=== 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: 672e1570b1b7b1d8caa570061018a37a7bb1d41b ################# 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/01 job id: 42111468 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/02 job id: 42111469 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/03 job id: 42111471 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/04 job id: 42111472 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/05 job id: 42111474 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/06 job id: 42111475 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/07 job id: 42111476 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/08 job id: 42111478 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/09 job id: 42111479 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/10 job id: 42111480 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/11 job id: 42111482 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/12 job id: 42111483 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/13 job id: 42111484 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/14 job id: 42111485 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/15 job id: 42111486 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/16 job id: 42111487 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/17 job id: 42111488 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/18 job id: 42111489 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/19 job id: 42111490 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/20 job id: 42111491 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/21 job id: 42111492 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/22 job id: 42111493 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/23 job id: 42111494 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/24 job id: 42111496 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/25 job id: 42111497 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/26 job id: 42111499 --- 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/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 133.023 133.024 farming_run 1 2.0 132.508 132.509 132.995 132.998 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460834E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 115.352 115.352 qs_energies 1 2.0 0.000 0.000 115.150 115.151 mp2_main 1 3.0 0.000 0.000 113.171 113.172 mp2_gpw_main 1 4.0 0.019 0.026 112.326 112.328 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.635 94.102 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.497 55.965 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.173 41.803 47.232 get_2c_integrals 1 6.0 0.000 0.000 36.940 37.966 integrate_v_rspace 273 8.0 0.436 0.449 25.146 30.305 pw_transfer 6555 10.6 0.372 0.385 27.478 28.206 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.126 26.714 grid_integrate_task_list 273 9.0 20.951 26.599 20.951 26.599 fft_wrap_pw1pw2_100 2178 12.4 1.216 1.445 23.689 24.277 compute_2c_integrals 1 7.0 0.003 0.003 19.696 19.697 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.029 19.428 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.451 19.026 19.427 rpa_ri_compute_en 1 5.0 0.001 0.001 18.581 18.743 cp_fm_cholesky_decompose 12 8.2 17.223 18.271 17.223 18.271 cholesky_decomp 1 7.0 0.000 0.000 16.096 17.143 fft3d_s 5443 13.4 16.200 16.618 16.222 16.639 ao_to_mo_and_store_B_mult_1 272 7.0 10.859 15.564 10.859 15.564 calculate_wavefunction 272 8.0 5.432 5.584 12.570 13.162 rpa_num_int 1 6.0 0.000 0.001 10.508 10.518 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.474 10.509 calc_mat_Q 8 8.0 0.000 0.000 9.304 9.412 contract_S_to_Q 8 9.0 0.000 0.000 8.726 8.835 calc_potential_gpw 544 9.5 0.006 0.007 8.322 8.643 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.296 8.553 potential_pw2rs 545 10.0 0.107 0.109 7.713 8.471 parallel_gemm_fm 14 9.1 0.000 0.000 8.320 8.417 parallel_gemm_fm_cosma 14 10.1 8.320 8.417 8.320 8.417 collocate_single_gaussian 272 10.0 0.041 0.043 7.518 7.760 create_integ_mat 1 6.0 0.014 0.026 7.509 7.518 array2fm 1 7.0 0.000 0.000 6.576 7.013 pw_scatter_s 2720 13.7 4.441 4.646 4.441 4.646 pw_gather_s 2722 13.2 3.898 4.323 3.898 4.323 array2fm_buffer_send 1 8.0 2.908 3.142 2.908 3.142 pw_poisson_solve 545 10.5 1.133 1.189 2.194 2.