=== 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: 9368eddea8879e44b245fe945652cfe39f9313dd ################# 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 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, 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 (15.09.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/01 job id: 41701524 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/02 job id: 41701526 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/03 job id: 41701529 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/04 job id: 41701533 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/05 job id: 41701535 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/06 job id: 41701536 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/07 job id: 41701537 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/08 job id: 41701538 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/09 job id: 41701539 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/10 job id: 41701540 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/11 job id: 41701542 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/12 job id: 41701543 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/13 job id: 41701544 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/14 job id: 41701545 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/15 job id: 41701546 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/16 job id: 41701547 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/17 job id: 41701548 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/18 job id: 41701549 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/19 job id: 41701550 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/20 job id: 41701551 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/21 job id: 41701552 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/22 job id: 41701553 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/23 job id: 41701554 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/24 job id: 41701555 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/25 job id: 41701556 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/26 job id: 41701557 --- 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/9368eddea8879e44b245fe945652cfe39f9313dd_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 0.038 136.664 136.665 farming_run 1 2.0 136.047 136.049 136.629 136.632 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455616E+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 0.014 0.031 117.315 117.315 qs_energies 1 2.0 0.000 0.000 117.108 117.111 mp2_main 1 3.0 0.000 0.000 115.158 115.161 mp2_gpw_main 1 4.0 0.021 0.035 114.285 114.288 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 95.462 95.652 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 56.399 56.589 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 42.710 47.846 get_2c_integrals 1 6.0 0.000 0.000 38.258 38.891 integrate_v_rspace 273 8.0 0.433 0.445 25.294 30.212 pw_transfer 6555 10.6 0.372 0.390 27.677 28.309 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.303 26.846 grid_integrate_task_list 273 9.0 21.061 26.508 21.061 26.508 fft_wrap_pw1pw2_100 2178 12.4 1.209 1.432 23.864 24.381 compute_2c_integrals 1 7.0 0.002 0.002 20.591 20.591 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 19.739 20.294 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.444 19.736 20.293 rpa_ri_compute_en 1 5.0 0.000 0.000 18.714 18.940 cp_fm_cholesky_decompose 12 8.2 17.673 18.291 17.673 18.291 cholesky_decomp 1 7.0 0.000 0.000 16.506 17.158 fft3d_s 5443 13.4 16.374 16.891 16.395 16.914 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.588 10.855 15.588 calculate_wavefunction 272 8.0 5.418 5.570 12.628 13.271 rpa_num_int 1 6.0 0.000 0.000 10.824 10.825 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.551 10.574 calc_potential_gpw 544 9.5 0.005 0.006 9.700 10.085 calc_mat_Q 8 8.0 0.000 0.000 9.337 9.425 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.991 9.314 contract_S_to_Q 8 9.0 0.000 0.000 8.759 8.842 potential_pw2rs 545 10.0 0.106 0.109 7.774 8.485 parallel_gemm_fm 14 9.1 0.000 0.000 8.334 8.433 parallel_gemm_fm_cosma 14 10.1 8.334 8.433 8.334 8.433 collocate_single_gaussian 272 10.0 0.039 0.042 7.548 7.876 create_integ_mat 1 6.0 0.022 0.028 7.605 7.605 array2fm 1 7.0 0.000 0.000 6.612 7.033 pw_scatter_s 2720 13.7 4.417 4.539 4.417 4.539 pw_gather_s 2722 13.2 3.935 4.375 3.935 4.375 pw_poisson_solve 545 10.5 2.481 2.534 3.563 3.766 array2fm_buffer_send 1 8.0 2.888 3.149 2.888 3.149 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.287791, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 0.028 0.038 399.839 399.841 farming_run 1 2.0 398.731 398.755 399.801 399.804 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222169E+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 10203830. 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 0.013 0.048 212.314 212.315 qs_energies 1 2.0 0.000 0.000 212.040 212.065 scf_env_do_scf 1 3.0 0.000 0.000 106.676 106.677 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.800 105.807 rebuild_ks_matrix 4 6.0 0.000 0.000 105.798 105.806 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 105.798 105.806 hfx_ks_matrix 4 8.0 0.001 0.001 105.418 105.422 integrate_four_center 4 9.0 0.143 0.458 105.417 105.421 mp2_main 1 3.0 0.002 0.019 105.068 105.093 mp2_gpw_main 1 4.0 0.051 0.249 104.227 104.252 integrate_four_center_main 4 10.0 0.086 0.450 96.780 99.522 integrate_four_center_bin 264 11.0 96.693 99.509 96.693 99.509 init_scf_loop 1 4.0 0.000 0.000 92.468 92.469 mp2_ri_gpw_compute_in 1 5.0 0.064 0.070 77.062 78.298 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 55.027 56.252 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.165 42.633 47.641 integrate_v_rspace 95 8.0 0.396 0.564 28.568 33.446 pw_transfer 2240 10.6 0.146 0.163 30.153 30.524 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 29.138 29.487 ao_to_mo_and_store_B_mult_1 91 7.0 10.700 29.427 10.700 29.427 grid_integrate_task_list 95 9.0 23.848 28.922 23.848 28.922 mp2_ri_gpw_compute_en 1 5.0 0.069 0.170 26.892 28.692 fft_wrap_pw1pw2_100 730 12.4 1.272 1.424 26.877 27.294 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.890 24.987 24.997 get_2c_integrals 1 6.0 0.036 0.295 20.998 21.980 compute_2c_integrals 1 7.0 0.005 0.027 19.791 19.805 compute_2c_integrals_loop_lm 1 8.0 0.002 0.011 19.315 19.650 mp2_eri_2c_integrate_gpw 1 9.0 1.749 1.830 19.313 19.648 fft3d_s 1823 13.4 18.539 18.903 18.552 18.916 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.206 14.206 calculate_wavefunction 91 8.0 2.023 2.054 9.803 10.148 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.584 8.816 9.454 potential_pw2rs 186 10.0 0.033 0.035 8.725 9.333 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.616 9.100 local_gemm 172 8.0 8.260 8.875 8.260 8.875 calc_potential_gpw 182 9.5 0.002 0.002 8.589 8.837 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.527 7.944 8.453 collocate_single_gaussian 91 10.0 0.031 0.318 7.949 8.246 mp_sync 38 10.4 4.256 6.720 4.256 6.720 mp_sendrecv_dm3 2068 8.0 5.963 6.467 5.963 6.467 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.440 6.345 6.440 pw_gather_s 912 13.2 5.014 5.480 5.014 5.480 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=104.220273, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1509.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 450.211840E+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 1767401. 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 0.011 0.029 52.569 52.570 qs_mol_dyn_low 1 2.0 0.003 0.003 52.353 52.361 qs_forces 11 3.9 0.002 0.002 52.290 52.291 qs_energies 11 4.9 0.001 0.001 50.829 50.844 scf_env_do_scf 11 5.9 0.000 0.001 43.794 43.794 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.650 41.651 dbcsr_multiply_generic 2286 12.5 0.092 0.095 34.255 34.686 qs_scf_new_mos 108 7.5 0.000 0.001 31.432 31.705 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.432 31.705 ot_scf_mini 108 9.5 0.002 0.002 29.768 29.946 multiply_cannon 2286 13.5 0.185 0.194 27.092 28.778 multiply_cannon_loop 2286 14.5 1.520 1.595 26.375 28.049 velocity_verlet 10 3.0 0.001 0.001 25.353 25.354 ot_mini 108 10.5 0.001 0.001 19.084 19.321 qs_ot_get_derivative 108 11.5 0.001 0.001 16.128 16.292 mp_waitall_1 267858 16.1 8.677 15.315 8.677 15.315 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 6.085 13.547 multiply_cannon_multrec 54864 15.5 4.214 6.574 8.480 12.224 rebuild_ks_matrix 119 8.3 0.000 0.000 8.026 8.149 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.026 8.148 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.091 7.205 multiply_cannon_sync_h2d 54864 15.5 5.896 6.791 5.896 6.791 qs_ot_get_p 119 10.4 0.001 0.001 6.170 6.478 mp_sum_l 7207 12.9 5.055 6.466 5.055 6.466 init_scf_run 11 5.9 0.000 0.001 5.810 5.810 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.810 5.810 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.337 5.797 dbcsr_mm_accdrv_process 76910 16.1 1.179 1.844 4.189 5.792 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.264 5.368 sum_up_and_integrate 119 10.3 0.008 0.009 4.619 4.628 integrate_v_rspace 119 11.3 0.002 0.002 4.611 4.621 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.177 4.345 calculate_rho_elec 119 8.7 0.011 0.016 4.177 4.344 jit_kernel_multiply 13 15.8 2.947 3.891 2.947 3.891 calculate_first_density_matrix 1 7.0 0.000 0.001 3.648 3.651 calculate_dm_sparse 119 9.5 0.000 0.000 3.473 3.631 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.404 3.446 rs_pw_transfer 974 11.9 0.012 0.013 3.051 3.174 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.909 3.108 apply_single 119 13.6 0.000 0.000 2.909 3.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.828 2.830 ot_diis_step 108 11.5 0.006 0.006 2.658 2.659 multiply_cannon_metrocomm1 54864 15.5 0.052 0.059 1.687 2.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.580 2.624 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.569 2.569 acc_transpose_blocks 54864 15.5 0.228 0.248 2.017 2.468 density_rs2pw 119 9.7 0.003 0.004 2.290 2.420 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.333 2.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.346 2.347 cp_fm_redistribute_end 50 14.0 2.129 2.322 2.134 2.324 cp_fm_diag_elpa_base 50 14.0 0.188 2.278 0.189 2.287 init_scf_loop 11 6.9 0.000 0.000 2.125 2.126 wfi_extrapolate 11 7.9 0.001 0.001 2.105 2.105 grid_integrate_task_list 119 12.3 2.010 2.097 2.010 2.097 potential_pw2rs 119 12.3 0.004 0.004 2.016 2.028 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.975 1.991 mp_sum_d 4129 12.0 1.302 1.914 1.302 1.914 make_m2s 4572 13.5 0.054 0.056 1.796 1.851 make_images 4572 14.5 0.133 0.139 1.714 1.767 pw_transfer 1439 11.6 0.052 0.056 1.677 1.753 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.603 1.678 mp_alltoall_d11v 2130 13.8 1.314 1.493 1.314 1.493 fft3d_ps 1201 14.6 0.357 0.460 1.378 1.450 grid_collocate_task_list 119 9.7 1.290 1.353 1.290 1.353 mp_waitany 12084 13.8 1.174 1.327 1.174 1.327 acc_transpose_blocks_kernels 54864 16.5 0.253 0.384 1.030 1.313 fft_wrap_pw1pw2_140 487 13.2 0.083 0.097 1.228 1.302 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.713 1.123 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.570000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 485.781504E+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 1722386. 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 0.013 0.031 39.765 39.765 qs_mol_dyn_low 1 2.0 0.003 0.003 39.585 39.593 qs_forces 11 3.9 0.002 0.006 39.521 39.522 qs_energies 11 4.9 0.001 0.002 37.824 37.828 scf_env_do_scf 11 5.9 0.000 0.001 32.026 32.027 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.465 29.466 dbcsr_multiply_generic 2286 12.5 0.100 0.104 22.518 22.910 qs_scf_new_mos 108 7.5 0.001 0.001 20.379 20.619 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.378 20.618 ot_scf_mini 108 9.5 0.002 0.003 19.470 19.640 multiply_cannon 2286 13.5 0.208 0.215 17.186 18.815 velocity_verlet 10 3.0 0.001 0.002 18.506 18.507 multiply_cannon_loop 2286 14.5 0.911 0.981 16.018 17.785 ot_mini 108 10.5 0.001 0.001 12.255 12.484 mp_waitall_1 217478 16.2 6.269 12.196 6.269 12.196 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.468 10.243 qs_ot_get_derivative 108 11.5 0.001 0.001 9.788 9.962 multiply_cannon_multrec 27432 15.5 1.974 4.495 6.561 9.946 rebuild_ks_matrix 119 8.3 0.000 0.000 7.214 7.361 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.213 7.361 dbcsr_mm_accdrv_process 47894 16.0 3.753 6.249 4.517 6.693 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.372 6.507 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.697 4.729 init_scf_run 11 5.9 0.000 0.001 4.559 4.559 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.559 4.559 qs_ot_get_p 119 10.4 0.001 0.001 4.286 4.522 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.234 4.375 apply_single 119 13.6 0.000 0.000 3.234 4.375 mp_sum_l 7207 12.9 2.296 4.272 2.296 4.272 sum_up_and_integrate 119 10.3 0.015 0.016 4.235 4.241 integrate_v_rspace 119 11.3 0.002 0.002 4.221 4.227 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.663 3.700 calculate_rho_elec 119 8.7 0.021 0.024 3.663 3.699 calculate_first_density_matrix 1 7.0 0.001 0.001 3.009 3.010 make_m2s 4572 13.5 0.052 0.053 2.588 2.836 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.791 2.812 multiply_cannon_sync_h2d 27432 15.5 2.155 2.785 2.155 2.785 rs_pw_transfer 974 11.9 0.010 0.011 2.679 2.776 make_images 4572 14.5 0.201 0.241 2.500 2.744 init_scf_loop 11 6.9 0.000 0.000 2.534 2.535 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.428 2.431 ot_diis_step 108 11.5 0.011 0.011 2.412 2.413 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.346 2.347 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.206 2.297 calculate_dm_sparse 119 9.5 0.000 0.000 2.213 2.