=== 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: 5c6c216e857c9fb3b35611c0159716e1fe51e750 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/01 job id: 42396342 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/02 job id: 42396343 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/03 job id: 42396345 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/04 job id: 42396348 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/05 job id: 42396350 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/06 job id: 42396351 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/07 job id: 42396354 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/08 job id: 42396367 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/09 job id: 42396368 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/10 job id: 42396370 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/11 job id: 42396372 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/12 job id: 42396374 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/13 job id: 42396377 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/14 job id: 42396379 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/15 job id: 42396380 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/16 job id: 42396382 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/17 job id: 42396383 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/18 job id: 42396385 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/19 job id: 42396386 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/20 job id: 42396387 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/21 job id: 42396388 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/22 job id: 42396389 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/23 job id: 42396390 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/24 job id: 42396391 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/25 job id: 42396392 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/26 job id: 42396393 --- 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 0.025 0.034 133.305 133.306 farming_run 1 2.0 132.696 132.697 133.274 133.277 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.451569E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 115.395 115.396 qs_energies 1 2.0 0.000 0.000 115.123 115.126 mp2_main 1 3.0 0.000 0.000 113.195 113.198 mp2_gpw_main 1 4.0 0.019 0.025 112.318 112.321 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.432 93.788 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.517 55.874 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.166 41.831 47.130 get_2c_integrals 1 6.0 0.000 0.000 37.377 37.743 integrate_v_rspace 273 8.0 0.439 0.453 25.182 30.306 pw_transfer 6555 10.6 0.377 0.391 27.462 28.191 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 26.097 26.724 grid_integrate_task_list 273 9.0 20.978 26.615 20.978 26.615 fft_wrap_pw1pw2_100 2178 12.4 1.201 1.409 23.662 24.290 compute_2c_integrals 1 7.0 0.002 0.002 19.614 19.615 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.942 19.342 mp2_eri_2c_integrate_gpw 1 9.0 2.365 2.427 18.939 19.341 rpa_ri_compute_en 1 5.0 0.000 0.000 18.776 19.075 cp_fm_cholesky_decompose 12 8.2 17.746 18.119 17.746 18.119 cholesky_decomp 1 7.0 0.000 0.000 16.605 16.976 fft3d_s 5443 13.4 16.207 16.667 16.229 16.689 ao_to_mo_and_store_B_mult_1 272 7.0 10.849 15.550 10.849 15.550 calculate_wavefunction 272 8.0 5.400 5.550 12.553 13.152 rpa_num_int 1 6.0 0.000 0.000 10.626 10.626 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.579 10.591 calc_mat_Q 8 8.0 0.000 0.000 9.391 9.486 contract_S_to_Q 8 9.0 0.000 0.000 8.812 8.905 calc_potential_gpw 544 9.5 0.005 0.006 8.314 8.651 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.263 8.559 parallel_gemm_fm 14 9.1 0.000 0.000 8.377 8.494 parallel_gemm_fm_cosma 14 10.1 8.377 8.494 8.377 8.494 potential_pw2rs 545 10.0 0.108 0.110 7.705 8.446 collocate_single_gaussian 272 10.0 0.041 0.044 7.495 7.746 create_integ_mat 1 6.0 0.022 0.027 7.699 7.699 array2fm 1 7.0 0.000 0.000 6.709 7.147 pw_scatter_s 2720 13.7 4.424 4.710 4.424 4.710 pw_gather_s 2722 13.2 3.897 4.368 3.897 4.368 array2fm_buffer_send 1 8.0 3.009 3.189 3.009 3.189 pw_poisson_solve 545 10.5 1.124 1.179 2.195 2.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.321005, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 0.053 0.117 396.862 396.864 farming_run 1 2.0 395.696 395.725 396.797 396.825 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225191E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.049 209.773 209.774 qs_energies 1 2.0 0.000 0.000 209.524 209.554 scf_env_do_scf 1 3.0 0.000 0.000 106.614 106.614 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.753 105.760 rebuild_ks_matrix 4 6.0 0.000 0.000 105.751 105.759 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.751 105.759 hfx_ks_matrix 4 8.0 0.001 0.001 105.364 105.368 integrate_four_center 4 9.0 0.144 0.454 105.364 105.367 mp2_main 1 3.0 0.001 0.009 102.624 102.655 mp2_gpw_main 1 4.0 0.035 0.056 101.799 101.822 integrate_four_center_main 4 10.0 0.122 0.672 96.644 99.573 integrate_four_center_bin 266 11.0 96.521 99.571 96.521 99.571 init_scf_loop 1 4.0 0.000 0.000 92.481 92.481 mp2_ri_gpw_compute_in 1 5.0 0.068 0.148 74.963 76.004 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.579 55.623 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.144 47.352 integrate_v_rspace 95 8.0 0.399 0.566 28.521 33.589 pw_transfer 2240 10.6 0.145 0.161 29.902 30.239 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.912 29.273 grid_integrate_task_list 95 9.0 23.793 29.002 23.793 29.002 mp2_ri_gpw_compute_en 1 5.0 0.072 0.123 26.687 28.546 fft_wrap_pw1pw2_100 730 12.4 1.268 1.480 26.669 27.069 ao_to_mo_and_store_B_mult_1 91 7.0 10.748 26.572 10.748 26.572 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.819 1.885 24.980 24.990 get_2c_integrals 1 6.0 0.000 0.005 20.299 20.317 compute_2c_integrals 1 7.0 0.003 0.005 19.286 19.290 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.858 19.160 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.865 18.857 19.158 fft3d_s 1823 13.4 18.431 18.721 18.445 18.734 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.131 14.131 calculate_wavefunction 91 8.0 2.016 2.059 9.748 9.963 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.581 8.773 9.366 potential_pw2rs 186 10.0 0.034 0.036 8.673 9.143 local_gemm 172 8.0 8.215 8.800 8.215 8.800 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.002 8.225 8.608 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.519 8.000 8.447 calc_potential_gpw 182 9.5 0.002 0.002 7.874 8.163 collocate_single_gaussian 91 10.0 0.017 0.024 7.885 8.134 mp_sendrecv_dm3 2068 8.0 6.024 6.463 6.024 6.463 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.449 6.345 6.449 mp_sync 38 10.4 3.300 5.731 3.300 5.731 pw_gather_s 912 13.2 4.966 5.373 4.966 5.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.799299, 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/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 451.780608E+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 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.060 52.717 52.720 qs_mol_dyn_low 1 2.0 0.003 0.003 52.440 52.458 qs_forces 11 3.9 0.001 0.002 52.377 52.378 qs_energies 11 4.9 0.001 0.002 50.906 50.942 scf_env_do_scf 11 5.9 0.000 0.001 44.090 44.090 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.968 41.968 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.155 34.596 qs_scf_new_mos 108 7.5 0.000 0.000 32.012 32.329 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.012 32.328 ot_scf_mini 108 9.5 0.002 0.002 30.338 30.529 multiply_cannon 2286 13.5 0.184 0.188 26.539 28.283 multiply_cannon_loop 2286 14.5 1.479 1.564 25.867 27.601 velocity_verlet 10 3.0 0.001 0.001 25.409 25.409 ot_mini 108 10.5 0.001 0.001 19.516 19.765 qs_ot_get_derivative 108 11.5 0.001 0.001 16.464 16.642 mp_waitall_1 267858 16.1 9.026 15.702 9.026 15.702 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.274 14.343 multiply_cannon_multrec 54864 15.5 4.207 6.565 7.526 10.958 rebuild_ks_matrix 119 8.3 0.000 0.000 7.954 8.085 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.954 8.085 multiply_cannon_sync_h2d 54864 15.5 5.889 7.864 5.889 7.864 mp_sum_l 7207 12.9 5.662 7.456 5.662 7.456 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.041 7.161 qs_ot_get_p 119 10.4 0.001 0.001 6.186 6.484 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.774 5.897 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.398 5.826 init_scf_run 11 5.9 0.000 0.001 5.604 5.605 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.604 5.604 sum_up_and_integrate 119 10.3 0.012 0.015 4.563 4.570 integrate_v_rspace 119 11.3 0.002 0.002 4.551 4.559 dbcsr_mm_accdrv_process 76910 16.1 1.173 1.824 3.242 4.539 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.031 4.112 calculate_rho_elec 119 8.7 0.011 0.016 4.031 4.111 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.899 3.554 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.418 3.482 calculate_first_density_matrix 1 7.0 0.000 0.000 3.332 3.339 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.929 3.132 apply_single 119 13.6 0.000 0.000 2.929 3.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.121 3.128 calculate_dm_sparse 119 9.5 0.000 0.000 2.891 3.042 rs_pw_transfer 974 11.9 0.012 0.013 2.806 2.929 ot_diis_step 108 11.5 0.006 0.006 2.780 2.782 acc_transpose_blocks 54864 15.5 0.223 0.243 2.103 2.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.603 2.665 jit_kernel_multiply 13 15.8 2.007 2.630 2.007 2.630 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.626 2.626 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.456 2.505 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.431 2.431 cp_fm_redistribute_end 50 14.0 2.212 2.412 2.216 2.413 cp_fm_diag_elpa_base 50 14.0 0.196 2.348 0.196 2.357 density_rs2pw 119 9.7 0.004 0.004 2.109 2.237 wfi_extrapolate 11 7.9 0.001 0.001 2.214 2.214 grid_integrate_task_list 119 12.3 2.024 2.147 2.024 2.147 init_scf_loop 11 6.9 0.000 0.000 2.106 2.106 mp_sum_d 4129 12.0 1.357 1.896 1.357 1.896 potential_pw2rs 119 12.3 0.004 0.004 1.876 1.886 make_m2s 4572 13.5 0.053 0.055 1.651 1.704 mp_alltoall_d11v 2130 13.8 1.525 1.667 1.525 1.667 make_images 4572 14.5 0.133 0.138 1.570 1.622 pw_transfer 1439 11.6 0.051 0.055 1.549 1.617 acc_transpose_blocks_kernels 54864 16.5 0.250 0.383 1.133 1.590 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.474 1.543 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.398 1.447 mp_waitany 12084 13.8 1.321 1.431 1.321 1.431 grid_collocate_task_list 119 9.7 1.286 1.352 1.286 1.352 fft3d_ps 1201 14.6 0.356 0.458 1.253 1.316 jit_kernel_transpose 5 15.5 0.883 1.213 0.883 1.213 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.131 1.201 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.785 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.720000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 488.005632E+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 1032117. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.090 39.562 39.563 qs_mol_dyn_low 1 2.0 0.006 0.021 39.125 39.131 qs_forces 11 3.9 0.005 0.017 38.773 38.779 qs_energies 11 4.9 0.015 0.058 37.077 37.090 scf_env_do_scf 11 5.9 0.001 0.001 31.709 31.710 scf_env_do_scf_inner_loop 108 6.5 0.005 0.013 29.068 29.069 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.581 21.920 qs_scf_new_mos 108 7.5 0.001 0.001 19.968 20.209 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.967 20.208 ot_scf_mini 108 9.5 0.003 0.003 19.049 19.222 velocity_verlet 10 3.0 0.001 0.001 18.767 18.773 multiply_cannon 2286 13.5 0.208 0.217 16.465 18.013 multiply_cannon_loop 2286 14.5 0.902 0.978 15.349 16.678 ot_mini 108 10.5 0.001 0.001 11.849 12.085 mp_waitall_1 217478 16.2 6.223 11.414 6.223 11.414 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.172 9.587 qs_ot_get_derivative 108 11.5 0.001 0.001 9.380 9.557 multiply_cannon_multrec 27432 15.5 1.987 4.600 5.965 8.825 rebuild_ks_matrix 119 8.3 0.000 0.000 7.290 7.428 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.290 7.428 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.466 6.589 dbcsr_mm_accdrv_process 47894 16.0 3.120 5.130 3.909 5.771 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.656 4.501 qs_ot_get_p 119 10.4 0.001 0.001 4.240 4.471 sum_up_and_integrate 119 10.3 0.024 0.027 4.303 4.309 integrate_v_rspace 119 11.3 0.002 0.002 4.279 4.