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.325207, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.039 394.716 394.717 farming_run 1 2.0 394.115 394.119 394.677 394.680 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225490E+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 0.012 0.044 208.835 208.836 qs_energies 1 2.0 0.000 0.000 208.616 208.622 scf_env_do_scf 1 3.0 0.000 0.000 106.445 106.446 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.597 105.605 rebuild_ks_matrix 4 6.0 0.000 0.000 105.596 105.604 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.596 105.604 hfx_ks_matrix 4 8.0 0.001 0.001 105.223 105.227 integrate_four_center 4 9.0 0.144 0.457 105.223 105.226 mp2_main 1 3.0 0.000 0.000 101.876 101.883 mp2_gpw_main 1 4.0 0.044 0.063 100.965 100.974 integrate_four_center_main 4 10.0 0.094 0.593 96.350 98.982 integrate_four_center_bin 265 11.0 96.256 98.939 96.256 98.939 init_scf_loop 1 4.0 0.000 0.000 92.034 92.035 mp2_ri_gpw_compute_in 1 5.0 0.069 0.096 74.730 75.811 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.437 55.505 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 42.146 47.147 integrate_v_rspace 95 8.0 0.398 0.566 28.521 33.492 pw_transfer 2240 10.6 0.147 0.162 29.843 30.188 ao_to_mo_and_store_B_mult_1 91 7.0 10.606 29.759 10.606 29.759 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.867 29.212 grid_integrate_task_list 95 9.0 23.819 28.992 23.819 28.992 mp2_ri_gpw_compute_en 1 5.0 0.056 0.062 26.077 27.795 fft_wrap_pw1pw2_100 730 12.4 1.247 1.412 26.616 27.051 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.887 24.362 24.373 get_2c_integrals 1 6.0 0.000 0.000 20.189 20.227 compute_2c_integrals 1 7.0 0.018 0.029 19.170 19.191 compute_2c_integrals_loop_lm 1 8.0 0.006 0.008 18.791 19.024 mp2_eri_2c_integrate_gpw 1 9.0 1.755 1.854 18.785 19.024 fft3d_s 1823 13.4 18.419 18.782 18.433 18.795 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.409 14.409 calculate_wavefunction 91 8.0 2.014 2.047 9.738 9.944 potential_pw2rs 186 10.0 0.033 0.035 8.600 9.188 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.580 8.738 9.173 local_gemm 172 8.0 8.181 8.605 8.181 8.605 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.187 8.494 calc_potential_gpw 182 9.5 0.002 0.002 7.883 8.098 collocate_single_gaussian 91 10.0 0.017 0.021 7.844 8.070 mp2_ri_gpw_compute_en_comm 22 7.0 0.535 0.557 7.393 7.825 mp_sync 38 10.4 3.527 6.776 3.527 6.776 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.417 6.343 6.417 mp_sendrecv_dm3 2068 8.0 5.335 5.770 5.335 5.770 pw_gather_s 912 13.2 4.919 5.450 4.919 5.450 pw_scatter_s 910 13.7 3.926 4.216 3.926 4.216 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.956265, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 450.048000E+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 415965. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 51.870 51.871 qs_mol_dyn_low 1 2.0 0.003 0.003 51.680 51.689 qs_forces 11 3.9 0.002 0.002 51.613 51.614 qs_energies 11 4.9 0.001 0.002 50.163 50.176 scf_env_do_scf 11 5.9 0.001 0.001 44.326 44.326 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.273 42.273 dbcsr_multiply_generic 2286 12.5 0.092 0.096 33.665 34.151 qs_scf_new_mos 108 7.5 0.000 0.001 32.337 32.610 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.337 32.609 ot_scf_mini 108 9.5 0.002 0.002 30.645 30.836 multiply_cannon 2286 13.5 0.180 0.186 25.970 27.716 multiply_cannon_loop 2286 14.5 1.515 1.587 25.275 27.039 velocity_verlet 10 3.0 0.002 0.002 25.801 25.802 ot_mini 108 10.5 0.001 0.001 19.696 19.944 qs_ot_get_derivative 108 11.5 0.001 0.002 16.715 16.881 mp_waitall_1 267858 16.1 8.836 15.252 8.836 15.252 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 6.084 13.424 multiply_cannon_multrec 54864 15.5 4.179 6.398 7.489 10.845 rebuild_ks_matrix 119 8.3 0.000 0.000 7.815 7.944 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.814 7.943 multiply_cannon_sync_h2d 54864 15.5 5.852 7.775 5.852 7.775 mp_sum_l 7207 12.9 5.737 7.441 5.737 7.441 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.902 7.018 qs_ot_get_p 119 10.4 0.001 0.001 6.331 6.664 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.714 6.173 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.762 5.860 init_scf_run 11 5.9 0.000 0.001 4.635 4.636 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.635 4.636 dbcsr_mm_accdrv_process 76910 16.1 1.197 1.854 3.234 4.515 sum_up_and_integrate 119 10.3 0.012 0.014 4.476 4.484 integrate_v_rspace 119 11.3 0.002 0.002 4.464 4.472 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.025 4.147 calculate_rho_elec 119 8.7 0.011 0.016 4.025 4.146 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.890 3.517 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.448 3.504 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.888 3.083 apply_single 119 13.6 0.000 0.000 2.888 3.083 calculate_dm_sparse 119 9.5 0.000 0.000 2.901 3.038 rs_pw_transfer 974 11.9 0.012 0.013 2.733 2.813 ot_diis_step 108 11.5 0.006 0.006 2.687 2.688 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.622 2.622 jit_kernel_multiply 13 15.8 1.972 2.573 1.972 2.573 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.431 2.505 calculate_first_density_matrix 1 7.0 0.000 0.001 2.479 2.484 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.400 2.400 cp_fm_redistribute_end 50 14.0 2.179 2.376 2.183 2.378 cp_fm_diag_elpa_base 50 14.0 0.193 2.332 0.194 2.