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.205 2.247 density_rs2pw 119 9.7 0.004 0.004 2.026 2.125 jit_kernel_multiply 8 16.3 0.711 2.053 0.711 2.053 potential_pw2rs 119 12.3 0.006 0.006 2.001 2.007 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.993 1.993 cp_fm_redistribute_end 50 14.0 1.642 1.965 1.647 1.967 grid_integrate_task_list 119 12.3 1.823 1.937 1.823 1.937 cp_fm_diag_elpa_base 50 14.0 0.310 1.895 0.318 1.924 pw_transfer 1439 11.6 0.064 0.066 1.813 1.846 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.724 1.760 acc_transpose_blocks 27432 15.5 0.109 0.113 1.200 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.628 1.655 make_preconditioner 11 8.9 0.000 0.000 1.628 1.655 make_images_data 4572 15.5 0.045 0.051 1.187 1.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.528 1.588 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.034 1.528 wfi_extrapolate 11 7.9 0.001 0.001 1.498 1.498 fft3d_ps 1201 14.6 0.498 0.550 1.419 1.458 mp_alltoall_d11v 2130 13.8 1.236 1.458 1.236 1.458 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.423 1.432 mp_allgather_i34 2286 14.5 0.611 1.372 0.611 1.372 fft_wrap_pw1pw2_140 487 13.2 0.087 0.096 1.307 1.341 grid_collocate_task_list 119 9.7 1.240 1.306 1.240 1.306 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.205 1.255 acc_transpose_blocks_kernels 27432 16.5 0.181 0.268 0.668 1.108 mp_sum_d 4129 12.0 0.571 1.012 0.571 1.012 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.956 0.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.898 0.913 make_images_sizes 4572 15.5 0.005 0.005 0.586 0.843 jit_kernel_transpose 5 15.5 0.486 0.840 0.486 0.840 mp_alltoall_i44 4572 16.5 0.581 0.838 0.581 0.838 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.605 0.775 0.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.765000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.727273, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.965568E+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 0.018 0.040 33.727 33.728 qs_mol_dyn_low 1 2.0 0.003 0.006 33.440 33.449 qs_forces 11 3.9 0.002 0.007 33.373 33.374 qs_energies 11 4.9 0.002 0.008 31.799 31.803 scf_env_do_scf 11 5.9 0.000 0.001 25.743 25.743 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.051 23.051 dbcsr_multiply_generic 2286 12.5 0.093 0.095 17.746 17.862 velocity_verlet 10 3.0 0.001 0.002 15.387 15.389 multiply_cannon 2286 13.5 0.195 0.204 14.336 15.092 qs_scf_new_mos 108 7.5 0.001 0.001 14.960 14.980 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.960 14.980 multiply_cannon_loop 2286 14.5 0.645 0.679 13.525 14.287 ot_scf_mini 108 9.5 0.003 0.006 14.226 14.237 ot_mini 108 10.5 0.001 0.001 8.821 8.839 multiply_cannon_multrec 18288 15.5 1.937 2.925 7.716 8.017 qs_ot_get_derivative 108 11.5 0.001 0.001 7.298 7.310 dbcsr_mm_accdrv_process 38222 16.0 4.734 6.166 5.694 6.589 rebuild_ks_matrix 119 8.3 0.000 0.000 6.430 6.460 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.044 6.430 6.460 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.697 5.717 init_scf_run 11 5.9 0.000 0.001 4.833 4.833 scf_env_initial_rho_setup 11 6.9 0.001 0.004 4.832 4.833 sum_up_and_integrate 119 10.3 0.018 0.025 4.019 4.024 integrate_v_rspace 119 11.3 0.002 0.003 4.000 4.007 mp_waitall_1 169478 16.3 3.021 3.936 3.021 3.936 calculate_first_density_matrix 1 7.0 0.001 0.002 3.609 3.611 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.722 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.003 3.310 3.317 calculate_rho_elec 119 8.7 0.030 0.031 3.309 3.317 qs_ot_get_p 119 10.4 0.001 0.001 3.239 3.266 jit_kernel_multiply 10 16.0 0.908 2.841 0.908 2.841 init_scf_loop 11 6.9 0.001 0.007 2.672 2.672 rs_pw_transfer 974 11.9 0.009 0.010 2.459 2.585 calculate_dm_sparse 119 9.5 0.000 0.001 2.483 2.494 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.058 2.368 apply_single 119 13.6 0.000 0.000 2.058 2.368 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.322 2.323 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.410 2.245 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.175 2.183 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.111 2.119 make_m2s 4572 13.5 0.044 0.045 1.852 2.010 density_rs2pw 119 9.7 0.003 0.004 1.901 2.005 make_images 4572 14.5 0.191 0.202 1.766 1.923 grid_integrate_task_list 119 12.3 1.800 1.893 1.800 1.893 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.842 1.847 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.840 1.840 potential_pw2rs 119 12.3 0.007 0.008 1.815 1.824 prepare_preconditioner 11 7.9 0.000 0.000 1.816 1.818 make_preconditioner 11 8.9 0.000 0.001 1.816 1.818 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.667 1.754 pw_transfer 1439 11.6 0.063 0.066 1.735 1.747 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.645 1.658 acc_transpose_blocks 18288 15.5 0.076 0.078 1.588 1.616 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.605 1.606 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.598 1.604 cp_fm_redistribute_end 50 14.0 1.193 1.584 1.194 1.584 multiply_cannon_sync_h2d 18288 15.5 1.368 1.574 1.368 1.574 cp_fm_diag_elpa_base 50 14.0 0.373 1.513 0.388 1.553 mp_sum_l 7207 12.9 1.182 1.525 1.182 1.525 ot_diis_step 108 11.5 0.011 0.013 1.495 1.495 fft3d_ps 1201 14.6 0.509 0.531 1.331 1.344 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.293 1.304 grid_collocate_task_list 119 9.7 1.204 1.281 1.204 1.281 wfi_extrapolate 11 7.9 0.001 0.002 1.170 1.170 acc_transpose_blocks_kernels 18288 16.5 0.210 0.220 1.146 1.160 make_images_data 4572 15.5 0.044 0.049 0.839 1.008 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.986 0.988 jit_kernel_transpose 5 15.6 0.937 0.941 0.937 0.941 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.897 0.921 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.722 0.899 mp_alltoall_d11v 2130 13.8 0.705 0.830 0.705 0.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.814 cp_fm_cholesky_invert 11 10.9 0.769 0.773 0.769 0.773 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.738 mp_alltoall_z22v 1201 16.6 0.665 0.727 0.665 0.727 rs_pw_transfer_RS2PW_140 130 11.5 0.175 0.180 0.597 0.727 mp_waitany 9880 13.7 0.549 0.702 0.549 0.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.728000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=2.004128 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 555.323392E+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 0.038 0.056 36.852 36.853 qs_mol_dyn_low 1 2.0 0.003 0.003 36.596 36.604 qs_forces 11 3.9 0.001 0.002 36.534 36.535 qs_energies 11 4.9 0.002 0.006 34.796 34.802 scf_env_do_scf 11 5.9 0.001 0.001 28.668 28.669 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.995 24.995 dbcsr_multiply_generic 2286 12.5 0.099 0.102 19.308 19.414 velocity_verlet 10 3.0 0.002 0.002 18.215 18.217 qs_scf_new_mos 108 7.5 0.001 0.001 16.660 16.714 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.659 16.714 multiply_cannon 2286 13.5 0.227 0.251 15.473 16.319 ot_scf_mini 108 9.5 0.002 0.003 15.706 15.760 multiply_cannon_loop 2286 14.5 0.944 0.976 14.465 15.142 multiply_cannon_multrec 27432 15.5 2.330 2.989 9.338 9.839 ot_mini 108 10.5 0.001 0.001 9.562 9.625 dbcsr_mm_accdrv_process 47916 15.9 5.798 8.168 6.915 8.225 qs_ot_get_derivative 108 11.5 0.001 0.001 7.731 7.785 rebuild_ks_matrix 119 8.3 0.000 0.000 6.593 6.646 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.593 6.645 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.849 5.894 init_scf_run 11 5.9 0.000 0.001 4.637 4.638 scf_env_initial_rho_setup 11 6.9 0.010 0.017 4.637 4.638 sum_up_and_integrate 119 10.3 0.020 0.022 3.841 3.850 integrate_v_rspace 119 11.3 0.002 0.003 3.821 3.830 init_scf_loop 11 6.9 0.000 0.002 3.649 3.651 qs_ot_get_p 119 10.4 0.001 0.001 3.484 3.561 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.406 3.439 calculate_rho_elec 119 8.7 0.040 0.046 3.405 3.439 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.968 3.359 calculate_first_density_matrix 1 7.0 0.002 0.004 3.218 3.229 mp_waitall_1 145218 16.4 2.436 3.128 2.436 3.128 prepare_preconditioner 11 7.9 0.000 0.000 2.791 2.798 make_preconditioner 11 8.9 0.000 0.001 2.791 2.798 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.370 2.707 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.595 apply_single 119 13.6 0.000 0.000 2.051 2.595 make_m2s 4572 13.5 0.054 0.056 2.374 2.477 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.475 2.476 make_images 4572 14.5 0.289 0.354 2.266 2.368 calculate_dm_sparse 119 9.5 0.000 0.000 2.282 2.340 rs_pw_transfer 974 11.9 0.009 0.009 2.225 2.292 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.258 2.269 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.129 2.138 jit_kernel_multiply 10 16.0 1.056 2.072 1.056 2.072 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.987 2.017 density_rs2pw 119 9.7 0.003 0.004 1.898 1.955 grid_integrate_task_list 119 12.3 1.819 1.934 1.819 1.934 pw_transfer 1439 11.6 0.063 0.067 1.863 1.892 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.774 1.806 ot_diis_step 108 11.5 0.012 0.012 1.790 1.790 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.783 1.784 mp_sum_l 7207 12.9 1.042 1.710 1.042 1.710 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.915 1.673 potential_pw2rs 119 12.3 0.009 0.010 1.641 1.648 acc_transpose_blocks 27432 15.5 0.113 0.114 1.569 1.590 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.521 1.522 cp_fm_redistribute_end 50 14.0 1.001 1.492 1.003 1.492 cp_fm_diag_elpa_base 50 14.0 0.465 1.421 0.486 1.460 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.451 1.460 fft3d_ps 1201 14.6 0.535 0.590 1.431 1.458 fft_wrap_pw1pw2_140 487 13.2 0.108 0.132 1.423 1.458 wfi_extrapolate 11 7.9 0.001 0.001 1.358 1.358 grid_collocate_task_list 119 9.7 1.221 1.310 1.221 1.310 qs_energies_init_hamiltonians 11 5.9 0.002 0.009 1.203 1.204 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.144 1.159 dbcsr_complete_redistribute 329 12.2 0.121 0.173 0.879 1.155 cp_fm_upper_to_full 72 13.5 0.822 1.143 0.822 1.143 multiply_cannon_sync_h2d 27432 15.5 0.982 1.052 0.982 1.052 make_images_data 4572 15.5 0.044 0.048 0.930 1.036 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.795 0.973 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.942 0.957 mp_alltoall_d11v 2130 13.8 0.804 0.929 0.804 0.929 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.638 0.908 cp_fm_cholesky_invert 11 10.9 0.883 0.887 0.883 0.887 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.836 0.841 mp_alltoall_z22v 1201 16.6 0.778 0.823 0.778 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.853000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=3.638636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 605.650944E+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 0.016 0.030 27.584 27.585 qs_mol_dyn_low 1 2.0 0.003 0.003 27.410 27.417 qs_forces 11 3.9 0.002 0.002 27.350 27.351 qs_energies 11 4.9 0.001 0.001 25.658 25.661 scf_env_do_scf 11 5.9 0.000 0.001 20.921 20.921 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.390 18.391 velocity_verlet 10 3.0 0.002 0.002 14.157 14.160 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.165 12.238 qs_scf_new_mos 108 7.5 0.001 0.001 10.839 10.860 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.838 10.860 ot_scf_mini 108 9.5 0.002 0.002 10.165 10.187 multiply_cannon 2286 13.5 0.229 0.234 9.631 10.058 multiply_cannon_loop 2286 14.5 0.332 0.344 8.714 8.885 multiply_cannon_multrec 9144 15.5 1.579 1.816 5.774 6.052 rebuild_ks_matrix 119 8.3 0.000 0.000 5.816 5.834 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.815 5.834 ot_mini 108 10.5 0.001 0.001 5.619 5.644 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.187 5.203 qs_ot_get_derivative 108 11.5 0.001 0.001 4.317 4.338 dbcsr_mm_accdrv_process 12550 15.8 3.122 3.803 4.092 4.182 sum_up_and_integrate 119 10.3 0.022 0.023 3.543 3.548 integrate_v_rspace 119 11.3 0.003 0.003 3.521 3.525 init_scf_run 11 5.9 0.000 0.001 3.321 3.321 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.321 3.321 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.198 3.206 calculate_rho_elec 119 8.7 0.059 0.061 3.198 3.205 qs_ot_get_p 119 10.4 0.001 0.001 2.753 2.790 init_scf_loop 11 6.9 0.000 0.000 2.511 2.512 calculate_first_density_matrix 1 7.0 0.001 0.001 2.206 2.207 mp_waitall_1 121218 16.5 1.657 2.197 1.657 2.197 make_m2s 4572 13.5 0.035 0.035 1.763 1.911 grid_integrate_task_list 119 12.3 1.851 1.907 1.851 1.907 make_images 4572 14.5 0.268 0.304 1.673 1.819 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.816 1.818 calculate_dm_sparse 119 9.5 0.000 0.000 1.790 1.811 jit_kernel_multiply 10 15.9 0.932 1.811 0.932 1.811 rs_pw_transfer 974 11.9 0.008 0.008 1.680 1.789 prepare_preconditioner 11 7.9 0.000 0.000 1.777 1.782 make_preconditioner 11 8.9 0.000 0.000 1.777 1.782 density_rs2pw 119 9.7 0.003 0.004 1.629 1.735 pw_transfer 1439 11.6 0.063 0.065 1.713 1.720 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.665 1.691 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.624 1.631 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.616 1.616 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.577 1.578 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.522 1.