 13.059 16.728 22.258 27.033 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.028 21.837 make_m2s 222 7.7 0.006 0.007 14.948 15.613 make_images 222 8.7 1.578 1.944 14.918 15.584 dbcsr_mm_accdrv_process 4041 10.4 0.282 0.445 8.804 10.325 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.892 8.399 9.892 make_images_data 222 9.7 0.004 0.004 8.413 9.316 hybrid_alltoall_any 227 10.6 0.522 2.438 7.879 8.755 mp_sum_l 807 5.4 5.425 8.165 5.425 8.165 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.239 7.696 mp_irecv_dv 3311 11.0 3.220 7.643 3.220 7.643 calculate_norms 2376 9.8 5.986 6.791 5.986 6.791 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.132 6.553 multiply_cannon_sync_h2d 1332 9.7 4.723 5.834 4.723 5.834 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.092 5.295 arnoldi_extremal 4 6.8 0.000 0.000 4.739 4.763 arnoldi_normal_ev 4 7.8 0.001 0.004 4.739 4.762 build_subspace 16 8.4 0.014 0.021 4.474 4.476 ls_scf_post 1 4.0 0.000 0.000 3.576 3.582 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.174 3.419 ls_scf_store_result 1 5.0 0.000 0.000 3.247 3.397 dbcsr_matrix_vector_mult_local 304 10.0 2.734 3.218 2.736 3.219 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.264 3.053 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.615 2.678 make_images_pack 222 9.7 2.024 2.457 2.026 2.459 mp_allgather_i34 111 8.7 1.003 2.233 1.003 2.233 dbcsr_sort_data 436 11.2 1.831 2.037 1.831 2.037 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.915 rebuild_ks_matrix 3 7.3 0.000 0.000 1.900 1.902 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.900 1.902 dbcsr_data_new 4174 10.1 1.614 1.850 1.614 1.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.976000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 2.719613E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_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 0.038 0.089 94.006 94.007 qs_energies 1 2.0 0.000 0.000 93.323 93.373 ls_scf 1 3.0 0.000 0.001 91.920 91.971 dbcsr_multiply_generic 111 6.7 0.015 0.018 76.488 76.794 ls_scf_main 1 4.0 0.001 0.014 57.581 57.586 multiply_cannon 111 7.7 0.039 0.071 52.458 56.433 multiply_cannon_loop 111 8.7 0.099 0.106 48.871 52.881 density_matrix_trs4 2 5.0 0.002 0.004 51.506 51.715 mp_waitall_1 7281 11.0 23.686 33.866 23.686 33.866 ls_scf_init_scf 1 4.0 0.000 0.003 30.716 30.718 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.502 29.582 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.040 27.054 multiply_cannon_multrec 888 9.7 12.572 15.215 21.087 24.215 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.846 22.781 make_m2s 222 7.7 0.006 0.007 17.308 18.547 make_images 222 8.7 1.968 2.290 17.269 18.508 hybrid_alltoall_any 227 10.6 0.619 2.876 9.499 11.124 make_images_data 222 9.7 0.003 0.004 9.899 10.879 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.414 8.041 9.226 mp_sum_l 807 5.4 5.194 8.946 5.194 8.946 dbcsr_mm_accdrv_process_sort 3754 11.4 7.671 8.812 7.671 8.812 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.459 7.228 multiply_cannon_sync_h2d 888 9.7 6.027 7.204 6.027 7.204 mp_irecv_dv 2335 11.1 2.445 7.181 2.445 7.181 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.872 6.867 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.559 6.685 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.993 5.241 arnoldi_extremal 4 6.8 0.000 0.000 5.138 5.155 arnoldi_normal_ev 4 7.8 0.001 0.005 5.138 5.155 build_subspace 16 8.4 0.014 0.020 4.836 4.841 calculate_norms 1584 9.8 4.281 4.674 4.281 4.674 mp_allgather_i34 111 8.7 1.426 3.868 1.426 3.868 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.459 3.796 ls_scf_post 1 4.0 0.004 0.030 3.623 3.676 dbcsr_matrix_vector_mult_local 304 10.0 3.009 3.577 3.012 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.339 3.432 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.981 make_images_sizes 222 9.7 0.000 0.000 1.086 2.221 mp_alltoall_i44 222 10.7 1.085 2.221 1.085 2.221 make_images_pack 222 9.7 1.808 2.131 1.810 2.134 dbcsr_sort_data 325 11.1 1.899 2.128 1.899 2.128 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.944 1.946 dbcsr_data_release 9322 10.9 1.311 1.944 1.311 1.944 rebuild_ks_matrix 3 7.3 0.000 0.000 1.926 1.928 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.926 1.