=== 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: c2dc0ed42209bbdc4f6d5b58e456999b747df445 ################# 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/01 job id: 41808573 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/02 job id: 41808574 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/03 job id: 41808575 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/04 job id: 41808576 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/05 job id: 41808578 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/06 job id: 41808579 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/07 job id: 41808582 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/08 job id: 41808584 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/09 job id: 41808585 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/10 job id: 41808586 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/11 job id: 41808588 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/12 job id: 41808590 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/13 job id: 41808591 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/14 job id: 41808593 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/15 job id: 41808594 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/16 job id: 41808595 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/17 job id: 41808596 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/18 job id: 41808597 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/19 job id: 41808598 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/20 job id: 41808599 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/21 job id: 41808600 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/22 job id: 41808601 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/23 job id: 41808603 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/24 job id: 41808605 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/25 job id: 41808607 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/26 job id: 41808609 --- 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/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 0.023 0.035 134.442 134.442 farming_run 1 2.0 133.737 133.740 134.413 134.415 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.456333E+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 0.008 0.024 115.070 115.070 qs_energies 1 2.0 0.000 0.000 114.862 114.865 mp2_main 1 3.0 0.000 0.000 112.954 112.956 mp2_gpw_main 1 4.0 0.056 0.109 112.107 112.110 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.302 93.827 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.496 55.994 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.174 41.808 47.271 get_2c_integrals 1 6.0 0.000 0.000 37.118 37.634 integrate_v_rspace 273 8.0 0.436 0.449 25.165 30.351 pw_transfer 6555 10.6 0.380 0.403 27.334 27.818 grid_integrate_task_list 273 9.0 20.984 26.671 20.984 26.671 fft_wrap_pw1pw2 5465 11.4 0.047 0.049 25.989 26.464 fft_wrap_pw1pw2_100 2178 12.4 1.180 1.357 23.557 24.030 compute_2c_integrals 1 7.0 0.002 0.003 19.566 19.594 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.897 19.319 mp2_eri_2c_integrate_gpw 1 9.0 2.393 2.447 18.894 19.315 rpa_ri_compute_en 1 5.0 0.000 0.000 18.659 18.880 cp_fm_cholesky_decompose 12 8.2 17.512 18.016 17.512 18.016 cholesky_decomp 1 7.0 0.000 0.000 16.395 16.915 fft3d_s 5443 13.4 16.152 16.504 16.173 16.525 ao_to_mo_and_store_B_mult_1 272 7.0 10.862 15.610 10.862 15.610 calculate_wavefunction 272 8.0 5.448 5.632 12.558 13.139 rpa_num_int 1 6.0 0.000 0.000 10.494 10.494 rpa_num_int_RPA_matrix_operati 8 7.0 0.001 0.001 10.469 10.489 calc_mat_Q 8 8.0 0.000 0.000 9.297 9.422 contract_S_to_Q 8 9.0 0.000 0.000 8.720 8.845 calc_potential_gpw 544 9.5 0.005 0.006 8.227 8.673 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.207 8.534 parallel_gemm_fm 14 9.1 0.000 0.000 8.292 8.412 parallel_gemm_fm_cosma 14 10.1 8.292 8.412 8.292 8.412 potential_pw2rs 545 10.0 0.107 0.110 7.703 8.330 collocate_single_gaussian 272 10.0 0.040 0.042 7.472 7.720 create_integ_mat 1 6.0 0.022 0.028 7.572 7.573 array2fm 1 7.0 0.000 0.000 6.564 7.058 pw_scatter_s 2720 13.7 4.419 4.548 4.419 4.548 pw_gather_s 2722 13.2 3.873 4.119 3.873 4.119 array2fm_buffer_send 1 8.0 2.959 3.171 2.959 3.171 pw_poisson_solve 545 10.5 1.127 1.188 2.175 2.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.110255, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 0.029 0.035 399.766 399.768 farming_run 1 2.0 398.644 398.657 399.723 399.726 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227756E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 210.169 210.169 qs_energies 1 2.0 0.000 0.000 209.844 209.904 scf_env_do_scf 1 3.0 0.000 0.000 106.417 106.417 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.532 105.540 rebuild_ks_matrix 4 6.0 0.000 0.000 105.530 105.538 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.530 105.538 hfx_ks_matrix 4 8.0 0.001 0.001 105.161 105.164 integrate_four_center 4 9.0 0.143 0.454 105.160 105.164 mp2_main 1 3.0 0.000 0.000 103.139 103.199 mp2_gpw_main 1 4.0 0.040 0.064 102.188 102.251 integrate_four_center_main 4 10.0 0.102 0.404 96.879 98.844 integrate_four_center_bin 266 11.0 96.777 98.837 96.777 98.837 init_scf_loop 1 4.0 0.000 0.000 91.787 91.787 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 74.983 76.150 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.588 55.768 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.153 42.122 47.378 integrate_v_rspace 95 8.0 0.394 0.559 28.473 33.511 pw_transfer 2240 10.6 0.146 0.167 29.836 30.346 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.861 29.319 grid_integrate_task_list 95 9.0 23.762 28.990 23.762 28.990 mp2_ri_gpw_compute_en 1 5.0 0.217 0.279 27.052 28.889 fft_wrap_pw1pw2_100 730 12.4 1.248 1.417 26.603 27.115 ao_to_mo_and_store_B_mult_1 91 7.0 10.782 27.047 10.782 27.047 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.912 25.065 25.074 get_2c_integrals 1 6.0 0.000 0.000 20.297 20.338 compute_2c_integrals 1 7.0 0.009 0.020 19.277 19.291 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.760 19.140 mp2_eri_2c_integrate_gpw 1 9.0 1.760 1.892 18.759 19.139 fft3d_s 1823 13.4 18.390 18.639 18.404 18.652 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.629 14.629 calculate_wavefunction 91 8.0 2.028 2.061 9.759 9.995 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.582 8.893 9.692 potential_pw2rs 186 10.0 0.033 0.035 8.595 9.225 local_gemm 172 8.0 8.338 9.133 8.338 9.133 mp2_ri_gpw_compute_en_comm 22 7.0 0.515 0.540 7.949 8.511 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.173 8.504 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.143 collocate_single_gaussian 91 10.0 0.017 0.021 7.836 8.092 mp_sendrecv_dm3 2068 8.0 5.943 6.524 5.943 6.524 mp2_ri_gpw_compute_en_ener 172 7.0 6.328 6.421 6.328 6.421 mp_sync 38 10.4 2.965 5.787 2.965 5.787 pw_gather_s 912 13.2 4.930 5.376 4.930 5.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.242884, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 449.835008E+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 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 0.012 0.029 54.604 54.605 qs_mol_dyn_low 1 2.0 0.003 0.003 54.160 54.168 qs_forces 11 3.9 0.002 0.002 54.099 54.100 qs_energies 11 4.9 0.003 0.003 52.625 52.644 scf_env_do_scf 11 5.9 0.000 0.001 44.874 44.874 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.705 42.706 dbcsr_multiply_generic 2286 12.5 0.092 0.098 35.232 35.812 qs_scf_new_mos 108 7.5 0.000 0.001 32.480 32.783 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.480 32.782 ot_scf_mini 108 9.5 0.002 0.002 30.790 30.966 multiply_cannon 2286 13.5 0.188 0.195 27.463 29.199 multiply_cannon_loop 2286 14.5 1.488 1.565 26.693 28.369 velocity_verlet 10 3.0 0.001 0.002 25.533 25.533 ot_mini 108 10.5 0.001 0.001 19.805 20.049 qs_ot_get_derivative 108 11.5 0.001 0.001 16.773 16.975 mp_waitall_1 267858 16.1 9.204 16.770 9.204 16.770 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.369 14.259 multiply_cannon_multrec 54864 15.5 4.206 6.520 8.418 12.161 rebuild_ks_matrix 119 8.3 0.000 0.000 8.002 8.140 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.002 8.140 multiply_cannon_sync_h2d 54864 15.5 5.896 7.922 5.896 7.922 mp_sum_l 7207 12.9 5.683 7.568 5.683 7.568 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.075 7.199 qs_ot_get_p 119 10.4 0.001 0.001 6.267 6.615 init_scf_run 11 5.9 0.000 0.001 6.117 6.118 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.117 6.117 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.909 6.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.408 5.861 dbcsr_mm_accdrv_process 76910 16.1 1.171 1.822 4.134 5.685 sum_up_and_integrate 119 10.3 0.013 0.015 4.584 4.591 integrate_v_rspace 119 11.3 0.002 0.003 4.571 4.579 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.207 4.331 calculate_rho_elec 119 8.7 0.011 0.016 4.206 4.331 calculate_first_density_matrix 1 7.0 0.000 0.000 3.956 3.961 jit_kernel_multiply 13 15.8 2.901 3.788 2.901 3.788 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.519 3.593 calculate_dm_sparse 119 9.5 0.000 0.000 3.305 3.454 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.333 3.335 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.849 3.169 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.908 3.124 apply_single 119 13.6 0.000 0.000 2.908 3.124 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.049 3.109 rs_pw_transfer 974 11.9 0.012 0.013 2.887 3.003 ot_diis_step 108 11.5 0.006 0.006 2.708 2.708 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.686 2.687 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.516 2.613 acc_transpose_blocks 54864 15.5 0.228 0.248 1.986 2.541 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.455 2.456 cp_fm_redistribute_end 50 14.0 2.223 2.425 2.228 2.427 density_rs2pw 119 9.7 0.004 0.004 2.304 2.404 cp_fm_diag_elpa_base 50 14.0 0.197 2.376 0.197 2.389 init_scf_loop 11 6.9 0.000 0.000 2.150 2.150 grid_integrate_task_list 119 12.3 2.031 2.124 2.031 2.124 wfi_extrapolate 11 7.9 0.001 0.001 2.098 2.098 mp_sum_d 4129 12.0 1.408 2.035 1.408 2.035 potential_pw2rs 119 12.3 0.004 0.005 1.934 1.951 pw_transfer 1439 11.6 0.052 0.058 1.764 1.842 make_m2s 4572 13.5 0.053 0.056 1.775 1.828 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.779 1.810 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.688 1.769 make_images 4572 14.5 0.133 0.139 1.693 1.746 fft3d_ps 1201 14.6 0.356 0.459 1.465 1.537 mp_alltoall_d11v 2130 13.8 1.355 1.515 1.355 1.515 mp_waitany 12084 13.8 1.275 1.475 1.275 1.475 acc_transpose_blocks_kernels 54864 16.5 0.251 0.383 0.991 1.379 fft_wrap_pw1pw2_140 487 13.2 0.081 0.095 1.297 1.375 grid_collocate_task_list 119 9.7 1.287 1.347 1.287 1.347 qs_energies_init_hamiltonians 11 5.9 0.390 0.394 1.195 1.196 dbcsr_dot_sd 1205 11.9 0.047 0.059 0.780 1.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.605000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 485.203968E+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 907179. 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 0.017 0.031 39.526 39.527 qs_mol_dyn_low 1 2.0 0.003 0.003 39.264 39.271 qs_forces 11 3.9 0.002 0.002 39.202 39.203 qs_energies 11 4.9 0.002 0.002 37.499 37.502 scf_env_do_scf 11 5.9 0.000 0.001 31.933 31.933 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.385 29.385 dbcsr_multiply_generic 2286 12.5 0.099 0.103 22.041 22.435 qs_scf_new_mos 108 7.5 0.001 0.001 20.198 20.438 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.198 20.438 ot_scf_mini 108 9.5 0.002 0.003 19.281 19.451 velocity_verlet 10 3.0 0.002 0.007 18.588 18.589 multiply_cannon 2286 13.5 0.208 0.220 16.885 18.583 multiply_cannon_loop 2286 14.5 0.902 0.982 15.717 17.296 ot_mini 108 10.5 0.001 0.001 11.989 12.229 mp_waitall_1 217478 16.2 6.478 11.789 6.478 11.789 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.373 9.804 qs_ot_get_derivative 108 11.5 0.001 0.001 9.502 9.677 multiply_cannon_multrec 27432 15.5 1.962 4.409 6.140 9.011 rebuild_ks_matrix 119 8.3 0.000 0.000 7.250 7.386 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.249 7.386 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.407 6.531 dbcsr_mm_accdrv_process 47894 16.0 3.314 5.370 4.109 6.016 qs_ot_get_p 119 10.4 0.001 0.001 4.344 4.574 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.690 4.493 init_scf_run 11 5.9 0.000 0.001 4.358 4.358 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.357 4.358 sum_up_and_integrate 119 10.3 0.024 0.027 4.274 4.280 integrate_v_rspace 119 11.3 0.002 0.003 4.250 4.257 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.024 4.138 apply_single 119 13.6 0.000 0.000 3.024 4.138 mp_sum_l 7207 12.9 2.098 4.069 2.098 4.069 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.735 3.798 calculate_rho_elec 119 8.7 0.021 0.024 3.735 3.797 make_m2s 4572 13.5 0.052 0.054 2.611 2.855 qs_ot_p2m_diag 50 11.0 0.009 0.012 2.826 2.848 rs_pw_transfer 974 11.9 0.010 0.011 2.700 2.796 calculate_first_density_matrix 1 7.0 0.000 0.000 2.793 2.795 make_images 4572 14.5 0.202 0.237 2.525 2.766 multiply_cannon_sync_h2d 27432 15.5 2.173 2.757 2.173 2.757 init_scf_loop 11 6.9 0.000 0.000 2.525 2.526 ot_diis_step 108 11.5 0.010 0.011 2.430 2.430 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.377 2.378 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.222 2.312 calculate_dm_sparse 119 9.5 0.000 0.001 2.228 2.