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.419 1.424 grid_collocate_task_list 119 9.7 1.271 1.357 1.271 1.357 potential_pw2rs 119 12.3 0.010 0.011 1.345 1.347 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.347 1.347 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.322 1.331 cp_fm_redistribute_end 50 14.0 0.670 1.326 0.671 1.327 cp_fm_diag_elpa_base 50 14.0 0.610 1.256 0.654 1.312 fft3d_ps 1201 14.6 0.539 0.550 1.288 1.294 ot_diis_step 108 11.5 0.012 0.013 1.287 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.210 1.216 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.208 1.209 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.166 1.189 apply_single 119 13.6 0.000 0.000 1.166 1.189 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.135 1.144 wfi_extrapolate 11 7.9 0.001 0.001 1.069 1.069 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.789 1.019 make_images_data 4572 15.5 0.038 0.042 0.836 1.013 acc_transpose_blocks 9144 15.5 0.039 0.040 0.998 1.007 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.920 cp_fm_cholesky_invert 11 10.9 0.901 0.904 0.901 0.904 mp_alltoall_d11v 2130 13.8 0.753 0.842 0.753 0.842 multiply_cannon_sync_h2d 9144 15.5 0.715 0.793 0.715 0.793 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.756 0.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.751 0.753 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.719 0.727 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.665 0.715 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.327 0.711 mp_allgather_i34 2286 14.5 0.251 0.676 0.251 0.676 mp_alltoall_z22v 1201 16.6 0.623 0.657 0.623 0.657 jit_kernel_transpose 5 15.6 0.640 0.647 0.640 0.647 qs_create_task_list 11 7.9 0.000 0.001 0.544 0.569 generate_qs_task_list 11 8.9 0.190 0.212 0.544 0.569 mp_waitany 5200 13.7 0.434 0.554 0.434 0.554 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.585000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.454545, yerr=7.878205 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 750.850048E+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 0.016 0.034 42.152 42.152 qs_mol_dyn_low 1 2.0 0.003 0.003 41.909 41.916 qs_forces 11 3.9 0.002 0.002 41.838 41.839 qs_energies 11 4.9 0.001 0.001 39.876 39.881 scf_env_do_scf 11 5.9 0.001 0.001 33.577 33.577 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.568 25.569 velocity_verlet 10 3.0 0.002 0.003 23.422 23.427 dbcsr_multiply_generic 2286 12.5 0.100 0.101 18.362 18.488 qs_scf_new_mos 108 7.5 0.001 0.001 16.448 16.545 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.448 16.544 ot_scf_mini 108 9.5 0.002 0.002 15.334 15.433 multiply_cannon 2286 13.5 0.303 0.308 14.268 15.130 multiply_cannon_loop 2286 14.5 0.347 0.351 12.973 13.871 ot_mini 108 10.5 0.001 0.001 9.113 9.232 multiply_cannon_multrec 9144 15.5 3.407 4.970 8.934 9.063 init_scf_loop 11 6.9 0.000 0.000 7.982 7.983 rebuild_ks_matrix 119 8.3 0.000 0.000 7.007 7.162 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.007 7.161 qs_ot_get_derivative 108 11.5 0.001 0.001 7.018 7.118 prepare_preconditioner 11 7.9 0.000 0.000 7.038 7.056 make_preconditioner 11 8.9 0.000 0.000 7.038 7.056 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.603 6.927 dbcsr_mm_accdrv_process 12550 15.8 4.261 5.920 5.401 6.762 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.331 6.472 cp_fm_upper_to_full 72 14.2 3.165 4.537 3.165 4.537 init_scf_run 11 5.9 0.000 0.001 4.266 4.266 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.265 4.266 mp_waitall_1 97218 16.6 2.993 3.893 2.993 3.893 sum_up_and_integrate 119 10.3 0.038 0.038 3.786 3.797 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.777 3.782 calculate_rho_elec 119 8.7 0.118 0.121 3.776 3.782 integrate_v_rspace 119 11.3 0.003 0.003 3.748 3.759 qs_ot_get_p 119 10.4 0.001 0.001 3.411 3.552 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.577 2.960 dbcsr_complete_redistribute 329 12.2 0.289 0.292 2.012 2.831 calculate_first_density_matrix 1 7.0 0.001 0.001 2.811 2.813 make_m2s 4572 13.5 0.038 0.038 2.422 2.587 calculate_dm_sparse 119 9.5 0.000 0.000 2.506 2.525 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.188 2.486 apply_single 119 13.6 0.000 0.000 2.188 2.486 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.673 2.471 mp_alltoall_i22 627 13.8 1.566 2.471 1.566 2.471 make_images 4572 14.5 0.354 0.386 2.302 2.466 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.477 2.415 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.430 2.225 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.187 2.188 pw_transfer 1439 11.6 0.067 0.068 2.089 2.094 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.092 2.094 grid_integrate_task_list 119 12.3 2.043 2.081 2.043 2.081 ot_diis_step 108 11.5 0.014 0.014 2.064 2.064 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.993 1.998 mp_sum_l 7207 12.9 1.135 1.922 1.135 1.922 density_rs2pw 119 9.7 0.003 0.003 1.894 1.917 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.791 1.894 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.790 1.842 jit_kernel_multiply 8 15.4 1.114 1.807 1.114 1.807 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.806 1.807 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.759 1.760 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.647 1.653 fft3d_ps 1201 14.6 0.569 0.579 1.626 1.630 cp_fm_cholesky_invert 11 10.9 1.621 1.624 1.621 1.624 rs_pw_transfer 974 11.9 0.009 0.009 1.549 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.546 1.554 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.507 1.508 cp_fm_diag_elpa_base 50 14.0 1.361 1.412 1.505 1.506 grid_collocate_task_list 119 9.7 1.477 1.498 1.477 1.498 wfi_extrapolate 11 7.9 0.001 0.001 1.398 1.398 make_images_data 4572 15.5 0.042 0.045 1.157 1.374 hybrid_alltoall_any 4725 16.4 0.088 0.148 1.134 1.339 potential_pw2rs 119 12.3 0.014 0.015 1.297 1.299 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.218 mp_alltoall_d11v 2130 13.8 1.078 1.119 1.078 1.119 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.094 1.109 multiply_cannon_sync_h2d 9144 15.5 1.040 1.047 1.040 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.024 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.827 0.956 qs_create_task_list 11 7.9 0.000 0.001 0.944 0.955 generate_qs_task_list 11 8.9 0.373 0.391 0.944 0.955 mp_alltoall_z22v 1201 16.6 0.921 0.942 0.921 0.942 acc_transpose_blocks 9144 15.5 0.039 0.039 0.902 0.911 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.152000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.727273, yerr=8.822398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 500.154368E+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 639035. 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 0.037 0.098 85.299 85.300 qs_mol_dyn_low 1 2.0 0.003 0.005 84.807 84.817 qs_forces 11 3.9 0.003 0.003 84.733 84.734 qs_energies 11 4.9 0.006 0.037 81.835 81.848 scf_env_do_scf 11 5.9 0.000 0.001 72.221 72.227 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.277 66.278 dbcsr_multiply_generic 2055 12.4 0.105 0.107 53.326 53.661 qs_scf_new_mos 99 7.5 0.000 0.001 48.780 48.896 qs_scf_loop_do_ot 99 8.5 0.002 0.009 48.780 48.895 ot_scf_mini 99 9.5 0.002 0.006 46.316 46.443 multiply_cannon 2055 13.4 0.177 0.181 43.507 44.419 multiply_cannon_loop 2055 14.4 1.579 1.619 42.502 43.451 velocity_verlet 10 3.0 0.003 0.010 43.231 43.233 ot_mini 99 10.5 0.001 0.002 28.201 28.339 qs_ot_get_derivative 99 11.5 0.001 0.001 21.380 21.511 multiply_cannon_multrec 49320 15.4 12.054 12.836 17.615 18.560 rebuild_ks_matrix 110 8.3 0.000 0.000 14.905 15.049 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.020 14.905 15.049 mp_waitall_1 241148 16.1 12.319 13.364 12.319 13.364 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.132 13.260 multiply_cannon_sync_h2d 49320 15.4 9.899 10.479 9.899 10.479 qs_ot_get_p 110 10.4 0.001 0.001 9.460 9.555 multiply_cannon_metrocomm3 49320 15.4 0.076 0.081 7.200 8.476 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.629 8.136 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.143 7.672 apply_single 110 13.6 0.000 0.001 7.143 7.671 init_scf_run 11 5.9 0.000 0.001 7.527 7.527 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.526 7.527 sum_up_and_integrate 110 10.3 0.023 0.028 7.344 7.359 integrate_v_rspace 110 11.3 0.002 0.003 7.320 7.341 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.498 6.639 calculate_rho_elec 110 8.6 0.020 0.024 6.497 6.639 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.543 6.595 ot_diis_step 99 11.5 0.006 0.008 6.558 6.558 mp_sum_l 6514 12.8 5.406 6.170 5.406 6.170 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.124 6.164 init_scf_loop 11 6.9 0.005 0.041 5.916 5.920 dbcsr_mm_accdrv_process 87628 16.1 2.092 2.219 5.441 5.786 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.102 5.102 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.588 4.588 cp_fm_redistribute_end 48 14.0 3.992 4.557 3.997 4.558 rs_pw_transfer 902 11.9 0.012 0.014 4.301 4.496 cp_fm_diag_elpa_base 48 14.0 0.555 4.450 0.558 4.472 calculate_dm_sparse 110 9.5 0.001 0.001 4.179 4.288 make_m2s 4110 13.4 0.061 0.065 4.004 4.129 wfi_extrapolate 11 7.9 0.001 0.001 4.064 4.064 make_images 4110 14.4 0.178 0.192 3.908 4.037 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.881 3.881 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.729 3.733 density_rs2pw 110 9.6 0.004 0.004 3.518 3.719 prepare_preconditioner 11 7.9 0.000 0.000 3.556 3.575 make_preconditioner 11 8.9 0.000 0.001 3.556 3.575 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.379 3.454 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.402 3.450 grid_integrate_task_list 110 12.3 3.262 3.395 3.262 3.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.336 3.388 calculate_first_density_matrix 1 7.0 0.001 0.003 3.345 3.349 jit_kernel_multiply 13 15.9 3.064 3.138 3.064 3.138 pw_transfer 1331 11.6 0.054 0.063 2.931 3.004 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.839 2.918 potential_pw2rs 110 12.3 0.006 0.006 2.892 2.917 mp_alltoall_d11v 2046 13.8 2.230 2.635 2.230 2.635 fft_wrap_pw1pw2_140 451 13.1 0.170 0.187 2.375 2.454 fft3d_ps 1111 14.6 0.753 0.850 2.377 2.444 acc_transpose_blocks 49320 15.4 0.216 0.229 2.259 2.428 grid_collocate_task_list 110 9.6 2.093 2.219 2.093 2.219 mp_sum_d 3883 11.9 1.569 2.201 1.569 2.201 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.998 2.028 mp_waitany 14300 13.8 1.779 2.001 1.779 2.001 cp_fm_cholesky_invert 11 10.9 1.963 1.967 1.963 1.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.957 make_images_data 4110 15.4 0.043 0.046 1.792 1.930 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.557 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.300000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.909091, yerr=2.234219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 586.985472E+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 2989734. 