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.305 2.307 density_rs2pw 119 9.7 0.004 0.004 2.159 2.272 acc_transpose_blocks 54864 15.5 0.225 0.244 1.750 2.179 grid_integrate_task_list 119 12.3 2.030 2.124 2.030 2.124 wfi_extrapolate 11 7.9 0.001 0.001 2.095 2.095 init_scf_loop 11 6.9 0.000 0.000 2.035 2.036 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.978 2.025 mp_sum_d 4129 12.0 1.322 1.947 1.322 1.947 potential_pw2rs 119 12.3 0.004 0.004 1.850 1.861 pw_transfer 1439 11.6 0.051 0.056 1.657 1.729 make_m2s 4572 13.5 0.053 0.055 1.645 1.685 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.582 1.654 make_images 4572 14.5 0.133 0.138 1.563 1.601 fft3d_ps 1201 14.6 0.356 0.460 1.360 1.427 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.375 1.394 mp_alltoall_d11v 2130 13.8 1.258 1.394 1.258 1.394 mp_waitany 12084 13.8 1.214 1.389 1.214 1.389 grid_collocate_task_list 119 9.7 1.291 1.362 1.291 1.362 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.203 1.276 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.727 1.114 mp_alltoall_i22 627 13.8 0.771 1.053 0.771 1.053 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.871000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.818182, yerr=1.641406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 485.273600E+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 1031550. 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 0.032 0.043 39.894 39.896 qs_mol_dyn_low 1 2.0 0.003 0.003 39.584 39.593 qs_forces 11 3.9 0.002 0.003 39.499 39.500 qs_energies 11 4.9 0.001 0.002 37.807 37.811 scf_env_do_scf 11 5.9 0.001 0.001 32.426 32.426 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.704 29.705 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.937 22.300 qs_scf_new_mos 108 7.5 0.001 0.001 20.454 20.698 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.454 20.698 ot_scf_mini 108 9.5 0.002 0.003 19.525 19.698 velocity_verlet 10 3.0 0.002 0.002 18.866 18.867 multiply_cannon 2286 13.5 0.210 0.218 16.641 18.298 multiply_cannon_loop 2286 14.5 0.912 0.986 15.431 16.842 ot_mini 108 10.5 0.001 0.001 12.093 12.332 mp_waitall_1 217478 16.2 6.432 11.663 6.432 11.663 qs_ot_get_derivative 108 11.5 0.001 0.002 9.602 9.777 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.233 9.554 multiply_cannon_multrec 27432 15.5 1.962 4.507 6.003 8.941 rebuild_ks_matrix 119 8.3 0.000 0.000 7.260 7.404 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.259 7.403 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.432 6.562 dbcsr_mm_accdrv_process 47894 16.0 3.117 5.391 3.972 5.871 qs_ot_get_p 119 10.4 0.001 0.001 4.473 4.703 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.671 4.524 sum_up_and_integrate 119 10.3 0.024 0.027 4.303 4.308 integrate_v_rspace 119 11.3 0.002 0.002 4.278 4.286 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.068 4.147 apply_single 119 13.6 0.000 0.000 3.068 4.147 init_scf_run 11 5.9 0.000 0.001 4.135 4.135 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.134 4.135 mp_sum_l 7207 12.9 2.162 4.117 2.162 4.117 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.811 3.850 calculate_rho_elec 119 8.7 0.021 0.024 3.811 3.849 rs_pw_transfer 974 11.9 0.010 0.011 2.840 2.964 make_m2s 4572 13.5 0.052 0.054 2.686 2.932 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.885 2.904 make_images 4572 14.5 0.201 0.237 2.598 2.842 multiply_cannon_sync_h2d 27432 15.5 2.160 2.840 2.160 2.840 init_scf_loop 11 6.9 0.000 0.000 2.691 2.692 calculate_first_density_matrix 1 7.0 0.000 0.001 2.540 2.542 ot_diis_step 108 11.5 0.011 0.011 2.437 2.438 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.427 2.427 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.248 2.339 calculate_dm_sparse 119 9.5 0.000 0.000 2.222 2.296 density_rs2pw 119 9.7 0.004 0.004 2.182 2.279 potential_pw2rs 119 12.3 0.006 0.006 2.059 2.068 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.022 2.024 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.009 2.009 cp_fm_redistribute_end 50 14.0 1.648 1.973 1.653 1.974 cp_fm_diag_elpa_base 50 14.0 0.312 1.900 0.320 1.930 jit_kernel_multiply 9 16.2 0.802 1.926 0.802 1.926 grid_integrate_task_list 119 12.3 1.823 1.918 1.823 1.918 pw_transfer 1439 11.6 0.063 0.066 1.843 1.876 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.787 1.827 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.754 1.789 prepare_preconditioner 11 7.9 0.000 0.000 1.741 1.768 make_preconditioner 11 8.9 0.000 0.000 1.740 1.