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.237 2.239 density_rs2pw 119 9.7 0.004 0.004 2.092 2.180 potential_pw2rs 119 12.3 0.006 0.007 2.009 2.023 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.021 2.021 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.965 2.006 cp_fm_redistribute_end 50 14.0 1.665 1.994 1.670 1.996 cp_fm_diag_elpa_base 50 14.0 0.316 1.928 0.324 1.959 grid_integrate_task_list 119 12.3 1.827 1.942 1.827 1.942 pw_transfer 1439 11.6 0.063 0.067 1.855 1.892 jit_kernel_multiply 8 16.3 0.743 1.873 0.743 1.873 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.765 1.804 acc_transpose_blocks 27432 15.5 0.108 0.112 1.205 1.714 make_images_data 4572 15.5 0.045 0.051 1.190 1.656 prepare_preconditioner 11 7.9 0.000 0.000 1.626 1.653 make_preconditioner 11 8.9 0.000 0.000 1.626 1.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.520 1.579 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.032 1.515 wfi_extrapolate 11 7.9 0.001 0.001 1.513 1.513 fft3d_ps 1201 14.6 0.497 0.551 1.472 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.431 1.439 mp_allgather_i34 2286 14.5 0.609 1.380 0.609 1.380 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.343 1.379 mp_alltoall_d11v 2130 13.8 1.258 1.365 1.258 1.365 grid_collocate_task_list 119 9.7 1.241 1.315 1.241 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.214 1.261 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.670 1.089 mp_sum_d 4129 12.0 0.586 1.032 0.586 1.032 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.933 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.898 0.911 make_images_sizes 4572 15.5 0.005 0.005 0.605 0.866 mp_alltoall_i44 4572 16.5 0.600 0.861 0.600 0.861 mp_alltoall_z22v 1201 16.6 0.738 0.824 0.738 0.824 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.611 0.778 0.822 jit_kernel_transpose 5 15.5 0.488 0.817 0.488 0.817 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.527000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.090909, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 521.007104E+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 0.185 0.275 36.981 36.982 qs_mol_dyn_low 1 2.0 0.004 0.007 36.240 36.247 qs_forces 11 3.9 0.008 0.057 36.131 36.145 qs_energies 11 4.9 0.033 0.156 34.068 34.444 scf_env_do_scf 11 5.9 0.001 0.001 27.306 27.306 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.548 24.552 dbcsr_multiply_generic 2286 12.5 0.095 0.117 18.857 19.056 qs_scf_new_mos 108 7.5 0.001 0.001 16.352 16.367 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.352 16.366 velocity_verlet 10 3.0 0.016 0.035 15.948 15.949 multiply_cannon 2286 13.5 0.197 0.205 14.748 15.689 ot_scf_mini 108 9.5 0.002 0.003 15.604 15.619 multiply_cannon_loop 2286 14.5 0.639 0.670 13.902 14.884 ot_mini 108 10.5 0.046 0.360 9.915 9.934 multiply_cannon_multrec 18288 15.5 2.086 3.178 8.215 8.546 qs_ot_get_derivative 108 11.5 0.001 0.001 8.011 8.026 dbcsr_mm_accdrv_process 38222 16.0 4.667 6.887 6.045 7.011 rebuild_ks_matrix 119 8.3 0.000 0.000 6.442 6.472 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.442 6.472 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.708 5.730 init_scf_run 11 5.9 0.000 0.001 5.104 5.104 scf_env_initial_rho_setup 11 6.9 0.016 0.125 5.103 5.104 mp_waitall_1 169478 16.3 3.271 4.556 3.271 4.556 sum_up_and_integrate 119 10.3 0.030 0.032 4.042 4.045 integrate_v_rspace 119 11.3 0.002 0.003 4.012 4.018 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.904 3.659 calculate_first_density_matrix 1 7.0 0.000 0.001 3.628 3.629 qs_rho_update_rho_low 119 7.7 0.001 0.003 3.503 3.527 calculate_rho_elec 119 8.7 0.030 0.031 3.503 3.526 qs_ot_get_p 119 10.4 0.001 0.001 3.468 3.493 jit_kernel_multiply 11 15.7 1.327 3.089 1.327 3.089 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.501 2.928 apply_single 119 13.6 0.000 0.000 2.501 2.928 init_scf_loop 11 6.9 0.000 0.002 2.740 2.747 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.570 2.707 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.652 2.660 rs_pw_transfer 974 11.9 0.009 0.010 2.506 2.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.367 2.379 calculate_dm_sparse 119 9.5 0.000 0.000 2.315 2.330 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.242 2.249 mp_sum_l 7207 12.9 1.762 2.216 1.762 2.216 make_m2s 4572 13.5 0.045 0.046 1.969 2.131 density_rs2pw 119 9.7 0.004 0.004 1.959 2.074 make_images 4572 14.5 0.192 0.205 1.884 2.046 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.969 1.970 prepare_preconditioner 11 7.9 0.000 0.000 1.906 1.908 make_preconditioner 11 8.9 0.000 0.001 1.905 1.908 grid_integrate_task_list 119 12.3 1.794 1.900 1.794 1.900 ot_diis_step 108 11.5 0.012 0.016 1.827 1.872 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.762 1.842 potential_pw2rs 119 12.3 0.007 0.008 1.809 1.819 pw_transfer 1439 11.6 0.063 0.066 1.777 1.787 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.714 1.718 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.687 1.699 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.687 1.688 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.664 1.669 cp_fm_redistribute_end 50 14.0 1.251 1.659 1.253 1.660 cp_fm_diag_elpa_base 50 14.0 0.389 1.580 0.405 1.624 multiply_cannon_sync_h2d 18288 15.5 1.370 1.565 1.370 1.565 fft3d_ps 1201 14.6 0.504 0.518 1.374 1.386 acc_transpose_blocks 18288 15.5 0.075 0.076 1.307 1.325 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.304 1.313 grid_collocate_task_list 119 9.7 1.204 1.311 1.204 1.311 wfi_extrapolate 11 7.9 0.008 0.063 1.292 1.293 mp_sum_d 4129 12.0 0.943 1.190 0.943 1.190 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.146 1.151 mp_alltoall_d11v 2130 13.8 0.917 1.033 0.917 1.033 make_images_data 4572 15.5 0.045 0.049 0.841 1.017 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.962 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.737 0.945 cp_fm_cholesky_invert 11 10.9 0.873 0.878 0.873 0.878 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.863 0.877 calculate_ecore_overlap 22 5.9 0.000 0.000 0.676 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.814 mp_alltoall_z22v 1201 16.6 0.703 0.769 0.703 0.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.982000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.454545, yerr=2.310594 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 558.985216E+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 0.022 0.050 36.844 36.845 qs_mol_dyn_low 1 2.0 0.003 0.003 36.564 36.571 qs_forces 11 3.9 0.003 0.012 36.493 36.494 qs_energies 11 4.9 0.002 0.003 34.757 34.763 scf_env_do_scf 11 5.9 0.000 0.001 28.900 28.902 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.309 25.309 dbcsr_multiply_generic 2286 12.5 0.098 0.100 19.412 19.579 velocity_verlet 10 3.0 0.008 0.019 18.207 18.208 qs_scf_new_mos 108 7.5 0.001 0.001 16.962 17.010 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.961 17.009 ot_scf_mini 108 9.5 0.002 0.003 16.010 16.059 multiply_cannon 2286 13.5 0.237 0.273 15.545 15.926 multiply_cannon_loop 2286 14.5 0.936 0.967 14.501 14.918 ot_mini 108 10.5 0.001 0.001 9.844 9.903 multiply_cannon_multrec 27432 15.5 2.339 3.144 9.065 9.498 qs_ot_get_derivative 108 11.5 0.001 0.001 8.002 8.051 dbcsr_mm_accdrv_process 47916 15.9 5.695 7.472 6.631 7.859 rebuild_ks_matrix 119 8.3 0.000 0.000 6.607 6.653 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.606 6.653 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.876 5.916 init_scf_run 11 5.9 0.000 0.001 4.501 4.501 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.500 4.501 sum_up_and_integrate 119 10.3 0.035 0.038 3.874 3.882 integrate_v_rspace 119 11.3 0.002 0.003 3.839 3.848 qs_ot_get_p 119 10.4 0.001 0.001 3.520 3.590 init_scf_loop 11 6.9 0.000 0.001 3.568 3.571 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.386 3.422 calculate_rho_elec 119 8.7 0.040 0.046 3.385 3.421 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.885 3.270 mp_waitall_1 145218 16.4 2.604 3.244 2.604 3.244 calculate_first_density_matrix 1 7.0 0.000 0.000 3.064 3.066 prepare_preconditioner 11 7.9 0.000 0.000 2.702 2.708 make_preconditioner 11 8.9 0.000 0.001 2.702 2.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.300 2.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.519 2.520 make_m2s 4572 13.5 0.054 0.056 2.368 2.513 make_images 4572 14.5 0.274 0.334 2.260 2.403 jit_kernel_multiply 10 16.0 0.875 2.311 0.875 2.311 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.294 2.304 rs_pw_transfer 974 11.9 0.009 0.009 2.191 2.292 calculate_dm_sparse 119 9.5 0.000 0.000 2.180 2.233 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.052 2.182 apply_single 119 13.6 0.000 0.000 2.052 2.182 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.106 2.115 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.050 2.076 density_rs2pw 119 9.7 0.003 0.004 1.868 1.979 grid_integrate_task_list 119 12.3 1.825 1.910 1.825 1.910 pw_transfer 1439 11.6 0.063 0.066 1.792 1.824 ot_diis_step 108 11.5 0.012 0.013 1.798 1.799 acc_transpose_blocks 27432 15.5 0.111 0.115 1.770 1.793 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.778 1.778 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.702 1.737 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.958 1.713 potential_pw2rs 119 12.3 0.009 0.009 1.660 1.664 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.501 1.502 cp_fm_redistribute_end 50 14.0 0.988 1.470 0.990 1.471 cp_fm_diag_elpa_base 50 14.0 0.458 1.402 0.478 1.443 fft3d_ps 1201 14.6 0.529 0.584 1.383 1.406 wfi_extrapolate 11 7.9 0.001 0.001 1.387 1.387 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.327 1.363 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.349 1.360 mp_sum_l 7207 12.9 1.105 1.337 1.105 1.337 grid_collocate_task_list 119 9.7 1.221 1.313 1.221 1.313 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.144 1.161 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 1.140 1.161 cp_fm_upper_to_full 72 13.5 0.816 1.135 0.816 1.135 dbcsr_complete_redistribute 329 12.2 0.131 0.165 0.851 1.132 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.098 1.098 make_images_data 4572 15.5 0.045 0.049 0.911 1.062 multiply_cannon_sync_h2d 27432 15.5 0.985 1.045 0.985 1.045 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.780 0.994 mp_alltoall_d11v 2130 13.8 0.816 0.987 0.816 0.987 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.622 0.897 jit_kernel_transpose 5 15.6 0.871 0.897 0.871 0.897 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.875 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.272 0.864 cp_fm_cholesky_invert 11 10.9 0.839 0.842 0.839 0.842 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.827 mp_alltoall_z22v 1201 16.6 0.737 0.762 0.737 0.762 mp_alltoall_i22 627 13.8 0.452 0.746 0.452 0.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.845000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.090909, yerr=3.871916 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 602.673152E+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 0.133 0.431 29.018 29.019 qs_mol_dyn_low 1 2.0 0.003 0.003 28.369 28.377 qs_forces 11 3.9 0.002 0.002 28.310 28.312 qs_energies 11 4.9 0.001 0.002 26.618 26.621 scf_env_do_scf 11 5.9 0.000 0.001 21.482 21.482 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.950 18.951 velocity_verlet 10 3.0 0.002 0.002 14.451 14.454 dbcsr_multiply_generic 2286 12.5 0.091 0.093 12.768 12.862 qs_scf_new_mos 108 7.5 0.001 0.001 11.262 11.288 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.262 11.287 ot_scf_mini 108 9.5 0.002 0.002 10.566 10.586 multiply_cannon 2286 13.5 0.230 0.241 10.055 10.549 multiply_cannon_loop 2286 14.5 0.331 0.342 9.085 9.309 multiply_cannon_multrec 9144 15.5 1.575 1.854 6.085 6.313 rebuild_ks_matrix 119 8.3 0.000 0.000 5.878 5.911 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.877 5.911 ot_mini 108 10.5 0.001 0.001 5.880 5.909 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.254 5.285 qs_ot_get_derivative 108 11.5 0.001 0.001 4.571 4.593 dbcsr_mm_accdrv_process 12550 15.8 3.377 4.496 4.407 4.559 init_scf_run 11 5.9 0.000 0.001 3.706 3.706 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.706 3.706 sum_up_and_integrate 119 10.3 0.038 0.041 3.619 3.623 integrate_v_rspace 119 11.3 0.003 0.003 3.581 3.