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 0.068 0.200 73.768 73.770 qs_mol_dyn_low 1 2.0 0.003 0.005 72.942 73.027 qs_forces 11 3.9 0.003 0.008 71.455 71.457 qs_energies 11 4.9 0.002 0.008 68.112 68.116 scf_env_do_scf 11 5.9 0.001 0.001 57.885 57.888 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.918 49.919 dbcsr_multiply_generic 2055 12.4 0.113 0.116 39.672 39.859 velocity_verlet 10 3.0 0.001 0.002 38.338 38.368 qs_scf_new_mos 99 7.5 0.001 0.001 33.495 33.597 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.494 33.596 multiply_cannon 2055 13.4 0.227 0.270 32.365 33.578 ot_scf_mini 99 9.5 0.003 0.004 31.814 31.902 multiply_cannon_loop 2055 14.4 0.936 0.954 30.945 31.790 ot_mini 99 10.5 0.046 0.366 18.839 18.935 multiply_cannon_multrec 24660 15.4 7.617 9.188 14.742 16.338 rebuild_ks_matrix 110 8.3 0.000 0.001 14.013 14.071 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 14.012 14.071 qs_ot_get_derivative 99 11.5 0.001 0.002 12.685 12.774 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.364 12.421 mp_waitall_1 186928 16.3 8.108 10.679 8.108 10.679 multiply_cannon_sync_h2d 24660 15.4 6.997 8.207 6.997 8.207 init_scf_loop 11 6.9 0.001 0.004 7.929 7.931 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.101 7.833 init_scf_run 11 5.9 0.000 0.001 7.750 7.751 scf_env_initial_rho_setup 11 6.9 0.001 0.003 7.750 7.751 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.804 7.531 apply_single 110 13.6 0.000 0.001 6.804 7.531 dbcsr_mm_accdrv_process 52282 16.1 5.203 6.315 6.963 7.284 sum_up_and_integrate 110 10.3 0.029 0.035 6.839 6.851 integrate_v_rspace 110 11.3 0.002 0.003 6.810 6.822 qs_ot_get_p 110 10.4 0.001 0.003 6.294 6.441 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.098 6.121 calculate_rho_elec 110 8.6 0.039 0.047 6.098 6.120 ot_diis_step 99 11.5 0.011 0.013 6.060 6.106 prepare_preconditioner 11 7.9 0.000 0.000 5.814 5.823 make_preconditioner 11 8.9 0.000 0.001 5.814 5.823 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.411 5.570 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.757 5.471 make_m2s 4110 13.4 0.057 0.061 4.505 5.016 make_images 4110 14.4 0.400 0.445 4.395 4.903 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.234 4.253 calculate_first_density_matrix 1 7.0 0.002 0.011 3.992 3.996 calculate_dm_sparse 110 9.5 0.001 0.001 3.919 3.952 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.777 3.778 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.763 3.766 wfi_extrapolate 11 7.9 0.001 0.003 3.666 3.666 pw_transfer 1331 11.6 0.067 0.073 3.513 3.660 rs_pw_transfer 902 11.9 0.012 0.013 3.410 3.593 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.402 3.550 density_rs2pw 110 9.6 0.004 0.004 3.361 3.527 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.347 3.387 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.318 3.372 grid_integrate_task_list 110 12.3 3.139 3.327 3.139 3.327 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.149 3.153 cp_fm_redistribute_end 48 14.0 2.345 3.116 2.347 3.119 cp_fm_diag_elpa_base 48 14.0 0.734 2.984 0.765 3.066 make_images_data 4110 15.4 0.047 0.051 2.471 2.995 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.170 2.914 fft_wrap_pw1pw2_140 451 13.1 0.202 0.217 2.754 2.893 fft3d_ps 1111 14.6 1.073 1.268 2.758 2.892 mp_sum_l 6514 12.8 2.232 2.817 2.232 2.817 cp_fm_cholesky_invert 11 10.9 2.742 2.750 2.742 2.750 potential_pw2rs 110 12.3 0.009 0.009 2.681 2.695 jit_kernel_multiply 9 16.5 1.407 2.636 1.407 2.636 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.476 2.489 grid_collocate_task_list 110 9.6 2.121 2.262 2.121 2.262 mp_alltoall_d11v 2046 13.8 1.944 2.185 1.944 2.185 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.985 2.000 acc_transpose_blocks 24660 15.4 0.108 0.112 1.905 1.929 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.922 1.924 mp_allgather_i34 2055 14.4 0.735 1.830 0.735 1.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.816 1.829 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.803 1.760 mp_irecv_dv 57340 16.2 0.679 1.643 0.679 1.643 mp_sum_dm 438 4.9 1.476 1.523 1.476 1.523 dbcsr_complete_redistribute 325 12.2 0.237 0.289 1.244 1.515 mp_alltoall_z22v 1111 16.6 1.409 1.489 1.409 1.489 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.770000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.454545, yerr=7.414527 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 660.148224E+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 0.064 0.204 63.628 63.629 qs_mol_dyn_low 1 2.0 0.003 0.003 63.174 63.183 qs_forces 11 3.9 0.013 0.043 62.963 62.964 qs_energies 11 4.9 0.003 0.008 59.711 59.726 scf_env_do_scf 11 5.9 0.001 0.001 51.015 51.016 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.714 41.716 velocity_verlet 10 3.0 0.001 0.002 34.395 34.402 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.813 30.201 qs_scf_new_mos 99 7.5 0.001 0.001 26.474 26.564 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.474 26.563 ot_scf_mini 99 9.5 0.002 0.003 25.186 25.296 multiply_cannon 2055 13.4 0.212 0.227 22.537 23.819 multiply_cannon_loop 2055 14.4 0.622 0.637 21.185 22.701 ot_mini 99 10.5 0.001 0.001 14.283 14.402 rebuild_ks_matrix 110 8.3 0.000 0.000 12.635 12.815 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.635 12.814 mp_waitall_1 146766 16.3 7.956 11.638 7.956 11.638 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.162 11.316 multiply_cannon_multrec 16440 15.4 3.869 5.041 9.617 10.699 qs_ot_get_derivative 99 11.5 0.001 0.001 9.704 9.809 init_scf_loop 11 6.9 0.002 0.007 9.261 9.264 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.616 8.099 prepare_preconditioner 11 7.9 0.000 0.000 7.411 7.426 make_preconditioner 11 8.9 0.000 0.001 7.411 7.426 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.713 7.076 sum_up_and_integrate 110 10.3 0.033 0.034 6.739 6.752 integrate_v_rspace 110 11.3 0.002 0.003 6.706 6.719 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.187 6.208 calculate_rho_elec 110 8.6 0.058 0.058 6.186 6.207 qs_ot_get_p 110 10.4 0.001 0.001 5.963 6.113 init_scf_run 11 5.9 0.000 0.001 6.082 6.083 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.082 6.083 dbcsr_mm_accdrv_process 34862 16.1 4.857 5.294 5.605 5.791 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.053 5.526 apply_single 110 13.6 0.000 0.000 5.052 5.525 make_m2s 4110 13.4 0.050 0.052 4.855 5.286 make_images 4110 14.4 0.553 0.832 4.739 5.171 ot_diis_step 99 11.5 0.011 0.011 4.534 4.535 multiply_cannon_sync_h2d 16440 15.4 3.635 4.113 3.635 4.113 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.104 4.112 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.293 3.959 rs_pw_transfer 902 11.9 0.010 0.011 3.346 3.549 pw_transfer 1331 11.6 0.066 0.073 3.391 3.400 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.395 3.396 grid_integrate_task_list 110 12.3 3.191 3.377 3.191 3.377 density_rs2pw 110 9.6 0.004 0.004 3.182 3.314 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.280 3.290 make_images_data 4110 15.4 0.044 0.047 2.620 3.164 wfi_extrapolate 11 7.9 0.001 0.001 3.149 3.150 cp_fm_cholesky_invert 11 10.9 3.063 3.070 3.063 3.070 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.312 3.003 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.955 2.957 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.792 2.794 cp_fm_redistribute_end 48 14.0 1.739 2.759 1.742 2.760 fft_wrap_pw1pw2_140 451 13.1 0.217 0.222 2.745 2.759 cp_fm_diag_elpa_base 48 14.0 0.955 2.597 1.013 2.705 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.569 2.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.559 2.635 calculate_first_density_matrix 1 7.0 0.000 0.001 2.610 2.612 fft3d_ps 1111 14.6 1.064 1.074 2.587 2.596 calculate_dm_sparse 110 9.5 0.001 0.001 2.567 2.593 potential_pw2rs 110 12.3 0.011 0.011 2.540 2.548 mp_sum_l 6514 12.8 1.803 2.539 1.803 2.539 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.895 2.487 mp_alltoall_d11v 2046 13.8 2.142 2.438 2.142 2.438 mp_irecv_dv 48980 15.7 0.825 2.362 0.825 2.362 grid_collocate_task_list 110 9.6 2.186 2.359 2.186 2.359 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 2.166 2.178 dbcsr_complete_redistribute 325 12.2 0.329 0.358 1.583 2.055 mp_allgather_i34 2055 14.4 0.674 1.851 0.674 1.851 cp_fm_upper_to_full 70 13.6 1.410 1.845 1.410 1.845 cp_fm_cholesky_decompose 22 10.9 1.785 1.809 1.785 1.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.679 1.700 mp_waitany 17072 13.8 1.337 1.533 1.337 1.533 jit_kernel_multiply 8 17.0 0.359 1.505 0.359 1.505 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.045 1.501 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.381 1.483 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.463 1.473 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.239 1.349 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.337 1.346 mp_alltoall_z22v 1111 16.6 1.308 1.329 1.308 1.329 acc_transpose_blocks 16440 15.4 0.073 0.076 1.247 1.296 rs_pw_transfer_RS2PW_140 121 11.5 0.263 0.284 1.076 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.629000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.000000, yerr=9.302981 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 0.263 0.482 69.110 69.111 qs_mol_dyn_low 1 2.0 0.004 0.006 68.335 68.345 qs_forces 11 3.9 0.016 0.054 68.260 68.263 qs_energies 11 4.9 0.003 0.007 64.828 64.849 scf_env_do_scf 11 5.9 0.001 0.001 55.308 55.311 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 42.783 42.785 velocity_verlet 10 3.0 0.002 0.002 38.405 38.407 dbcsr_multiply_generic 2055 12.4 0.115 0.120 30.659 30.898 qs_scf_new_mos 99 7.5 0.001 0.001 27.869 27.980 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.869 27.980 ot_scf_mini 99 9.5 0.003 0.003 26.207 26.294 multiply_cannon 2055 13.4 0.243 0.263 23.023 24.171 multiply_cannon_loop 2055 14.4 0.887 0.903 21.432 22.094 ot_mini 99 10.5 0.001 0.001 14.619 14.726 multiply_cannon_multrec 24660 15.4 4.201 6.898 13.041 14.308 init_scf_loop 11 6.9 0.002 0.005 12.479 12.481 rebuild_ks_matrix 110 8.3 0.000 0.000 12.279 12.384 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.279 12.384 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.880 10.971 prepare_preconditioner 11 7.9 0.000 0.000 10.690 10.704 make_preconditioner 11 8.9 0.000 0.001 10.690 10.704 qs_ot_get_derivative 99 11.5 0.001 0.001 10.371 10.461 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.825 10.338 dbcsr_mm_accdrv_process 52304 16.0 7.370 8.902 8.691 9.648 mp_waitall_1 126806 16.4 4.944 7.003 4.944 7.003 qs_ot_get_p 110 10.4 0.001 0.001 6.433 6.562 sum_up_and_integrate 110 10.3 0.038 0.039 6.547 6.562 integrate_v_rspace 110 11.3 0.003 0.003 6.509 6.523 make_m2s 4110 13.4 0.060 0.062 5.826 6.263 make_images 4110 14.4 0.579 0.700 5.684 6.117 init_scf_run 11 5.9 0.000 0.001 6.108 6.109 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.108 6.109 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.917 5.923 calculate_rho_elec 110 8.6 0.077 0.081 5.916 5.923 cp_fm_upper_to_full 70 13.8 3.396 4.835 3.396 4.835 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.448 4.462 ot_diis_step 99 11.5 0.011 0.012 4.200 4.200 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.087 4.196 apply_single 110 13.6 0.000 0.000 4.086 4.196 dbcsr_complete_redistribute 325 12.2 0.409 0.444 2.854 4.027 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.906 3.908 rs_pw_transfer 902 11.9 0.010 0.011 3.247 3.540 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.448 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.421 3.470 pw_transfer 1331 11.6 0.066 0.073 3.415 3.440 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.269 3.438 make_images_data 4110 15.4 0.046 0.051 2.979 3.423 grid_integrate_task_list 110 12.3 3.258 3.417 3.258 3.417 calculate_dm_sparse 110 9.5 0.001 0.001 3.369 3.405 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.512 3.391 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.305 3.332 multiply_cannon_sync_h2d 24660 15.4 3.157 3.320 3.157 3.320 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.264 3.269 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.199 3.250 cp_fm_redistribute_end 48 14.0 1.628 3.220 1.631 3.222 density_rs2pw 110 9.6 0.004 0.004 3.043 3.193 cp_fm_diag_elpa_base 48 14.0 1.503 3.064 1.585 3.175 wfi_extrapolate 11 7.9 0.001 0.001 3.137 3.137 cp_fm_cholesky_invert 11 10.9 3.082 3.091 3.082 3.091 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.855 2.993 mp_alltoall_i22 605 13.7 1.739 2.954 1.739 2.954 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.887 2.905 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.877 2.878 calculate_first_density_matrix 1 7.0 0.001 0.001 2.849 2.851 fft_wrap_pw1pw2_140 451 13.1 0.207 0.221 2.723 2.751 fft3d_ps 1111 14.6 1.061 1.090 2.601 2.622 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.504 jit_kernel_multiply 11 15.9 0.993 2.408 0.993 2.408 grid_collocate_task_list 110 9.6 2.228 2.368 2.228 2.368 potential_pw2rs 110 12.3 0.013 0.013 2.328 2.340 mp_alltoall_d11v 2046 13.8 1.991 2.228 1.991 2.228 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.949 1.967 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.687 1.952 cp_fm_cholesky_decompose 22 10.9 1.888 1.938 1.888 1.938 mp_waitany 13376 13.8 1.534 1.895 1.534 1.895 mp_allgather_i34 2055 14.4 0.697 1.875 0.697 1.875 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.800 