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.638 1.695 make_images_data 4572 15.5 0.045 0.051 1.176 1.615 acc_transpose_blocks 27432 15.5 0.107 0.112 1.256 1.575 wfi_extrapolate 11 7.9 0.001 0.001 1.534 1.534 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.037 1.521 fft3d_ps 1201 14.6 0.496 0.551 1.461 1.494 mp_allgather_i34 2286 14.5 0.650 1.419 0.650 1.419 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.411 1.419 mp_alltoall_d11v 2130 13.8 1.293 1.381 1.293 1.381 grid_collocate_task_list 119 9.7 1.240 1.312 1.240 1.312 fft_wrap_pw1pw2_140 487 13.2 0.075 0.081 1.254 1.288 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.221 1.269 mp_sum_d 4129 12.0 0.605 1.043 0.605 1.043 make_images_sizes 4572 15.5 0.005 0.005 0.699 0.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.954 0.955 mp_alltoall_i44 4572 16.5 0.695 0.954 0.695 0.954 acc_transpose_blocks_kernels 27432 16.5 0.184 0.271 0.719 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.901 rs_pw_transfer_PW2RS_50 119 14.3 0.585 0.605 0.791 0.824 cp_fm_cholesky_invert 11 10.9 0.795 0.800 0.795 0.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.896000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.000000, yerr=0.953463 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 525.455360E+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 0.014 0.031 32.137 32.138 qs_mol_dyn_low 1 2.0 0.003 0.003 31.899 31.908 qs_forces 11 3.9 0.002 0.003 31.836 31.838 qs_energies 11 4.9 0.003 0.014 30.263 30.265 scf_env_do_scf 11 5.9 0.001 0.001 25.551 25.551 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.862 22.863 dbcsr_multiply_generic 2286 12.5 0.093 0.095 16.426 16.528 velocity_verlet 10 3.0 0.001 0.002 15.345 15.347 qs_scf_new_mos 108 7.5 0.001 0.001 14.871 14.891 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.870 14.890 ot_scf_mini 108 9.5 0.002 0.003 14.146 14.163 multiply_cannon 2286 13.5 0.196 0.203 13.044 13.922 multiply_cannon_loop 2286 14.5 0.644 0.675 12.239 13.144 ot_mini 108 10.5 0.001 0.001 8.812 8.829 qs_ot_get_derivative 108 11.5 0.001 0.002 7.310 7.325 multiply_cannon_multrec 18288 15.5 1.941 2.865 6.773 7.095 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.362 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.348 6.362 dbcsr_mm_accdrv_process 38222 16.0 4.203 5.317 4.747 5.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.609 5.621 sum_up_and_integrate 119 10.3 0.030 0.031 3.978 3.984 integrate_v_rspace 119 11.3 0.002 0.003 3.948 3.958 mp_waitall_1 169478 16.3 2.950 3.911 2.950 3.911 init_scf_run 11 5.9 0.000 0.001 3.526 3.526 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.526 3.526 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.714 3.354 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.271 3.282 calculate_rho_elec 119 8.7 0.030 0.031 3.270 3.282 qs_ot_get_p 119 10.4 0.001 0.001 3.185 3.208 init_scf_loop 11 6.9 0.000 0.002 2.669 2.671 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.423 2.410 rs_pw_transfer 974 11.9 0.009 0.010 2.319 2.410 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.055 2.356 apply_single 119 13.6 0.000 0.000 2.055 2.356 calculate_first_density_matrix 1 7.0 0.001 0.006 2.297 2.298 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.066 2.073 make_m2s 4572 13.5 0.044 0.045 1.834 1.959 density_rs2pw 119 9.7 0.004 0.004 1.852 1.936 grid_integrate_task_list 119 12.3 1.807 1.912 1.807 1.912 make_images 4572 14.5 0.191 0.203 1.749 1.872 calculate_dm_sparse 119 9.5 0.000 0.000 1.819 1.839 prepare_preconditioner 11 7.9 0.000 0.000 1.836 1.839 make_preconditioner 11 8.9 0.000 0.000 1.836 1.839 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.799 1.800 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.690 1.774 potential_pw2rs 119 12.3 0.007 0.008 1.759 1.773 pw_transfer 1439 11.6 0.063 0.066 1.728 1.744 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.739 1.740 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.638 1.656 multiply_cannon_sync_h2d 18288 15.5 1.390 1.626 1.390 1.626 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.593 1.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.563 1.569 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.559 1.560 cp_fm_redistribute_end 50 14.0 1.156 1.536 1.157 1.536 jit_kernel_multiply 8 16.6 0.491 1.530 0.491 1.530 cp_fm_diag_elpa_base 50 14.0 0.362 1.471 0.377 1.509 mp_sum_l 7207 12.9 1.162 1.481 1.162 1.481 ot_diis_step 108 11.5 0.011 0.012 1.479 1.479 fft3d_ps 1201 14.6 0.505 0.523 1.326 1.343 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.267 1.283 grid_collocate_task_list 119 9.7 1.204 1.279 1.204 1.279 acc_transpose_blocks 18288 15.5 0.074 0.076 1.243 1.261 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.186 1.190 wfi_extrapolate 11 7.9 0.001 0.001 1.180 1.180 make_images_data 4572 15.5 0.044 0.048 0.804 0.960 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 