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.263 3.273 calculate_rho_elec 119 8.7 0.059 0.061 3.263 3.272 qs_ot_get_p 119 10.4 0.001 0.001 2.865 2.906 calculate_first_density_matrix 1 7.0 0.000 0.000 2.547 2.549 init_scf_loop 11 6.9 0.000 0.000 2.510 2.512 mp_waitall_1 121218 16.5 1.873 2.447 1.873 2.447 make_m2s 4572 13.5 0.034 0.035 1.856 2.078 jit_kernel_multiply 10 15.8 0.993 2.077 0.993 2.077 make_images 4572 14.5 0.269 0.306 1.768 1.988 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.980 1.982 grid_integrate_task_list 119 12.3 1.851 1.906 1.851 1.906 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.889 1.892 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.803 1.811 rs_pw_transfer 974 11.9 0.008 0.009 1.726 1.794 prepare_preconditioner 11 7.9 0.000 0.000 1.779 1.785 make_preconditioner 11 8.9 0.000 0.000 1.779 1.785 calculate_dm_sparse 119 9.5 0.000 0.000 1.761 1.780 pw_transfer 1439 11.6 0.063 0.066 1.747 1.755 density_rs2pw 119 9.7 0.003 0.004 1.660 1.726 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.691 1.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.655 1.691 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.657 1.664 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.615 1.629 potential_pw2rs 119 12.3 0.010 0.011 1.404 1.411 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.393 1.393 cp_fm_redistribute_end 50 14.0 0.694 1.368 0.695 1.369 cp_fm_diag_elpa_base 50 14.0 0.627 1.289 0.672 1.346 grid_collocate_task_list 119 9.7 1.273 1.332 1.273 1.332 fft_wrap_pw1pw2_140 487 13.2 0.083 0.088 1.324 1.332 fft3d_ps 1201 14.6 0.539 0.552 1.322 1.329 ot_diis_step 108 11.5 0.012 0.013 1.296 1.296 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.266 1.282 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.218 1.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.169 1.189 apply_single 119 13.6 0.000 0.000 1.169 1.189 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.159 1.164 make_images_data 4572 15.5 0.038 0.042 0.873 1.117 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.816 1.109 wfi_extrapolate 11 7.9 0.001 0.001 1.102 1.103 acc_transpose_blocks 9144 15.5 0.039 0.040 0.958 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.923 mp_alltoall_d11v 2130 13.8 0.839 0.909 0.839 0.909 cp_fm_cholesky_invert 11 10.9 0.903 0.906 0.903 0.906 mp_sum_l 7207 12.9 0.721 0.853 0.721 0.853 multiply_cannon_sync_h2d 9144 15.5 0.709 0.780 0.709 0.780 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.765 0.772 mp_allgather_i34 2286 14.5 0.302 0.754 0.302 0.754 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.748 0.751 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.715 0.722 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.672 0.720 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.345 0.704 mp_alltoall_z22v 1201 16.6 0.657 0.685 0.657 0.685 jit_kernel_transpose 5 15.6 0.598 0.604 0.598 0.604 dbcsr_complete_redistribute 329 12.2 0.162 0.169 0.556 0.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.019000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.818182, yerr=7.094405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 749.576192E+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 0.015 0.036 42.620 42.620 qs_mol_dyn_low 1 2.0 0.003 0.003 42.353 42.361 qs_forces 11 3.9 0.010 0.011 42.097 42.098 qs_energies 11 4.9 0.034 0.035 40.140 40.146 scf_env_do_scf 11 5.9 0.001 0.001 33.971 33.971 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.694 25.695 velocity_verlet 10 3.0 0.160 0.161 23.241 23.252 dbcsr_multiply_generic 2286 12.5 0.100 0.101 18.549 18.747 qs_scf_new_mos 108 7.5 0.001 0.001 16.666 16.767 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.665 16.766 ot_scf_mini 108 9.5 0.002 0.002 15.569 15.671 multiply_cannon 2286 13.5 0.305 0.313 14.337 15.465 multiply_cannon_loop 2286 14.5 0.345 0.351 13.060 14.182 ot_mini 108 10.5 0.001 0.001 9.524 9.644 multiply_cannon_multrec 9144 15.5 3.516 4.989 8.781 9.220 init_scf_loop 11 6.9 0.000 0.000 8.248 8.249 qs_ot_get_derivative 108 11.5 0.001 0.001 7.468 7.569 rebuild_ks_matrix 119 8.3 0.000 0.000 7.255 7.401 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.255 7.401 prepare_preconditioner 11 7.9 0.000 0.000 6.964 6.977 make_preconditioner 11 8.9 0.000 0.000 6.964 6.977 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.480 6.855 dbcsr_mm_accdrv_process 12550 15.8 4.151 5.627 5.141 6.852 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.601 6.734 cp_fm_upper_to_full 72 14.2 3.160 4.588 3.160 4.588 mp_waitall_1 97218 16.6 2.943 4.236 2.943 4.236 init_scf_run 11 5.9 0.000 0.001 4.113 4.113 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.112 4.112 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.752 3.755 calculate_rho_elec 119 8.7 0.118 0.121 3.751 3.754 sum_up_and_integrate 119 10.3 0.064 0.066 3.742 3.749 integrate_v_rspace 119 11.3 0.003 0.004 3.678 3.686 qs_ot_get_p 119 10.4 0.001 0.001 3.310 3.455 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.734 3.391 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.449 2.981 apply_single 119 13.6 0.000 0.000 2.449 2.981 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.565 2.859 dbcsr_complete_redistribute 329 12.2 0.289 0.293 2.045 2.849 calculate_first_density_matrix 1 7.0 0.000 0.000 2.678 2.683 make_m2s 4572 13.5 0.038 0.038 2.346 2.524 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.719 2.523 make_images 4572 14.5 0.354 0.386 2.227 2.404 calculate_dm_sparse 119 9.5 0.000 0.000 2.326 2.342 mp_alltoall_i22 627 13.8 1.501 2.309 1.501 2.309 mp_sum_l 7207 12.9 1.333 2.309 1.333 2.309 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.478 2.274 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.130 2.131 pw_transfer 1439 11.6 0.066 0.066 2.123 2.128 grid_integrate_task_list 119 12.3 2.015 2.041 2.015 2.041 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.027 2.034 ot_diis_step 108 11.5 0.014 0.014 2.028 2.029 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.013 2.014 density_rs2pw 119 9.7 0.003 0.003 1.844 1.873 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.820 1.831 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.701 1.755 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.751 1.754 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.736 1.737 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.672 1.679 fft3d_ps 1201 14.6 0.570 0.584 1.660 1.667 jit_kernel_multiply 6 15.6 0.963 1.550 0.963 1.550 cp_fm_cholesky_invert 11 10.9 1.529 1.533 1.529 1.533 grid_collocate_task_list 119 9.7 1.452 1.480 1.452 1.480 hybrid_alltoall_any 4725 16.4 0.087 0.150 1.202 1.459 rs_pw_transfer 974 11.9 0.009 0.009 1.426 1.457 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.425 1.426 cp_fm_diag_elpa_base 50 14.0 1.277 1.330 1.423 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.380 1.382 make_images_data 4572 15.5 0.042 0.046 1.150 1.381 wfi_extrapolate 11 7.9 0.001 0.001 1.371 1.371 potential_pw2rs 119 12.3 0.014 0.015 1.270 1.272 mp_alltoall_d11v 2130 13.8 1.166 1.182 1.166 1.182 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.147 1.169 acc_transpose_blocks 9144 15.5 0.039 0.039 1.128 1.150 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.086 1.108 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 0.989 1.029 mp_alltoall_z22v 1201 16.6 0.954 0.974 0.954 0.974 qs_create_task_list 11 7.9 0.000 0.000 0.944 0.956 generate_qs_task_list 11 8.9 0.374 0.393 0.944 0.955 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 0.876 0.897 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.869 0.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.620000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.636364, yerr=13.633333 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 499.658752E+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 1185660. 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 0.013 0.030 82.349 82.350 qs_mol_dyn_low 1 2.0 0.003 0.003 82.075 82.085 qs_forces 11 3.9 0.009 0.010 82.007 82.008 qs_energies 11 4.9 0.001 0.002 79.252 79.263 scf_env_do_scf 11 5.9 0.000 0.001 69.592 69.594 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.225 64.226 dbcsr_multiply_generic 2055 12.4 0.102 0.104 52.587 52.794 qs_scf_new_mos 99 7.5 0.000 0.001 47.872 47.963 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.871 47.962 ot_scf_mini 99 9.5 0.002 0.002 45.469 45.534 multiply_cannon 2055 13.4 0.181 0.188 43.095 44.158 multiply_cannon_loop 2055 14.4 1.558 1.596 42.184 43.254 velocity_verlet 10 3.0 0.001 0.002 41.415 41.415 ot_mini 99 10.5 0.001 0.001 27.894 27.959 qs_ot_get_derivative 99 11.5 0.001 0.001 21.080 21.137 multiply_cannon_multrec 49320 15.4 12.145 12.898 17.610 18.275 rebuild_ks_matrix 110 8.3 0.000 0.001 14.066 14.174 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 14.065 14.174 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.372 12.461 mp_waitall_1 241148 16.1 11.392 12.274 11.392 12.274 multiply_cannon_sync_h2d 49320 15.4 9.967 10.621 9.967 10.621 qs_ot_get_p 110 10.4 0.001 0.001 9.197 9.292 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.930 8.009 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.334 7.893 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.188 7.689 apply_single 110 13.6 0.000 0.001 7.187 7.689 init_scf_run 11 5.9 0.000 0.001 7.612 7.612 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.612 7.612 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.755 6.791 sum_up_and_integrate 110 10.3 0.038 0.044 6.754 6.771 integrate_v_rspace 110 11.3 0.003 0.003 6.717 6.742 ot_diis_step 99 11.5 0.005 0.005 6.559 6.559 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.015 6.048 mp_sum_l 6514 12.8 5.326 6.045 5.326 6.045 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.888 6.036 calculate_rho_elec 110 8.6 0.020 0.024 5.888 6.036 dbcsr_mm_accdrv_process 87628 16.1 2.070 2.172 5.344 5.646 init_scf_loop 11 6.9 0.000 0.000 5.341 5.341 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.094 5.095 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.634 4.635 cp_fm_redistribute_end 48 14.0 4.039 4.612 4.044 4.613 cp_fm_diag_elpa_base 48 14.0 0.563 4.516 0.566 4.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.196 4.200 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.890 3.928 wfi_extrapolate 11 7.9 0.001 0.001 3.918 3.918 calculate_dm_sparse 110 9.5 0.000 0.001 3.760 3.848 make_m2s 4110 13.4 0.061 0.065 3.718 3.819 make_images 4110 14.4 0.179 0.192 3.623 3.728 rs_pw_transfer 902 11.9 0.013 0.014 3.543 3.694 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.617 3.629 calculate_first_density_matrix 1 7.0 0.000 0.000 3.605 3.614 grid_integrate_task_list 110 12.3 3.243 3.397 3.243 3.397 prepare_preconditioner 11 7.9 0.000 0.000 3.234 3.252 make_preconditioner 11 8.9 0.000 0.000 3.234 3.252 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.147 3.196 density_rs2pw 110 9.6 0.004 0.004 3.019 3.162 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.034 3.086 jit_kernel_multiply 13 15.9 2.992 3.005 2.992 3.005 pw_transfer 1331 11.6 0.054 0.061 2.575 2.628 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.488 2.543 acc_transpose_blocks 49320 15.4 0.218 0.228 2.435 2.509 potential_pw2rs 110 12.3 0.007 0.007 2.427 2.445 mp_alltoall_d11v 2046 13.8 1.922 2.392 1.922 2.392 grid_collocate_task_list 110 9.6 2.083 2.197 2.083 2.197 fft_wrap_pw1pw2_140 451 13.1 0.170 0.190 2.102 2.157 fft3d_ps 1111 14.6 0.745 0.836 2.024 2.070 mp_waitany 14300 13.8 1.768 1.979 1.768 1.979 mp_sum_d 3883 11.9 1.405 1.863 1.405 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.833 1.856 cp_fm_cholesky_invert 11 10.9 1.794 1.797 1.794 1.797 make_images_data 4110 15.4 0.044 0.047 1.679 1.795 hybrid_alltoall_any 4261 16.3 0.083 0.480 1.467 1.692 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.350000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.727273, yerr=2.178027 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 584.179712E+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 1935295. 