1.831 mp_sum_l 6514 12.8 1.134 1.791 1.134 1.791 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.604 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.648 1.662 rs_pw_transfer_RS2PW_140 121 11.5 0.244 0.253 1.309 1.593 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.844 1.537 acc_transpose_blocks 24660 15.4 0.106 0.108 1.504 1.524 build_core_hamiltonian_matrix 11 6.9 0.001 0.003 0.981 1.464 mp_irecv_dv 62702 16.1 0.746 1.461 0.746 1.461 mp_alltoall_z22v 1111 16.6 1.324 1.388 1.324 1.388 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.111000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.818182, yerr=9.173263 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 826.990592E+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 0.014 0.030 56.494 56.495 qs_mol_dyn_low 1 2.0 0.003 0.003 56.143 56.153 qs_forces 11 3.9 0.002 0.002 56.071 56.072 qs_energies 11 4.9 0.001 0.001 52.419 52.425 scf_env_do_scf 11 5.9 0.000 0.001 44.065 44.065 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.863 35.864 velocity_verlet 10 3.0 0.002 0.002 31.621 31.625 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.287 23.451 qs_scf_new_mos 99 7.5 0.001 0.001 21.398 21.438 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.398 21.438 ot_scf_mini 99 9.5 0.002 0.002 20.124 20.155 multiply_cannon 2055 13.4 0.249 0.264 17.533 19.063 multiply_cannon_loop 2055 14.4 0.325 0.338 16.141 16.553 rebuild_ks_matrix 110 8.3 0.000 0.000 11.713 11.750 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.713 11.749 ot_mini 99 10.5 0.001 0.001 10.728 10.755 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.417 10.450 multiply_cannon_multrec 8220 15.4 3.209 4.422 7.499 8.289 init_scf_loop 11 6.9 0.000 0.000 8.152 8.154 mp_waitall_1 106626 16.5 6.466 7.928 6.466 7.928 qs_ot_get_derivative 99 11.5 0.001 0.001 6.907 6.939 prepare_preconditioner 11 7.9 0.000 0.000 6.481 6.486 make_preconditioner 11 8.9 0.000 0.000 6.481 6.486 sum_up_and_integrate 110 10.3 0.048 0.049 6.355 6.368 integrate_v_rspace 110 11.3 0.003 0.003 6.308 6.320 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.050 6.126 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.799 5.816 calculate_rho_elec 110 8.6 0.115 0.116 5.799 5.815 qs_ot_get_p 110 10.4 0.001 0.001 5.422 5.455 init_scf_run 11 5.9 0.000 0.001 5.183 5.183 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.183 5.183 dbcsr_mm_accdrv_process 17442 15.9 2.881 3.682 4.159 5.054 make_m2s 4110 13.4 0.039 0.040 4.348 4.616 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.086 4.496 make_images 4110 14.4 0.641 0.701 4.217 4.485 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.968 3.972 ot_diis_step 99 11.5 0.012 0.012 3.785 3.786 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.710 3.738 apply_single 110 13.6 0.000 0.000 3.710 3.738 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.644 3.645 grid_integrate_task_list 110 12.3 3.361 3.492 3.361 3.492 pw_transfer 1331 11.6 0.066 0.071 3.381 3.395 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.269 3.288 cp_fm_cholesky_invert 11 10.9 3.210 3.214 3.210 3.214 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.069 3.070 cp_fm_redistribute_end 48 14.0 0.783 3.038 0.789 3.040 density_rs2pw 110 9.6 0.004 0.004 2.822 3.012 cp_fm_diag_elpa_base 48 14.0 2.066 2.831 2.242 2.994 multiply_cannon_sync_h2d 8220 15.4 2.916 2.953 2.916 2.953 make_images_data 4110 15.4 0.038 0.043 2.463 2.896 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.383 2.805 wfi_extrapolate 11 7.9 0.001 0.001 2.783 2.784 fft_wrap_pw1pw2_140 451 13.1 0.224 0.231 2.763 2.779 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.705 2.706 rs_pw_transfer 902 11.9 0.010 0.010 2.420 2.616 calculate_dm_sparse 110 9.5 0.001 0.001 2.561 2.602 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.559 2.560 fft3d_ps 1111 14.6 1.109 1.130 2.515 2.532 grid_collocate_task_list 110 9.6 2.324 2.455 2.324 2.455 calculate_first_density_matrix 1 7.0 0.001 0.001 2.306 2.307 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.170 2.189 potential_pw2rs 110 12.3 0.015 0.015 2.094 2.101 mp_alltoall_d11v 2046 13.8 1.804 1.995 1.804 1.995 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.977 1.993 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.987 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.888 1.904 cp_fm_cholesky_decompose 22 10.9 1.785 1.795 1.785 1.795 mp_allgather_i34 2055 14.4 0.576 1.701 0.576 1.701 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.547 1.671 dbcsr_complete_redistribute 325 12.2 0.560 0.590 1.552 1.643 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.631 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.493 1.505 jit_kernel_multiply 8 15.9 0.970 1.365 0.970 1.365 qs_create_task_list 11 7.9 0.014 0.017 1.255 1.355 generate_qs_task_list 11 8.9 0.378 0.445 1.242 1.355 mp_waitany 9240 13.8 1.059 1.261 1.059 1.261 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.217 1.255 mp_alltoall_z22v 1111 16.6 1.174 1.199 1.174 1.199 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.870 1.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.495000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.454545, yerr=11.340704 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 1.213702E+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 0.018 0.038 87.470 87.471 qs_mol_dyn_low 1 2.0 0.003 0.003 87.121 87.130 qs_forces 11 3.9 0.002 0.002 87.035 87.037 qs_energies 11 4.9 0.001 0.001 82.974 82.975 scf_env_do_scf 11 5.9 0.001 0.001 73.032 73.032 velocity_verlet 10 3.0 0.002 0.002 55.272 55.281 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.360 44.362 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.575 29.657 init_scf_loop 11 6.9 0.000 0.000 28.595 28.598 qs_scf_new_mos 99 7.5 0.001 0.001 27.780 27.822 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.779 27.821 prepare_preconditioner 11 7.9 0.000 0.000 26.627 26.636 make_preconditioner 11 8.9 0.000 0.000 26.627 26.636 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.943 26.103 ot_scf_mini 99 9.5 0.002 0.002 26.025 26.059 multiply_cannon 2055 13.4 0.347 0.375 21.942 22.708 multiply_cannon_loop 2055 14.4 0.343 0.346 20.138 21.019 cp_fm_upper_to_full 70 14.2 12.669 18.071 12.669 18.071 ot_mini 99 10.5 0.001 0.001 14.478 14.515 rebuild_ks_matrix 110 8.3 0.001 0.001 13.365 13.399 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.364 13.399 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.163 12.195 dbcsr_complete_redistribute 325 12.2 1.011 1.028 7.288 10.413 multiply_cannon_multrec 8220 15.4 4.473 4.666 9.756 10.205 qs_ot_get_derivative 99 11.5 0.001 0.001 9.872 9.907 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.253 9.381 mp_waitall_1 87304 16.6 8.186 9.083 8.186 9.083 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.669 8.777 mp_alltoall_i22 605 13.7 5.297 8.437 5.297 8.437 sum_up_and_integrate 110 10.3 0.090 0.091 6.706 6.721 integrate_v_rspace 110 11.3 0.003 0.003 6.615 6.631 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.482 6.519 calculate_rho_elec 110 8.6 0.226 0.227 6.481 6.519 qs_ot_get_p 110 10.4 0.001 0.001 6.266 6.315 make_m2s 4110 13.4 0.043 0.044 5.331 5.849 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.223 5.757 apply_single 110 13.6 0.000 0.000 5.222 5.757 init_scf_run 11 5.9 0.000 0.001 5.692 5.693 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.692 5.692 make_images 4110 14.4 0.881 0.930 5.143 5.662 cp_fm_cholesky_invert 11 10.9 5.642 5.646 5.642 5.646 dbcsr_mm_accdrv_process 11614 15.7 3.335 3.830 5.140 5.463 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.037 5.393 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.643 4.651 ot_diis_step 99 11.5 0.015 0.015 4.578 4.578 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.183 4.184 multiply_cannon_sync_h2d 8220 15.4 3.946 3.950 3.946 3.950 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.971 3.915 pw_transfer 1331 11.6 0.073 0.074 3.883 3.885 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.758 3.760 grid_integrate_task_list 110 12.3 3.664 3.728 3.664 3.728 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.636 3.638 make_images_data 4110 15.4 0.041 0.044 2.917 3.609 hybrid_alltoall_any 4261 16.3 0.256 0.556 2.873 3.591 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.582 3.583 cp_fm_diag_elpa_base 48 14.0 3.045 3.236 3.579 3.580 wfi_extrapolate 11 7.9 0.001 0.001 3.248 3.248 fft_wrap_pw1pw2_140 451 13.1 0.217 0.218 3.210 3.216 calculate_dm_sparse 110 9.5 0.001 0.001 3.121 3.149 density_rs2pw 110 9.6 0.004 0.004 3.086 3.101 fft3d_ps 1111 14.6 1.266 1.270 2.967 2.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.893 2.895 grid_collocate_task_list 110 9.6 2.631 2.650 2.631 2.650 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.501 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.404 2.418 calculate_first_density_matrix 1 7.0 0.001 0.001 2.338 2.339 rs_pw_transfer 902 11.9 0.011 0.011 2.282 2.324 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.181 2.237 potential_pw2rs 110 12.3 0.021 0.021 2.179 2.181 cp_fm_cholesky_decompose 22 10.9 2.150 2.171 2.150 2.171 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.165 qs_create_task_list 11 7.9 0.001 0.001 1.898 1.941 generate_qs_task_list 11 8.9 0.738 0.791 1.897 1.940 mp_alltoall_d11v 2046 13.8 1.873 1.927 1.873 1.927 mp_sum_l 6514 12.8 1.416 1.829 1.416 1.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.807 1.810 jit_kernel_multiply 10 15.3 1.604 1.784 1.604 1.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.471000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1126.727273, yerr=32.091038 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.273728E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 56701. MP_Allreduce 11263 785. MP_Sync 168 MP_Alltoall 2210 1969584. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 205.781 205.782 qs_mol_dyn_low 1 2.0 0.003 0.003 205.107 205.121 qs_forces 11 3.9 0.003 0.004 204.977 204.977 qs_energies 11 4.9 0.001 0.002 199.362 199.374 scf_env_do_scf 11 5.9 0.001 0.001 182.211 182.215 scf_env_do_scf_inner_loop 116 6.6 0.002 0.007 160.344 160.346 dbcsr_multiply_generic 2485 12.5 0.174 0.177 123.887 124.695 velocity_verlet 10 3.0 0.002 0.003 122.792 122.794 qs_scf_new_mos 116 7.6 0.001 0.001 120.677 120.971 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.676 120.970 ot_scf_mini 116 9.6 0.003 0.003 114.133 114.365 multiply_cannon 2485 13.5 0.236 0.243 100.382 103.082 multiply_cannon_loop 2485 14.5 2.111 2.177 98.078 100.362 ot_mini 116 10.6 0.001 0.001 65.615 65.887 multiply_cannon_multrec 59640 15.5 32.847 35.022 41.458 43.343 qs_ot_get_derivative 116 11.6 0.001 0.001 40.847 41.103 rebuild_ks_matrix 127 8.3 0.001 0.001 33.921 34.257 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.921 34.256 mp_waitall_1 288940 16.2 28.941 32.258 28.941 32.258 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.484 30.787 multiply_cannon_sync_h2d 59640 15.5 27.113 29.794 27.113 29.794 qs_ot_get_p 127 10.4 0.001 0.001 27.109 27.469 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.187 24.829 apply_single 127 13.6 0.001 0.001 24.187 24.828 ot_diis_step 116 11.6 0.007 0.008 24.404 24.405 init_scf_loop 11 6.9 0.000 0.001 21.781 21.782 qs_ot_p2m_diag 82 11.4 0.076 0.090 19.974 20.056 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.933 19.145 multiply_cannon_metrocomm3 59640 15.5 0.110 0.115 15.457 17.551 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.440 17.441 prepare_preconditioner 11 7.9 0.000 0.000 16.768 16.817 make_preconditioner 11 8.9 0.000 0.000 16.768 16.817 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.968 16.181 sum_up_and_integrate 127 10.3 0.052 0.065 14.594 14.608 integrate_v_rspace 127 11.3 0.003 0.004 14.541 14.559 make_m2s 4970 13.5 0.102 0.110 14.177 14.517 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.481 14.486 cp_fm_redistribute_end 82 14.4 11.428 14.397 11.440 14.400 make_images 4970 14.5 0.399 0.418 13.997 14.345 cp_fm_diag_elpa_base 82 14.4 2.917 14.098 2.948 14.210 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.175 13.279 calculate_rho_elec 127 8.7 0.044 0.063 13.174 13.278 init_scf_run 11 5.9 0.000 0.001 12.946 12.946 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.946 12.946 cp_fm_cholesky_invert 11 10.9 9.627 9.636 9.627 9.636 wfi_extrapolate 11 7.9 0.001 0.001 9.353 9.353 mp_sum_l 7804 13.0 8.076 8.909 8.076 8.909 dbcsr_mm_accdrv_process 123452 16.2 3.291 3.456 8.174 8.730 multiply_cannon_metrocomm1 59640 15.5 0.085 0.092 5.999 8.539 calculate_dm_sparse 127 9.5 0.001 0.001 8.421 8.517 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.097 8.204 make_images_data 4970 15.5 0.069 0.074 6.997 7.922 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.790 7.909 density_rs2pw 127 9.7 0.006 0.006 7.013 7.567 grid_integrate_task_list 127 12.3 7.034 7.548 7.034 7.548 rs_pw_transfer 1038 11.9 0.017 0.019 6.735 7.356 pw_transfer 1535 11.6 0.075 0.105 7.046 7.332 hybrid_alltoall_any 5155 16.4 0.288 2.239 6.125 7.305 fft_wrap_pw1pw2 1281 12.7 0.010 0.014 6.832 7.