0.943 0.944 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.888 0.912 hybrid_alltoall_any 4725 16.4 0.054 0.113 0.701 0.886 mp_alltoall_d11v 2130 13.8 0.721 0.839 0.721 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.813 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.797 0.806 cp_fm_cholesky_invert 11 10.9 0.779 0.784 0.779 0.784 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.743 mp_alltoall_z22v 1201 16.6 0.662 0.735 0.662 0.735 mp_waitany 9880 13.7 0.537 0.644 0.537 0.644 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.138000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=4.187743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 559.529984E+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 0.017 0.031 34.839 34.840 qs_mol_dyn_low 1 2.0 0.003 0.003 34.598 34.607 qs_forces 11 3.9 0.002 0.002 34.534 34.535 qs_energies 11 4.9 0.003 0.004 32.812 32.818 scf_env_do_scf 11 5.9 0.001 0.001 27.759 27.760 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.280 24.282 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.096 18.212 velocity_verlet 10 3.0 0.002 0.002 17.772 17.774 qs_scf_new_mos 108 7.5 0.001 0.001 16.069 16.125 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.068 16.124 ot_scf_mini 108 9.5 0.002 0.003 15.134 15.189 multiply_cannon 2286 13.5 0.225 0.269 14.361 14.871 multiply_cannon_loop 2286 14.5 0.943 0.977 13.382 13.776 ot_mini 108 10.5 0.001 0.001 9.244 9.316 multiply_cannon_multrec 27432 15.5 2.327 3.018 8.603 8.989 qs_ot_get_derivative 108 11.5 0.001 0.001 7.432 7.488 dbcsr_mm_accdrv_process 47916 15.9 5.299 7.221 6.183 7.392 rebuild_ks_matrix 119 8.3 0.000 0.000 6.511 6.566 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.510 6.565 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.781 5.828 sum_up_and_integrate 119 10.3 0.035 0.038 3.785 3.791 integrate_v_rspace 119 11.3 0.002 0.003 3.751 3.757 init_scf_run 11 5.9 0.000 0.001 3.698 3.698 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.698 3.698 init_scf_loop 11 6.9 0.000 0.003 3.458 3.462 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.332 3.367 calculate_rho_elec 119 8.7 0.040 0.046 3.332 3.366 qs_ot_get_p 119 10.4 0.001 0.002 3.272 3.353 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.695 3.183 mp_waitall_1 145218 16.4 2.082 2.687 2.082 2.687 prepare_preconditioner 11 7.9 0.000 0.000 2.606 2.615 make_preconditioner 11 8.9 0.000 0.001 2.606 2.615 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.126 2.574 apply_single 119 13.6 0.000 0.000 2.126 2.574 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.207 2.536 make_m2s 4572 13.5 0.054 0.056 2.229 2.344 calculate_first_density_matrix 1 7.0 0.000 0.002 2.300 2.302 make_images 4572 14.5 0.274 0.334 2.122 2.234 rs_pw_transfer 974 11.9 0.009 0.009 2.060 2.157 calculate_dm_sparse 119 9.5 0.000 0.000 2.023 2.082 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.955 1.984 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.931 1.942 density_rs2pw 119 9.7 0.003 0.004 1.813 1.896 grid_integrate_task_list 119 12.3 1.822 1.883 1.822 1.883 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.877 1.878 jit_kernel_multiply 10 16.2 0.822 1.830 0.822 1.830 pw_transfer 1439 11.6 0.063 0.067 1.788 1.821 ot_diis_step 108 11.5 0.012 0.012 1.773 1.773 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.698 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.672 1.684 mp_sum_l 7207 12.9 1.087 1.668 1.087 1.668 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.607 1.607 potential_pw2rs 119 12.3 0.008 0.010 1.576 1.585 acc_transpose_blocks 27432 15.5 0.111 0.114 1.455 1.477 fft_wrap_pw1pw2_140 487 13.2 0.085 0.096 1.376 1.416 fft3d_ps 1201 14.6 0.535 0.589 1.379 1.402 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.351 1.352 wfi_extrapolate 11 7.9 0.001 0.001 1.346 1.346 cp_fm_redistribute_end 50 14.0 0.889 1.325 0.890 1.325 grid_collocate_task_list 119 9.7 1.220 1.305 1.220 1.305 cp_fm_diag_elpa_base 50 14.0 0.413 1.260 0.432 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.207 1.218 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.713 1.212 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.116 1.135 cp_fm_upper_to_full 72 13.5 0.807 1.119 0.807 1.119 dbcsr_complete_redistribute 329 12.2 0.128 0.165 0.833 1.106 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.093 1.095 multiply_cannon_sync_h2d 27432 15.5 0.981 1.027 0.981 1.027 make_images_data 4572 15.5 0.045 0.048 0.865 0.985 mp_alltoall_d11v 2130 13.8 0.781 0.927 0.781 0.927 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.743 0.913 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.608 0.876 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.823 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.820 0.826 