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.047 4.105 apply_single 119 13.6 0.000 0.000 3.047 4.105 mp_sum_l 7207 12.9 2.079 4.027 2.079 4.027 init_scf_run 11 5.9 0.000 0.001 3.972 3.972 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.972 3.972 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.680 3.708 calculate_rho_elec 119 8.7 0.021 0.024 3.679 3.707 rs_pw_transfer 974 11.9 0.010 0.011 2.739 2.858 multiply_cannon_sync_h2d 27432 15.5 2.190 2.846 2.190 2.846 make_m2s 4572 13.5 0.054 0.071 2.608 2.828 make_images 4572 14.5 0.201 0.240 2.517 2.736 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.702 2.720 init_scf_loop 11 6.9 0.001 0.006 2.619 2.619 calculate_first_density_matrix 1 7.0 0.002 0.011 2.462 2.465 ot_diis_step 108 11.5 0.011 0.011 2.419 2.421 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.275 2.275 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.143 2.234 calculate_dm_sparse 119 9.5 0.000 0.000 2.131 2.205 density_rs2pw 119 9.7 0.004 0.004 1.999 2.075 potential_pw2rs 119 12.3 0.006 0.007 2.012 2.021 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.944 1.945 cp_fm_redistribute_end 50 14.0 1.606 1.922 1.610 1.924 grid_integrate_task_list 119 12.3 1.830 1.923 1.830 1.923 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.003 1.908 1.910 cp_fm_diag_elpa_base 50 14.0 0.305 1.843 0.312 1.874 pw_transfer 1439 11.6 0.063 0.067 1.793 1.827 jit_kernel_multiply 9 16.1 0.736 1.783 0.736 1.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.707 1.747 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.704 1.740 prepare_preconditioner 11 7.9 0.000 0.000 1.608 1.635 make_preconditioner 11 8.9 0.000 0.000 1.608 1.635 make_images_data 4572 15.5 0.045 0.052 1.180 1.607 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.512 1.568 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.018 1.512 acc_transpose_blocks 27432 15.5 0.108 0.112 1.192 1.495 fft3d_ps 1201 14.6 0.507 0.561 1.412 1.444 wfi_extrapolate 11 7.9 0.001 0.001 1.441 1.441 mp_alltoall_d11v 2130 13.8 1.278 1.435 1.278 1.435 fft_wrap_pw1pw2_140 487 13.2 0.075 0.083 1.314 1.350 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.320 1.327 grid_collocate_task_list 119 9.7 1.238 1.320 1.238 1.320 mp_allgather_i34 2286 14.5 0.560 1.316 0.560 1.316 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.211 1.257 mp_sum_d 4129 12.0 0.633 1.086 0.633 1.086 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.045 1.057 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.900 make_images_sizes 4572 15.5 0.005 0.005 0.611 0.885 mp_alltoall_i44 4572 16.5 0.606 0.880 0.606 0.880 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.661 0.875 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.821 0.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.563000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.363636, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.665984E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.070 34.391 34.392 qs_mol_dyn_low 1 2.0 0.003 0.004 34.124 34.131 qs_forces 11 3.9 0.002 0.003 34.036 34.039 qs_energies 11 4.9 0.001 0.003 32.403 32.409 scf_env_do_scf 11 5.9 0.001 0.001 26.648 26.648 scf_env_do_scf_inner_loop 108 6.5 0.005 0.027 23.874 23.875 dbcsr_multiply_generic 2286 12.5 0.095 0.096 17.931 18.061 velocity_verlet 10 3.0 0.001 0.001 16.284 16.287 qs_scf_new_mos 108 7.5 0.001 0.001 15.493 15.513 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.492 15.513 multiply_cannon 2286 13.5 0.194 0.201 14.151 15.106 ot_scf_mini 108 9.5 0.002 0.003 14.743 14.759 multiply_cannon_loop 2286 14.5 0.636 0.664 13.319 14.346 ot_mini 108 10.5 0.001 0.001 9.290 9.309 qs_ot_get_derivative 108 11.5 0.001 0.001 7.714 7.730 multiply_cannon_multrec 18288 15.5 1.919 2.856 6.968 7.305 rebuild_ks_matrix 119 8.3 0.000 0.000 6.687 6.715 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.687 6.715 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.894 5.919 dbcsr_mm_accdrv_process 38222 16.0 4.084 5.598 4.966 5.718 mp_waitall_1 169478 16.3 3.449 4.598 3.449 4.598 init_scf_run 11 5.9 0.000 0.001 4.532 4.532 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.532 4.532 sum_up_and_integrate 119 10.3 0.029 0.030 4.165 4.170 integrate_v_rspace 119 11.3 0.002 0.003 4.135 4.143 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.961 3.602 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.445 3.458 calculate_rho_elec 119 8.7 0.030 0.031 3.445 3.458 qs_ot_get_p 119 10.4 0.001 0.001 3.232 3.265 calculate_first_density_matrix 1 7.0 0.001 0.004 3.181 3.182 init_scf_loop 11 6.9 0.000 0.002 2.753 2.753 rs_pw_transfer 974 11.9 0.009 0.010 2.430 2.561 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.560 2.541 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.427 2.429 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.110 2.415 apply_single 119 13.6 0.000 0.000 2.110 2.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.239 2.255 make_m2s 4572 13.5 0.044 0.045 2.056 2.157 density_rs2pw 119 9.7 0.004 0.004 1.992 2.090 make_images 4572 14.5 0.191 0.203 1.970 2.071 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.060 2.067 jit_kernel_multiply 10 15.8 0.830 2.034 0.830 2.034 calculate_dm_sparse 119 9.5 0.000 0.000 2.005 2.012 pw_transfer 1439 11.6 0.063 0.066 1.904 1.919 grid_integrate_task_list 119 12.3 1.803 1.906 1.803 1.906 acc_transpose_blocks 18288 15.5 0.075 0.076 1.867 1.892 prepare_preconditioner 11 7.9 0.000 0.000 1.877 1.880 make_preconditioner 11 8.9 0.000 0.000 1.877 1.880 potential_pw2rs 119 12.3 0.007 0.008 1.841 1.849 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.814 1.831 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.711 1.796 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.773 1.774 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.694 1.706 mp_sum_l 7207 12.9 1.336 1.676 1.336 1.676 ot_diis_step 108 11.5 0.011 0.011 1.550 1.551 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.535 1.536 multiply_cannon_sync_h2d 18288 15.5 1.386 1.531 1.386 1.531 fft3d_ps 1201 14.6 0.508 0.524 1.501 1.520 cp_fm_redistribute_end 50 14.0 1.136 1.509 1.137 1.509 cp_fm_diag_elpa_base 50 14.0 0.357 1.446 0.371 1.483 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.451 1.465 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 1.429 1.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.354 1.358 grid_collocate_task_list 119 9.7 1.206 1.329 1.206 1.329 wfi_extrapolate 11 7.9 0.001 0.001 1.300 1.300 jit_kernel_transpose 5 15.6 1.219 1.236 1.219 1.236 make_images_data 4572 15.5 0.045 0.049 0.971 1.101 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.505 1.078 mp_alltoall_d11v 2130 13.8 0.904 1.001 0.904 1.001 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.843 0.991 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 0.960 0.962 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.913 0.940 mp_alltoall_z22v 1201 16.6 0.833 0.910 0.833 0.910 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.867 cp_fm_cholesky_invert 11 10.9 0.803 0.807 0.803 0.807 mp_allgather_i34 2286 14.5 0.294 0.736 0.294 0.736 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.641 0.730 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.392000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=2.405228 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 549.081088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.043 36.654 36.655 qs_mol_dyn_low 1 2.0 0.003 0.004 36.291 36.298 qs_forces 11 3.9 0.002 0.003 35.929 35.930 qs_energies 11 4.9 0.004 0.026 34.177 34.183 scf_env_do_scf 11 5.9 0.000 0.001 28.120 28.121 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.600 24.601 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.113 19.232 velocity_verlet 10 3.0 0.001 0.001 18.536 18.551 qs_scf_new_mos 108 7.5 0.001 0.001 16.353 16.419 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.353 16.419 multiply_cannon 2286 13.5 0.229 0.271 15.397 16.271 ot_scf_mini 108 9.5 0.002 0.003 15.379 15.433 multiply_cannon_loop 2286 14.5 0.934 0.962 14.426 15.151 ot_mini 108 10.5 0.001 0.001 9.358 9.431 multiply_cannon_multrec 27432 15.5 2.330 3.013 8.862 9.188 dbcsr_mm_accdrv_process 47916 15.9 5.603 7.017 6.438 7.650 qs_ot_get_derivative 108 11.5 0.001 0.001 7.512 7.571 rebuild_ks_matrix 119 8.3 0.000 0.000 6.565 6.629 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.565 6.629 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.790 5.847 init_scf_run 11 5.9 0.000 0.001 4.689 4.689 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.688 4.689 sum_up_and_integrate 119 10.3 0.034 0.037 3.792 3.800 integrate_v_rspace 119 11.3 0.002 0.003 3.758 3.766 init_scf_loop 11 6.9 0.001 0.004 3.500 3.501 qs_ot_get_p 119 10.4 0.001 0.001 3.367 3.439 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.408 3.433 calculate_rho_elec 119 8.7 0.040 0.046 3.407 3.433 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.858 3.273 calculate_first_density_matrix 1 7.0 0.000 0.003 3.193 3.199 mp_waitall_1 145218 16.4 2.321 2.944 2.321 2.944 prepare_preconditioner 11 7.9 0.000 0.000 2.664 2.672 make_preconditioner 11 8.9 0.000 0.002 2.664 2.672 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.592 apply_single 119 13.6 0.000 0.000 2.060 2.591 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.211 2.568 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.492 2.495 make_m2s 4572 13.5 0.054 0.055 2.275 2.404 calculate_dm_sparse 119 9.5 0.000 0.000 2.276 2.329 make_images 4572 14.5 0.274 0.334 2.167 2.295 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.274 2.294 rs_pw_transfer 974 11.9 0.009 0.009 2.078 2.209 jit_kernel_multiply 10 16.1 0.774 2.125 0.774 2.125 acc_transpose_blocks 27432 15.5 0.112 0.114 2.038 2.062 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.047 2.055 density_rs2pw 119 9.7 0.004 0.004 1.856 1.985 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.938 1.966 grid_integrate_task_list 119 12.3 1.827 1.899 1.827 1.899 pw_transfer 1439 11.6 0.063 0.066 1.795 1.823 ot_diis_step 108 11.5 0.012 0.012 1.807 1.807 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.737 1.738 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.705 1.736 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.882 1.655 mp_sum_l 7207 12.9 1.021 1.605 1.021 1.605 potential_pw2rs 119 12.3 0.008 0.010 1.537 1.541 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.480 1.482 cp_fm_redistribute_end 50 14.0 0.976 1.457 0.979 1.460 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.426 1.436 cp_fm_diag_elpa_base 50 14.0 0.457 1.391 0.477 1.432 acc_transpose_blocks_kernels 27432 16.5 0.269 0.279 1.412 1.429 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.392 1.426 fft3d_ps 1201 14.6 0.534 0.587 1.386 1.412 wfi_extrapolate 11 7.9 0.001 0.001 1.407 1.407 grid_collocate_task_list 119 9.7 1.221 1.310 1.221 1.310 cp_fm_upper_to_full 72 13.5 0.848 1.191 0.848 1.191 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.145 1.166 jit_kernel_transpose 5 15.6 1.143 1.157 1.143 1.157 dbcsr_complete_redistribute 329 12.2 0.118 0.143 0.841 1.125 multiply_cannon_sync_h2d 27432 15.5 0.995 1.103 0.995 1.103 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.084 1.085 make_images_data 4572 15.5 0.045 0.049 0.907 1.052 mp_alltoall_d11v 2130 13.8 0.864 1.003 0.864 1.003 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.800 0.980 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.630 0.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.865 0.871 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 mp_alltoall_i22 627 13.8 0.474 0.789 0.474 0.789 cp_fm_cholesky_invert 11 10.9 0.780 0.783 0.780 0.783 mp_alltoall_z22v 1201 16.6 0.734 0.762 0.734 0.