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 0.227 0.652 70.071 70.074 qs_mol_dyn_low 1 2.0 0.003 0.005 69.111 69.120 qs_forces 11 3.9 0.005 0.023 69.015 69.016 qs_energies 11 4.9 0.002 0.004 65.686 65.690 scf_env_do_scf 11 5.9 0.001 0.001 56.063 56.066 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.437 48.440 dbcsr_multiply_generic 2055 12.4 0.111 0.115 38.239 38.416 velocity_verlet 10 3.0 0.006 0.033 36.011 36.012 qs_scf_new_mos 99 7.5 0.001 0.001 32.755 32.887 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.755 32.887 multiply_cannon 2055 13.4 0.234 0.263 31.682 32.829 multiply_cannon_loop 2055 14.4 0.931 0.953 30.459 31.570 ot_scf_mini 99 9.5 0.003 0.003 31.101 31.219 ot_mini 99 10.5 0.001 0.001 18.407 18.534 multiply_cannon_multrec 24660 15.4 7.674 9.610 14.267 16.348 rebuild_ks_matrix 110 8.3 0.000 0.001 13.555 13.705 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 13.555 13.705 qs_ot_get_derivative 99 11.5 0.001 0.001 12.582 12.706 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.982 12.121 mp_waitall_1 186928 16.3 7.979 10.528 7.979 10.528 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.389 8.122 multiply_cannon_sync_h2d 24660 15.4 7.074 8.119 7.074 8.119 init_scf_loop 11 6.9 0.000 0.001 7.590 7.593 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.539 7.257 apply_single 110 13.6 0.000 0.001 6.538 7.257 init_scf_run 11 5.9 0.000 0.001 6.916 6.917 scf_env_initial_rho_setup 11 6.9 0.002 0.015 6.916 6.917 dbcsr_mm_accdrv_process 52282 16.1 4.704 6.003 6.433 6.761 sum_up_and_integrate 110 10.3 0.053 0.060 6.464 6.477 integrate_v_rspace 110 11.3 0.002 0.003 6.411 6.426 qs_ot_get_p 110 10.4 0.001 0.001 6.085 6.219 ot_diis_step 99 11.5 0.010 0.011 5.772 5.772 qs_rho_update_rho_low 110 7.6 0.002 0.010 5.692 5.707 calculate_rho_elec 110 8.6 0.039 0.047 5.690 5.706 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.923 5.659 prepare_preconditioner 11 7.9 0.000 0.000 5.521 5.535 make_preconditioner 11 8.9 0.000 0.001 5.521 5.535 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.114 5.266 make_m2s 4110 13.4 0.056 0.059 4.151 4.566 make_images 4110 14.4 0.403 0.450 4.047 4.460 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.215 4.234 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.782 3.789 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.718 3.719 wfi_extrapolate 11 7.9 0.006 0.045 3.519 3.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.344 3.396 pw_transfer 1331 11.6 0.066 0.071 3.243 3.380 grid_integrate_task_list 110 12.3 3.145 3.306 3.145 3.306 calculate_first_density_matrix 1 7.0 0.000 0.001 3.290 3.294 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.136 3.273 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.246 3.248 density_rs2pw 110 9.6 0.004 0.004 2.982 3.229 cp_fm_redistribute_end 48 14.0 2.421 3.222 2.424 3.223 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.141 3.214 cp_fm_diag_elpa_base 48 14.0 0.763 3.099 0.795 3.188 rs_pw_transfer 902 11.9 0.013 0.014 2.909 3.114 calculate_dm_sparse 110 9.5 0.001 0.001 2.957 2.991 jit_kernel_multiply 13 16.3 1.385 2.903 1.385 2.903 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 2.665 2.800 make_images_data 4110 15.4 0.047 0.051 2.345 2.754 hybrid_alltoall_any 4261 16.3 0.103 0.450 2.049 2.667 fft3d_ps 1111 14.6 1.072 1.272 2.492 2.615 cp_fm_cholesky_invert 11 10.9 2.573 2.580 2.573 2.580 mp_sum_l 6514 12.8 1.856 2.573 1.856 2.573 potential_pw2rs 110 12.3 0.009 0.009 2.344 2.359 grid_collocate_task_list 110 9.6 2.119 2.267 2.119 2.267 qs_energies_init_hamiltonians 11 5.9 0.003 0.012 2.187 2.188 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.952 1.974 mp_alltoall_d11v 2046 13.8 1.745 1.960 1.745 1.960 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.778 1.883 acc_transpose_blocks 24660 15.4 0.108 0.110 1.599 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.763 1.774 mp_irecv_dv 57340 16.2 0.655 1.672 0.655 1.672 mp_sum_d 3883 11.9 1.136 1.619 1.136 1.619 mp_allgather_i34 2055 14.4 0.537 1.553 0.537 1.553 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.551 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.380 1.484 dbcsr_complete_redistribute 325 12.2 0.268 0.328 1.184 1.449 mp_waitany 10164 13.8 1.199 1.422 1.199 1.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.074000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=551.727273, yerr=7.161652 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 660.041728E+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 0.037 0.097 59.664 59.664 qs_mol_dyn_low 1 2.0 0.003 0.005 59.290 59.299 qs_forces 11 3.9 0.002 0.004 59.222 59.223 qs_energies 11 4.9 0.002 0.005 56.074 56.078 scf_env_do_scf 11 5.9 0.000 0.001 47.711 47.711 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.256 39.256 velocity_verlet 10 3.0 0.002 0.003 32.057 32.059 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.478 28.747 qs_scf_new_mos 99 7.5 0.001 0.001 24.792 24.882 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.792 24.881 ot_scf_mini 99 9.5 0.002 0.003 23.577 23.701 multiply_cannon 2055 13.4 0.213 0.226 22.256 23.493 multiply_cannon_loop 2055 14.4 0.619 0.634 21.045 22.087 ot_mini 99 10.5 0.001 0.001 13.410 13.535 rebuild_ks_matrix 110 8.3 0.000 0.000 12.154 12.316 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.154 12.315 multiply_cannon_multrec 16440 15.4 3.990 5.397 9.920 11.217 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.721 10.868 mp_waitall_1 146766 16.3 6.985 10.226 6.985 10.226 qs_ot_get_derivative 99 11.5 0.030 0.460 9.008 9.134 init_scf_loop 11 6.9 0.000 0.000 8.420 8.420 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.146 7.290 prepare_preconditioner 11 7.9 0.000 0.000 6.684 6.699 make_preconditioner 11 8.9 0.000 0.000 6.684 6.699 sum_up_and_integrate 110 10.3 0.060 0.061 6.402 6.416 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.378 integrate_v_rspace 110 11.3 0.003 0.003 6.341 6.355 dbcsr_mm_accdrv_process 34862 16.1 4.685 5.451 5.783 5.951 init_scf_run 11 5.9 0.000 0.001 5.927 5.927 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.926 5.927 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.499 5.508 calculate_rho_elec 110 8.6 0.058 0.058 5.499 5.508 qs_ot_get_p 110 10.4 0.001 0.001 5.359 5.481 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.910 5.379 apply_single 110 13.6 0.000 0.000 4.910 5.379 make_m2s 4110 13.4 0.050 0.052 4.175 4.558 multiply_cannon_sync_h2d 16440 15.4 3.708 4.458 3.708 4.458 make_images 4110 14.4 0.395 0.514 4.061 4.444 ot_diis_step 99 11.5 0.011 0.012 4.372 4.372 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.780 3.784 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.073 3.713 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.446 3.447 grid_integrate_task_list 110 12.3 3.192 3.365 3.192 3.365 pw_transfer 1331 11.6 0.065 0.072 3.158 3.165 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.144 3.145 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.051 3.061 density_rs2pw 110 9.6 0.004 0.004 2.775 3.049 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.940 2.941 make_images_data 4110 15.4 0.044 0.047 2.453 2.941 rs_pw_transfer 902 11.9 0.011 0.012 2.660 2.928 cp_fm_redistribute_end 48 14.0 1.828 2.915 1.830 2.916 calculate_first_density_matrix 1 7.0 0.000 0.001 2.894 2.896 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.166 2.894 cp_fm_diag_elpa_base 48 14.0 1.022 2.775 1.081 2.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.696 2.750 cp_fm_cholesky_invert 11 10.9 2.632 2.637 2.632 2.637 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.613 2.626 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.553 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.320 2.390 fft3d_ps 1111 14.6 1.064 1.076 2.365 2.374 grid_collocate_task_list 110 9.6 2.169 2.364 2.169 2.364 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.868 2.352 mp_irecv_dv 48980 15.7 0.798 2.230 0.798 2.230 potential_pw2rs 110 12.3 0.011 0.011 2.214 2.221 mp_alltoall_d11v 2046 13.8 1.798 2.206 1.798 2.206 mp_sum_l 6514 12.8 1.521 2.142 1.521 2.142 qs_energies_init_hamiltonians 11 5.9 0.016 0.037 1.992 2.014 dbcsr_complete_redistribute 325 12.2 0.400 0.475 1.503 1.955 cp_fm_upper_to_full 70 13.6 1.390 1.867 1.390 1.867 jit_kernel_multiply 8 16.6 0.717 1.694 0.717 1.694 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.647 mp_allgather_i34 2055 14.4 0.529 1.628 0.529 1.628 cp_fm_cholesky_decompose 22 10.9 1.567 1.601 1.567 1.601 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.088 1.546 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.486 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.475 mp_waitany 17072 13.8 1.165 1.440 1.165 1.440 acc_transpose_blocks 16440 15.4 0.074 0.076 1.274 1.336 rs_gather_matrices 110 12.3 0.139 0.151 0.877 1.281 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.249 1.255 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.664000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.636364, yerr=8.466756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 726.388736E+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 0.020 0.065 65.181 65.182 qs_mol_dyn_low 1 2.0 0.003 0.005 64.570 64.589 qs_forces 11 3.9 0.003 0.011 64.470 64.471 qs_energies 11 4.9 0.002 0.007 61.122 61.130 scf_env_do_scf 11 5.9 0.000 0.001 52.893 52.895 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.202 41.203 velocity_verlet 10 3.0 0.002 0.003 36.310 36.314 dbcsr_multiply_generic 2055 12.4 0.113 0.115 30.239 30.542 qs_scf_new_mos 99 7.5 0.001 0.001 26.945 27.043 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.944 27.043 ot_scf_mini 99 9.5 0.003 0.003 25.338 25.444 multiply_cannon 2055 13.4 0.244 0.264 23.036 24.217 multiply_cannon_loop 2055 14.4 0.881 0.904 21.633 22.267 ot_mini 99 10.5 0.057 0.454 14.994 15.106 multiply_cannon_multrec 24660 15.4 4.246 6.952 13.023 14.194 rebuild_ks_matrix 110 8.3 0.000 0.000 11.841 11.944 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 11.840 11.944 init_scf_loop 11 6.9 0.000 0.000 11.650 11.651 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.493 10.584 qs_ot_get_derivative 99 11.5 0.001 0.001 10.414 10.517 prepare_preconditioner 11 7.9 0.000 0.000 9.930 9.947 make_preconditioner 11 8.9 0.000 0.000 9.930 9.947 dbcsr_mm_accdrv_process 52304 16.0 6.937 8.444 8.630 9.663 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.111 9.619 mp_waitall_1 126806 16.4 4.569 6.650 4.569 6.650 sum_up_and_integrate 110 10.3 0.068 0.072 6.258 6.272 integrate_v_rspace 110 11.3 0.003 0.003 6.190 6.204 make_m2s 4110 13.4 0.059 0.062 5.255 5.610 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.526 5.535 calculate_rho_elec 110 8.6 0.077 0.081 5.525 5.534 qs_ot_get_p 110 10.4 0.001 0.001 5.355 5.491 make_images 4110 14.4 0.582 0.712 5.114 5.465 init_scf_run 11 5.9 0.000 0.001 5.440 5.441 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.440 5.441 cp_fm_upper_to_full 70 13.8 3.382 4.819 3.382 4.819 ot_diis_step 99 11.5 0.011 0.013 4.488 4.545 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.394 4.542 apply_single 110 13.6 0.000 0.000 4.393 4.541 dbcsr_complete_redistribute 325 12.2 0.417 0.471 2.686 3.865 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.641 3.690 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.632 3.646 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.586 3.402 grid_integrate_task_list 110 12.3 3.282 3.391 3.282 3.391 multiply_cannon_sync_h2d 24660 15.4 3.194 3.342 3.194 3.342 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.163 3.336 pw_transfer 1331 11.6 0.065 0.073 3.177 3.199 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.161 3.161 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.071 3.097 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.985 3.041 hybrid_alltoall_any 4261 16.3 0.121 0.457 2.222 3.020 make_images_data 4110 15.4 0.046 0.050 2.650 2.981 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.811 2.961 calculate_dm_sparse 110 9.5 0.001 0.001 2.905 2.934 mp_alltoall_i22 605 13.7 1.691 2.923 1.691 2.923 wfi_extrapolate 11 7.9 0.001 0.003 2.916 2.916 density_rs2pw 110 9.6 0.004 0.004 2.684 2.901 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.804 2.805 cp_fm_cholesky_invert 11 10.9 2.655 2.663 2.655 2.663 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.616 2.642 rs_pw_transfer 902 11.9 0.011 0.011 2.392 2.636 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.633 2.636 cp_fm_redistribute_end 48 14.0 1.313 2.609 1.315 2.610 cp_fm_diag_elpa_base 48 14.0 1.210 2.477 1.292 2.