cp_fm_cholesky_invert 11 10.9 0.777 0.780 0.777 0.780 mp_alltoall_z22v 1201 16.6 0.725 0.751 0.725 0.751 mp_alltoall_i22 627 13.8 0.436 0.734 0.436 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.840000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.909091, yerr=6.052115 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 606.654464E+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 0.015 0.031 28.177 28.178 qs_mol_dyn_low 1 2.0 0.003 0.003 27.914 27.921 qs_forces 11 3.9 0.002 0.002 27.849 27.850 qs_energies 11 4.9 0.001 0.001 26.164 26.166 scf_env_do_scf 11 5.9 0.000 0.001 21.410 21.410 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.769 18.769 velocity_verlet 10 3.0 0.004 0.005 14.402 14.405 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.312 12.382 qs_scf_new_mos 108 7.5 0.001 0.001 11.278 11.303 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.278 11.303 ot_scf_mini 108 9.5 0.002 0.002 10.593 10.622 multiply_cannon 2286 13.5 0.232 0.239 9.658 10.136 multiply_cannon_loop 2286 14.5 0.334 0.346 8.705 8.867 multiply_cannon_multrec 9144 15.5 1.682 1.930 5.898 6.118 ot_mini 108 10.5 0.001 0.001 5.877 5.908 rebuild_ks_matrix 119 8.3 0.000 0.000 5.751 5.774 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.751 5.773 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.133 5.153 qs_ot_get_derivative 108 11.5 0.001 0.001 4.568 4.596 dbcsr_mm_accdrv_process 12550 15.8 3.210 4.165 4.116 4.206 sum_up_and_integrate 119 10.3 0.037 0.041 3.523 3.528 integrate_v_rspace 119 11.3 0.003 0.003 3.486 3.491 init_scf_run 11 5.9 0.000 0.001 3.303 3.304 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.303 3.303 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.176 3.184 calculate_rho_elec 119 8.7 0.059 0.061 3.176 3.183 qs_ot_get_p 119 10.4 0.001 0.001 2.916 2.950 init_scf_loop 11 6.9 0.000 0.000 2.613 2.614 mp_waitall_1 121218 16.5 1.645 2.171 1.645 2.171 calculate_first_density_matrix 1 7.0 0.000 0.000 2.140 2.141 make_m2s 4572 13.5 0.034 0.035 1.814 1.974 grid_integrate_task_list 119 12.3 1.853 1.925 1.853 1.925 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.907 1.909 prepare_preconditioner 11 7.9 0.000 0.000 1.885 1.889 make_preconditioner 11 8.9 0.000 0.000 1.885 1.889 make_images 4572 14.5 0.269 0.302 1.725 1.884 jit_kernel_multiply 10 15.5 0.867 1.814 0.867 1.814 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.776 1.803 rs_pw_transfer 974 11.9 0.008 0.008 1.665 1.781 calculate_dm_sparse 119 9.5 0.000 0.000 1.757 1.778 density_rs2pw 119 9.7 0.004 0.004 1.602 1.712 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.704 1.704 pw_transfer 1439 11.6 0.063 0.065 1.666 1.674 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.601 1.614 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.576 1.584 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.575 1.576 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.407 1.408 cp_fm_redistribute_end 50 14.0 0.699 1.381 0.700 1.382 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.359 1.371 grid_collocate_task_list 119 9.7 1.272 1.368 1.272 1.368 cp_fm_diag_elpa_base 50 14.0 0.636 1.304 0.680 1.362 potential_pw2rs 119 12.3 0.010 0.010 1.326 1.330 ot_diis_step 108 11.5 0.013 0.013 1.296 1.296 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.254 1.265 fft3d_ps 1201 14.6 0.535 0.547 1.243 1.251 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.223 1.223 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.191 1.204 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.165 1.181 apply_single 119 13.6 0.000 0.000 1.165 1.181 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.168 1.173 wfi_extrapolate 11 7.9 0.001 0.001 1.116 1.116 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.784 1.033 make_images_data 4572 15.5 0.038 0.042 0.820 1.016 cp_fm_cholesky_invert 11 10.9 0.976 0.980 0.976 0.980 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.915 acc_transpose_blocks 9144 15.5 0.038 0.040 0.901 0.908 mp_alltoall_d11v 2130 13.8 0.787 0.882 0.787 0.882 multiply_cannon_sync_h2d 9144 15.5 0.709 0.782 0.709 0.782 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.757 0.765 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.741 0.743 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.727 mp_allgather_i34 2286 14.5 0.282 0.703 0.282 0.703 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.658 0.661 dbcsr_complete_redistribute 329 12.2 0.179 0.204 0.602 0.632 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.309 0.631 mp_alltoall_z22v 1201 16.6 0.583 0.612 0.583 0.612 qs_create_task_list 11 7.9 0.009 0.010 0.556 0.583 generate_qs_task_list 11 8.9 0.190 0.212 0.547 0.