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.007000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2203.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 3.354296E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_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 0.035 0.050 97.329 97.330 qs_energies 1 2.0 0.000 0.000 96.685 96.690 ls_scf 1 3.0 0.000 0.000 94.973 94.980 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.700 78.975 ls_scf_main 1 4.0 0.000 0.002 59.101 59.101 multiply_cannon 111 7.7 0.051 0.125 51.485 56.314 density_matrix_trs4 2 5.0 0.002 0.004 52.912 53.021 multiply_cannon_loop 111 8.7 0.115 0.127 46.431 50.221 ls_scf_init_scf 1 4.0 0.000 0.002 32.615 32.619 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.386 31.452 mp_waitall_1 6369 11.0 22.703 29.297 22.703 29.297 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.924 28.938 multiply_cannon_multrec 1332 9.7 14.267 17.683 22.118 24.892 make_m2s 222 7.7 0.007 0.008 21.316 22.784 make_images 222 8.7 3.145 3.599 21.265 22.736 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.077 17.308 make_images_data 222 9.7 0.004 0.004 11.928 13.749 hybrid_alltoall_any 227 10.6 0.796 3.762 11.187 12.993 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.401 7.477 8.980 dbcsr_mm_accdrv_process_sort 3641 11.4 7.104 8.563 7.104 8.563 mp_sum_l 807 5.4 4.311 7.646 4.311 7.646 multiply_cannon_sync_h2d 1332 9.7 5.524 6.149 5.524 6.149 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.337 5.984 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.060 5.903 mp_irecv_dv 3229 10.9 2.038 5.855 2.038 5.855 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.525 5.482 arnoldi_extremal 4 6.8 0.000 0.000 5.204 5.217 arnoldi_normal_ev 4 7.8 0.001 0.004 5.204 5.217 build_subspace 16 8.4 0.014 0.021 4.864 4.870 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.743 calculate_norms 2376 9.8 4.187 4.568 4.187 4.568 mp_allgather_i34 111 8.7 2.181 4.497 2.181 4.497 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.574 3.858 dbcsr_matrix_vector_mult_local 304 10.0 3.167 3.672 3.169 3.674 dbcsr_sort_data 658 11.4 3.032 3.470 3.032 3.470 ls_scf_post 1 4.0 0.000 0.001 3.258 3.263 dbcsr_special_finalize 555 9.7 0.006 0.007 2.767 3.244 dbcsr_merge_single_wm 555 10.7 0.528 0.664 2.759 3.236 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.079 3.134 ls_scf_store_result 1 5.0 0.000 0.000 2.997 3.058 dbcsr_data_release 10477 10.7 1.604 2.402 1.604 2.402 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.984 1.986 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.968 rebuild_ks_matrix 3 7.3 0.000 0.000 1.961 1.962 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.961 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.330000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 4.631388E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_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 0.044 0.057 99.677 99.678 qs_energies 1 2.0 0.000 0.000 98.934 98.939 ls_scf 1 3.0 0.000 0.000 96.953 96.963 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.027 78.208 ls_scf_main 1 4.0 0.000 0.000 62.395 62.396 multiply_cannon 111 7.7 0.104 0.188 55.375 60.639 density_matrix_trs4 2 5.0 0.002 0.003 55.047 55.127 multiply_cannon_loop 111 8.7 0.070 0.079 50.830 52.211 mp_waitall_1 5436 11.0 26.778 32.453 26.778 32.453 ls_scf_init_scf 1 4.0 0.000 0.000 30.907 30.912 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.513 29.557 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.340 27.348 multiply_cannon_multrec 444 9.7 14.073 16.637 21.098 24.592 make_m2s 222 7.7 0.004 0.005 17.891 20.410 make_images 222 8.7 3.728 4.435 17.829 20.349 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.254 15.871 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.148 14.717 make_images_data 222 9.7 0.003 0.004 10.176 12.575 hybrid_alltoall_any 227 10.6 0.789 3.784 9.850 12.389 multiply_cannon_sync_h2d 444 9.7 6.514 9.082 6.514 9.082 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.345 6.727 7.860 dbcsr_mm_accdrv_process_sort 3003 11.4 6.412 7.513 6.412 7.513 mp_allgather_i34 111 8.7 2.778 7.060 2.778 7.060 arnoldi_extremal 