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.969 6.977 mp_alltoall_d11v 2401 14.1 4.388 6.256 4.388 6.256 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.078 6.178 fft_wrap_pw1pw2_140 519 13.2 0.432 0.490 5.754 5.974 fft3d_ps 1281 14.7 2.079 2.572 5.643 5.876 potential_pw2rs 127 12.3 0.009 0.010 5.141 5.188 grid_collocate_task_list 127 9.7 4.689 5.088 4.689 5.088 cp_fm_cholesky_decompose 22 10.9 4.906 4.920 4.906 4.920 mp_sum_d 4448 12.1 3.791 4.688 3.791 4.688 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.782000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.636364, yerr=6.399897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 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 826.228736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 4818336. 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 0.018 0.048 199.076 199.077 qs_mol_dyn_low 1 2.0 0.003 0.003 198.694 198.708 qs_forces 11 3.9 0.003 0.003 193.317 193.318 qs_energies 11 4.9 0.002 0.005 186.526 186.543 scf_env_do_scf 11 5.9 0.001 0.001 169.183 169.193 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.945 134.947 velocity_verlet 10 3.0 0.002 0.002 126.611 126.722 dbcsr_multiply_generic 2507 12.6 0.185 0.190 98.920 100.206 qs_scf_new_mos 117 7.6 0.001 0.001 95.518 96.026 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.517 96.025 ot_scf_mini 117 9.6 0.004 0.004 90.669 91.100 multiply_cannon 2507 13.6 0.477 0.528 78.236 82.008 multiply_cannon_loop 2507 14.6 1.266 1.299 74.906 77.660 ot_mini 117 10.6 0.001 0.001 50.955 51.456 mp_waitall_1 226760 16.4 25.693 40.202 25.693 40.202 multiply_cannon_multrec 30084 15.6 21.800 27.308 32.083 37.785 init_scf_loop 11 6.9 0.000 0.000 34.139 34.140 rebuild_ks_matrix 128 8.3 0.001 0.001 33.226 33.801 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.225 33.800 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.930 30.470 prepare_preconditioner 11 7.9 0.000 0.000 29.578 29.641 make_preconditioner 11 8.9 0.000 0.000 29.578 29.641 qs_ot_get_derivative 117 11.6 0.001 0.002 28.947 29.391 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 16.089 29.270 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.271 28.830 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.139 23.384 apply_single 128 13.6 0.001 0.001 22.139 23.384 qs_ot_get_p 128 10.4 0.001 0.001 21.599 22.294 multiply_cannon_sync_h2d 30084 15.6 19.150 22.101 19.150 22.101 ot_diis_step 117 11.6 0.014 0.015 21.829 21.831 cp_fm_cholesky_invert 11 10.9 17.269 17.282 17.269 17.282 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.452 16.504 make_m2s 5014 13.6 0.089 0.096 14.557 16.036 make_images 5014 14.6 1.155 1.351 14.347 15.827 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.250 15.251 sum_up_and_integrate 128 10.3 0.061 0.071 14.757 14.781 integrate_v_rspace 128 11.3 0.003 0.003 14.696 14.722 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.429 13.466 calculate_rho_elec 128 8.7 0.088 0.105 13.428 13.466 init_scf_run 11 5.9 0.000 0.001 12.088 12.089 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.088 12.089 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.053 12.062 cp_fm_redistribute_end 83 14.4 7.051 11.986 7.063 11.988 cp_fm_diag_elpa_base 83 14.4 4.691 11.428 4.904 11.874 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.300 11.642 multiply_cannon_metrocomm4 27577 15.6 0.097 0.113 3.798 10.880 make_images_data 5014 15.6 0.066 0.073 8.757 10.533 mp_irecv_dv 69486 16.3 3.603 10.487 3.603 10.487 dbcsr_mm_accdrv_process 62242 16.2 5.046 6.107 9.747 10.340 hybrid_alltoall_any 5200 16.5 0.341 1.498 7.468 9.579 wfi_extrapolate 11 7.9 0.001 0.001 8.507 8.507 pw_transfer 1547 11.6 0.084 0.096 8.018 8.089 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.789 7.860 grid_integrate_task_list 128 12.3 7.156 7.604 7.156 7.604 density_rs2pw 128 9.7 0.006 0.006 7.182 7.562 cp_fm_cholesky_decompose 22 10.9 7.399 7.479 7.399 7.479 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.350 7.198 calculate_dm_sparse 128 9.5 0.001 0.001 6.911 7.031 fft_wrap_pw1pw2_140 523 13.2 0.460 0.496 6.673 6.754 rs_pw_transfer 1046 11.9 0.014 0.016 6.243 6.652 mp_sum_l 7870 13.0 4.366 6.339 4.366 6.339 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.207 6.215 fft3d_ps 1291 14.7 2.778 2.940 6.166 6.213 mp_sum_dm 438 4.9 5.362 5.543 5.362 5.543 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.400 5.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.256 5.416 md_output 10 3.0 0.000 0.000 0.117 5.362 md_write_output 11 3.9 0.112 5.327 0.113 5.359 update_particle_set 20 4.0 0.000 0.000 5.243 5.356 grid_collocate_task_list 128 9.7 4.904 5.308 4.904 5.308 potential_pw2rs 128 12.3 0.016 0.019 5.221 5.249 mp_alltoall_d11v 2415 14.1 4.351 5.069 4.351 5.069 mp_allgather_i34 2507 14.6 1.908 4.886 1.908 4.886 dbcsr_complete_redistribute 395 12.7 0.788 0.903 3.323 4.182 mp_sum_d 4464 12.1 2.738 4.074 2.738 4.074 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=199.077000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.454545, yerr=4.292989 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 967.479296E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.215 185.411 185.413 qs_mol_dyn_low 1 2.0 0.004 0.005 184.758 184.788 qs_forces 11 3.9 0.007 0.016 182.074 182.083 qs_energies 11 4.9 0.002 0.005 175.487 175.501 scf_env_do_scf 11 5.9 0.001 0.001 158.642 158.645 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 122.524 122.527 velocity_verlet 10 3.0 0.002 0.003 118.296 118.378 qs_scf_new_mos 117 7.6 0.001 0.001 85.053 85.380 qs_scf_loop_do_ot 117 8.6 0.001 0.002 85.053 85.379 dbcsr_multiply_generic 2507 12.6 0.182 0.186 83.400 84.547 ot_scf_mini 117 9.6 0.004 0.004 80.884 81.251 multiply_cannon 2507 13.6 0.534 0.817 62.944 67.379 multiply_cannon_loop 2507 14.6 0.866 0.906 59.781 62.262 ot_mini 117 10.6 0.001 0.002 43.506 43.881 mp_waitall_1 178456 16.5 26.666 36.419 26.666 36.419 init_scf_loop 11 6.9 0.001 0.005 36.007 36.010 prepare_preconditioner 11 7.9 0.000 0.000 31.839 31.892 make_preconditioner 11 8.9 0.000 0.001 31.839 31.892 rebuild_ks_matrix 128 8.3 0.001 0.001 30.706 31.196 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.023 30.706 31.196 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.375 30.819 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.716 28.160 multiply_cannon_multrec 20056 15.6 13.589 16.388 22.616 25.408 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.666 25.093 qs_ot_get_derivative 117 11.6 0.002 0.005 23.802 24.162 qs_ot_get_p 128 10.4 0.001 0.001 22.720 23.175 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.793 20.786 apply_single 128 13.6 0.001 0.001 19.793 20.786 ot_diis_step 117 11.6 0.020 0.029 19.600 19.606 qs_ot_p2m_diag 83 11.4 0.266 0.273 18.290 18.298 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.253 17.255 make_m2s 5014 13.6 0.081 0.087 15.258 16.285 make_images 5014 14.6 1.181 1.276 15.024 16.040 multiply_cannon_sync_h2d 20056 15.6 14.217 15.790 14.217 15.790 cp_fm_cholesky_invert 11 10.9 14.780 14.789 14.780 14.789 sum_up_and_integrate 128 10.3 0.072 0.082 14.537 14.562 integrate_v_rspace 128 11.3 0.003 0.004 14.465 14.486 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.973 13.975 cp_fm_redistribute_end 83 14.4 5.310 13.899 5.335 13.903 cp_fm_diag_elpa_base 83 14.4 8.122 13.187 8.542 13.755 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.514 13.564 calculate_rho_elec 128 8.7 0.132 0.146 13.513 13.563 init_scf_run 11 5.9 0.000 0.001 11.645 11.646 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.645 11.646 make_images_data 5014 15.6 0.060 0.067 9.434 10.915 hybrid_alltoall_any 5200 16.5 0.429 1.932 8.113 10.214 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.184 9.438 multiply_cannon_metrocomm4 17549 15.6 0.063 0.074 3.409 9.137 mp_irecv_dv 50230 16.2 3.286 8.894 3.286 8.894 dbcsr_mm_accdrv_process 41502 16.2 4.637 5.572 8.478 8.878 pw_transfer 1547 11.6 0.083 0.102 8.064 8.177 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.837 7.957 cp_fm_cholesky_decompose 22 10.9 7.721 7.837 7.721 7.837 grid_integrate_task_list 128 12.3 7.267 7.764 7.267 7.764 wfi_extrapolate 11 7.9 0.030 0.117 7.618 7.618 cp_fm_upper_to_full 105 14.5 5.823 7.580 5.823 7.580 density_rs2pw 128 9.7 0.006 0.006 6.876 7.231 fft_wrap_pw1pw2_140 523 13.2 0.470 0.513 6.825 6.953 dbcsr_complete_redistribute 395 12.7 1.166 1.196 4.836 6.652 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.427 6.437 fft3d_ps 1291 14.7 2.698 2.910 6.117 6.200 rs_pw_transfer 1046 11.9 0.013 0.014 5.671 6.038 calculate_dm_sparse 128 9.5 0.001 0.002 5.876 5.969 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.047 5.764 grid_collocate_task_list 128 9.7 5.073 5.604 5.073 5.604 mp_alltoall_d11v 2415 14.1 4.714 5.555 4.714 5.555 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.228 5.529 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.590 5.400 mp_sum_l 7870 13.0 3.413 4.950 3.413 4.950 potential_pw2rs 128 12.3 0.021 0.023 4.861 4.879 mp_allgather_i34 2507 14.6 1.654 4.856 1.654 4.856 mp_alltoall_i22 716 14.1 2.269 4.245 2.269 4.245 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.445 4.240 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.017 4.061 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.857 3.860 calculate_first_density_matrix 1 7.0 0.002 0.004 3.847 3.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=185.413000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=909.272727, yerr=21.063069 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.155793E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.223 197.611 197.612 qs_mol_dyn_low 1 2.0 0.003 0.003 196.888 196.901 qs_forces 11 3.9 0.003 0.003 192.091 192.098 qs_energies 11 4.9 0.005 0.024 185.009 185.017 scf_env_do_scf 11 5.9 0.001 0.001 167.145 167.154 velocity_verlet 10 3.0 0.002 0.002 130.809 131.007 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.176 119.178 qs_scf_new_mos 117 7.6 0.001 0.001 82.968 83.277 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.967 83.277 dbcsr_multiply_generic 2507 12.6 0.200 0.206 80.470 81.409 ot_scf_mini 117 9.6 0.003 0.004 78.396 78.666 multiply_cannon 2507 13.6 0.553 0.591 55.126 57.990 multiply_cannon_loop 2507 14.6 1.186 1.216 51.333 53.560 init_scf_loop 11 6.9 0.000 0.001 47.836 47.837 prepare_preconditioner 11 7.9 0.000 0.000 43.396 43.416 make_preconditioner 11 8.9 0.000 0.003 43.396 43.416 ot_mini 117 10.6 0.001 0.001 42.954 43.211 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.052 42.057 multiply_cannon_multrec 30084 15.6 14.061 19.035 26.013 30.461 rebuild_ks_matrix 128 8.3 0.001 0.001 29.781 30.137 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.780 30.137 mp_waitall_1 153770 16.5 18.209 29.075 18.209 29.075 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.950 27.269 qs_ot_get_derivative 117 11.6 0.001 0.002 23.152 23.423 make_m2s 5014 13.6 0.095 0.100 20.717 21.745 qs_ot_get_p 128 10.4 0.001 0.001 21.464 21.743 make_images 5014 14.6 1.953 2.276 20.410 21.438 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.135 19.876 apply_single 128 13.6 0.001 0.001 19.135 19.876 ot_diis_step 117 11.6 0.018 0.019 19.659 19.660 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.186 17.243 cp_fm_cholesky_invert 11 10.9 16.947 16.956 16.947 16.956 cp_fm_upper_to_full 105 14.7 11.141 16.502 11.141 16.502 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.709 15.865 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.840 15.843 sum_up_and_integrate 128 10.3 0.079 0.086 14.582 14.605 integrate_v_rspace 128 11.3 0.003 0.003 14.503 14.529 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.444 13.482 calculate_rho_elec 128 8.7 0.174 0.190 13.443 13.482 dbcsr_complete_redistribute 395 12.7 1.569 1.754 9.296 13.090 multiply_cannon_sync_h2d 30084 15.6 11.784 12.976 11.784 12.976 make_images_data 5014 15.6 0.063 0.068 11.237 12.905 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.643 12.645 cp_fm_redistribute_end 83 14.4 2.170 12.544 2.186 12.549 cp_fm_diag_elpa_base 83 14.4 9.739 11.921 10.325 12.402 dbcsr_mm_accdrv_process 62264 16.2 7.365 8.315 11.527 12.074 copy_fm_to_dbcsr 209 11.7 0.003 0.015 7.847 11.641 hybrid_alltoall_any 5200 16.5 0.527 2.215 9.972 11.613 init_scf_run 11 5.9 0.000 0.001 11.192 11.194 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.192 11.193 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.320 10.004 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.659 9.854 mp_alltoall_i22 716 14.1 5.552 9.414 5.552 9.414 grid_integrate_task_list 128 12.3 7.531 8.186 7.531 8.186 pw_transfer 1547 11.6 0.083 0.095 8.012 8.106 cp_fm_cholesky_decompose 22 10.9 7.870 7.959 7.870 7.959 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.784 7.882 wfi_extrapolate 11 7.9 0.001 0.001 7.698 7.698 multiply_cannon_metrocomm4 25070 15.6 0.075 0.086 2.758 7.286 mp_irecv_dv 76098 16.2 2.615 7.012 2.615 7.012 density_rs2pw 128 9.7 0.005 0.006 6.533 6.990 fft_wrap_pw1pw2_140 523 13.2 0.472 0.481 6.823 6.931 calculate_dm_sparse 128 9.5 0.001 0.001 6.351 6.450 fft3d_ps 1291 14.7 2.792 2.914 6.028 6.114 mp_alltoall_d11v 2415 14.1 5.354 5.889 5.354 5.889 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.763 5.813 rs_pw_transfer 1046 11.9 0.013 0.014 5.189 5.659 grid_collocate_task_list 128 9.7 5.206 5.568 5.206 5.568 qs_energies_init_hamiltonians 11 5.9 0.293 0.339 5.078 5.079 mp_sum_dm 438 4.9 4.652 4.881 4.652 4.881 update_particle_set 20 4.0 0.000 0.000 4.561 4.759 md_output 10 3.0 0.000 0.000 0.203 4.756 md_write_output 11 3.9 0.197 4.715 0.200 4.753 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.471 4.555 potential_pw2rs 128 12.3 0.024 0.024 4.514 4.