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.178000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=577.000000, yerr=1.858641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 760.229888E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 41.680 41.681 qs_mol_dyn_low 1 2.0 0.003 0.003 41.469 41.478 qs_forces 11 3.9 0.002 0.002 41.404 41.404 qs_energies 11 4.9 0.001 0.001 39.450 39.453 scf_env_do_scf 11 5.9 0.001 0.001 33.663 33.664 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.535 25.537 velocity_verlet 10 3.0 0.002 0.002 23.358 23.364 dbcsr_multiply_generic 2286 12.5 0.101 0.103 17.891 18.078 qs_scf_new_mos 108 7.5 0.001 0.001 16.574 16.672 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.574 16.671 ot_scf_mini 108 9.5 0.002 0.002 15.479 15.579 multiply_cannon 2286 13.5 0.304 0.312 13.765 14.748 multiply_cannon_loop 2286 14.5 0.346 0.350 12.447 13.409 ot_mini 108 10.5 0.001 0.001 9.249 9.369 multiply_cannon_multrec 9144 15.5 3.520 5.032 8.663 8.874 init_scf_loop 11 6.9 0.000 0.000 8.099 8.103 qs_ot_get_derivative 108 11.5 0.001 0.001 7.210 7.310 prepare_preconditioner 11 7.9 0.000 0.000 7.170 7.183 make_preconditioner 11 8.9 0.000 0.000 7.170 7.183 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.687 7.061 rebuild_ks_matrix 119 8.3 0.000 0.000 6.854 6.997 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.853 6.997 dbcsr_mm_accdrv_process 12550 15.8 4.092 5.682 5.019 6.454 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.202 6.333 cp_fm_upper_to_full 72 14.2 3.241 4.667 3.241 4.667 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.792 3.851 calculate_rho_elec 119 8.7 0.118 0.121 3.792 3.851 mp_waitall_1 97218 16.6 2.708 3.806 2.708 3.806 init_scf_run 11 5.9 0.000 0.001 3.756 3.756 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.756 3.756 sum_up_and_integrate 119 10.3 0.064 0.065 3.712 3.717 integrate_v_rspace 119 11.3 0.003 0.003 3.648 3.655 qs_ot_get_p 119 10.4 0.001 0.001 3.466 3.601 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.516 2.996 dbcsr_complete_redistribute 329 12.2 0.290 0.293 2.080 2.972 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.733 2.612 make_m2s 4572 13.5 0.037 0.038 2.388 2.581 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.255 2.527 apply_single 119 13.6 0.000 0.000 2.255 2.527 make_images 4572 14.5 0.353 0.383 2.267 2.460 mp_alltoall_i22 627 13.8 1.503 2.389 1.503 2.389 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.477 2.352 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.339 2.328 calculate_first_density_matrix 1 7.0 0.000 0.000 2.244 2.308 calculate_dm_sparse 119 9.5 0.000 0.000 2.218 2.277 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.133 2.133 pw_transfer 1439 11.6 0.066 0.067 2.069 2.072 grid_integrate_task_list 119 12.3 2.013 2.032 2.013 2.032 ot_diis_step 108 11.5 0.014 0.014 2.011 2.011 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.974 1.976 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.937 1.938 mp_sum_l 7207 12.9 1.197 1.908 1.197 1.908 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.853 1.853 density_rs2pw 119 9.7 0.003 0.003 1.834 1.847 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.739 1.789 jit_kernel_multiply 10 15.4 0.899 1.773 0.899 1.773 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.754 1.755 cp_fm_cholesky_invert 11 10.9 1.634 1.637 1.634 1.637 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.628 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.620 1.623 fft3d_ps 1201 14.6 0.567 0.579 1.607 1.611 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.525 1.526 cp_fm_diag_elpa_base 50 14.0 1.375 1.431 1.523 1.523 grid_collocate_task_list 119 9.7 1.453 1.478 1.453 1.478 rs_pw_transfer 974 11.9 0.009 0.009 1.434 1.476 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.180 1.471 make_images_data 4572 15.5 0.042 0.045 1.142 1.391 wfi_extrapolate 11 7.9 0.001 0.001 1.390 1.390 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.275 1.334 mp_alltoall_d11v 2130 13.8 1.246 1.320 1.246 1.320 potential_pw2rs 119 12.3 0.014 0.014 1.251 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.164 1.185 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.094 1.111 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.047 multiply_cannon_sync_h2d 9144 15.5 1.037 1.041 1.037 1.041 qs_create_task_list 11 7.9 0.008 0.010 0.952 0.963 generate_qs_task_list 11 8.9 0.372 0.391 0.943 0.954 mp_alltoall_z22v 1201 16.6 0.904 0.926 0.904 0.926 acc_transpose_blocks 9144 15.5 0.038 0.038 0.898 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.864 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.681000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.909091, yerr=14.575537 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.264960E+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 343226. 