4 6.8 0.000 0.000 5.891 5.900 arnoldi_normal_ev 4 7.8 0.002 0.006 5.891 5.900 build_subspace 16 8.4 0.015 0.019 5.496 5.507 mp_sum_l 807 5.4 2.804 5.280 2.804 5.280 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.583 4.729 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.184 4.404 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.628 4.280 mp_irecv_dv 1241 11.2 1.610 4.261 1.610 4.261 dbcsr_matrix_vector_mult_local 304 10.0 3.664 4.128 3.666 4.130 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.933 3.949 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.783 3.868 calculate_norms 792 9.8 3.551 3.701 3.551 3.701 ls_scf_post 1 4.0 0.000 0.000 3.651 3.656 make_images_sizes 222 9.7 0.000 0.000 1.070 3.642 mp_alltoall_i44 222 10.7 1.070 3.642 1.070 3.642 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.475 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.285 2.286 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.260 rebuild_ks_matrix 3 7.3 0.000 0.000 2.252 2.253 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.252 2.253 dbcsr_merge_all 275 8.9 0.472 0.524 2.046 2.085 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.678000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9ee8bdf25fd706f00eae9f0d758bddb493a004a9_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 8.715522E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_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 0.080 0.096 108.812 108.813 qs_energies 1 2.0 0.000 0.000 107.405 107.425 ls_scf 1 3.0 0.000 0.000 104.454 104.473 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.811 77.934 ls_scf_main 1 4.0 0.000 0.000 66.398 66.399 density_matrix_trs4 2 5.0 0.002 0.003 57.072 57.132 multiply_cannon 111 7.7 0.131 0.321 50.213 52.025 multiply_cannon_loop 111 8.7 0.067 0.069 46.758 47.733 ls_scf_init_scf 1 4.0 0.000 0.000 34.366 34.366 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.681 32.700 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.855 29.864 mp_waitall_1 4527 11.1 22.590 26.306 22.590 26.306 make_m2s 222 7.7 0.005 0.005 23.900 24.885 make_images 222 8.7 4.603 4.976 23.794 24.778 multiply_cannon_multrec 444 9.7 17.835 18.719 22.434 23.090 hybrid_alltoall_any 227 10.6 1.661 3.623 12.941 15.680 make_images_data 222 9.7 0.003 0.004 13.128 15.563 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.846 11.352 multiply_cannon_sync_h2d 444 9.7 8.867 9.028 8.867 9.028 arnoldi_extremal 4 6.8 0.000 0.000 7.398 7.413 arnoldi_normal_ev 4 7.8 0.003 0.009 7.398 7.413 build_subspace 16 8.4 0.026 0.036 6.837 6.844 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.458 5.606 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.283 5.376 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.053 5.304 dbcsr_matrix_vector_mult_local 304 10.0 4.964 5.257 4.967 5.260 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.324 4.423 4.552 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.257 4.122 4.257 ls_scf_post 1 4.0 0.000 0.000 3.689 3.708 make_images_sizes 222 9.7 0.000 0.000 1.436 3.487 mp_alltoall_i44 222 10.7 1.436 3.486 1.436 3.486 mp_allgather_i34 111 8.7 1.111 3.486 1.111 3.486 ls_scf_store_result 1 5.0 0.000 0.000 3.437 3.447 calculate_norms 792 9.8 3.219 3.276 3.219 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.083 3.157 dbcsr_merge_all 275 8.9 0.892 0.919 2.868 2.938 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.921 2.921 dbcsr_complete_redistribute 5 7.6 1.432 1.467 2.764 2.889 dbcsr_data_release 12724 10.6 2.326 2.866 2.326 2.866 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.598 2.601 matrix_ls_to_qs 2 6.0 0.000 0.000 2.443 2.577 rebuild_ks_matrix 3 7.3 0.000 0.000 2.532 2.535 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.532 2.535 dbcsr_sort_data 325 11.1 2.442 2.492 2.442 2.492 dbcsr_new_transposed 4 7.5 0.246 0.255 2.306 2.319 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.193 2.235 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.204 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.813000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6850.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9ee8bdf25fd706f00eae9f0d758bddb493a004a9 Summary: empty Status: OK