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.655000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=520.454545, yerr=3.026058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 611.434496E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 27.375 27.375 qs_mol_dyn_low 1 2.0 0.003 0.004 27.213 27.220 qs_forces 11 3.9 0.002 0.002 27.140 27.141 qs_energies 11 4.9 0.001 0.001 25.465 25.468 scf_env_do_scf 11 5.9 0.000 0.001 20.718 20.718 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.240 18.241 velocity_verlet 10 3.0 0.001 0.001 13.934 13.937 dbcsr_multiply_generic 2286 12.5 0.124 0.127 12.146 12.234 qs_scf_new_mos 108 7.5 0.001 0.001 10.861 10.887 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.860 10.886 ot_scf_mini 108 9.5 0.002 0.002 10.200 10.230 multiply_cannon 2286 13.5 0.231 0.242 9.643 10.099 multiply_cannon_loop 2286 14.5 0.331 0.344 8.747 8.940 multiply_cannon_multrec 9144 15.5 1.657 1.891 5.891 6.100 rebuild_ks_matrix 119 8.3 0.000 0.000 5.786 5.809 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.786 5.809 ot_mini 108 10.5 0.001 0.001 5.741 5.775 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.164 5.185 qs_ot_get_derivative 108 11.5 0.001 0.001 4.460 4.488 dbcsr_mm_accdrv_process 12550 15.8 3.390 3.885 4.134 4.211 sum_up_and_integrate 119 10.3 0.037 0.040 3.522 3.526 integrate_v_rspace 119 11.3 0.003 0.003 3.484 3.489 init_scf_run 11 5.9 0.000 0.001 3.314 3.314 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.314 3.314 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.066 3.073 calculate_rho_elec 119 8.7 0.059 0.061 3.065 3.073 qs_ot_get_p 119 10.4 0.001 0.001 2.690 2.731 init_scf_loop 11 6.9 0.000 0.000 2.458 2.459 calculate_first_density_matrix 1 7.0 0.000 0.000 2.201 2.201 mp_waitall_1 121218 16.5 1.575 2.034 1.575 2.034 grid_integrate_task_list 119 12.3 1.845 1.927 1.845 1.927 make_m2s 4572 13.5 0.034 0.036 1.719 1.884 jit_kernel_multiply 10 16.1 0.705 1.813 0.705 1.813 make_images 4572 14.5 0.269 0.300 1.630 1.793 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.745 1.748 calculate_dm_sparse 119 9.5 0.000 0.000 1.699 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.702 1.706 make_preconditioner 11 8.9 0.000 0.000 1.702 1.706 pw_transfer 1439 11.6 0.063 0.065 1.681 1.688 rs_pw_transfer 974 11.9 0.008 0.008 1.572 1.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.594 1.623 density_rs2pw 119 9.7 0.004 0.004 1.538 1.612 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.601 1.602 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.590 1.597 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.577 1.593 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.524 1.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.390 1.400 grid_collocate_task_list 119 9.7 1.271 1.342 1.271 1.342 potential_pw2rs 119 12.3 0.010 0.011 1.311 1.313 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.301 1.310 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.271 1.274 ot_diis_step 108 11.5 0.012 0.013 1.268 1.269 fft3d_ps 1201 14.6 0.537 0.549 1.257 1.266 cp_fm_redistribute_end 50 14.0 0.634 1.253 0.635 1.254 cp_fm_diag_elpa_base 50 14.0 0.573 1.176 0.615 1.231 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.209 1.209 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.138 1.163 apply_single 119 13.6 0.000 0.000 1.138 1.163 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.151 1.156 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.113 1.127 wfi_extrapolate 11 7.9 0.001 0.001 1.070 1.070 make_images_data 4572 15.5 0.039 0.043 0.804 1.006 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.722 0.980 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.915 acc_transpose_blocks 9144 15.5 0.039 0.040 0.907 0.913 cp_fm_cholesky_invert 11 10.9 0.851 0.853 0.851 0.853 multiply_cannon_sync_h2d 9144 15.5 0.707 0.787 0.707 0.787 mp_alltoall_d11v 2130 13.8 0.679 0.773 0.679 0.773 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.743 0.745 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.712 0.722 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.665 0.716 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.665 0.668 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.327 0.648 mp_allgather_i34 2286 14.5 0.230 0.633 0.230 0.633 mp_alltoall_z22v 1201 16.6 0.593 0.625 0.593 0.625 dbcsr_complete_redistribute 329 12.2 0.183 0.229 0.582 0.608 qs_create_task_list 11 7.9 0.000 0.000 0.544 0.569 generate_qs_task_list 11 8.9 0.190 0.212 0.544 0.569 jit_kernel_transpose 5 15.6 0.548 0.552 0.548 0.552 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.375000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=579.363636, yerr=3.471251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 770.289664E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.042 42.714 42.714 qs_mol_dyn_low 1 2.0 0.003 0.003 42.459 42.466 qs_forces 11 3.9 0.005 0.006 42.398 42.399 qs_energies 11 4.9 0.016 0.031 40.399 40.403 scf_env_do_scf 11 5.9 0.000 0.001 33.770 33.770 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.706 25.708 velocity_verlet 10 3.0 0.001 0.001 23.643 23.649 dbcsr_multiply_generic 2286 12.5 0.206 0.253 18.690 18.870 qs_scf_new_mos 108 7.5 0.001 0.001 16.481 16.576 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.480 16.576 ot_scf_mini 108 9.5 0.002 0.002 15.363 15.464 multiply_cannon 2286 13.5 0.302 0.310 14.011 14.937 multiply_cannon_loop 2286 14.5 0.345 0.353 12.687 13.596 ot_mini 108 10.5 0.001 0.001 9.226 9.342 multiply_cannon_multrec 9144 15.5 3.400 4.720 8.682 8.759 init_scf_loop 11 6.9 0.000 0.000 8.036 8.039 qs_ot_get_derivative 108 11.5 0.001 0.001 7.156 7.258 rebuild_ks_matrix 119 8.3 0.000 0.000 7.095 7.238 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.095 7.238 prepare_preconditioner 11 7.9 0.000 0.000 7.039 7.053 make_preconditioner 11 8.9 0.000 0.000 7.039 7.053 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.613 6.925 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.394 6.523 dbcsr_mm_accdrv_process 12550 15.8 4.201 5.611 5.161 6.450 cp_fm_upper_to_full 72 14.2 3.176 4.539 3.176 4.539 init_scf_run 11 5.9 0.000 0.001 4.525 4.525 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.525 4.525 mp_waitall_1 97218 16.6 2.912 4.007 2.912 4.007 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.943 3.949 calculate_rho_elec 119 8.7 0.118 0.121 3.942 3.948 sum_up_and_integrate 119 10.3 0.064 0.065 3.856 3.863 integrate_v_rspace 119 11.3 0.003 0.003 3.792 3.798 qs_ot_get_p 119 10.4 0.001 0.001 3.260 3.402 make_m2s 4572 13.5 0.038 0.038 2.851 3.088 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.645 3.064 calculate_first_density_matrix 1 7.0 0.000 0.000 3.027 3.029 make_images 4572 14.5 0.354 0.384 2.730 2.966 dbcsr_complete_redistribute 329 12.2 0.343 0.372 2.069 2.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.653 2.654 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.742 2.560 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.206 2.493 apply_single 119 13.6 0.000 0.000 2.206 2.493 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.494 2.442 mp_alltoall_i22 627 13.8 1.483 2.339 1.483 2.339 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.236 2.276 calculate_dm_sparse 119 9.5 0.000 0.000 2.240 2.263 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.421 2.234 pw_transfer 1439 11.6 0.066 0.066 2.221 2.228 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.125 2.132 grid_integrate_task_list 119 12.3 2.026 2.055 2.026 2.055 ot_diis_step 108 11.5 0.014 0.014 2.042 2.042 density_rs2pw 119 9.7 0.003 0.004 2.006 2.019 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.945 1.947 mp_sum_l 7207 12.9 1.181 1.849 1.181 1.849 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.751 1.803 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.787 1.794 qs_energies_init_hamiltonians 11 5.9 0.022 0.024 1.781 1.781 fft3d_ps 1201 14.6 0.567 0.579 1.759 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.671 1.672 cp_fm_cholesky_invert 11 10.9 1.630 1.633 1.630 1.633 rs_pw_transfer 974 11.9 0.009 0.009 1.497 1.519 hybrid_alltoall_any 4725 16.4 0.087 0.150 1.211 1.489 make_images_data 4572 15.5 0.043 0.046 1.207 1.489 grid_collocate_task_list 119 9.7 1.461 1.475 1.461 1.475 jit_kernel_multiply 8 15.5 0.932 1.461 0.932 1.461 wfi_extrapolate 11 7.9 0.001 0.001 1.440 1.440 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.373 1.373 cp_fm_diag_elpa_base 50 14.0 1.230 1.283 1.371 1.371 mp_alltoall_d11v 2130 13.8 1.316 1.341 1.316 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.274 1.283 potential_pw2rs 119 12.3 0.014 0.015 1.260 1.263 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.218 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.100 1.115 mp_alltoall_z22v 1201 16.6 1.057 1.076 1.057 1.076 multiply_cannon_sync_h2d 9144 15.5 1.039 1.044 1.039 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.030 acc_transpose_blocks 9144 15.5 0.039 0.039 0.958 0.996 qs_create_task_list 11 7.9 0.000 0.000 0.949 0.959 generate_qs_task_list 11 8.9 0.373 0.393 0.949 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.914 0.927 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.714000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.636364, yerr=19.165655 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.314112E+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 797052. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.143 0.176 84.400 84.403 qs_mol_dyn_low 1 2.0 0.003 0.004 83.903 83.912 qs_forces 11 3.9 0.003 0.004 83.820 83.821 qs_energies 11 4.9 0.004 0.021 80.970 81.010 scf_env_do_scf 11 5.9 0.000 0.001 71.583 71.585 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.852 65.852 dbcsr_multiply_generic 2055 12.4 0.206 0.216 53.246 53.550 qs_scf_new_mos 99 7.5 0.000 0.001 48.959 49.077 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.959 49.077 ot_scf_mini 99 9.5 0.002 0.003 46.493 46.563 multiply_cannon 2055 13.4 0.178 0.184 42.975 43.935 multiply_cannon_loop 2055 14.4 1.548 1.601 41.926 42.836 velocity_verlet 10 3.0 0.001 0.001 42.452 42.454 ot_mini 99 10.5 0.001 0.001 28.347 28.426 qs_ot_get_derivative 99 11.5 0.001 0.001 21.387 21.455 multiply_cannon_multrec 49320 15.4 12.057 12.798 17.101 17.945 rebuild_ks_matrix 110 8.3 0.000 0.001 14.312 14.393 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.311 14.393 mp_waitall_1 241148 16.1 12.236 13.375 12.236 13.375 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.571 12.645 multiply_cannon_sync_h2d 49320 15.4 9.923 10.559 9.923 10.559 qs_ot_get_p 110 10.4 0.001 0.001 9.401 9.507 multiply_cannon_metrocomm3 49320 15.4 0.075 0.079 7.028 8.133 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.277 7.753 apply_single 110 13.6 0.000 0.001 7.277 7.753 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.141 7.670 init_scf_run 11 5.9 0.000 0.001 7.343 7.346 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.343 7.346 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.024 7.067 sum_up_and_integrate 110 10.3 0.037 0.043 6.865 6.877 integrate_v_rspace 110 11.3 0.003 0.003 6.828 6.848 ot_diis_step 99 11.5 0.005 0.006 6.679 6.683 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.297 6.417 calculate_rho_elec 110 8.6 0.020 0.024 6.296 6.417 mp_sum_l 6514 12.8 5.443 6.281 5.443 6.281 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.078 6.097 init_scf_loop 11 6.9 0.000 0.002 5.705 5.706 dbcsr_mm_accdrv_process 87628 16.1 2.090 2.237 4.923 5.274 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.053 5.053 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.583 4.585 cp_fm_redistribute_end 48 14.0 3.993 4.563 3.996 4.563 cp_fm_diag_elpa_base 48 14.0 0.561 4.491 0.564 4.515 make_m2s 4110 13.4 0.060 0.065 4.293 4.402 make_images 4110 14.4 0.178 0.191 4.198 4.309 wfi_extrapolate 11 7.9 0.001 0.001 4.126 4.126 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.081 4.085 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 3.040 3.972 calculate_dm_sparse 110 9.5 0.000 0.001 3.866 3.971 rs_pw_transfer 902 11.9 0.012 0.014 3.658 3.921 prepare_preconditioner 11 7.9 0.000 0.000 3.585 3.606 make_preconditioner 11 8.9 0.000 0.000 3.585 3.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.439 3.487 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.354 3.427 density_rs2pw 110 9.6 0.004 0.004 3.215 3.413 grid_integrate_task_list 110 12.3 3.246 3.391 3.246 3.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.332 3.376 calculate_first_density_matrix 1 7.0 0.001 0.010 3.133 3.141 mp_alltoall_d11v 2046 13.8 2.391 2.846 2.391 2.846 pw_transfer 1331 11.6 0.053 0.063 2.787 2.842 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.701 2.758 jit_kernel_multiply 13 15.9 2.547 2.581 2.547 2.581 potential_pw2rs 110 12.3 0.006 0.007 2.475 2.491 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.302 2.359 acc_transpose_blocks 49320 15.4 0.214 0.228 2.208 2.285 fft3d_ps 1111 14.6 0.753 0.836 2.238 2.285 grid_collocate_task_list 110 9.6 2.082 2.195 2.082 2.195 mp_waitany 14300 13.8 1.826 2.175 1.826 2.175 make_images_data 4110 15.4 0.043 0.046 1.913 2.083 mp_sum_d 3883 11.9 1.581 2.060 1.581 2.060 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.689 1.965 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.886 1.899 cp_fm_cholesky_invert 11 10.9 1.821 1.824 1.821 1.824 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.706 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.403000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.727273, yerr=2.048987 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 589.242368E+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 2857111. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.114 68.476 68.477 qs_mol_dyn_low 1 2.0 0.004 0.011 68.084 68.113 qs_forces 11 3.9 0.009 0.055 67.949 67.951 qs_energies 11 4.9 0.003 0.011 64.609 64.616 scf_env_do_scf 11 5.9 0.001 0.001 55.941 55.945 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.225 48.228 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.673 37.898 velocity_verlet 10 3.0 0.001 0.002 36.001 36.007 qs_scf_new_mos 99 7.5 0.001 0.001 32.578 32.715 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.577 32.715 multiply_cannon 2055 13.4 0.221 0.243 31.008 32.194 ot_scf_mini 99 9.5 0.003 0.004 30.913 31.032 multiply_cannon_loop 2055 14.4 0.931 0.953 29.768 30.887 ot_mini 99 10.5 0.001 0.003 18.323 18.452 multiply_cannon_multrec 24660 15.4 7.639 9.226 13.878 15.541 rebuild_ks_matrix 110 8.3 0.000 0.001 13.520 13.624 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.022 13.519 13.624 qs_ot_get_derivative 99 11.5 0.001 0.002 12.468 12.588 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.903 11.988 mp_waitall_1 186928 16.3 7.824 10.371 7.824 10.371 multiply_cannon_sync_h2d 24660 15.4 7.007 8.165 7.007 8.165 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.159 7.928 init_scf_loop 11 6.9 0.001 0.005 7.681 7.683 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.520 7.234 apply_single 110 13.6 0.000 0.001 6.519 7.234 sum_up_and_integrate 110 10.3 0.052 0.062 6.470 6.482 integrate_v_rspace 110 11.3 0.002 0.004 6.417 6.432 dbcsr_mm_accdrv_process 52282 16.1 4.633 5.679 6.079 6.364 init_scf_run 11 5.9 0.000 0.001 6.163 6.164 scf_env_initial_rho_setup 11 6.9 0.000 0.003 6.163 6.163 qs_ot_get_p 110 10.4 0.001 0.002 5.988 6.134 ot_diis_step 99 11.5 0.010 0.012 5.802 5.803 qs_rho_update_rho_low 110 7.6 0.001 0.002 5.664 5.672 calculate_rho_elec 110 8.6 0.039 0.047 5.663 5.672 prepare_preconditioner 11 7.9 0.000 0.000 5.648 5.671 make_preconditioner 11 8.9 0.000 0.002 5.648 5.671 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.759 5.484 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.206 5.371 make_m2s 4110 13.4 0.057 0.061 4.210 4.624 make_images 4110 14.4 0.401 0.447 4.102 4.515 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.084 4.103 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.661 3.662 wfi_extrapolate 11 7.9 0.003 0.017 3.486 3.486 pw_transfer 1331 11.6 0.065 0.071 3.237 3.376 grid_integrate_task_list 110 12.3 3.154 3.365 3.154 3.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.283 3.285 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.131 3.268 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.178 3.242 density_rs2pw 110 9.6 0.004 0.004 2.976 3.230 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.135 3.137 cp_fm_redistribute_end 48 14.0 2.334 3.110 2.336 3.111 cp_fm_diag_elpa_base 48 14.0 0.740 3.001 0.771 3.087 rs_pw_transfer 902 11.9 0.012 0.013 2.888 3.083 calculate_dm_sparse 110 9.5 0.001 0.001 2.990 3.023 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.946 2.992 make_images_data 4110 15.4 0.048 0.051 2.396 2.817 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.644 2.781 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.082 2.687 fft3d_ps 1111 14.6 1.073 1.268 2.489 2.614 calculate_first_density_matrix 1 7.0 0.001 0.007 2.587 2.589 cp_fm_cholesky_invert 11 10.9 2.556 2.563 2.556 2.563 mp_sum_l 6514 12.8 1.876 2.524 1.876 2.524 potential_pw2rs 110 12.3 0.008 0.009 2.343 2.356 jit_kernel_multiply 12 16.2 1.090 2.330 1.090 2.330 grid_collocate_task_list 110 9.6 2.122 2.284 2.122 2.284 mp_alltoall_d11v 2046 13.8 1.755 2.005 1.755 2.005 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.955 1.980 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.873 1.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.795 multiply_cannon_metrocomm4 22605 15.4 0.073 0.079 0.794 1.717 mp_allgather_i34 2055 14.4 0.574 1.631 0.574 1.631 acc_transpose_blocks 24660 15.4 0.107 0.111 1.557 1.588 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.561 1.572 mp_irecv_dv 57340 16.2 0.671 1.505 0.671 1.505 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.478 cp_fm_cholesky_decompose 22 10.9 1.443 1.449 1.443 1.449 dbcsr_complete_redistribute 325 12.2 0.239 0.299 1.154 1.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.477000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.000000, yerr=8.090398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 658.108416E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.194 60.594 60.596 qs_mol_dyn_low 1 2.0 0.003 0.003 59.960 59.969 qs_forces 11 3.9 0.003 0.004 59.896 59.897 qs_energies 11 4.9 0.004 0.007 56.726 56.731 scf_env_do_scf 11 5.9 0.001 0.002 48.199 48.199 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.674 39.675 velocity_verlet 10 3.0 0.001 0.001 32.509 32.510 dbcsr_multiply_generic 2055 12.4 0.216 0.251 28.924 29.195 qs_scf_new_mos 99 7.5 0.001 0.001 25.166 25.253 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.165 25.252 ot_scf_mini 99 9.5 0.003 0.003 23.949 24.052 multiply_cannon 2055 13.4 0.213 0.228 22.114 23.185 multiply_cannon_loop 2055 14.4 0.618 0.633 20.944 21.994 ot_mini 99 10.5 0.001 0.001 13.478 13.581 rebuild_ks_matrix 110 8.3 0.000 0.001 12.153 12.282 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.152 12.282 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.718 10.836 multiply_cannon_multrec 16440 15.4 3.929 4.876 9.702 10.538 mp_waitall_1 146766 16.3 7.191 9.885 7.191 9.885 qs_ot_get_derivative 99 11.5 0.001 0.001 9.031 9.134 init_scf_loop 11 6.9 0.001 0.004 8.490 8.492 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.189 6.934 prepare_preconditioner 11 7.9 0.000 0.000 6.705 6.720 make_preconditioner 11 8.9 0.001 0.002 6.705 6.720 sum_up_and_integrate 110 10.3 0.060 0.060 6.417 6.433 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.067 6.414 integrate_v_rspace 110 11.3 0.003 0.003 6.357 6.373 init_scf_run 11 5.9 0.001 0.004 5.981 5.982 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.981 5.982 dbcsr_mm_accdrv_process 34862 16.1 4.715 5.427 5.626 5.827 qs_ot_get_p 110 10.4 0.001 0.002 5.625 5.762 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.592 5.600 calculate_rho_elec 110 8.6 0.058 0.058 5.592 5.600 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.940 5.354 apply_single 110 13.6 0.000 0.000 4.940 5.353 make_m2s 4110 13.4 0.050 0.052 4.541 4.911 make_images 4110 14.4 0.395 0.512 4.426 4.794 ot_diis_step 99 11.5 0.010 0.011 4.419 4.421 multiply_cannon_sync_h2d 16440 15.4 3.715 4.352 3.715 4.352 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.811 3.815 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.088 3.739 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.455 3.455 grid_integrate_task_list 110 12.3 3.204 3.364 3.204 3.364 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.300 3.302 pw_transfer 1331 11.6 0.064 0.070 3.186 3.193 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.080 3.088 density_rs2pw 110 9.6 0.004 0.004 2.881 3.029 calculate_first_density_matrix 1 7.0 0.001 0.004 2.967 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.945 2.946 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.880 2.926 wfi_extrapolate 11 7.9 0.001 0.001 2.922 2.922 rs_pw_transfer 902 11.9 0.011 0.012 2.784 2.920 cp_fm_redistribute_end 48 14.0 1.824 2.914 1.826 2.914 cp_fm_diag_elpa_base 48 14.0 1.027 2.787 1.084 2.894 make_images_data 4110 15.4 0.044 0.051 2.433 2.889 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.170 2.775 cp_fm_cholesky_invert 11 10.9 2.698 2.704 2.698 2.704 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.619 2.627 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.552 fft3d_ps 1111 14.6 1.066 1.079 2.394 2.404 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.314 2.369 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.865 2.350 grid_collocate_task_list 110 9.6 2.178 2.321 2.178 2.321 mp_sum_l 6514 12.8 1.597 2.249 1.597 2.249 potential_pw2rs 110 12.3 0.010 0.010 2.237 2.245 mp_irecv_dv 48980 15.7 0.795 2.232 0.795 2.232 qs_energies_init_hamiltonians 11 5.9 0.002 0.009 2.103 2.105 mp_alltoall_d11v 2046 13.8 1.718 1.995 1.718 1.995 dbcsr_complete_redistribute 325 12.2 0.346 0.424 1.439 1.892 cp_fm_upper_to_full 70 13.6 1.368 1.835 1.368 1.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.647 cp_fm_cholesky_decompose 22 10.9 1.572 1.586 1.572 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.487 1.499 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.364 1.488 jit_kernel_multiply 8 16.4 0.516 1.481 0.516 1.481 mp_allgather_i34 2055 14.4 0.495 1.478 0.495 1.478 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.017 1.471 mp_waitany 17072 13.8 1.185 1.369 1.185 1.369 acc_transpose_blocks 16440 15.4 0.073 0.075 1.274 1.293 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.193 1.287 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.277 1.284 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.596000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.545455, yerr=9.257563 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 727.015424E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.071 65.350 65.351 qs_mol_dyn_low 1 2.0 0.003 0.003 64.847 64.858 qs_forces 11 3.9 0.002 0.003 64.733 64.734 qs_energies 11 4.9 0.005 0.019 61.385 61.389 scf_env_do_scf 11 5.9 0.001 0.001 52.826 52.829 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.934 40.935 velocity_verlet 10 3.0 0.001 0.001 37.068 37.072 dbcsr_multiply_generic 2055 12.4 0.117 0.120 29.894 30.126 qs_scf_new_mos 99 7.5 0.001 0.001 26.550 26.666 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.550 26.665 ot_scf_mini 99 9.5 0.003 0.003 24.920 25.029 multiply_cannon 2055 13.4 0.245 0.265 22.638 23.874 multiply_cannon_loop 2055 14.4 0.881 0.908 21.183 21.814 ot_mini 99 10.5 0.001 0.001 14.460 14.594 multiply_cannon_multrec 24660 15.4 4.201 6.801 12.641 13.786 rebuild_ks_matrix 110 8.3 0.000 0.000 11.985 12.096 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.984 12.095 init_scf_loop 11 6.9 0.001 0.002 11.849 11.850 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.622 10.722 qs_ot_get_derivative 99 11.5 0.001 0.001 10.286 10.400 prepare_preconditioner 11 7.9 0.000 0.000 10.103 10.124 make_preconditioner 11 8.9 0.000 0.001 10.103 10.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.231 9.787 dbcsr_mm_accdrv_process 52304 16.0 6.817 8.676 8.294 9.161 mp_waitall_1 126806 16.4 4.747 6.841 4.747 6.841 sum_up_and_integrate 110 10.3 0.067 0.070 6.339 6.353 integrate_v_rspace 110 11.3 0.003 0.003 6.272 6.285 make_m2s 4110 13.4 0.060 0.062 5.578 5.960 make_images 4110 14.4 0.580 0.707 5.437 5.816 init_scf_run 11 5.9 0.000 0.001 5.718 5.718 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.718 5.718 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.549 5.557 calculate_rho_elec 110 8.6 0.077 0.081 5.549 5.556 qs_ot_get_p 110 10.4 0.001 0.001 5.392 5.542 cp_fm_upper_to_full 70 13.8 3.345 4.836 3.345 4.836 dbcsr_complete_redistribute 325 12.2 0.619 0.684 3.018 4.187 ot_diis_step 99 11.5 0.011 0.011 4.135 4.136 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.051 4.115 apply_single 110 13.6 0.000 0.000 4.050 4.115 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.641 3.655 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.487 3.644 grid_integrate_task_list 110 12.3 3.270 3.452 3.270 3.452 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.383 3.435 multiply_cannon_sync_h2d 24660 15.4 3.181 3.361 3.181 3.361 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.522 3.334 pw_transfer 1331 11.6 0.064 0.071 3.253 3.276 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.409 3.275 make_images_data 4110 15.4 0.046 0.050 2.865 3.238 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.174 3.174 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.148 3.173 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.067 3.069 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.986 3.054 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.864 3.003 mp_alltoall_i22 605 13.7 1.762 2.984 1.762 2.984 calculate_dm_sparse 110 9.5 0.001 0.001 2.916 2.946 wfi_extrapolate 11 7.9 0.001 0.001 2.943 2.943 density_rs2pw 110 9.6 0.004 0.004 2.706 2.856 calculate_first_density_matrix 1 7.0 0.001 0.010 2.673 2.675 cp_fm_cholesky_invert 11 10.9 2.663 2.672 2.663 2.672 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 2.642 2.669 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.651 2.653 cp_fm_redistribute_end 48 14.0 1.319 2.627 1.321 2.628 cp_fm_diag_elpa_base 48 14.0 1.224 2.495 1.304 2.610 rs_pw_transfer 902 11.9 0.010 0.011 2.407 2.