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.447 2.484 calculate_first_density_matrix 1 7.0 0.000 0.001 2.440 2.442 fft3d_ps 1111 14.6 1.063 1.093 2.372 2.385 grid_collocate_task_list 110 9.6 2.224 2.372 2.224 2.372 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 2.286 2.287 jit_kernel_multiply 10 15.7 1.365 2.259 1.365 2.259 potential_pw2rs 110 12.3 0.012 0.013 2.050 2.059 mp_alltoall_d11v 2046 13.8 1.762 1.930 1.762 1.930 mp_sum_l 6514 12.8 1.282 1.929 1.282 1.929 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.679 1.713 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.603 1.706 cp_fm_cholesky_decompose 22 10.9 1.603 1.661 1.603 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.594 1.606 acc_transpose_blocks 24660 15.4 0.105 0.107 1.527 1.555 mp_allgather_i34 2055 14.4 0.493 1.552 0.493 1.552 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.546 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.847 1.451 mp_irecv_dv 62702 16.1 0.748 1.377 0.748 1.377 mp_waitany 13376 13.8 1.075 1.369 1.075 1.369 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.238 1.343 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.182000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.181818, yerr=8.798948 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 842.711040E+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 0.018 0.076 54.502 54.503 qs_mol_dyn_low 1 2.0 0.003 0.003 54.192 54.201 qs_forces 11 3.9 0.002 0.002 54.126 54.126 qs_energies 11 4.9 0.001 0.002 50.527 50.532 scf_env_do_scf 11 5.9 0.000 0.001 42.321 42.321 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.233 34.233 velocity_verlet 10 3.0 0.002 0.002 30.323 30.326 dbcsr_multiply_generic 2055 12.4 0.103 0.104 22.676 22.819 qs_scf_new_mos 99 7.5 0.001 0.001 20.243 20.285 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.242 20.284 ot_scf_mini 99 9.5 0.002 0.002 19.013 19.031 multiply_cannon 2055 13.4 0.248 0.265 17.289 18.708 multiply_cannon_loop 2055 14.4 0.323 0.337 15.933 16.263 rebuild_ks_matrix 110 8.3 0.000 0.000 11.831 11.845 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.831 11.844 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.594 10.607 ot_mini 99 10.5 0.001 0.001 10.297 10.308 multiply_cannon_multrec 8220 15.4 3.233 4.492 7.433 8.380 init_scf_loop 11 6.9 0.000 0.000 8.041 8.042 mp_waitall_1 106626 16.5 6.098 7.645 6.098 7.645 qs_ot_get_derivative 99 11.5 0.001 0.001 6.532 6.547 sum_up_and_integrate 110 10.3 0.080 0.082 6.135 6.149 integrate_v_rspace 110 11.3 0.003 0.003 6.054 6.069 prepare_preconditioner 11 7.9 0.000 0.000 5.979 5.983 make_preconditioner 11 8.9 0.000 0.000 5.979 5.983 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.566 5.634 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.541 5.557 calculate_rho_elec 110 8.6 0.115 0.115 5.541 5.556 init_scf_run 11 5.9 0.000 0.001 5.094 5.094 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.094 5.094 dbcsr_mm_accdrv_process 17442 15.9 2.808 3.773 4.070 4.968 qs_ot_get_p 110 10.4 0.001 0.001 4.807 4.832 make_m2s 4110 13.4 0.039 0.040 4.168 4.444 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.006 4.412 make_images 4110 14.4 0.641 0.693 4.040 4.315 ot_diis_step 99 11.5 0.012 0.012 3.745 3.745 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.662 3.674 apply_single 110 13.6 0.000 0.000 3.662 3.674 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.500 3.503 grid_integrate_task_list 110 12.3 3.369 3.484 3.369 3.484 pw_transfer 1331 11.6 0.065 0.070 3.203 3.214 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.201 3.201 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.096 3.112 multiply_cannon_sync_h2d 8220 15.4 2.919 3.020 2.919 3.020 cp_fm_cholesky_invert 11 10.9 2.870 2.874 2.870 2.874 make_images_data 4110 15.4 0.038 0.042 2.387 2.805 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.320 2.744 density_rs2pw 110 9.6 0.004 0.004 2.601 2.699 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.694 2.695 wfi_extrapolate 11 7.9 0.001 0.001 2.688 2.688 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.660 2.681 cp_fm_redistribute_end 48 14.0 0.683 2.670 0.688 2.672 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.658 2.658 cp_fm_diag_elpa_base 48 14.0 1.799 2.477 1.976 2.640 calculate_dm_sparse 110 9.5 0.001 0.001 2.469 2.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.498 2.499 grid_collocate_task_list 110 9.6 2.328 2.497 2.328 2.497 fft3d_ps 1111 14.6 1.117 1.149 2.352 2.360 calculate_first_density_matrix 1 7.0 0.000 0.000 2.322 2.322 rs_pw_transfer 902 11.9 0.010 0.011 2.141 2.274 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.152 2.164 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.995 potential_pw2rs 110 12.3 0.015 0.016 1.901 1.905 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.875 1.888 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.771 1.780 jit_kernel_multiply 10 15.5 0.958 1.746 0.958 1.746 mp_alltoall_d11v 2046 13.8 1.600 1.710 1.600 1.710 cp_fm_cholesky_decompose 22 10.9 1.660 1.684 1.660 1.684 mp_allgather_i34 2055 14.4 0.538 1.634 0.538 1.634 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.500 1.620 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.570 1.574 dbcsr_complete_redistribute 325 12.2 0.564 0.594 1.427 1.511 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.445 1.458 qs_create_task_list 11 7.9 0.000 0.001 1.224 1.326 generate_qs_task_list 11 8.9 0.379 0.447 1.224 1.325 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.778 1.216 mp_waitany 9240 13.8 1.053 1.191 1.053 1.191 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.122 1.138 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.481 1.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.503000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=789.909091, yerr=14.487955 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 1.357894E+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 0.023 0.068 86.393 86.394 qs_mol_dyn_low 1 2.0 0.003 0.003 86.038 86.047 qs_forces 11 3.9 0.002 0.002 85.689 85.690 qs_energies 11 4.9 0.002 0.002 81.679 81.681 scf_env_do_scf 11 5.9 0.001 0.001 71.414 71.414 velocity_verlet 10 3.0 0.290 0.291 55.314 55.320 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.722 42.723 dbcsr_multiply_generic 2055 12.4 0.123 0.125 29.040 29.196 init_scf_loop 11 6.9 0.000 0.000 28.617 28.621 prepare_preconditioner 11 7.9 0.000 0.000 26.656 26.668 make_preconditioner 11 8.9 0.000 0.000 26.656 26.668 qs_scf_new_mos 99 7.5 0.001 0.001 26.432 26.504 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.431 26.504 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.806 26.114 ot_scf_mini 99 9.5 0.002 0.002 24.648 24.703 multiply_cannon 2055 13.4 0.356 0.386 21.817 22.519 multiply_cannon_loop 2055 14.4 0.344 0.354 19.953 20.325 cp_fm_upper_to_full 70 14.2 12.889 18.444 12.889 18.444 ot_mini 99 10.5 0.001 0.001 13.702 13.763 rebuild_ks_matrix 110 8.3 0.001 0.001 13.126 13.199 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.126 13.199 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.941 12.007 dbcsr_complete_redistribute 325 12.2 1.027 1.057 7.441 10.754 multiply_cannon_multrec 8220 15.4 4.377 4.568 9.617 9.728 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.407 9.720 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.835 9.129 qs_ot_get_derivative 99 11.5 0.001 0.001 9.072 9.129 mp_waitall_1 87304 16.6 7.765 8.858 7.765 8.858 mp_alltoall_i22 605 13.7 5.470 8.759 5.470 8.759 sum_up_and_integrate 110 10.3 0.152 0.153 6.602 6.617 integrate_v_rspace 110 11.3 0.004 0.004 6.450 6.466 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.381 6.415 calculate_rho_elec 110 8.6 0.227 0.227 6.380 6.415 init_scf_run 11 5.9 0.000 0.001 6.036 6.036 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.036 6.036 make_m2s 4110 13.4 0.043 0.044 5.340 5.918 qs_ot_get_p 110 10.4 0.001 0.001 5.725 5.802 make_images 4110 14.4 0.885 0.929 5.153 5.731 cp_fm_cholesky_invert 11 10.9 5.365 5.370 5.365 5.370 dbcsr_mm_accdrv_process 11614 15.7 3.311 3.682 5.097 5.329 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.704 5.238 apply_single 110 13.6 0.000 0.000 4.704 5.238 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.714 5.035 ot_diis_step 99 11.5 0.016 0.016 4.605 4.605 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.105 4.113 multiply_cannon_sync_h2d 8220 15.4 3.946 3.950 3.946 3.950 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.940 3.723 grid_integrate_task_list 110 12.3 3.655 3.718 3.655 3.718 pw_transfer 1331 11.6 0.073 0.074 3.695 3.700 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.662 3.662 make_images_data 4110 15.4 0.041 0.044 2.938 3.636 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.632 3.633 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.579 3.584 calculate_dm_sparse 110 9.5 0.001 0.001 3.447 3.465 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.895 3.359 wfi_extrapolate 11 7.9 0.001 0.001 3.222 3.222 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.097 3.101 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.056 3.057 cp_fm_diag_elpa_base 48 14.0 2.520 2.717 3.054 3.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.923 2.926 density_rs2pw 110 9.6 0.004 0.004 2.885 2.901 fft3d_ps 1111 14.6 1.262 1.267 2.787 2.792 calculate_first_density_matrix 1 7.0 0.000 0.000 2.711 2.711 grid_collocate_task_list 110 9.6 2.625 2.647 2.625 2.647 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.426 2.459 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.428 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.239 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.158 rs_pw_transfer 902 11.9 0.011 0.011 2.090 2.135 mp_alltoall_d11v 2046 13.8 2.039 2.087 2.039 2.087 potential_pw2rs 110 12.3 0.021 0.021 2.048 2.051 cp_fm_cholesky_decompose 22 10.9 2.016 2.030 2.016 2.030 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.963 1.978 qs_create_task_list 11 7.9 0.001 0.001 1.904 1.949 generate_qs_task_list 11 8.9 0.741 0.796 1.903 1.949 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.796 1.828 jit_kernel_multiply 10 15.3 1.586 1.815 1.586 1.815 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.799 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.394000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1218.272727, yerr=59.671968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 630.583296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+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 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 56839. MP_Allreduce 11238 786. MP_Sync 168 MP_Alltoall 2210 1965128. 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 0.035 0.097 203.275 203.277 qs_mol_dyn_low 1 2.0 0.013 0.052 202.679 202.689 qs_forces 11 3.9 0.008 0.041 202.532 202.534 qs_energies 11 4.9 0.010 0.020 196.999 197.018 scf_env_do_scf 11 5.9 0.001 0.002 179.404 179.409 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 158.770 158.772 dbcsr_multiply_generic 2485 12.5 0.170 0.175 124.416 125.261 velocity_verlet 10 3.0 0.004 0.012 120.978 120.982 qs_scf_new_mos 116 7.6 0.001 0.001 120.359 120.683 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.358 120.682 ot_scf_mini 116 9.6 0.003 0.004 113.833 114.114 multiply_cannon 2485 13.5 0.252 0.279 100.941 103.597 multiply_cannon_loop 2485 14.5 2.082 2.152 98.741 100.965 ot_mini 116 10.6 0.001 0.002 65.481 65.851 multiply_cannon_multrec 59640 15.5 32.920 35.697 41.628 43.457 qs_ot_get_derivative 116 11.6 0.001 0.002 40.661 40.979 rebuild_ks_matrix 127 8.3 0.001 0.001 33.032 33.537 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 33.031 33.537 mp_waitall_1 288940 16.2 28.635 31.401 28.635 31.401 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.707 30.184 multiply_cannon_sync_h2d 59640 15.5 27.173 29.964 27.173 29.964 qs_ot_get_p 127 10.4 0.001 0.001 26.466 26.717 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.184 24.835 apply_single 127 13.6 0.001 0.001 24.184 24.835 ot_diis_step 116 11.6 0.054 0.194 24.473 24.611 init_scf_loop 11 6.9 0.000 0.000 20.558 20.560 qs_ot_p2m_diag 82 11.4 0.076 0.091 19.875 19.956 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.897 19.105 multiply_cannon_metrocomm3 59640 15.5 0.114 0.119 15.693 18.096 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.275 17.276 prepare_preconditioner 11 7.9 0.000 0.000 15.960 16.016 make_preconditioner 11 8.9 0.000 0.000 15.960 16.016 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.187 15.356 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.474 14.479 cp_fm_redistribute_end 82 14.4 11.434 14.406 11.445 14.408 cp_fm_diag_elpa_base 82 14.4 2.921 14.112 2.952 14.230 make_m2s 4970 13.5 0.102 0.111 13.883 14.187 make_images 4970 14.5 0.408 0.429 13.711 14.025 sum_up_and_integrate 127 10.3 0.091 0.110 13.886 13.902 integrate_v_rspace 127 11.3 0.004 0.006 13.795 13.812 init_scf_run 11 5.9 0.000 0.001 13.381 13.381 scf_env_initial_rho_setup 11 6.9 0.002 0.003 13.381 13.381 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.547 12.653 calculate_rho_elec 127 8.7 0.044 0.063 12.546 12.652 mp_sum_l 7804 13.0 8.376 10.240 8.376 10.240 wfi_extrapolate 11 7.9 0.017 0.065 9.150 9.151 cp_fm_cholesky_invert 11 10.9 9.054 9.062 9.054 9.062 dbcsr_mm_accdrv_process 123452 16.2 3.250 3.413 8.272 8.789 calculate_dm_sparse 127 9.5 0.001 0.001 8.376 8.470 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.012 8.164 multiply_cannon_metrocomm1 59640 15.5 0.087 0.092 5.947 8.008 make_images_data 4970 15.5 0.072 0.078 6.881 7.861 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.774 7.