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.418 4.497 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.281 4.346 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=197.612000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.181818, yerr=22.172129 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+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 6806382528 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 1.515799E+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 5.526746E+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 726592214928 ------------------------------------------------------------------------------- - - - 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 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848081. 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 0.018 0.052 178.904 178.905 qs_mol_dyn_low 1 2.0 0.003 0.003 178.508 178.521 qs_forces 11 3.9 0.003 0.003 175.509 175.513 qs_energies 11 4.9 0.002 0.002 168.175 168.183 scf_env_do_scf 11 5.9 0.001 0.001 150.638 150.653 velocity_verlet 10 3.0 0.008 0.015 117.372 117.557 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 114.187 114.189 qs_scf_new_mos 118 7.6 0.001 0.001 77.847 77.921 qs_scf_loop_do_ot 118 8.6 0.001 0.001 77.846 77.920 dbcsr_multiply_generic 2529 12.6 0.182 0.188 75.620 76.039 ot_scf_mini 118 9.6 0.004 0.004 73.385 73.416 multiply_cannon 2529 13.6 0.586 0.615 55.536 59.850 multiply_cannon_loop 2529 14.6 0.454 0.462 50.712 51.338 ot_mini 118 10.6 0.001 0.001 40.474 40.512 init_scf_loop 11 6.9 0.000 0.000 36.295 36.297 mp_waitall_1 130746 16.6 26.670 33.868 26.670 33.868 prepare_preconditioner 11 7.9 0.000 0.000 32.434 32.463 make_preconditioner 11 8.9 0.000 0.000 32.434 32.463 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.255 30.522 rebuild_ks_matrix 129 8.3 0.001 0.001 29.068 29.119 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.067 29.118 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.544 26.593 multiply_cannon_multrec 10116 15.6 10.521 14.497 17.947 21.308 qs_ot_get_derivative 118 11.6 0.002 0.002 20.391 20.415 ot_diis_step 118 11.6 0.020 0.021 20.014 20.015 multiply_cannon_metrocomm3 10116 15.6 0.022 0.024 12.583 19.872 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.561 19.862 apply_single 129 13.6 0.001 0.001 19.560 19.862 qs_ot_get_p 129 10.4 0.001 0.001 19.787 19.837 make_m2s 5058 13.6 0.067 0.071 16.380 18.725 cp_fm_cholesky_invert 11 10.9 18.661 18.668 18.661 18.668 make_images 5058 14.6 2.325 2.883 16.071 18.405 qs_ot_p2m_diag 84 11.4 0.502 0.508 15.907 15.925 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.701 14.702 sum_up_and_integrate 129 10.3 0.104 0.110 14.268 14.314 integrate_v_rspace 129 11.3 0.004 0.004 14.164 14.215 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.417 13.447 calculate_rho_elec 129 8.7 0.259 0.270 13.416 13.446 make_images_data 5058 15.6 0.052 0.060 9.942 12.398 hybrid_alltoall_any 5245 16.5 0.814 3.597 9.688 12.355 multiply_cannon_sync_h2d 10116 15.6 11.692 12.155 11.692 12.155 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.498 11.503 cp_fm_diag_elpa_base 84 14.4 11.243 11.321 11.486 11.491 init_scf_run 11 5.9 0.000 0.001 10.548 10.549 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.548 10.549 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.263 8.289 cp_fm_cholesky_decompose 22 10.9 8.048 8.138 8.048 8.138 grid_integrate_task_list 129 12.3 7.803 8.110 7.803 8.110 pw_transfer 1559 11.6 0.083 0.089 7.866 7.881 dbcsr_mm_accdrv_process 20934 16.1 2.731 3.631 7.053 7.750 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.639 7.657 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.536 7.470 mp_allgather_i34 2529 14.6 3.025 7.342 3.025 7.342 wfi_extrapolate 11 7.9 0.001 0.001 7.267 7.267 fft_wrap_pw1pw2_140 527 13.2 0.496 0.513 6.694 6.717 calculate_dm_sparse 129 9.5 0.001 0.001 6.445 6.529 density_rs2pw 129 9.7 0.005 0.006 6.167 6.519 fft3d_ps 1301 14.7 2.739 2.812 5.798 5.814 grid_collocate_task_list 129 9.7 5.527 5.749 5.527 5.749 dbcsr_complete_redistribute 397 12.7 2.122 2.181 5.318 5.709 mp_alltoall_d11v 2429 14.1 4.878 5.450 4.878 5.450 qs_energies_init_hamiltonians 11 5.9 0.225 0.229 5.446 5.446 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.179 5.185 rs_pw_transfer 1054 12.0 0.013 0.013 4.444 4.807 multiply_cannon_metrocomm4 7587 15.6 0.024 0.026 1.867 4.505 mp_irecv_dv 29102 15.9 1.828 4.428 1.828 4.428 potential_pw2rs 129 12.3 0.027 0.028 4.279 4.290 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.083 4.108 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.809 3.848 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.834 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.461 3.778 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.625 3.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.905000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1406.818182, yerr=59.648696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992004743168 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237285E+12 0.0% 0.0% 100.0% flops max/rank 11.786603E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806448352 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 2.260484E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592178544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58052. MP_Allreduce 11197 1509. MP_Sync 87 MP_Alltoall 1724 36993608. 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 0.059 0.323 299.205 299.205 qs_mol_dyn_low 1 2.0 0.003 0.003 298.100 298.112 qs_forces 11 3.9 0.004 0.005 296.155 296.159 qs_energies 11 4.9 0.002 0.002 287.510 287.520 scf_env_do_scf 11 5.9 0.001 0.001 265.354 265.365 velocity_verlet 10 3.0 0.005 0.006 215.247 215.485 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 137.954 137.956 init_scf_loop 11 6.9 0.000 0.000 127.139 127.141 prepare_preconditioner 11 7.9 0.000 0.000 122.396 122.434 make_preconditioner 11 8.9 0.000 0.000 122.396 122.434 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.522 119.520 qs_scf_new_mos 118 7.6 0.001 0.001 95.123 95.259 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.122 95.258 ot_scf_mini 118 9.6 0.004 0.004 90.287 90.331 dbcsr_multiply_generic 2531 12.6 0.216 0.224 85.229 85.817 cp_fm_upper_to_full 105 14.8 53.624 77.547 53.624 77.547 multiply_cannon 2531 13.6 0.701 0.746 60.252 60.972 multiply_cannon_loop 2531 14.6 0.478 0.482 56.565 58.084 ot_mini 118 10.6 0.001 0.001 45.645 45.686 dbcsr_complete_redistribute 395 12.7 3.991 4.022 30.622 44.386 copy_fm_to_dbcsr 209 11.7 0.001 0.001 27.180 40.879 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.828 38.392 mp_alltoall_i22 718 14.1 22.683 36.140 22.683 36.140 cp_fm_cholesky_invert 11 10.9 34.469 34.477 34.469 34.477 rebuild_ks_matrix 129 8.3 0.001 0.001 34.225 34.267 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.224 34.267 mp_waitall_1 106466 16.7 29.561 33.427 29.561 33.427 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.979 32.019 qs_ot_get_p 129 10.4 0.001 0.001 28.753 28.774 qs_ot_get_derivative 118 11.6 0.002 0.002 25.027 25.072 qs_ot_p2m_diag 83 11.4 0.879 0.885 24.478 24.511 cp_dbcsr_syevd 83 12.4 0.005 0.006 22.692 22.694 make_m2s 5062 13.6 0.077 0.081 20.242 21.258 multiply_cannon_metrocomm3 10124 15.6 0.023 0.023 19.664 21.024 make_images 5062 14.6 3.774 3.877 19.763 20.781 ot_diis_step 118 11.6 0.023 0.028 20.567 20.568 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.547 19.761 apply_single 129 13.6 0.001 0.001 19.547 19.761 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.367 19.371 cp_fm_diag_elpa_base 83 14.4 14.866 16.574 19.361 19.366 multiply_cannon_multrec 10124 15.6 10.586 12.481 18.072 18.153 sum_up_and_integrate 129 10.3 0.195 0.196 15.815 15.907 multiply_cannon_sync_h2d 10124 15.6 15.777 15.787 15.777 15.787 integrate_v_rspace 129 11.3 0.004 0.004 15.619 15.711 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.124 15.133 calculate_rho_elec 129 8.7 0.487 0.487 15.123 15.133 hybrid_alltoall_any 5248 16.5 1.313 3.064 11.012 12.940 make_images_data 5062 15.6 0.059 0.064 10.923 12.691 init_scf_run 11 5.9 0.000 0.001 12.235 12.235 scf_env_initial_rho_setup 11 6.9 0.003 0.003 12.234 12.235 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.725 9.768 cp_fm_cholesky_decompose 22 10.9 9.257 9.280 9.257 9.280 dbcsr_mm_accdrv_process 20942 16.1 3.888 5.774 7.246 9.138 grid_integrate_task_list 129 12.3 8.599 8.778 8.599 8.778 wfi_extrapolate 11 7.9 0.001 0.001 8.768 8.768 pw_transfer 1559 11.6 0.093 0.095 8.453 8.464 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.213 8.224 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.912 7.912 fft_wrap_pw1pw2_140 527 13.2 0.529 0.531 7.224 7.245 mp_alltoall_d11v 2423 14.1 6.907 7.032 6.907 7.032 calculate_dm_sparse 129 9.5 0.001 0.001 6.688 6.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.685 6.737 grid_collocate_task_list 129 9.7 6.387 6.417 6.387 6.417 fft3d_ps 1301 14.7 2.786 2.803 6.255 6.264 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.111 6.214 density_rs2pw 129 9.7 0.005 0.005 6.147 6.171 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.205000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2127.272727, yerr=49.883666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 1.260278E+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 0.351 0.398 85.923 85.924 qs_energies 1 2.0 0.000 0.000 85.155 85.168 ls_scf 1 3.0 0.000 0.000 84.228 84.241 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.242 73.490 multiply_cannon 111 7.7 0.017 0.020 56.152 57.229 multiply_cannon_loop 111 8.7 0.212 0.229 52.649 54.027 ls_scf_main 1 4.0 0.000 0.000 52.595 52.596 density_matrix_trs4 2 5.0 0.002 0.004 47.043 47.147 ls_scf_init_scf 1 4.0 0.000 0.000 28.568 28.569 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.485 27.534 mp_waitall_1 11316 10.9 22.684 26.344 22.684 26.344 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.372 25.399 multiply_cannon_multrec 2664 9.7 8.178 8.893 15.474 17.213 multiply_cannon_sync_h2d 2664 9.7 13.788 15.918 13.788 15.918 make_m2s 222 7.7 0.008 0.011 13.316 13.857 make_images 222 8.7 0.098 0.108 13.294 13.838 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.669 12.793 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.453 8.544 make_images_data 222 9.7 0.004 0.005 7.880 8.512 hybrid_alltoall_any 227 10.6 0.215 1.843 6.683 8.190 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.622 6.914 7.910 dbcsr_mm_accdrv_process_sort 4760 11.4 6.205 7.140 6.205 7.140 calculate_norms 4752 9.8 5.533 6.185 5.533 6.185 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.058 5.273 mp_sum_l 807 5.4 3.254 4.992 3.254 4.992 make_images_sizes 222 9.7 0.000 0.000 0.767 3.620 mp_alltoall_i44 222 10.7 0.767 3.619 0.767 3.619 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.067 3.618 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.378 3.596 mp_irecv_dv 6231 10.9 2.049 3.590 2.049 3.590 arnoldi_extremal 4 6.8 0.000 0.000 3.283 3.322 arnoldi_normal_ev 4 7.8 0.001 0.003 3.283 3.322 build_subspace 16 8.4 0.009 0.012 3.178 3.180 ls_scf_post 1 4.0 0.000 0.000 3.065 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.882 2.916 dbcsr_special_finalize 555 9.7 0.005 0.006 2.316 2.874 dbcsr_merge_single_wm 555 10.7 0.460 0.594 2.308 2.866 make_images_pack 222 9.7 2.208 2.651 2.210 2.652 dbcsr_sort_data 658 11.4 2.107 2.593 2.107 2.593 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.332 2.575 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.460 2.067 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.292 2.409 buffer_matrices_ensure_size 222 8.7 1.752 2.176 1.752 2.176 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.766 1.767 rebuild_ks_matrix 3 7.3 0.000 0.000 1.756 1.758 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.756 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.924000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1124.