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.512 fft3d_ps 1111 14.6 1.068 1.101 2.451 2.469 grid_collocate_task_list 110 9.6 2.222 2.316 2.222 2.316 qs_energies_init_hamiltonians 11 5.9 0.004 0.009 2.309 2.310 potential_pw2rs 110 12.3 0.013 0.013 2.101 2.109 mp_alltoall_d11v 2046 13.8 1.841 2.108 1.841 2.108 jit_kernel_multiply 10 15.8 1.143 1.860 1.143 1.860 cp_fm_cholesky_decompose 22 10.9 1.733 1.778 1.733 1.778 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.697 1.728 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.607 1.618 mp_allgather_i34 2055 14.4 0.552 1.615 0.552 1.615 mp_sum_l 6514 12.8 1.004 1.597 1.004 1.597 acc_transpose_blocks 24660 15.4 0.105 0.107 1.522 1.546 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.536 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.853 1.493 mp_irecv_dv 62702 16.1 0.754 1.418 0.754 1.418 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.256 1.338 mp_waitany 13376 13.8 1.077 1.325 1.077 1.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.351000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=685.545455, yerr=7.024161 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 847.761408E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 55.964 55.964 qs_mol_dyn_low 1 2.0 0.003 0.003 55.725 55.734 qs_forces 11 3.9 0.002 0.002 55.647 55.648 qs_energies 11 4.9 0.001 0.001 52.010 52.020 scf_env_do_scf 11 5.9 0.000 0.001 43.466 43.467 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.451 35.452 velocity_verlet 10 3.0 0.001 0.001 31.899 31.903 dbcsr_multiply_generic 2055 12.4 0.106 0.109 23.242 23.381 qs_scf_new_mos 99 7.5 0.001 0.001 20.934 20.986 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.934 20.985 ot_scf_mini 99 9.5 0.002 0.002 19.632 19.652 multiply_cannon 2055 13.4 0.248 0.262 17.446 18.912 multiply_cannon_loop 2055 14.4 0.325 0.340 16.073 16.387 rebuild_ks_matrix 110 8.3 0.000 0.000 11.763 11.786 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.762 11.785 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.484 10.505 ot_mini 99 10.5 0.001 0.001 10.455 10.474 multiply_cannon_multrec 8220 15.4 3.198 4.429 7.462 8.426 mp_waitall_1 106626 16.5 6.518 8.173 6.518 8.173 init_scf_loop 11 6.9 0.000 0.000 7.969 7.970 qs_ot_get_derivative 99 11.5 0.001 0.001 6.651 6.672 sum_up_and_integrate 110 10.3 0.079 0.081 6.433 6.451 integrate_v_rspace 110 11.3 0.003 0.003 6.353 6.370 prepare_preconditioner 11 7.9 0.000 0.000 6.339 6.344 make_preconditioner 11 8.9 0.000 0.000 6.339 6.344 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.915 6.004 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.754 5.769 calculate_rho_elec 110 8.6 0.115 0.115 5.753 5.768 init_scf_run 11 5.9 0.000 0.001 5.251 5.251 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.251 5.251 qs_ot_get_p 110 10.4 0.001 0.001 5.094 5.110 dbcsr_mm_accdrv_process 17442 15.9 2.811 4.018 4.135 5.097 make_m2s 4110 13.4 0.039 0.040 4.460 4.755 make_images 4110 14.4 0.639 0.690 4.330 4.624 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.898 4.214 ot_diis_step 99 11.5 0.012 0.012 3.784 3.785 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.731 3.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.681 3.707 apply_single 110 13.6 0.000 0.000 3.681 3.707 grid_integrate_task_list 110 12.3 3.361 3.516 3.361 3.516 pw_transfer 1331 11.6 0.065 0.072 3.437 3.463 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.428 3.428 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.330 3.360 make_images_data 4110 15.4 0.038 0.044 2.553 3.011 multiply_cannon_sync_h2d 8220 15.4 2.908 3.007 2.908 3.007 hybrid_alltoall_any 4261 16.3 0.199 0.858 2.526 3.000 cp_fm_cholesky_invert 11 10.9 2.972 2.976 2.972 2.976 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.914 2.914 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.858 2.899 density_rs2pw 110 9.6 0.004 0.004 2.773 2.896 cp_fm_redistribute_end 48 14.0 0.735 2.891 0.739 2.892 wfi_extrapolate 11 7.9 0.001 0.001 2.873 2.873 cp_fm_diag_elpa_base 48 14.0 1.974 2.711 2.146 2.865 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.693 2.695 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.679 2.680 fft3d_ps 1111 14.6 1.116 1.168 2.590 2.618 calculate_dm_sparse 110 9.5 0.001 0.001 2.543 2.577 grid_collocate_task_list 110 9.6 2.333 2.521 2.333 2.521 rs_pw_transfer 902 11.9 0.010 0.010 2.359 2.505 calculate_first_density_matrix 1 7.0 0.000 0.000 2.294 2.296 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.227 2.248 potential_pw2rs 110 12.3 0.015 0.016 2.123 2.131 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.061 2.074 mp_alltoall_d11v 2046 13.8 1.847 2.056 1.847 2.056 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.767 1.990 cp_fm_cholesky_decompose 22 10.9 1.818 1.839 1.818 1.839 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.759 1.770 mp_allgather_i34 2055 14.4 0.561 1.721 0.561 1.721 dbcsr_complete_redistribute 325 12.2 0.569 0.595 1.561 1.676 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.523 1.636 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.613 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.489 1.503 jit_kernel_multiply 8 15.8 1.011 1.445 1.011 1.445 qs_create_task_list 11 7.9 0.000 0.000 1.256 1.352 generate_qs_task_list 11 8.9 0.379 0.446 1.256 1.352 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.006 1.351 mp_waitany 9240 13.8 1.146 1.302 1.146 1.302 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.275 1.298 mp_alltoall_z22v 1111 16.6 1.242 1.260 1.242 1.260 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.489 1.120 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.964000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=800.545455, yerr=11.187360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 1.382044E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.042 85.106 85.107 qs_mol_dyn_low 1 2.0 0.003 0.003 84.767 84.775 qs_forces 11 3.9 0.002 0.002 84.704 84.705 qs_energies 11 4.9 0.002 0.008 80.647 80.649 scf_env_do_scf 11 5.9 0.000 0.001 70.357 70.358 velocity_verlet 10 3.0 0.001 0.001 54.512 54.518 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.897 41.898 dbcsr_multiply_generic 2055 12.4 0.408 0.417 28.510 28.605 init_scf_loop 11 6.9 0.000 0.000 28.387 28.388 prepare_preconditioner 11 7.9 0.000 0.000 26.414 26.421 make_preconditioner 11 8.9 0.000 0.000 26.414 26.421 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.630 25.874 qs_scf_new_mos 99 7.5 0.001 0.001 25.643 25.704 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.642 25.704 ot_scf_mini 99 9.5 0.002 0.002 23.908 23.955 multiply_cannon 2055 13.4 0.341 0.361 21.331 22.033 multiply_cannon_loop 2055 14.4 0.342 0.344 19.576 19.996 cp_fm_upper_to_full 70 14.2 12.832 18.319 12.832 18.319 ot_mini 99 10.5 0.001 0.001 13.360 13.409 rebuild_ks_matrix 110 8.3 0.001 0.001 13.086 13.117 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.086 13.116 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.903 11.930 dbcsr_complete_redistribute 325 12.2 1.123 1.154 7.449 10.682 multiply_cannon_multrec 8220 15.4 4.352 4.536 9.588 9.685 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.432 9.670 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.768 8.981 qs_ot_get_derivative 99 11.5 0.001 0.001 8.888 8.940 mp_alltoall_i22 605 13.7 5.372 8.594 5.372 8.594 mp_waitall_1 87304 16.6 7.489 8.272 7.489 8.272 sum_up_and_integrate 110 10.3 0.150 0.152 6.578 6.595 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.418 6.451 calculate_rho_elec 110 8.6 0.227 0.227 6.418 6.451 integrate_v_rspace 110 11.3 0.004 0.004 6.427 6.444 init_scf_run 11 5.9 0.000 0.001 6.033 6.033 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.033 6.033 make_m2s 4110 13.4 0.043 0.044 5.152 5.622 qs_ot_get_p 110 10.4 0.001 0.001 5.419 5.474 make_images 4110 14.4 0.883 0.935 4.965 5.434 dbcsr_mm_accdrv_process 11614 15.7 3.420 4.026 5.093 5.320 cp_fm_cholesky_invert 11 10.9 5.284 5.288 5.284 5.288 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.648 5.118 apply_single 110 13.6 0.000 0.000 4.648 5.118 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.588 4.969 ot_diis_step 99 11.5 0.015 0.016 4.455 4.455 multiply_cannon_sync_h2d 8220 15.4 3.947 3.955 3.947 3.955 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.921 3.928 pw_transfer 1331 11.6 0.073 0.074 3.715 3.723 grid_integrate_task_list 110 12.3 3.664 3.719 3.664 3.719 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.664 3.665 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.599 3.608 hybrid_alltoall_any 4261 16.3 0.256 0.553 2.798 3.528 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.488 3.488 make_images_data 4110 15.4 0.042 0.045 2.788 3.384 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.831 3.331 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.249 3.253 calculate_dm_sparse 110 9.5 0.001 0.001 3.167 3.195 wfi_extrapolate 11 7.9 0.001 0.001 3.178 3.178 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.131 3.142 density_rs2pw 110 9.6 0.004 0.004 2.924 2.944 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.902 2.903 cp_fm_diag_elpa_base 48 14.0 2.365 2.552 2.900 2.900 fft3d_ps 1111 14.6 1.265 1.281 2.809 2.818 calculate_first_density_matrix 1 7.0 0.000 0.000 2.751 2.753 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.664 2.675 grid_collocate_task_list 110 9.6 2.627 2.650 2.627 2.650 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.346 2.373 qs_env_update_s_mstruct 11 6.9 0.008 0.018 2.203 2.275 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.196 rs_pw_transfer 902 11.9 0.010 0.011 2.091 2.144 mp_alltoall_d11v 2046 13.8 2.013 2.061 2.013 2.061 potential_pw2rs 110 12.3 0.022 0.022 2.015 2.020 cp_fm_cholesky_decompose 22 10.9 1.999 2.017 1.999 2.017 qs_create_task_list 11 7.9 0.000 0.000 1.893 1.939 generate_qs_task_list 11 8.9 0.739 0.791 1.893 1.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.790 jit_kernel_multiply 10 15.3 1.470 1.776 1.470 1.776 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.732 1.766 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.724 1.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.107000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.363636, yerr=71.228535 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 628.932608E+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 2777576. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.035 202.999 203.000 qs_mol_dyn_low 1 2.0 0.003 0.004 202.629 202.643 qs_forces 11 3.9 0.003 0.003 202.468 202.469 qs_energies 11 4.9 0.002 0.003 196.991 197.006 scf_env_do_scf 11 5.9 0.001 0.001 180.667 180.671 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 159.970 159.971 dbcsr_multiply_generic 2507 12.6 0.177 0.183 124.683 125.350 velocity_verlet 10 3.0 0.001 0.001 122.993 122.995 qs_scf_new_mos 117 7.6 0.001 0.001 121.547 121.844 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.546 121.843 ot_scf_mini 117 9.6 0.003 0.003 114.932 115.177 multiply_cannon 2507 13.6 0.238 0.245 101.027 102.607 multiply_cannon_loop 2507 14.6 2.086 2.159 98.867 100.484 ot_mini 117 10.6 0.001 0.001 66.155 66.445 multiply_cannon_multrec 60168 15.6 33.224 35.553 41.565 43.415 qs_ot_get_derivative 117 11.6 0.001 0.001 41.207 41.473 rebuild_ks_matrix 128 8.3 0.001 0.001 33.091 33.347 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.090 33.347 mp_waitall_1 291448 16.2 28.804 31.565 28.804 31.565 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.807 30.052 multiply_cannon_sync_h2d 60168 15.6 27.244 29.179 27.244 29.179 qs_ot_get_p 128 10.4 0.001 0.001 26.763 27.082 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.407 25.308 apply_single 128 13.6 0.001 0.001 24.407 25.308 ot_diis_step 117 11.6 0.007 0.008 24.540 24.542 init_scf_loop 11 6.9 0.000 0.000 20.625 20.626 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.105 20.228 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.067 19.252 multiply_cannon_metrocomm3 60168 15.6 0.109 0.114 15.996 18.221 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.483 17.484 prepare_preconditioner 11 7.9 0.000 0.000 16.051 16.100 make_preconditioner 11 8.9 0.000 0.000 16.051 16.100 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.271 15.443 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.653 14.658 cp_fm_redistribute_end 83 14.4 11.551 14.585 11.562 14.588 cp_fm_diag_elpa_base 83 14.4 2.984 14.361 3.015 14.480 make_m2s 5014 13.6 0.104 0.112 13.698 14.101 make_images 5014 14.6 0.403 0.422 13.517 13.927 sum_up_and_integrate 128 10.3 0.089 0.107 13.911 13.927 integrate_v_rspace 128 11.3 0.003 0.004 13.821 13.838 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.384 12.484 calculate_rho_elec 128 8.7 0.045 0.063 12.384 12.483 init_scf_run 11 5.9 0.000 0.001 12.231 12.232 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.231 12.232 mp_sum_l 7870 13.0 8.695 10.112 8.695 10.112 cp_fm_cholesky_invert 11 10.9 9.116 9.124 9.116 9.124 wfi_extrapolate 11 7.9 0.001 0.001 9.001 9.001 calculate_dm_sparse 128 9.5 0.001 0.001 8.420 8.563 dbcsr_mm_accdrv_process 124484 16.2 3.280 3.479 7.907 8.525 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.192 8.333 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.999 8.103 make_images_data 5014 15.6 0.069 0.075 6.714 7.712 grid_integrate_task_list 128 12.3 7.065 7.503 7.065 7.503 multiply_cannon_metrocomm1 60168 15.6 0.085 0.089 6.069 7.458 hybrid_alltoall_any 5200 16.5 0.289 2.231 5.868 7.198 pw_transfer 1547 11.6 0.074 0.106 6.592 6.876 density_rs2pw 128 9.7 0.006 0.007 6.233 6.