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.503 7.513 grid_integrate_task_list 127 12.3 7.040 7.380 7.040 7.380 hybrid_alltoall_any 5155 16.4 0.292 2.251 5.996 7.345 pw_transfer 1535 11.6 0.075 0.108 6.782 7.065 density_rs2pw 127 9.7 0.006 0.007 6.426 6.954 fft_wrap_pw1pw2 1281 12.7 0.010 0.014 6.579 6.836 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.571 6.647 rs_pw_transfer 1038 11.9 0.018 0.020 5.681 6.335 fft_wrap_pw1pw2_140 519 13.2 0.443 0.509 5.656 5.830 fft3d_ps 1281 14.7 2.078 2.548 5.376 5.578 mp_alltoall_d11v 2401 14.1 4.159 5.550 4.159 5.550 grid_collocate_task_list 127 9.7 4.688 5.096 4.688 5.096 cp_fm_cholesky_decompose 22 10.9 4.716 4.732 4.716 4.732 mp_sum_d 4443 12.1 3.646 4.567 3.646 4.567 potential_pw2rs 127 12.3 0.010 0.011 4.443 4.464 calculate_first_density_matrix 1 7.0 0.001 0.006 4.073 4.079 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.277000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.090909, yerr=7.166266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 832.950272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57228. MP_Allreduce 11247 947. MP_Sync 168 MP_Alltoall 1955 5502961. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.136 190.175 190.177 qs_mol_dyn_low 1 2.0 0.004 0.011 189.634 189.646 qs_forces 11 3.9 0.013 0.030 189.535 189.536 qs_energies 11 4.9 0.002 0.008 182.788 182.809 scf_env_do_scf 11 5.9 0.001 0.001 165.885 165.895 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.717 132.728 velocity_verlet 10 3.0 0.002 0.005 118.155 118.157 dbcsr_multiply_generic 2485 12.5 0.182 0.196 97.538 98.728 qs_scf_new_mos 116 7.6 0.001 0.001 94.564 95.115 qs_scf_loop_do_ot 116 8.6 0.001 0.002 94.563 95.114 ot_scf_mini 116 9.6 0.004 0.005 89.799 90.297 multiply_cannon 2485 13.5 0.475 0.531 77.871 81.483 multiply_cannon_loop 2485 14.5 1.246 1.289 74.688 77.349 ot_mini 116 10.6 0.001 0.002 50.052 50.587 mp_waitall_1 224796 16.4 24.919 39.521 24.919 39.521 multiply_cannon_multrec 29820 15.5 22.003 26.419 31.659 36.342 init_scf_loop 11 6.9 0.001 0.004 33.067 33.080 rebuild_ks_matrix 127 8.3 0.001 0.001 32.462 33.039 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.022 32.462 33.038 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.219 29.758 multiply_cannon_metrocomm3 29820 15.5 0.095 0.101 16.012 29.246 qs_ot_get_derivative 116 11.6 0.001 0.002 28.246 28.736 prepare_preconditioner 11 7.9 0.000 0.000 28.538 28.604 make_preconditioner 11 8.9 0.000 0.002 28.538 28.604 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.233 27.780 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.868 23.000 apply_single 127 13.6 0.001 0.001 21.868 22.999 qs_ot_get_p 127 10.4 0.001 0.002 21.853 22.463 multiply_cannon_sync_h2d 29820 15.5 19.251 21.724 19.251 21.724 ot_diis_step 116 11.6 0.015 0.023 21.635 21.640 qs_ot_p2m_diag 82 11.4 0.186 0.217 16.984 17.022 cp_fm_cholesky_invert 11 10.9 16.675 16.689 16.675 16.689 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.835 15.837 make_m2s 4970 13.5 0.090 0.094 14.000 15.485 make_images 4970 14.5 1.162 1.349 13.802 15.283 sum_up_and_integrate 127 10.3 0.114 0.134 14.242 14.269 integrate_v_rspace 127 11.3 0.004 0.004 14.128 14.157 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.790 12.825 calculate_rho_elec 127 8.7 0.087 0.104 12.790 12.825 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.761 12.770 cp_fm_redistribute_end 82 14.4 7.445 12.687 7.459 12.692 cp_fm_diag_elpa_base 82 14.4 4.997 12.237 5.214 12.591 init_scf_run 11 5.9 0.000 0.001 12.000 12.001 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.000 12.001 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 10.762 11.106 multiply_cannon_metrocomm4 27335 15.5 0.096 0.111 3.685 10.447 dbcsr_mm_accdrv_process 61726 16.2 4.499 5.796 9.122 10.246 make_images_data 4970 15.5 0.067 0.073 8.381 10.082 mp_irecv_dv 68888 16.3 3.494 10.071 3.494 10.071 hybrid_alltoall_any 5155 16.4 0.342 1.519 7.213 9.579 wfi_extrapolate 11 7.9 0.001 0.003 8.338 8.339 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.943 7.787 pw_transfer 1535 11.6 0.084 0.100 7.702 7.771 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.479 7.542 grid_integrate_task_list 127 12.3 7.118 7.462 7.118 7.462 density_rs2pw 127 9.7 0.006 0.006 6.659 7.159 cp_fm_cholesky_decompose 22 10.9 6.995 7.074 6.995 7.074 calculate_dm_sparse 127 9.5 0.001 0.001 6.792 6.927 fft_wrap_pw1pw2_140 519 13.2 0.465 0.515 6.493 6.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.173 6.183 rs_pw_transfer 1038 11.9 0.015 0.017 5.620 6.110 fft3d_ps 1281 14.7 2.757 2.906 5.855 5.898 mp_sum_l 7804 13.0 3.932 5.766 3.932 5.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.292 5.424 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.272 5.344 grid_collocate_task_list 127 9.7 4.872 5.292 4.872 5.292 potential_pw2rs 127 12.3 0.015 0.018 4.823 4.839 mp_allgather_i34 2485 14.5 1.759 4.592 1.759 4.592 mp_alltoall_d11v 2401 14.1 4.017 4.581 4.017 4.581 mp_sum_d 4445 12.1 2.736 4.185 2.736 4.185 dbcsr_complete_redistribute 393 12.7 0.775 0.862 3.199 4.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.177000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.727273, yerr=5.310212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 934.146048E+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 0.054 0.109 177.909 177.910 qs_mol_dyn_low 1 2.0 0.004 0.005 177.251 177.324 qs_forces 11 3.9 0.009 0.020 177.051 177.054 qs_energies 11 4.9 0.006 0.020 170.565 170.575 scf_env_do_scf 11 5.9 0.001 0.001 154.692 154.694 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.929 119.929 velocity_verlet 10 3.0 0.010 0.039 112.506 112.518 dbcsr_multiply_generic 2507 12.6 0.179 0.184 82.922 84.083 qs_scf_new_mos 117 7.6 0.001 0.001 83.655 83.966 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.654 83.965 ot_scf_mini 117 9.6 0.004 0.004 79.517 79.853 multiply_cannon 2507 13.6 0.501 0.528 62.819 66.854 multiply_cannon_loop 2507 14.6 0.861 0.889 59.721 62.567 ot_mini 117 10.6 0.001 0.002 44.226 44.697 mp_waitall_1 178456 16.5 25.730 35.024 25.730 35.024 init_scf_loop 11 6.9 0.000 0.000 34.659 34.660 prepare_preconditioner 11 7.9 0.000 0.000 30.699 30.757 make_preconditioner 11 8.9 0.000 0.000 30.699 30.757 rebuild_ks_matrix 128 8.3 0.001 0.001 29.796 30.234 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.029 29.796 30.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.396 29.765 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.920 27.323 multiply_cannon_multrec 20056 15.6 13.559 17.160 22.517 26.341 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 15.846 25.215 qs_ot_get_derivative 117 11.6 0.002 0.002 24.478 24.815 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.123 21.221 apply_single 128 13.6 0.001 0.001 20.122 21.221 qs_ot_get_p 128 10.4 0.001 0.001 20.638 21.088 ot_diis_step 117 11.6 0.061 0.193 19.643 19.771 qs_ot_p2m_diag 83 11.4 0.277 0.314 16.377 16.387 multiply_cannon_sync_h2d 20056 15.6 14.365 16.194 14.365 16.194 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.347 15.359 make_m2s 5014 13.6 0.083 0.087 14.482 15.343 make_images 5014 14.6 1.188 1.290 14.253 15.114 cp_fm_cholesky_invert 11 10.9 14.326 14.334 14.326 14.334 sum_up_and_integrate 128 10.3 0.133 0.145 14.071 14.095 integrate_v_rspace 128 11.3 0.004 0.004 13.939 13.968 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.967 13.017 calculate_rho_elec 128 8.7 0.132 0.147 12.966 13.016 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.184 12.184 cp_fm_redistribute_end 83 14.4 4.608 12.117 4.623 12.120 cp_fm_diag_elpa_base 83 14.4 7.067 11.495 7.475 12.014 init_scf_run 11 5.9 0.000 0.001 10.750 10.750 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.750 10.750 make_images_data 5014 15.6 0.062 0.069 8.838 10.075 hybrid_alltoall_any 5200 16.5 0.435 1.997 7.584 9.350 dbcsr_mm_accdrv_process 41502 16.2 4.753 5.926 8.405 9.294 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.989 9.248 multiply_cannon_metrocomm4 17549 15.6 0.062 0.071 3.442 9.032 mp_irecv_dv 50230 16.2 3.320 8.789 3.320 8.789 grid_integrate_task_list 128 12.3 7.345 7.713 7.345 7.713 pw_transfer 1547 11.6 0.085 0.105 7.595 7.705 wfi_extrapolate 11 7.9 0.061 0.241 7.598 7.599 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.371 7.488 cp_fm_upper_to_full 105 14.5 5.736 7.452 5.736 7.452 cp_fm_cholesky_decompose 22 10.9 7.288 7.322 7.288 7.322 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.810 6.867 density_rs2pw 128 9.7 0.006 0.006 6.418 6.817 fft_wrap_pw1pw2_140 523 13.2 0.479 0.526 6.477 6.602 dbcsr_complete_redistribute 395 12.7 1.171 1.201 4.528 6.293 calculate_dm_sparse 128 9.5 0.001 0.001 5.765 5.887 fft3d_ps 1291 14.7 2.695 2.898 5.634 5.692 rs_pw_transfer 1046 11.9 0.014 0.015 5.145 5.558 grid_collocate_task_list 128 9.7 5.063 5.451 5.063 5.451 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.427 5.434 mp_sum_l 7870 13.0 3.853 5.156 3.853 5.156 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.340 5.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.532 4.683 mp_alltoall_d11v 2415 14.1 4.190 4.541 4.190 4.541 potential_pw2rs 128 12.3 0.020 0.022 4.452 4.468 mp_allgather_i34 2507 14.6 1.609 4.323 1.609 4.323 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.031 4.068 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.284 3.984 mp_alltoall_i22 716 14.1 1.925 3.849 1.925 3.849 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.827 3.834 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.910000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.181818, yerr=11.574409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.142391E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57301. MP_Allreduce 11231 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.033 186.503 186.504 qs_mol_dyn_low 1 2.0 0.003 0.003 186.099 186.112 qs_forces 11 3.9 0.012 0.014 185.976 185.983 qs_energies 11 4.9 0.014 0.016 178.966 178.987 scf_env_do_scf 11 5.9 0.001 0.001 161.637 161.652 velocity_verlet 10 3.0 0.002 0.002 122.628 122.631 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.084 115.085 qs_scf_new_mos 116 7.6 0.001 0.001 80.224 80.472 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.224 80.471 dbcsr_multiply_generic 2485 12.5 0.186 0.190 79.347 80.007 ot_scf_mini 116 9.6 0.003 0.003 75.789 76.049 multiply_cannon 2485 13.5 0.556 0.592 55.180 58.434 multiply_cannon_loop 2485 14.5 1.168 1.197 51.330 52.801 init_scf_loop 11 6.9 0.000 0.000 46.433 46.434 prepare_preconditioner 11 7.9 0.000 0.000 42.385 42.427 make_preconditioner 11 8.9 0.000 0.000 42.385 42.427 ot_mini 116 10.6 0.001 0.001 42.026 42.275 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.034 41.082 multiply_cannon_multrec 29820 15.5 14.140 19.100 26.512 31.113 rebuild_ks_matrix 127 8.3 0.001 0.001 28.603 28.791 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 28.602 28.790 mp_waitall_1 152434 16.5 16.938 27.297 16.938 27.297 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.841 26.013 qs_ot_get_derivative 116 11.6 0.002 0.002 22.477 22.735 make_m2s 4970 13.5 0.096 0.100 19.880 21.197 make_images 4970 14.5 1.954 2.254 19.579 20.896 qs_ot_get_p 127 10.4 0.001 0.001 20.137 20.422 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.057 19.551 apply_single 127 13.6 0.001 0.001 19.057 19.550 ot_diis_step 116 11.6 0.017 0.018 19.424 19.426 cp_fm_upper_to_full 104 14.7 11.157 16.559 11.157 16.559 cp_fm_cholesky_invert 11 10.9 16.241 16.250 16.241 16.250 qs_ot_p2m_diag 82 11.4 0.339 0.386 16.153 16.204 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.193 14.926 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.892 14.893 sum_up_and_integrate 127 10.3 0.141 0.152 13.856 13.884 integrate_v_rspace 127 11.3 0.004 0.004 13.715 13.749 dbcsr_complete_redistribute 393 12.7 1.509 1.630 9.026 12.866 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.816 12.849 calculate_rho_elec 127 8.7 0.174 0.189 12.816 12.848 dbcsr_mm_accdrv_process 61748 16.2 7.274 8.230 11.953 12.506 multiply_cannon_sync_h2d 29820 15.5 11.698 12.438 11.698 12.438 make_images_data 4970 15.5 0.064 0.069 10.628 12.233 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.700 11.702 cp_fm_redistribute_end 82 14.4 2.024 11.618 2.040 11.625 hybrid_alltoall_any 5155 16.4 0.515 2.082 9.469 11.561 cp_fm_diag_elpa_base 82 14.4 8.959 10.977 9.545 11.500 init_scf_run 11 5.9 0.000 0.001 11.472 11.473 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.471 11.473 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.663 11.422 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.329 9.969 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.152 9.330 mp_alltoall_i22 712 14.1 5.556 9.305 5.556 9.305 pw_transfer 1535 11.6 0.084 0.096 7.638 7.716 grid_integrate_task_list 127 12.3 7.513 7.706 7.513 7.706 cp_fm_cholesky_decompose 22 10.9 7.568 7.647 7.568 7.647 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 7.416 7.498 wfi_extrapolate 11 7.9 0.001 0.001 7.425 7.425 multiply_cannon_metrocomm4 24850 15.5 0.074 0.083 2.706 7.130 mp_irecv_dv 75445 16.2 2.566 6.868 2.566 6.868 fft_wrap_pw1pw2_140 519 13.2 0.478 0.487 6.563 6.661 density_rs2pw 127 9.7 0.005 0.006 6.072 6.470 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.131 6.174 calculate_dm_sparse 127 9.5 0.001 0.001 6.079 6.165 fft3d_ps 1281 14.7 2.777 2.856 5.658 5.724 grid_collocate_task_list 127 9.7 5.166 5.568 5.166 5.