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 2.086015E+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 0.064 0.131 94.125 94.126 qs_energies 1 2.0 0.000 0.000 93.240 93.275 ls_scf 1 3.0 0.282 2.259 91.870 91.908 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.490 78.051 multiply_cannon 111 7.7 0.028 0.041 53.159 56.673 ls_scf_main 1 4.0 0.000 0.007 55.263 55.269 multiply_cannon_loop 111 8.7 0.117 0.124 49.908 53.284 density_matrix_trs4 2 5.0 0.002 0.003 49.578 49.775 mp_waitall_1 9246 10.9 21.322 30.993 21.322 30.993 ls_scf_init_scf 1 4.0 0.000 0.002 30.827 30.829 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.607 29.717 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.232 27.242 multiply_cannon_multrec 1332 9.7 13.112 16.397 22.320 26.769 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.740 21.376 make_m2s 222 7.7 0.006 0.008 15.673 16.412 make_images 222 8.7 1.578 1.972 15.643 16.383 mp_sum_l 807 5.4 7.401 10.918 7.401 10.918 dbcsr_mm_accdrv_process 4041 10.4 0.283 0.452 8.813 10.422 make_images_data 222 9.7 0.004 0.004 9.050 10.005 dbcsr_mm_accdrv_process_sort 4041 11.4 8.407 9.970 8.407 9.970 hybrid_alltoall_any 227 10.6 0.522 2.476 8.437 9.488 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.252 7.768 mp_irecv_dv 3311 11.0 3.232 7.703 3.232 7.703 apply_matrix_preconditioner 6 5.3 0.000 0.000 7.009 7.506 calculate_norms 2376 9.8 6.024 6.841 6.024 6.841 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.122 6.759 multiply_cannon_sync_h2d 1332 9.7 4.819 6.158 4.819 6.158 ls_scf_post 1 4.0 0.014 0.051 5.498 5.817 ls_scf_store_result 1 5.0 0.000 0.000 5.188 5.585 arnoldi_extremal 4 6.8 0.000 0.003 4.744 4.769 arnoldi_normal_ev 4 7.8 0.001 0.006 4.744 4.766 build_subspace 16 8.4 0.014 0.021 4.482 4.487 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.173 3.421 dbcsr_matrix_vector_mult_local 304 10.0 2.760 3.225 2.762 3.227 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.556 2.657 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.162 2.571 make_images_pack 222 9.7 2.024 2.421 2.026 2.423 mp_allgather_i34 111 8.7 0.954 2.410 0.954 2.410 dbcsr_sort_data 436 11.2 1.838 2.085 1.838 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.906 rebuild_ks_matrix 3 7.3 0.000 0.000 1.891 1.894 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.891 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.126000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 2.697732E+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 0.055 0.103 96.219 96.220 qs_energies 1 2.0 0.000 0.000 95.364 95.366 ls_scf 1 3.0 0.097 0.776 93.906 93.908 dbcsr_multiply_generic 111 6.7 0.017 0.024 77.848 78.193 ls_scf_main 1 4.0 0.029 0.248 58.259 58.266 multiply_cannon 111 7.7 0.039 0.074 53.269 57.389 multiply_cannon_loop 111 8.7 0.100 0.105 49.574 53.447 density_matrix_trs4 2 5.0 0.013 0.356 52.158 52.302 mp_waitall_1 7374 11.0 24.135 34.533 24.135 34.533 ls_scf_init_scf 1 4.0 0.000 0.002 31.028 31.029 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.542 29.619 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.144 27.161 multiply_cannon_multrec 888 9.7 12.603 15.339 21.106 24.437 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.246 23.516 make_m2s 222 7.7 0.006 0.008 16.922 18.296 make_images 222 8.7 1.971 2.308 16.884 18.255 make_images_data 222 9.7 0.004 0.004 9.689 10.888 hybrid_alltoall_any 227 10.6 0.621 2.863 9.269 10.860 mp_sum_l 807 5.4 6.122 9.429 6.122 9.429 dbcsr_mm_accdrv_process 3754 10.4 0.243 0.425 8.038 9.241 dbcsr_mm_accdrv_process_sort 3754 11.4 7.665 8.816 7.665 8.816 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.819 7.320 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.911 7.178 multiply_cannon_sync_h2d 888 9.7 6.069 7.141 6.069 7.141 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.440 6.943 mp_irecv_dv 2335 11.1 2.424 6.898 2.424 6.898 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.623 5.915 arnoldi_extremal 4 6.8 0.000 0.003 5.156 5.176 arnoldi_normal_ev 4 7.8 0.002 0.009 5.156 5.176 build_subspace 16 8.4 0.015 0.024 4.826 4.836 ls_scf_post 1 4.0 0.000 0.002 4.522 4.621 calculate_norms 1584 9.8 4.257 4.589 4.257 4.589 ls_scf_store_result 1 5.0 0.000 0.000 4.014 4.194 mp_allgather_i34 111 8.7 1.542 3.933 1.542 3.933 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.463 3.776 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.608 3.039 3.610 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.917 3.026 make_images_sizes 222 9.7 0.000 0.000 0.958 2.185 mp_alltoall_i44 222 10.7 0.958 2.184 0.958 2.184 dbcsr_sort_data 325 11.1 1.850 2.164 1.850 2.164 make_images_pack 222 9.7 1.810 2.137 1.813 2.139 dbcsr_data_release 9322 10.9 1.310 1.940 1.310 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.220000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2188.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 3.353108E+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 0.046 0.082 97.572 97.573 qs_energies 1 2.0 0.000 0.000 96.561 96.568 ls_scf 1 3.0 0.000 0.003 94.513 94.519 dbcsr_multiply_generic 111 6.7 0.018 0.023 78.243 78.470 ls_scf_main 1 4.0 0.001 0.006 58.665 58.667 multiply_cannon 111 7.7 0.053 0.109 51.702 56.269 density_matrix_trs4 2 5.0 0.002 0.003 52.587 52.693 multiply_cannon_loop 111 8.7 0.115 0.126 46.633 49.439 ls_scf_init_scf 1 4.0 0.000 0.002 32.609 32.612 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.398 31.463 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.867 28.880 mp_waitall_1 6438 11.0 22.738 28.800 22.738 28.800 multiply_cannon_multrec 1332 9.7 14.162 17.328 21.991 24.331 make_m2s 222 7.7 0.007 0.008 21.041 22.500 make_images 222 8.7 3.136 3.582 20.990 22.452 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.321 17.306 make_images_data 222 9.7 0.004 0.005 11.702 13.319 hybrid_alltoall_any 227 10.6 0.800 3.841 11.048 12.721 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.407 7.465 8.994 dbcsr_mm_accdrv_process_sort 3641 11.4 7.093 8.577 7.093 8.577 mp_sum_l 807 5.4 3.926 7.604 3.926 7.604 multiply_cannon_sync_h2d 1332 9.7 5.566 6.161 5.566 6.161 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.089 6.094 mp_irecv_dv 3229 10.9 2.065 6.047 2.065 6.047 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.966 5.971 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.543 5.383 arnoldi_extremal 4 6.8 0.001 0.003 5.337 5.353 arnoldi_normal_ev 4 7.8 0.002 0.005 5.336 5.353 build_subspace 16 8.4 0.015 0.021 4.978 4.989 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.639 4.795 mp_allgather_i34 111 8.7 2.201 4.689 2.201 4.689 calculate_norms 2376 9.8 4.182 4.566 4.182 4.566 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.645 3.950 dbcsr_matrix_vector_mult_local 304 10.0 3.226 3.713 3.228 3.715 dbcsr_sort_data 658 11.4 3.094 3.417 3.094 3.417 dbcsr_special_finalize 555 9.7 0.006 0.008 2.847 3.253 ls_scf_post 1 4.0 0.000 0.002 3.239 3.245 dbcsr_merge_single_wm 555 10.7 0.540 0.663 2.838 3.245 ls_scf_store_result 1 5.0 0.000 0.000 2.985 3.039 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.910 2.939 dbcsr_data_release 10477 10.7 1.577 2.375 1.577 2.375 qs_energies_init_hamiltonians 1 3.0 0.237 0.384 2.038 2.038 dbcsr_finalize 304 7.8 0.049 0.061 1.800 1.979 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.957 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.573000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2700.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 4.628083E+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 0.045 0.059 99.565 99.565 qs_energies 1 2.0 0.000 0.000 98.802 98.808 ls_scf 1 3.0 0.000 0.000 96.874 96.879 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.324 78.541 ls_scf_main 1 4.0 0.000 0.000 62.038 62.039 multiply_cannon 111 7.7 0.102 0.192 55.627 60.821 density_matrix_trs4 2 5.0 0.002 0.004 54.982 55.095 multiply_cannon_loop 111 8.7 0.070 0.078 51.081 52.692 mp_waitall_1 5481 11.0 26.871 32.695 26.871 32.695 ls_scf_init_scf 1 4.0 0.000 0.000 31.237 31.243 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.049 30.089 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.507 27.513 multiply_cannon_multrec 444 9.7 13.876 16.090 20.965 22.922 make_m2s 222 7.7 0.004 0.005 17.645 20.194 make_images 222 8.7 3.706 4.390 17.583 20.133 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.508 16.852 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.246 15.200 make_images_data 222 9.7 0.003 0.004 9.911 12.433 hybrid_alltoall_any 227 10.6 0.788 3.771 9.772 12.121 dbcsr_mm_accdrv_process 3003 10.4 0.161 0.667 6.801 8.219 multiply_cannon_sync_h2d 444 9.7 6.506 7.945 6.506 7.945 dbcsr_mm_accdrv_process_sort 3003 11.4 6.402 7.508 6.402 7.508 mp_allgather_i34 111 8.7 2.782 6.962 2.782 6.962 arnoldi_extremal 4 6.8 0.000 0.000 5.809 5.814 arnoldi_normal_ev 4 7.8 0.004 0.007 5.809 5.814 mp_sum_l 807 5.4 3.109 5.543 3.109 5.543 build_subspace 16 8.4 0.015 0.020 5.422 5.431 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.024 5.152 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.201 4.363 dbcsr_matrix_vector_mult_local 304 10.0 3.711 4.160 3.713 4.162 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.973 3.971 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.655 3.934 mp_irecv_dv 1241 11.2 1.635 3.908 1.635 3.908 calculate_norms 792 9.8 3.554 3.703 3.554 3.703 ls_scf_post 1 4.0 0.000 0.000 3.598 3.604 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.381 3.488 make_images_sizes 222 9.7 0.000 0.000 1.115 3.464 mp_alltoall_i44 222 10.7 1.114 3.464 1.114 3.464 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.426 dbcsr_finalize 304 7.8 0.062 0.078 2.194 2.249 dbcsr_merge_all 275 8.9 0.473 0.527 2.043 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.565000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3616.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9368eddea8879e44b245fe945652cfe39f9313dd_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 8.755315E+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 0.127 0.158 108.449 108.450 qs_energies 1 2.0 0.000 0.000 106.768 106.783 ls_scf 1 3.0 0.000 0.000 103.882 103.896 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.888 78.024 ls_scf_main 1 4.0 0.000 0.000 65.847 65.848 density_matrix_trs4 2 5.0 0.002 0.004 56.983 57.060 multiply_cannon 111 7.7 0.172 0.236 50.201 51.697 multiply_cannon_loop 111 8.7 0.067 0.070 46.830 47.889 ls_scf_init_scf 1 4.0 0.000 0.000 34.260 34.261 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.912 32.925 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.093 30.106 mp_waitall_1 4569 11.1 22.665 26.463 22.665 26.463 make_m2s 222 7.7 0.005 0.005 23.891 24.788 make_images 222 8.7 4.569 4.957 23.785 24.680 multiply_cannon_multrec 444 9.7 17.767 18.514 22.456 23.081 hybrid_alltoall_any 227 10.6 1.657 3.623 12.923 15.781 make_images_data 222 9.7 0.003 0.003 13.161 15.511 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.848 11.928 multiply_cannon_sync_h2d 444 9.7 8.830 8.871 8.830 8.871 arnoldi_extremal 4 6.8 0.000 0.000 7.456 7.472 arnoldi_normal_ev 4 7.8 0.003 0.009 7.456 7.472 build_subspace 16 8.4 0.026 0.036 6.901 6.913 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.527 5.645 dbcsr_matrix_vector_mult_local 304 10.0 5.050 5.330 5.052 5.332 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.063 5.309 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.788 4.879 dbcsr_mm_accdrv_process 1814 10.4 0.282 0.512 4.517 4.728 dbcsr_mm_accdrv_process_sort 1814 11.4 4.118 4.247 4.118 4.247 ls_scf_post 1 4.0 0.000 0.000 3.775 3.789 make_images_sizes 222 9.7 0.000 0.000 1.439 3.613 mp_alltoall_i44 222 10.7 1.439 3.612 1.439 3.612 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.453 calculate_norms 792 9.8 3.257 3.298 3.257 3.298 mp_allgather_i34 111 8.7 1.031 3.218 1.031 3.218 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.126 dbcsr_merge_all 275 8.9 0.892 0.913 2.865 2.907 dbcsr_complete_redistribute 5 7.6 1.430 1.464 2.768 2.878 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 2.856 2.856 dbcsr_data_release 12724 10.6 2.312 2.802 2.312 2.802 matrix_ls_to_qs 2 6.0 0.000 0.000 2.409 2.531 dbcsr_sort_data 325 11.1 2.437 2.510 2.437 2.510 mp_sum_l 807 5.4 1.485 2.509 1.485 2.509 dbcsr_new_transposed 4 7.5 0.243 0.251 2.309 2.329 dbcsr_frobenius_norm 74 6.6 2.056 2.133 2.192 2.242 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.232 2.234 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.204 dbcsr_add_anytype 103 7.2 0.860 0.894 2.132 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.450000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6817.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9368eddea8879e44b245fe945652cfe39f9313dd Summary: empty Status: OK