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.694 6.701 fft_wrap_pw1pw2 1291 12.7 0.009 0.013 6.389 6.648 rs_pw_transfer 1046 11.9 0.017 0.019 5.476 6.051 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.819 5.919 fft_wrap_pw1pw2_140 523 13.2 0.446 0.509 5.495 5.661 mp_alltoall_d11v 2415 14.1 4.225 5.534 4.225 5.534 fft3d_ps 1291 14.7 2.091 2.577 5.183 5.386 grid_collocate_task_list 128 9.7 4.721 5.091 4.721 5.091 cp_fm_cholesky_decompose 22 10.9 4.745 4.758 4.745 4.758 potential_pw2rs 128 12.3 0.009 0.010 4.323 4.345 mp_sum_d 4464 12.1 3.603 4.242 3.603 4.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.000000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.818182, yerr=7.755563 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 825.696256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5329886. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.040 191.041 191.042 qs_mol_dyn_low 1 2.0 0.005 0.019 190.624 190.637 qs_forces 11 3.9 0.003 0.005 190.501 190.503 qs_energies 11 4.9 0.006 0.031 183.809 183.820 scf_env_do_scf 11 5.9 0.001 0.001 167.392 167.402 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.945 133.948 velocity_verlet 10 3.0 0.001 0.001 120.361 120.364 dbcsr_multiply_generic 2507 12.6 0.189 0.195 98.037 99.220 qs_scf_new_mos 117 7.6 0.001 0.001 95.179 95.677 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.178 95.676 ot_scf_mini 117 9.6 0.004 0.004 90.336 90.969 multiply_cannon 2507 13.6 0.474 0.527 77.544 82.126 multiply_cannon_loop 2507 14.6 1.257 1.304 74.160 76.544 ot_mini 117 10.6 0.001 0.001 49.957 50.508 mp_waitall_1 226760 16.4 25.201 39.356 25.201 39.356 multiply_cannon_multrec 30084 15.6 21.934 26.843 31.598 36.864 rebuild_ks_matrix 128 8.3 0.001 0.001 32.885 33.472 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.885 33.472 init_scf_loop 11 6.9 0.000 0.000 33.358 33.361 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.648 30.191 prepare_preconditioner 11 7.9 0.000 0.000 28.952 29.004 make_preconditioner 11 8.9 0.000 0.000 28.952 29.004 qs_ot_get_derivative 117 11.6 0.001 0.002 28.005 28.640 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.610 28.192 multiply_cannon_metrocomm3 30084 15.6 0.094 0.099 15.753 28.112 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.003 23.281 apply_single 128 13.6 0.001 0.001 22.002 23.280 qs_ot_get_p 128 10.4 0.001 0.001 22.020 22.784 multiply_cannon_sync_h2d 30084 15.6 19.225 22.207 19.225 22.207 ot_diis_step 117 11.6 0.014 0.015 21.772 21.774 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.865 16.899 cp_fm_cholesky_invert 11 10.9 16.874 16.886 16.874 16.886 make_m2s 5014 13.6 0.090 0.097 14.692 16.421 make_images 5014 14.6 1.162 1.359 14.483 16.207 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.710 15.711 sum_up_and_integrate 128 10.3 0.114 0.131 14.340 14.368 integrate_v_rspace 128 11.3 0.003 0.004 14.226 14.260 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.947 12.986 calculate_rho_elec 128 8.7 0.088 0.105 12.946 12.986 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.552 12.566 cp_fm_redistribute_end 83 14.4 7.305 12.488 7.324 12.502 cp_fm_diag_elpa_base 83 14.4 4.939 12.007 5.154 12.378 init_scf_run 11 5.9 0.000 0.001 11.637 11.638 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.636 11.638 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.059 11.523 make_images_data 5014 15.6 0.066 0.074 8.913 10.915 multiply_cannon_metrocomm4 27577 15.6 0.095 0.112 3.768 10.414 hybrid_alltoall_any 5200 16.5 0.339 1.483 7.503 10.318 mp_irecv_dv 69486 16.3 3.575 10.027 3.575 10.027 dbcsr_mm_accdrv_process 62242 16.2 4.604 5.247 9.123 9.684 wfi_extrapolate 11 7.9 0.001 0.001 8.355 8.355 pw_transfer 1547 11.6 0.084 0.102 7.740 7.803 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.517 7.576 grid_integrate_task_list 128 12.3 7.179 7.549 7.179 7.549 cp_fm_cholesky_decompose 22 10.9 7.025 7.109 7.025 7.109 density_rs2pw 128 9.7 0.006 0.007 6.660 7.040 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.182 6.952 calculate_dm_sparse 128 9.5 0.001 0.001 6.518 6.682 fft_wrap_pw1pw2_140 523 13.2 0.469 0.515 6.595 6.668 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.273 6.279 rs_pw_transfer 1046 11.9 0.015 0.016 5.477 5.987 fft3d_ps 1291 14.7 2.792 2.945 5.890 5.927 mp_sum_l 7870 13.0 4.023 5.814 4.023 5.814 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.359 5.481 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.331 5.396 grid_collocate_task_list 128 9.7 4.909 5.295 4.909 5.295 mp_allgather_i34 2507 14.6 1.973 5.124 1.973 5.124 mp_alltoall_d11v 2415 14.1 4.332 4.811 4.332 4.811 potential_pw2rs 128 12.3 0.015 0.016 4.718 4.731 dbcsr_complete_redistribute 395 12.7 0.787 0.889 3.370 4.252 mp_sum_d 4459 12.1 2.686 4.067 2.686 4.067 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.042000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.363636, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 931.033088E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.091 178.256 178.258 qs_mol_dyn_low 1 2.0 0.003 0.003 177.523 177.546 qs_forces 11 3.9 0.003 0.004 177.421 177.425 qs_energies 11 4.9 0.007 0.025 170.884 170.895 scf_env_do_scf 11 5.9 0.001 0.001 155.181 155.182 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.550 119.551 velocity_verlet 10 3.0 0.001 0.001 114.127 114.129 dbcsr_multiply_generic 2507 12.6 0.186 0.199 82.238 83.336 qs_scf_new_mos 117 7.6 0.001 0.001 82.229 82.526 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.229 82.525 ot_scf_mini 117 9.6 0.004 0.004 78.066 78.437 multiply_cannon 2507 13.6 0.500 0.526 61.945 66.511 multiply_cannon_loop 2507 14.6 0.863 0.888 58.752 61.323 ot_mini 117 10.6 0.001 0.001 42.750 43.127 init_scf_loop 11 6.9 0.002 0.010 35.530 35.533 mp_waitall_1 178456 16.5 25.925 35.339 25.925 35.339 prepare_preconditioner 11 7.9 0.000 0.000 31.416 31.464 make_preconditioner 11 8.9 0.000 0.002 31.416 31.464 rebuild_ks_matrix 128 8.3 0.001 0.001 30.401 30.849 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.024 30.400 30.848 make_full_inverse_cholesky 11 9.9 0.000 0.001 29.001 30.477 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.430 27.849 multiply_cannon_multrec 20056 15.6 13.494 16.174 22.173 24.816 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.169 24.196 qs_ot_get_derivative 117 11.6 0.001 0.002 23.122 23.484 qs_ot_get_p 128 10.4 0.001 0.002 20.507 20.965 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.759 20.709 apply_single 128 13.6 0.001 0.001 19.759 20.709 ot_diis_step 117 11.6 0.018 0.018 19.526 19.529 make_m2s 5014 13.6 0.081 0.086 15.186 16.311 multiply_cannon_sync_h2d 20056 15.6 14.284 16.240 14.284 16.240 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.186 16.194 make_images 5014 14.6 1.177 1.302 14.953 16.076 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.182 15.183 cp_fm_cholesky_invert 11 10.9 14.708 14.717 14.708 14.717 sum_up_and_integrate 128 10.3 0.131 0.141 14.225 14.259 integrate_v_rspace 128 11.3 0.003 0.004 14.094 14.121 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.717 13.762 calculate_rho_elec 128 8.7 0.132 0.147 13.716 13.761 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.045 12.045 cp_fm_redistribute_end 83 14.4 4.522 11.987 4.534 11.990 cp_fm_diag_elpa_base 83 14.4 7.029 11.419 7.437 11.912 make_images_data 5014 15.6 0.060 0.069 9.445 10.953 init_scf_run 11 5.9 0.000 0.001 10.601 10.602 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.601 10.601 hybrid_alltoall_any 5200 16.5 0.432 1.960 8.149 10.257 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.215 9.478 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.442 8.856 mp_irecv_dv 50230 16.2 3.318 8.614 3.318 8.614 pw_transfer 1547 11.6 0.084 0.106 8.271 8.380 dbcsr_mm_accdrv_process 41502 16.2 4.339 5.236 8.130 8.293 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.049 8.162 grid_integrate_task_list 128 12.3 7.292 7.744 7.292 7.744 density_rs2pw 128 9.7 0.006 0.006 7.135 7.567 wfi_extrapolate 11 7.9 0.001 0.001 7.506 7.506 cp_fm_cholesky_decompose 22 10.9 7.438 7.477 7.438 7.477 cp_fm_upper_to_full 105 14.5 5.719 7.443 5.719 7.443 fft_wrap_pw1pw2_140 523 13.2 0.478 0.527 7.021 7.155 dbcsr_complete_redistribute 395 12.7 1.236 1.280 4.797 6.561 fft3d_ps 1291 14.7 2.696 2.906 6.319 6.383 calculate_dm_sparse 128 9.5 0.001 0.001 5.771 5.865 rs_pw_transfer 1046 11.9 0.014 0.014 5.329 5.824 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.585 5.590 grid_collocate_task_list 128 9.7 5.066 5.348 5.066 5.348 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.546 5.315 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.600 5.296 mp_alltoall_d11v 2415 14.1 4.494 5.290 4.494 5.290 mp_allgather_i34 2507 14.6 1.727 5.100 1.727 5.100 mp_sum_l 7870 13.0 3.290 4.840 3.290 4.840 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.580 4.706 potential_pw2rs 128 12.3 0.020 0.023 4.494 4.516 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.395 4.147 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.102 4.125 mp_alltoall_i22 716 14.1 2.036 3.990 2.036 3.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.803 3.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.258000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.727273, yerr=8.410933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.149641E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.047 186.861 186.862 qs_mol_dyn_low 1 2.0 0.003 0.003 186.434 186.446 qs_forces 11 3.9 0.005 0.006 186.315 186.319 qs_energies 11 4.9 0.002 0.005 179.317 179.332 scf_env_do_scf 11 5.9 0.001 0.001 162.751 162.764 velocity_verlet 10 3.0 0.001 0.001 124.296 124.299 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.873 115.874 qs_scf_new_mos 117 7.6 0.001 0.001 80.331 80.631 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.331 80.631 dbcsr_multiply_generic 2507 12.6 0.190 0.195 79.553 80.461 ot_scf_mini 117 9.6 0.003 0.004 75.810 76.094 multiply_cannon 2507 13.6 0.560 0.608 54.642 57.833 multiply_cannon_loop 2507 14.6 1.182 1.213 51.071 53.120 init_scf_loop 11 6.9 0.001 0.002 46.756 46.759 ot_mini 117 10.6 0.001 0.001 42.492 42.788 prepare_preconditioner 11 7.9 0.000 0.000 42.656 42.672 make_preconditioner 11 8.9 0.000 0.001 42.656 42.672 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.051 41.132 multiply_cannon_multrec 30084 15.6 14.109 18.866 26.062 30.764 rebuild_ks_matrix 128 8.3 0.001 0.001 29.040 29.321 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.040 29.320 mp_waitall_1 153770 16.5 17.764 27.356 17.764 27.356 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.254 26.489 qs_ot_get_derivative 117 11.6 0.001 0.002 22.900 23.186 make_m2s 5014 13.6 0.096 0.100 20.442 21.399 make_images 5014 14.6 1.970 2.253 20.135 21.095 qs_ot_get_p 128 10.4 0.001 0.002 19.543 19.893 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.950 19.627 apply_single 128 13.6 0.001 0.001 18.949 19.627 ot_diis_step 117 11.6 0.017 0.018 19.463 19.465 cp_fm_upper_to_full 105 14.7 11.214 16.595 11.214 16.595 cp_fm_cholesky_invert 11 10.9 16.121 16.130 16.121 16.130 multiply_cannon_metrocomm3 30084 15.6 0.046 0.048 6.385 15.486 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.378 15.431 sum_up_and_integrate 128 10.3 0.140 0.152 14.071 14.104 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.097 14.098 integrate_v_rspace 128 11.3 0.003 0.004 13.930 13.969 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.121 13.171 calculate_rho_elec 128 8.7 0.175 0.190 13.121 13.171 dbcsr_complete_redistribute 395 12.7 1.510 1.626 9.207 13.016 multiply_cannon_sync_h2d 30084 15.6 11.743 12.988 11.743 12.988 make_images_data 5014 15.6 0.063 0.068 11.063 12.664 dbcsr_mm_accdrv_process 62264 16.2 7.549 8.450 11.530 11.999 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.848 11.654 hybrid_alltoall_any 5200 16.5 0.524 2.214 9.776 11.361 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.055 11.057 cp_fm_redistribute_end 83 14.4 1.882 10.994 1.895 10.999 cp_fm_diag_elpa_base 83 14.4 8.485 10.404 9.073 10.897 init_scf_run 11 5.9 0.000 0.001 10.665 10.666 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.664 10.665 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.582 10.309 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.531 9.744 mp_alltoall_i22 716 14.1 5.708 9.652 5.708 9.652 pw_transfer 1547 11.6 0.084 0.097 7.902 7.992 grid_integrate_task_list 128 12.3 7.503 7.904 7.503 7.904 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.678 7.772 cp_fm_cholesky_decompose 22 10.9 7.635 7.738 7.635 7.738 wfi_extrapolate 11 7.9 0.001 0.001 7.613 7.613 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.762 7.086 fft_wrap_pw1pw2_140 523 13.2 0.480 0.490 6.807 6.906 mp_irecv_dv 76098 16.2 2.617 6.817 2.617 6.817 density_rs2pw 128 9.7 0.006 0.006 6.208 6.614 calculate_dm_sparse 128 9.5 0.001 0.001 6.188 6.278 fft3d_ps 1291 14.7 2.795 2.901 5.912 6.009 mp_alltoall_d11v 2415 14.1 5.056 5.754 5.056 5.754 grid_collocate_task_list 128 9.7 5.207 5.635 5.207 5.