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.107 5.190 mp_alltoall_d11v 2401 14.1 4.695 4.963 4.695 4.963 rs_pw_transfer 1038 11.9 0.014 0.015 4.452 4.836 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.448 4.577 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.393 4.402 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.131 4.194 potential_pw2rs 127 12.3 0.023 0.023 4.158 4.173 calculate_first_density_matrix 1 7.0 0.000 0.000 3.862 3.865 mp_allgather_i34 2485 14.5 1.954 3.793 1.954 3.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.504000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.363636, yerr=17.613387 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.490534E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58222. MP_Allreduce 11000 1175. MP_Sync 86 MP_Alltoall 1700 18828189. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.054 169.354 169.355 qs_mol_dyn_low 1 2.0 0.003 0.003 168.937 168.950 qs_forces 11 3.9 0.021 0.027 168.800 168.804 qs_energies 11 4.9 0.002 0.002 161.489 161.498 scf_env_do_scf 11 5.9 0.001 0.001 144.403 144.416 velocity_verlet 10 3.0 0.030 0.040 110.371 110.374 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.426 108.427 dbcsr_multiply_generic 2485 12.5 0.177 0.182 73.126 73.622 qs_scf_new_mos 116 7.6 0.001 0.001 73.498 73.594 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.497 73.593 ot_scf_mini 116 9.6 0.004 0.004 69.108 69.176 multiply_cannon 2485 13.5 0.581 0.624 53.763 57.737 multiply_cannon_loop 2485 14.5 0.446 0.458 49.177 50.083 ot_mini 116 10.6 0.001 0.001 39.013 39.078 init_scf_loop 11 6.9 0.000 0.000 35.828 35.829 mp_waitall_1 128490 16.6 25.503 32.603 25.503 32.603 prepare_preconditioner 11 7.9 0.000 0.000 31.658 31.689 make_preconditioner 11 8.9 0.000 0.000 31.658 31.689 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.495 29.753 rebuild_ks_matrix 127 8.3 0.001 0.001 28.090 28.171 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.090 28.171 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.640 25.713 multiply_cannon_multrec 9940 15.5 10.369 14.751 17.595 20.786 ot_diis_step 116 11.6 0.019 0.020 19.685 19.685 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.287 19.644 apply_single 127 13.6 0.001 0.001 19.287 19.643 qs_ot_get_derivative 116 11.6 0.002 0.002 19.259 19.321 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.125 18.949 make_m2s 4970 13.5 0.067 0.070 15.853 18.070 cp_fm_cholesky_invert 11 10.9 18.030 18.036 18.030 18.036 qs_ot_get_p 127 10.4 0.001 0.001 17.713 17.801 make_images 4970 14.5 2.322 2.878 15.552 17.769 sum_up_and_integrate 127 10.3 0.180 0.190 13.820 13.868 integrate_v_rspace 127 11.3 0.004 0.004 13.640 13.698 qs_ot_p2m_diag 82 11.4 0.490 0.496 13.666 13.682 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.888 12.932 calculate_rho_elec 127 8.7 0.256 0.267 12.887 12.932 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.544 12.545 make_images_data 4970 15.5 0.053 0.060 9.599 12.116 hybrid_alltoall_any 5155 16.4 0.803 3.546 9.383 12.024 multiply_cannon_sync_h2d 9940 15.5 11.488 11.971 11.488 11.971 init_scf_run 11 5.9 0.000 0.001 10.254 10.254 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.253 10.254 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.542 9.548 cp_fm_diag_elpa_base 82 14.4 9.301 9.375 9.539 9.544 grid_integrate_task_list 127 12.3 7.726 8.163 7.726 8.163 cp_fm_cholesky_decompose 22 10.9 7.923 8.022 7.923 8.022 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.685 7.735 pw_transfer 1535 11.6 0.083 0.090 7.565 7.589 dbcsr_mm_accdrv_process 20590 16.1 2.659 3.431 6.861 7.459 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.345 7.365 wfi_extrapolate 11 7.9 0.001 0.001 7.247 7.247 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.331 7.116 mp_allgather_i34 2485 14.5 2.793 6.997 2.793 6.997 fft_wrap_pw1pw2_140 519 13.2 0.501 0.522 6.473 6.501 calculate_dm_sparse 127 9.5 0.001 0.001 6.074 6.160 density_rs2pw 127 9.7 0.005 0.006 5.771 6.018 grid_collocate_task_list 127 9.7 5.452 5.692 5.452 5.692 dbcsr_complete_redistribute 393 12.7 2.134 2.224 5.258 5.625 fft3d_ps 1281 14.7 2.708 2.804 5.513 5.553 mp_alltoall_d11v 2401 14.1 4.641 5.293 4.641 5.293 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.259 5.266 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 5.203 5.210 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.858 4.556 mp_irecv_dv 28618 15.9 1.823 4.481 1.823 4.481 rs_pw_transfer 1038 11.9 0.013 0.014 3.970 4.233 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.069 4.095 potential_pw2rs 127 12.3 0.026 0.027 3.960 3.974 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.574 3.887 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.499 3.823 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.663 3.701 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.536 3.619 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.409 3.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.355000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1389.090909, yerr=52.456830 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.138343E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.070 297.473 297.473 qs_mol_dyn_low 1 2.0 0.003 0.003 296.877 296.889 qs_forces 11 3.9 0.003 0.003 296.765 296.768 qs_energies 11 4.9 0.003 0.003 288.126 288.137 scf_env_do_scf 11 5.9 0.001 0.001 265.613 265.622 velocity_verlet 10 3.0 0.021 0.022 212.136 212.145 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 139.616 139.617 init_scf_loop 11 6.9 0.000 0.000 125.736 125.738 prepare_preconditioner 11 7.9 0.000 0.000 121.037 121.060 make_preconditioner 11 8.9 0.000 0.000 121.037 121.060 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.993 118.190 qs_scf_new_mos 118 7.6 0.001 0.001 96.728 96.800 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.727 96.799 ot_scf_mini 118 9.6 0.004 0.004 91.900 92.022 dbcsr_multiply_generic 2534 12.6 0.213 0.219 85.398 85.653 cp_fm_upper_to_full 106 14.8 53.263 76.263 53.263 76.263 multiply_cannon 2534 13.6 0.706 0.791 59.855 61.204 multiply_cannon_loop 2534 14.6 0.478 0.486 56.010 56.578 ot_mini 118 10.6 0.001 0.001 45.899 46.047 dbcsr_complete_redistribute 397 12.7 4.003 4.040 29.832 42.908 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.379 39.432 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.999 36.939 mp_alltoall_i22 720 14.1 21.846 34.837 21.846 34.837 cp_fm_cholesky_invert 11 10.9 34.306 34.312 34.306 34.312 rebuild_ks_matrix 129 8.3 0.001 0.001 34.041 34.277 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.041 34.276 mp_waitall_1 106586 16.7 28.952 32.027 28.952 32.027 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.785 32.001 qs_ot_get_p 129 10.4 0.001 0.001 30.201 30.413 qs_ot_p2m_diag 84 11.4 0.889 0.895 25.718 25.751 qs_ot_get_derivative 118 11.6 0.002 0.002 25.123 25.236 cp_dbcsr_syevd 84 12.4 0.006 0.006 23.917 23.919 make_m2s 5068 13.6 0.079 0.082 20.243 21.277 make_images 5068 14.6 3.803 3.913 19.764 20.800 ot_diis_step 118 11.6 0.022 0.023 20.711 20.711 cp_fm_diag_elpa 84 13.4 0.000 0.000 20.518 20.523 cp_fm_diag_elpa_base 84 14.4 16.064 17.679 20.512 20.516 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.815 20.166 apply_single 129 13.6 0.001 0.001 19.814 20.166 multiply_cannon_metrocomm3 10136 15.6 0.023 0.023 19.170 19.793 multiply_cannon_multrec 10136 15.6 10.542 12.313 18.069 18.133 sum_up_and_integrate 129 10.3 0.324 0.326 16.110 16.207 integrate_v_rspace 129 11.3 0.004 0.004 15.785 15.882 multiply_cannon_sync_h2d 10136 15.6 15.778 15.794 15.778 15.794 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.334 15.368 calculate_rho_elec 129 8.7 0.487 0.488 15.333 15.368 make_images_data 5068 15.6 0.061 0.064 10.759 12.625 init_scf_run 11 5.9 0.000 0.001 12.536 12.537 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.535 12.537 hybrid_alltoall_any 5255 16.5 1.311 3.058 10.762 12.399 qs_ot_get_derivative_diag 78 12.4 0.003 0.003 10.022 10.105 cp_fm_cholesky_decompose 22 10.9 9.231 9.247 9.231 9.247 dbcsr_mm_accdrv_process 20954 16.1 3.829 5.635 7.289 9.124 pw_transfer 1559 11.6 0.091 0.092 8.787 8.796 grid_integrate_task_list 129 12.3 8.618 8.787 8.618 8.787 wfi_extrapolate 11 7.9 0.001 0.001 8.774 8.774 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.551 8.562 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.992 7.993 fft_wrap_pw1pw2_140 527 13.2 0.544 0.547 7.502 7.521 mp_alltoall_d11v 2429 14.1 7.158 7.276 7.158 7.276 calculate_dm_sparse 129 9.5 0.001 0.001 7.059 7.180 fft3d_ps 1301 14.7 2.784 2.823 6.577 6.589 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.474 6.530 grid_collocate_task_list 129 9.7 6.387 6.423 6.387 6.423 density_rs2pw 129 9.7 0.005 0.005 6.293 6.319 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.200 6.313 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.473000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2829.363636, yerr=180.998379 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 1.252848E+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 0.079 0.176 86.828 86.831 qs_energies 1 2.0 0.000 0.000 85.865 85.877 ls_scf 1 3.0 0.003 0.012 84.884 84.890 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.623 73.873 multiply_cannon 111 7.7 0.018 0.021 56.076 57.454 multiply_cannon_loop 111 8.7 0.212 0.225 52.615 53.987 ls_scf_main 1 4.0 0.080 0.327 53.352 53.367 density_matrix_trs4 2 5.0 0.002 0.003 47.412 47.538 ls_scf_init_scf 1 4.0 0.000 0.000 28.458 28.477 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.291 27.359 mp_waitall_1 11316 10.9 23.007 26.389 23.007 26.389 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.163 25.187 multiply_cannon_multrec 2664 9.7 8.204 8.928 15.529 17.213 multiply_cannon_sync_h2d 2664 9.7 13.586 14.739 13.586 14.739 make_m2s 222 7.7 0.009 0.011 13.527 14.036 make_images 222 8.7 0.101 0.112 13.505 14.017 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.847 12.330 make_images_data 222 9.7 0.005 0.006 7.998 8.651 hybrid_alltoall_any 227 10.6 0.218 1.857 6.732 8.550 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.621 6.943 7.914 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.460 7.607 dbcsr_mm_accdrv_process_sort 4760 11.4 6.234 7.141 6.234 7.141 calculate_norms 4752 9.8 5.519 6.127 5.519 6.127 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.390 5.603 mp_sum_l 807 5.4 3.548 5.185 3.548 5.185 make_images_sizes 222 9.7 0.000 0.000 0.930 3.747 mp_alltoall_i44 222 10.7 0.930 3.747 0.930 3.747 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.061 3.600 mp_irecv_dv 6231 10.9 2.043 3.571 2.043 3.571 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.414 3.393 arnoldi_extremal 4 6.8 0.000 0.000 3.313 3.342 arnoldi_normal_ev 4 7.8 0.001 0.002 3.313 3.342 build_subspace 16 8.4 0.009 0.012 3.218 3.219 ls_scf_post 1 4.0 0.000 0.001 3.071 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.892 2.936 dbcsr_special_finalize 555 9.7 0.005 0.006 2.237 2.692 dbcsr_merge_single_wm 555 10.7 0.455 0.590 2.229 2.684 make_images_pack 222 9.7 2.212 2.641 2.214 2.643 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.356 2.591 dbcsr_sort_data 658 11.4 2.032 2.485 2.032 2.485 dbcsr_matrix_vector_mult_local 304 10.0 2.057 2.447 2.059 2.449 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.277 2.385 buffer_matrices_ensure_size 222 8.7 1.762 2.076 1.762 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.779 1.800 rebuild_ks_matrix 3 7.3 0.000 0.000 1.770 