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.422 5.446 rs_pw_transfer 1046 11.9 0.013 0.014 4.494 4.905 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.441 4.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.340 4.464 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.415 4.418 potential_pw2rs 128 12.3 0.023 0.023 4.243 4.261 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.165 4.247 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.862000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.454545, yerr=17.737293 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.511293E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57686. MP_Allreduce 11094 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 171.806 171.807 qs_mol_dyn_low 1 2.0 0.003 0.010 171.416 171.427 qs_forces 11 3.9 0.004 0.005 171.317 171.326 qs_energies 11 4.9 0.001 0.002 164.092 164.099 scf_env_do_scf 11 5.9 0.001 0.001 147.116 147.128 velocity_verlet 10 3.0 0.001 0.001 113.316 113.319 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.078 111.080 dbcsr_multiply_generic 2529 12.6 0.183 0.189 75.114 75.584 qs_scf_new_mos 118 7.6 0.001 0.001 75.173 75.262 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.172 75.262 ot_scf_mini 118 9.6 0.004 0.004 70.669 70.735 multiply_cannon 2529 13.6 0.591 0.629 55.208 59.299 multiply_cannon_loop 2529 14.6 0.455 0.464 50.213 51.106 ot_mini 118 10.6 0.001 0.001 40.222 40.282 init_scf_loop 11 6.9 0.000 0.000 35.880 35.881 mp_waitall_1 130746 16.6 26.195 33.265 26.195 33.265 prepare_preconditioner 11 7.9 0.000 0.000 32.099 32.123 make_preconditioner 11 8.9 0.000 0.000 32.099 32.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.901 30.219 rebuild_ks_matrix 129 8.3 0.001 0.001 28.733 28.810 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.733 28.809 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.306 26.370 multiply_cannon_multrec 10116 15.6 10.546 14.381 17.969 21.274 ot_diis_step 118 11.6 0.020 0.021 20.263 20.264 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.727 20.032 apply_single 129 13.6 0.001 0.001 19.727 20.032 qs_ot_get_derivative 118 11.6 0.002 0.002 19.889 19.951 multiply_cannon_metrocomm3 10116 15.6 0.022 0.023 12.281 19.209 make_m2s 5058 13.6 0.068 0.072 16.234 18.561 cp_fm_cholesky_invert 11 10.9 18.417 18.423 18.417 18.423 make_images 5058 14.6 2.339 2.888 15.926 18.244 qs_ot_get_p 129 10.4 0.001 0.001 17.387 17.467 sum_up_and_integrate 129 10.3 0.181 0.191 14.185 14.234 integrate_v_rspace 129 11.3 0.004 0.004 14.004 14.062 qs_ot_p2m_diag 84 11.4 0.502 0.507 13.691 13.708 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.176 13.214 calculate_rho_elec 129 8.7 0.260 0.271 13.175 13.213 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.481 12.482 make_images_data 5058 15.6 0.053 0.061 9.825 12.434 multiply_cannon_sync_h2d 10116 15.6 11.762 12.271 11.762 12.271 hybrid_alltoall_any 5245 16.5 0.786 3.564 9.621 12.238 init_scf_run 11 5.9 0.000 0.001 10.131 10.131 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.131 10.131 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.367 9.369 cp_fm_diag_elpa_base 84 14.4 9.119 9.199 9.364 9.364 grid_integrate_task_list 129 12.3 7.820 8.183 7.820 8.183 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.137 8.181 cp_fm_cholesky_decompose 22 10.9 7.929 8.026 7.929 8.026 pw_transfer 1559 11.6 0.084 0.091 7.805 7.821 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.553 7.719 dbcsr_mm_accdrv_process 20934 16.1 2.582 3.316 7.049 7.702 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.583 7.606 mp_allgather_i34 2529 14.6 3.199 7.420 3.199 7.420 wfi_extrapolate 11 7.9 0.001 0.001 7.185 7.185 fft_wrap_pw1pw2_140 527 13.2 0.507 0.528 6.632 6.663 calculate_dm_sparse 129 9.5 0.001 0.001 6.174 6.275 density_rs2pw 129 9.7 0.005 0.006 5.928 6.226 grid_collocate_task_list 129 9.7 5.531 5.802 5.531 5.802 fft3d_ps 1301 14.7 2.762 2.836 5.727 5.768 dbcsr_complete_redistribute 397 12.7 2.124 2.187 5.274 5.635 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.242 5.242 mp_alltoall_d11v 2429 14.1 4.727 5.213 4.727 5.213 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.148 5.177 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.877 4.557 mp_irecv_dv 29102 15.9 1.838 4.484 1.838 4.484 rs_pw_transfer 1054 12.0 0.013 0.013 4.125 4.427 potential_pw2rs 129 12.3 0.027 0.028 4.161 4.177 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.089 4.124 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.846 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.495 3.804 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.746 3.786 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.507 3.568 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.449 3.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.807000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1406.181818, yerr=49.133988 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.141505E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.062 288.726 288.727 qs_mol_dyn_low 1 2.0 0.003 0.004 288.094 288.106 qs_forces 11 3.9 0.003 0.003 288.005 288.007 qs_energies 11 4.9 0.002 0.002 279.430 279.447 scf_env_do_scf 11 5.9 0.001 0.001 257.773 257.785 velocity_verlet 10 3.0 0.001 0.001 209.600 209.607 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.446 133.449 init_scf_loop 11 6.9 0.000 0.000 124.069 124.074 prepare_preconditioner 11 7.9 0.000 0.000 119.391 119.424 make_preconditioner 11 8.9 0.000 0.000 119.391 119.424 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.588 116.559 qs_scf_new_mos 118 7.6 0.001 0.001 91.261 91.393 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.260 91.392 ot_scf_mini 118 9.6 0.004 0.004 86.430 86.469 dbcsr_multiply_generic 2529 12.6 0.216 0.223 83.947 84.566 cp_fm_upper_to_full 106 14.8 52.897 75.675 52.897 75.675 multiply_cannon 2529 13.6 0.699 0.728 59.597 60.093 multiply_cannon_loop 2529 14.6 0.481 0.491 55.889 57.375 ot_mini 118 10.6 0.001 0.001 45.117 45.159 dbcsr_complete_redistribute 397 12.7 3.994 4.029 29.518 42.332 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.069 38.811 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.758 36.346 mp_alltoall_i22 720 14.1 21.575 34.514 21.575 34.514 rebuild_ks_matrix 129 8.3 0.001 0.001 33.599 33.653 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.599 33.652 cp_fm_cholesky_invert 11 10.9 33.537 33.542 33.537 33.542 mp_waitall_1 106386 16.7 28.648 32.778 28.648 32.778 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.415 31.476 qs_ot_get_p 129 10.4 0.001 0.001 25.671 25.699 qs_ot_get_derivative 118 11.6 0.002 0.002 24.609 24.650 qs_ot_p2m_diag 84 11.4 0.890 0.896 21.670 21.698 make_m2s 5058 13.6 0.076 0.078 19.752 20.805 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 19.261 20.733 ot_diis_step 118 11.6 0.022 0.023 20.480 20.481 make_images 5058 14.6 3.799 3.872 19.271 20.326 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.920 19.921 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.666 19.917 apply_single 129 13.6 0.001 0.001 19.665 19.917 multiply_cannon_multrec 10116 15.6 10.644 12.407 18.210 18.467 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.723 16.724 cp_fm_diag_elpa_base 84 14.4 12.358 13.961 16.720 16.720 sum_up_and_integrate 129 10.3 0.324 0.326 15.736 15.833 multiply_cannon_sync_h2d 10116 15.6 15.768 15.779 15.768 15.779 integrate_v_rspace 129 11.3 0.004 0.004 15.412 15.510 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.997 15.015 calculate_rho_elec 129 8.7 0.487 0.487 14.997 15.014 make_images_data 5058 15.6 0.061 0.067 10.474 12.347 hybrid_alltoall_any 5245 16.5 1.310 3.066 10.537 12.251 init_scf_run 11 5.9 0.000 0.001 11.764 11.765 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.764 11.765 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.534 9.574 dbcsr_mm_accdrv_process 20934 16.1 3.946 5.851 7.328 9.368 cp_fm_cholesky_decompose 22 10.9 8.885 8.904 8.885 8.904 grid_integrate_task_list 129 12.3 8.621 8.790 8.621 8.790 wfi_extrapolate 11 7.9 0.001 0.001 8.706 8.706 pw_transfer 1559 11.6 0.091 0.091 8.297 8.306 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.061 8.071 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 7.899 7.900 fft_wrap_pw1pw2_140 527 13.2 0.544 0.546 7.157 7.169 mp_alltoall_d11v 2429 14.1 6.934 7.053 6.934 7.053 calculate_dm_sparse 129 9.5 0.001 0.001 6.668 6.789 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.380 6.473 grid_collocate_task_list 129 9.7 6.387 6.423 6.387 6.423 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.107 6.173 fft3d_ps 1301 14.7 2.779 2.789 6.089 6.102 density_rs2pw 129 9.7 0.005 0.005 5.993 6.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.727000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2770.636364, yerr=200.052390 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 1.261535E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.135 85.494 85.496 qs_energies 1 2.0 0.000 0.000 84.776 84.811 ls_scf 1 3.0 0.000 0.001 83.793 83.827 dbcsr_multiply_generic 111 6.7 0.015 0.020 72.720 72.890 multiply_cannon 111 7.7 0.017 0.020 55.961 57.154 multiply_cannon_loop 111 8.7 0.206 0.225 52.491 53.678 ls_scf_main 1 4.0 0.000 0.000 52.329 52.329 density_matrix_trs4 2 5.0 0.002 0.003 46.868 46.945 ls_scf_init_scf 1 4.0 0.000 0.001 28.438 28.439 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.330 27.378 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.195 25.214 mp_waitall_1 11316 10.9 22.217 24.523 22.217 24.523 multiply_cannon_multrec 2664 9.7 8.173 8.988 15.563 17.382 multiply_cannon_sync_h2d 2664 9.7 13.818 15.513 13.818 15.513 make_m2s 222 7.7 0.009 0.013 13.098 13.589 make_images 222 8.7 0.099 0.112 13.076 13.565 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.448 12.247 make_images_data 222 9.7 0.004 0.005 7.652 8.187 dbcsr_mm_accdrv_process 4760 10.4 0.514 0.637 7.007 7.967 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.441 7.877 hybrid_alltoall_any 227 10.6 0.215 1.847 6.582 7.722 dbcsr_mm_accdrv_process_sort 4760 11.4 6.292 7.203 6.292 7.203 calculate_norms 4752 9.8 5.540 6.211 5.540 6.211 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.174 mp_sum_l 807 5.4 3.129 4.418 3.129 4.418 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.047 3.876 mp_irecv_dv 6231 10.9 2.030 3.864 2.030 3.864 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.329 3.414 arnoldi_extremal 4 6.8 0.000 0.000 3.351 3.373 arnoldi_normal_ev 4 7.8 0.001 0.003 3.351 3.373 make_images_sizes 222 9.7 0.000 0.000 0.750 3.351 mp_alltoall_i44 222 10.7 0.749 3.351 0.749 3.351 build_subspace 16 8.4 0.009 0.012 3.222 3.227 ls_scf_post 1 4.0 0.001 0.021 3.026 3.061 ls_scf_store_result 1 5.0 0.000 0.000 2.844 2.883 dbcsr_special_finalize 555 9.7 0.005 0.006 2.342 2.842 dbcsr_merge_single_wm 555 10.7 0.472 0.591 2.334 2.834 make_images_pack 222 9.7 2.207 2.624 2.209 2.626 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.362 2.606 dbcsr_sort_data 658 11.4 2.120 2.531 2.120 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.056 2.451 2.058 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.236 2.365 buffer_matrices_ensure_size 222 8.7 1.783 2.084 1.783 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.781 1.782 rebuild_ks_matrix 3 7.3 0.000 0.000 1.771 1.773 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.025 1.771 1.773 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.496000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 2.118779E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.094 93.618 93.627 qs_energies 1 2.0 0.000 0.000 92.839 92.846 ls_scf 1 3.0 0.000 0.001 91.533 91.539 dbcsr_multiply_generic 111 6.7 0.015 0.015 77.207 77.517 multiply_cannon 111 7.7 0.028 0.042 53.994 57.893 ls_scf_main 1 4.0 0.000 0.000 56.498 56.510 multiply_cannon_loop 111 8.7 0.115 0.123 50.698 54.298 density_matrix_trs4 2 5.0 0.002 0.003 50.696 50.871 ls_scf_init_scf 1 4.0 0.000 0.001 31.391 31.393 mp_waitall_1 9246 10.9 22.998 30.745 22.998 30.745 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.186 30.270 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.765 27.777 multiply_cannon_multrec 1332 9.7 13.014 16.179 22.291 26.537 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.595 21.434 make_m2s 222 7.7 0.006 0.007 16.308 16.939 make_images 222 8.7 1.583 1.973 16.278 16.911 make_images_data 222 9.7 0.004 0.004 9.722 10.718 dbcsr_mm_accdrv_process 4041 10.4 0.278 0.451 8.882 10.412 hybrid_alltoall_any 227 10.6 0.517 2.457 9.166 10.181 dbcsr_mm_accdrv_process_sort 4041 11.4 8.476 9.967 8.476 9.967 mp_sum_l 807 5.4 5.652 8.574 5.652 8.574 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.225 7.555 mp_irecv_dv 3311 11.0 3.205 7.500 3.205 7.500 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.386 6.919 calculate_norms 2376 9.8 5.977 6.768 5.977 6.768 multiply_cannon_sync_h2d 1332 9.7 4.676 5.962 4.676 5.962 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.175 5.381 arnoldi_extremal 4 6.8 0.000 0.000 4.825 4.853 arnoldi_normal_ev 4 7.8 0.001 0.005 4.825 4.853 build_subspace 16 8.4 0.014 0.021 4.561 4.563 ls_scf_post 1 4.0 0.001 0.020 3.644 3.651 ls_scf_store_result 1 5.0 0.000 0.000 3.328 3.469 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.244 3.446 dbcsr_matrix_vector_mult_local 304 10.0 2.782 3.276 2.784 3.277 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.301 2.728 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.584 2.662 mp_allgather_i34 111 8.7 1.005 2.638 1.005 2.638 make_images_pack 222 9.7 2.025 2.426 2.028 2.428 dbcsr_sort_data 436 11.2 1.824 2.059 1.824 2.059 make_images_sizes 222 9.7 0.000 0.000 0.625 2.042 mp_alltoall_i44 222 10.7 0.625 2.041 0.625 2.041 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.627000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1719.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 2.717598E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.050 95.271 95.271 qs_energies 1 2.0 0.000 0.000 94.568 94.571 ls_scf 1 3.0 0.000 0.001 93.123 93.126 dbcsr_multiply_generic 111 6.7 0.016 0.022 77.880 78.196 ls_scf_main 1 4.0 0.000 0.000 58.320 58.325 multiply_cannon 111 7.7 0.037 0.066 53.396 57.505 multiply_cannon_loop 111 8.7 0.099 0.105 49.809 54.154 density_matrix_trs4 2 5.0 0.002 0.003 52.353 52.553 mp_waitall_1 7374 11.0 24.583 35.161 24.583 35.161 ls_scf_init_scf 1 4.0 0.000 0.002 31.248 31.250 ls_scf_init_matrix_S 1 5.0 0.000 0.002 30.100 30.186 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.623 27.640 multiply_cannon_multrec 888 9.7 12.533 15.257 21.139 24.461 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.322 24.228 make_m2s 222 7.7 0.006 0.007 17.423 18.733 make_images 222 8.7 1.976 2.286 17.384 18.692 hybrid_alltoall_any 227 10.6 0.622 2.857 9.604 11.075 make_images_data 222 9.7 0.004 0.004 9.964 11.040 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.437 8.130 9.372 mp_sum_l 807 5.4 5.528 9.305 5.528 9.305 dbcsr_mm_accdrv_process_sort 3754 11.4 7.755 8.935 7.755 8.935 multiply_cannon_sync_h2d 888 9.7 6.062 7.568 6.062 7.568 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.281 7.392 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.921 6.994 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.478 6.829 mp_irecv_dv 2335 11.1 2.465 6.789 2.465 6.789 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.966 5.180 arnoldi_extremal 4 6.8 0.000 0.000 5.100 5.119 arnoldi_normal_ev 4 7.8 0.001 0.007 5.100 5.119 build_subspace 16 8.4 0.014 0.020 4.776 4.782 calculate_norms 1584 9.8 4.249 4.556 4.249 4.556 mp_allgather_i34 111 8.7 1.423 3.930 1.423 3.930 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.467 3.787 dbcsr_matrix_vector_mult_local 304 10.0 3.063 3.631 3.065 3.633 ls_scf_post 1 4.0 0.000 0.001 3.555 3.559 ls_scf_store_result 1 5.0 0.000 0.000 3.294 3.381 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.786 2.879 make_images_sizes 222 9.7 0.000 0.000 1.163 2.496 mp_alltoall_i44 222 10.7 1.162 2.496 1.162 2.496 make_images_pack 222 9.7 1.812 2.138 1.815 2.141 dbcsr_sort_data 325 11.1 1.841 2.130 1.841 2.130 dbcsr_data_release 9322 10.9 1.315 1.951 1.315 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.271000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 3.320316E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.063 96.401 96.402 qs_energies 1 2.0 0.000 0.000 95.599 95.606 ls_scf 1 3.0 0.000 0.001 93.951 93.956 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.870 78.143 ls_scf_main 1 4.0 0.000 0.000 58.328 58.329 multiply_cannon 111 7.7 0.046 0.099 51.227 55.359 density_matrix_trs4 2 5.0 0.002 0.003 52.321 52.449 multiply_cannon_loop 111 8.7 0.115 0.128 46.305 48.982 ls_scf_init_scf 1 4.0 0.000 0.002 32.405 32.407 ls_scf_init_matrix_S 1 5.0 0.000 0.003 31.267 31.321 mp_waitall_1 6438 11.0 22.348 28.992 22.348 28.992 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.796 28.808 multiply_cannon_multrec 1332 9.7 14.283 17.308 22.205 25.383 make_m2s 222 7.7 0.007 0.007 21.164 22.477 make_images 222 8.7 3.138 3.618 21.113 22.429 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.837 15.831 make_images_data 222 9.7 0.004 0.004 11.798 13.384 hybrid_alltoall_any 227 10.6 0.797 3.831 11.221 12.865 dbcsr_mm_accdrv_process 3641 10.4 0.203 0.401 7.560 9.084 dbcsr_mm_accdrv_process_sort 3641 11.4 7.187 8.675 7.187 8.675 mp_sum_l 807 5.4 3.902 7.539 3.902 7.539 multiply_cannon_sync_h2d 1332 9.7 5.527 6.096 5.527 6.096 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.080 5.922 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.944 5.890 mp_irecv_dv 3229 10.9 2.058 5.847 2.058 5.847 arnoldi_extremal 4 6.8 0.000 0.000 5.283 5.295 arnoldi_normal_ev 4 7.8 0.001 0.005 5.283 5.295 build_subspace 16 8.4 0.014 0.021 4.931 4.936 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.542 4.849 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.527 4.730 calculate_norms 2376 9.8 4.176 4.565 4.176 4.565 mp_allgather_i34 111 8.7 2.059 4.123 2.059 4.123 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.603 3.913 dbcsr_matrix_vector_mult_local 304 10.0 3.226 3.734 3.228 3.736 dbcsr_sort_data 658 11.4 3.113 3.431 3.113 3.431 ls_scf_post 1 4.0 0.000 0.001 3.218 3.228 dbcsr_special_finalize 555 9.7 0.006 0.007 2.856 3.186 dbcsr_merge_single_wm 555 10.7 0.537 0.644 2.848 3.178 ls_scf_store_result 1 5.0 0.000 0.000 2.968 3.023 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.880 2.920 dbcsr_data_release 10477 10.7 1.592 2.561 1.592 2.561 dbcsr_finalize 304 7.8 0.049 0.061 1.806 2.051 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.402000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2672.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 4.623905E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.060 99.656 99.656 qs_energies 1 2.0 0.000 0.000 98.870 98.876 ls_scf 1 3.0 0.000 0.000 96.967 96.971 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.619 78.864 ls_scf_main 1 4.0 0.000 0.000 62.557 62.557 multiply_cannon 111 7.7 0.075 0.192 55.722 60.541 density_matrix_trs4 2 5.0 0.002 0.003 55.574 55.657 multiply_cannon_loop 111 8.7 0.069 0.077 51.289 53.752 mp_waitall_1 5481 11.0 27.190 31.947 27.190 31.947 ls_scf_init_scf 1 4.0 0.000 0.000 30.738 30.743 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.562 29.603 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.384 27.396 multiply_cannon_multrec 444 9.7 14.072 16.470 21.157 23.882 make_m2s 222 7.7 0.004 0.005 17.925 20.348 make_images 222 8.7 3.720 4.428 17.864 20.288 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.690 16.448 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.044 15.368 make_images_data 222 9.7 0.003 0.004 10.252 12.523 hybrid_alltoall_any 227 10.6 0.789 3.785 10.004 12.364 multiply_cannon_sync_h2d 444 9.7 6.567 8.010 6.567 8.010 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.340 6.797 7.952 dbcsr_mm_accdrv_process_sort 3003 11.4 6.482 7.602 6.482 7.602 mp_allgather_i34 111 8.7 2.677 7.016 2.677 7.016 arnoldi_extremal 4 6.8 0.000 0.000 5.914 5.928 arnoldi_normal_ev 4 7.8 0.002 0.005 5.914 5.928 build_subspace 16 8.4 0.015 0.020 5.508 5.518 mp_sum_l 807 5.4 3.008 4.810 3.008 4.810 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.611 4.756 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.229 4.439 dbcsr_matrix_vector_mult_local 304 10.0 3.746 4.216 3.748 4.218 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.640 3.980 mp_irecv_dv 1241 11.2 1.622 3.954 1.622 3.954 calculate_norms 792 9.8 3.554 3.745 3.554 3.745 ls_scf_post 1 4.0 0.000 0.000 3.672 3.677 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.107 3.614 make_images_sizes 222 9.7 0.000 0.000 1.032 3.599 mp_alltoall_i44 222 10.7 1.032 3.598 1.032 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.443 3.493 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.375 3.484 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.270 dbcsr_merge_all 275 8.9 0.474 0.536 2.048 2.103 dbcsr_data_release 10123 10.8 1.334 2.014 1.334 2.014 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.656000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3610.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5c6c216e857c9fb3b35611c0159716e1fe51e750_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 8.770986E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.086 0.105 108.513 108.514 qs_energies 1 2.0 0.000 0.000 107.049 107.064 ls_scf 1 3.0 0.000 0.000 104.158 104.173 dbcsr_multiply_generic 111 6.7 0.023 0.028 78.059 78.215 ls_scf_main 1 4.0 0.000 0.000 66.410 66.411 density_matrix_trs4 2 5.0 0.002 0.003 57.586 57.648 multiply_cannon 111 7.7 0.158 0.287 50.087 51.874 multiply_cannon_loop 111 8.7 0.068 0.071 46.576 47.938 ls_scf_init_scf 1 4.0 0.000 0.000 34.092 34.093 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.759 32.776 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.927 29.937 mp_waitall_1 4569 11.1 22.524 26.814 22.524 26.814 make_m2s 222 7.7 0.005 0.005 24.062 25.097 make_images 222 8.7 4.588 4.992 23.956 24.988 multiply_cannon_multrec 444 9.7 17.838 18.563 22.477 23.227 make_images_data 222 9.7 0.003 0.003 13.283 15.775 hybrid_alltoall_any 227 10.6 1.661 3.640 13.048 15.674 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.594 11.290 multiply_cannon_sync_h2d 444 9.7 8.849 8.887 8.849 8.887 arnoldi_extremal 4 6.8 0.000 0.000 7.647 7.657 arnoldi_normal_ev 4 7.8 0.003 0.009 7.647 7.657 build_subspace 16 8.4 0.026 0.036 7.086 7.097 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.637 5.782 dbcsr_matrix_vector_mult_local 304 10.0 5.164 5.490 5.167 5.492 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.014 5.273 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.797 4.886 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.315 4.474 4.601 dbcsr_mm_accdrv_process_sort 1814 11.4 4.174 4.310 4.174 4.310 ls_scf_post 1 4.0 0.000 0.000 3.656 3.671 mp_allgather_i34 111 8.7 1.160 3.656 1.160 3.656 make_images_sizes 222 9.7 0.000 0.000 1.453 3.499 mp_alltoall_i44 222 10.7 1.453 3.498 1.453 3.498 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.419 calculate_norms 792 9.8 3.239 3.270 3.239 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.081 3.129 dbcsr_merge_all 275 8.9 0.890 0.914 2.865 2.909 dbcsr_complete_redistribute 5 7.6 1.439 1.472 2.773 2.903 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.860 2.860 dbcsr_data_release 12724 10.6 2.314 2.816 2.314 2.816 mp_sum_l 807 5.4 1.592 2.569 1.592 2.569 matrix_ls_to_qs 2 6.0 0.000 0.000 2.418 2.559 dbcsr_sort_data 325 11.1 2.441 2.509 2.441 2.509 dbcsr_new_transposed 4 7.5 0.245 0.255 2.296 2.309 dbcsr_frobenius_norm 74 6.6 2.058 2.137 2.210 2.257 dbcsr_add_d 103 6.2 0.000 0.000 2.153 2.222 dbcsr_add_anytype 103 7.2 0.860 0.892 2.153 2.222 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.194 2.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.514000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6910.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5c6c216e857c9fb3b35611c0159716e1fe51e750 Summary: empty Status: OK