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.017 0.045 1.770 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.831000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 2.111336E+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 0.097 92.809 92.812 qs_energies 1 2.0 0.000 0.000 92.029 92.045 ls_scf 1 3.0 0.002 0.005 90.716 90.736 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.153 76.530 multiply_cannon 111 7.7 0.029 0.046 53.658 57.903 ls_scf_main 1 4.0 0.019 0.146 55.905 55.911 multiply_cannon_loop 111 8.7 0.117 0.125 50.351 54.039 density_matrix_trs4 2 5.0 0.002 0.003 50.110 50.299 ls_scf_init_scf 1 4.0 0.001 0.005 31.248 31.250 mp_waitall_1 9246 10.9 21.726 30.534 21.726 30.534 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.060 30.165 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 27.486 27.647 multiply_cannon_multrec 1332 9.7 13.159 16.184 22.415 26.626 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.097 21.461 make_m2s 222 7.7 0.008 0.009 15.479 16.120 make_images 222 8.7 1.586 1.979 15.448 16.091 dbcsr_mm_accdrv_process 4041 10.4 0.280 0.454 8.861 10.474 dbcsr_mm_accdrv_process_sort 4041 11.4 8.454 10.020 8.454 10.020 make_images_data 222 9.7 0.004 0.005 8.970 9.885 hybrid_alltoall_any 227 10.6 0.523 2.456 8.338 9.526 mp_sum_l 807 5.4 5.750 9.104 5.750 9.104 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.220 7.957 mp_irecv_dv 3311 11.0 3.200 7.899 3.200 7.899 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.410 7.105 calculate_norms 2376 9.8 6.005 6.714 6.005 6.714 multiply_cannon_sync_h2d 1332 9.7 4.748 5.799 4.748 5.799 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.108 5.369 arnoldi_extremal 4 6.8 0.000 0.003 4.763 4.796 arnoldi_normal_ev 4 7.8 0.001 0.007 4.763 4.796 build_subspace 16 8.4 0.014 0.021 4.465 4.470 ls_scf_post 1 4.0 0.003 0.021 3.562 3.579 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.205 3.412 ls_scf_store_result 1 5.0 0.000 0.000 3.262 3.384 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.216 2.735 3.218 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.276 2.887 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.557 2.669 make_images_pack 222 9.7 2.030 2.395 2.033 2.397 mp_allgather_i34 111 8.7 0.993 2.395 0.993 2.395 dbcsr_sort_data 436 11.2 1.790 2.009 1.790 2.009 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.857 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.812000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1697.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 2.713424E+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 0.069 0.166 95.171 95.172 qs_energies 1 2.0 0.000 0.000 93.847 93.850 ls_scf 1 3.0 0.000 0.003 92.450 92.453 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.241 76.600 ls_scf_main 1 4.0 0.000 0.002 57.393 57.398 multiply_cannon 111 7.7 0.038 0.073 52.774 57.024 multiply_cannon_loop 111 8.7 0.101 0.106 49.162 52.708 density_matrix_trs4 2 5.0 0.002 0.003 51.445 51.632 mp_waitall_1 7374 11.0 23.201 33.110 23.201 33.110 ls_scf_init_scf 1 4.0 0.000 0.001 31.479 31.482 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.998 30.059 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.592 27.608 multiply_cannon_multrec 888 9.7 12.688 15.346 21.223 24.536 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.939 22.828 make_m2s 222 7.7 0.007 0.008 16.767 18.065 make_images 222 8.7 1.986 2.307 16.729 18.025 make_images_data 222 9.7 0.004 0.005 9.350 10.337 hybrid_alltoall_any 227 10.6 0.625 2.883 9.030 10.330 dbcsr_mm_accdrv_process 3754 10.4 0.234 0.413 8.077 9.291 dbcsr_mm_accdrv_process_sort 3754 11.4 7.707 8.878 7.707 8.878 mp_sum_l 807 5.4 5.158 8.827 5.158 8.827 multiply_cannon_sync_h2d 888 9.7 6.063 7.854 6.063 7.854 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.484 7.064 mp_irecv_dv 2335 11.1 2.468 7.022 2.468 7.022 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.876 6.822 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.535 6.375 arnoldi_extremal 4 6.8 0.001 0.002 5.703 5.717 arnoldi_normal_ev 4 7.8 0.004 0.019 5.702 5.717 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.964 5.126 build_subspace 16 8.4 0.014 0.020 4.948 4.959 calculate_norms 1584 9.8 4.282 4.616 4.282 4.616 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.539 3.870 mp_allgather_i34 111 8.7 1.437 3.680 1.437 3.680 dbcsr_matrix_vector_mult_local 304 10.0 3.012 3.584 3.014 3.586 ls_scf_post 1 4.0 0.000 0.002 3.577 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.332 3.403 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.758 2.853 dbcsr_sort_data 325 11.1 1.898 2.181 1.898 2.181 make_images_pack 222 9.7 1.826 2.152 1.829 2.155 make_images_sizes 222 9.7 0.000 0.000 1.057 2.069 mp_alltoall_i44 222 10.7 1.057 2.069 1.057 2.069 dbcsr_data_release 9322 10.9 1.307 1.968 1.307 1.968 dbcsr_finalize 304 7.8 0.027 0.032 1.636 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.172000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2163.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 3.368468E+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 0.034 0.048 97.440 97.441 qs_energies 1 2.0 0.000 0.000 96.767 96.773 ls_scf 1 3.0 0.000 0.002 95.124 95.129 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.858 79.087 ls_scf_main 1 4.0 0.000 0.001 59.172 59.174 multiply_cannon 111 7.7 0.081 0.195 51.757 56.052 density_matrix_trs4 2 5.0 0.002 0.003 53.074 53.195 multiply_cannon_loop 111 8.7 0.115 0.127 46.636 49.501 ls_scf_init_scf 1 4.0 0.000 0.001 32.681 32.684 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.548 31.609 mp_waitall_1 6438 11.0 22.747 29.892 22.747 29.892 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.050 29.065 multiply_cannon_multrec 1332 9.7 14.290 17.853 22.185 24.703 make_m2s 222 7.7 0.008 0.009 21.303 22.707 make_images 222 8.7 3.154 3.618 21.252 22.658 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.188 17.485 make_images_data 222 9.7 0.004 0.005 11.876 13.499 hybrid_alltoall_any 227 10.6 0.804 3.815 11.230 12.817 dbcsr_mm_accdrv_process 3641 10.4 0.213 0.399 7.541 9.052 dbcsr_mm_accdrv_process_sort 3641 11.4 7.167 8.633 7.167 8.633 mp_sum_l 807 5.4 4.211 7.410 4.211 7.410 multiply_cannon_sync_h2d 1332 9.7 5.534 6.349 5.534 6.349 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.084 6.010 mp_irecv_dv 3229 10.9 2.061 5.956 2.061 5.956 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.179 5.825 arnoldi_extremal 4 6.8 0.000 0.000 5.347 5.368 arnoldi_normal_ev 4 7.8 0.001 0.004 5.347 5.368 build_subspace 16 8.4 0.014 0.021 5.005 5.014 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.610 4.786 mp_allgather_i34 111 8.7 2.222 4.725 2.222 4.725 calculate_norms 2376 9.8 4.183 4.534 4.183 4.534 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.516 4.484 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.622 3.943 dbcsr_matrix_vector_mult_local 304 10.0 3.180 3.674 3.182 3.676 dbcsr_sort_data 658 11.4 3.121 3.408 3.121 3.408 ls_scf_post 1 4.0 0.000 0.001 3.270 3.277 dbcsr_special_finalize 555 9.7 0.006 0.007 2.878 3.250 dbcsr_merge_single_wm 555 10.7 0.543 0.672 2.869 3.242 ls_scf_store_result 1 5.0 0.000 0.000 3.013 3.074 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.920 2.978 dbcsr_data_release 10477 10.7 1.581 2.423 1.581 2.423 dbcsr_finalize 304 7.8 0.050 0.062 1.809 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.441000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 4.629918E+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 0.044 0.058 100.049 100.050 qs_energies 1 2.0 0.000 0.000 98.843 98.850 ls_scf 1 3.0 0.000 0.000 96.926 96.931 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.481 78.709 ls_scf_main 1 4.0 0.000 0.000 62.214 62.215 multiply_cannon 111 7.7 0.087 0.180 55.859 61.177 density_matrix_trs4 2 5.0 0.002 0.003 55.164 55.262 multiply_cannon_loop 111 8.7 0.070 0.077 51.279 52.958 mp_waitall_1 5481 11.0 26.757 31.878 26.757 31.878 ls_scf_init_scf 1 4.0 0.000 0.000 31.099 31.104 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.924 29.962 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.745 27.754 multiply_cannon_multrec 444 9.7 14.121 17.024 21.261 24.251 make_m2s 222 7.7 0.006 0.007 17.670 20.276 make_images 222 8.7 3.742 4.425 17.607 20.214 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.220 16.633 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.460 14.987 make_images_data 222 9.7 0.003 0.004 9.875 12.413 hybrid_alltoall_any 227 10.6 0.788 3.752 9.645 12.166 dbcsr_mm_accdrv_process 3003 10.4 0.252 0.749 6.852 8.189 multiply_cannon_sync_h2d 444 9.7 6.536 8.160 6.536 8.160 dbcsr_mm_accdrv_process_sort 3003 11.4 6.436 7.549 6.436 7.549 mp_allgather_i34 111 8.7 2.797 7.057 2.797 7.057 arnoldi_extremal 4 6.8 0.000 0.000 6.008 6.017 arnoldi_normal_ev 4 7.8 0.001 0.004 6.008 6.017 build_subspace 16 8.4 0.015 0.020 5.603 5.612 mp_sum_l 807 5.4 2.997 4.904 2.997 4.904 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.644 4.759 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.255 4.429 dbcsr_matrix_vector_mult_local 304 10.0 3.685 4.161 3.687 4.163 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.105 3.893 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.600 3.834 mp_irecv_dv 1241 11.2 1.581 3.798 1.581 3.798 calculate_norms 792 9.8 3.550 3.713 3.550 3.713 ls_scf_post 1 4.0 0.000 0.000 3.613 3.619 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.386 3.472 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.439 make_images_sizes 222 9.7 0.000 0.000 1.105 3.385 mp_alltoall_i44 222 10.7 1.104 3.384 1.104 3.384 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.270 dbcsr_merge_all 275 8.9 0.475 0.521 2.051 2.101 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.050000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3661.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c2dc0ed42209bbdc4f6d5b58e456999b747df445_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 8.770925E+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 0.104 0.186 108.972 108.973 qs_energies 1 2.0 0.000 0.000 107.088 107.102 ls_scf 1 3.0 0.000 0.000 103.881 103.895 dbcsr_multiply_generic 111 6.7 0.024 0.026 77.870 77.966 ls_scf_main 1 4.0 0.000 0.000 65.814 65.815 density_matrix_trs4 2 5.0 0.002 0.003 56.939 56.989 multiply_cannon 111 7.7 0.125 0.228 50.257 52.356 multiply_cannon_loop 111 8.7 0.068 0.072 46.722 47.536 ls_scf_init_scf 1 4.0 0.000 0.000 34.420 34.421 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.075 33.092 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.223 30.236 mp_waitall_1 4569 11.1 22.205 26.022 22.205 26.022 make_m2s 222 7.7 0.007 0.007 24.081 25.131 make_images 222 8.7 4.606 4.981 23.971 25.020 multiply_cannon_multrec 444 9.7 17.978 18.540 22.605 23.200 hybrid_alltoall_any 227 10.6 1.663 3.628 12.985 16.112 make_images_data 222 9.7 0.003 0.004 13.272 15.762 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.331 11.307 multiply_cannon_sync_h2d 444 9.7 8.850 8.889 8.850 8.889 arnoldi_extremal 4 6.8 0.000 0.000 7.529 7.542 arnoldi_normal_ev 4 7.8 0.002 0.008 7.529 7.542 build_subspace 16 8.4 0.026 0.036 6.970 6.983 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.546 5.720 dbcsr_matrix_vector_mult_local 304 10.0 5.034 5.354 5.037 5.357 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.031 5.277 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.836 4.926 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.324 4.450 4.581 dbcsr_mm_accdrv_process_sort 1814 11.4 4.149 4.269 4.149 4.269 ls_scf_post 1 4.0 0.000 0.000 3.647 3.661 make_images_sizes 222 9.7 0.000 0.000 1.440 3.593 mp_alltoall_i44 222 10.7 1.440 3.593 1.440 3.593 mp_allgather_i34 111 8.7 1.148 3.589 1.148 3.589 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.412 calculate_norms 792 9.8 3.239 3.273 3.239 3.273 qs_energies_init_hamiltonians 1 3.0 0.319 0.327 3.174 3.175 dbcsr_finalize 304 7.8 0.082 0.090 3.089 3.135 dbcsr_merge_all 275 8.9 0.893 0.912 2.873 2.924 dbcsr_complete_redistribute 5 7.6 1.443 1.479 2.782 2.913 dbcsr_data_release 12724 10.6 2.332 2.895 2.332 2.895 matrix_ls_to_qs 2 6.0 0.000 0.000 2.433 2.583 dbcsr_sort_data 325 11.1 2.440 2.497 2.440 2.497 dbcsr_new_transposed 4 7.5 0.243 0.251 2.294 2.309 dbcsr_frobenius_norm 74 6.6 2.059 2.138 2.197 2.240 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.207 dbcsr_add_anytype 103 7.2 0.860 0.892 2.136 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.204 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.973000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6911.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c2dc0ed42209bbdc4f6d5b58e456999b747df445 Summary: empty Status: OK