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 0.014 0.029 83.241 83.242 qs_mol_dyn_low 1 2.0 0.003 0.004 82.966 82.977 qs_forces 11 3.9 0.002 0.003 82.889 82.890 qs_energies 11 4.9 0.001 0.002 80.018 80.037 scf_env_do_scf 11 5.9 0.000 0.001 71.226 71.229 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.590 65.591 dbcsr_multiply_generic 2055 12.4 0.105 0.108 52.586 52.818 qs_scf_new_mos 99 7.5 0.000 0.001 48.745 48.887 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.745 48.887 ot_scf_mini 99 9.5 0.002 0.002 46.282 46.380 multiply_cannon 2055 13.4 0.177 0.183 42.638 43.381 velocity_verlet 10 3.0 0.001 0.002 42.650 42.651 multiply_cannon_loop 2055 14.4 1.575 1.623 41.652 42.455 ot_mini 99 10.5 0.001 0.001 28.179 28.280 qs_ot_get_derivative 99 11.5 0.001 0.001 21.335 21.448 multiply_cannon_multrec 49320 15.4 12.024 12.865 17.065 17.935 rebuild_ks_matrix 110 8.3 0.000 0.001 14.360 14.476 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.360 14.476 mp_waitall_1 241148 16.1 12.020 12.961 12.020 12.961 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.630 12.738 multiply_cannon_sync_h2d 49320 15.4 9.854 10.479 9.854 10.479 qs_ot_get_p 110 10.4 0.001 0.001 9.505 9.652 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.121 8.398 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.409 7.937 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.137 7.629 apply_single 110 13.6 0.000 0.001 7.136 7.629 sum_up_and_integrate 110 10.3 0.036 0.042 6.975 6.987 integrate_v_rspace 110 11.3 0.002 0.003 6.938 6.959 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.728 6.778 init_scf_run 11 5.9 0.000 0.001 6.726 6.727 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.726 6.727 ot_diis_step 99 11.5 0.006 0.006 6.507 6.507 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.179 6.337 calculate_rho_elec 110 8.6 0.020 0.024 6.178 6.337 mp_sum_l 6514 12.8 5.563 6.294 5.563 6.294 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.126 6.148 init_scf_loop 11 6.9 0.000 0.000 5.604 5.605 dbcsr_mm_accdrv_process 87628 16.1 2.104 2.202 4.923 5.221 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.109 5.109 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.581 4.582 cp_fm_redistribute_end 48 14.0 3.987 4.550 3.991 4.552 cp_fm_diag_elpa_base 48 14.0 0.554 4.445 0.558 4.472 rs_pw_transfer 902 11.9 0.012 0.014 3.796 4.054 wfi_extrapolate 11 7.9 0.001 0.001 4.046 4.046 make_m2s 4110 13.4 0.061 0.065 3.928 4.023 make_images 4110 14.4 0.178 0.192 3.832 3.930 calculate_dm_sparse 110 9.5 0.000 0.001 3.823 3.925 multiply_cannon_metrocomm1 49320 15.4 0.061 0.065 2.773 3.795 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.539 3.543 prepare_preconditioner 11 7.9 0.000 0.000 3.464 3.483 make_preconditioner 11 8.9 0.000 0.000 3.464 3.483 density_rs2pw 110 9.6 0.004 0.004 3.293 3.477 grid_integrate_task_list 110 12.3 3.261 3.454 3.261 3.454 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.376 3.421 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.273 3.318 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.182 3.218 pw_transfer 1331 11.6 0.054 0.064 2.863 2.931 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.776 2.847 potential_pw2rs 110 12.3 0.006 0.007 2.618 2.642 calculate_first_density_matrix 1 7.0 0.000 0.000 2.586 2.598 mp_alltoall_d11v 2046 13.8 2.004 2.583 2.004 2.583 jit_kernel_multiply 13 15.9 2.531 2.573 2.531 2.573 fft3d_ps 1111 14.6 0.752 0.844 2.314 2.373 fft_wrap_pw1pw2_140 451 13.1 0.168 0.187 2.283 2.355 acc_transpose_blocks 49320 15.4 0.214 0.224 2.184 2.257 grid_collocate_task_list 110 9.6 2.086 2.220 2.086 2.220 mp_waitany 14300 13.8 1.805 2.030 1.805 2.030 mp_sum_d 3883 11.9 1.600 2.016 1.600 2.016 cp_fm_cholesky_invert 11 10.9 1.951 1.955 1.951 1.955 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.871 1.885 make_images_data 4110 15.4 0.043 0.046 1.717 1.838 hybrid_alltoall_any 4261 16.3 0.082 0.477 1.515 1.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.242000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.454545, yerr=2.641062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.727424E+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 1819761. 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 0.020 0.060 67.641 67.642 qs_mol_dyn_low 1 2.0 0.003 0.003 67.322 67.333 qs_forces 11 3.9 0.002 0.002 67.251 67.252 qs_energies 11 4.9 0.002 0.009 63.991 63.997 scf_env_do_scf 11 5.9 0.001 0.001 55.610 55.614 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 47.984 47.985 dbcsr_multiply_generic 2055 12.4 0.112 0.117 37.484 37.659 velocity_verlet 10 3.0 0.001 0.002 35.599 35.600 qs_scf_new_mos 99 7.5 0.001 0.001 32.562 32.721 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.562 32.721 multiply_cannon 2055 13.4 0.223 0.244 30.981 32.273 ot_scf_mini 99 9.5 0.003 0.003 30.920 31.073 multiply_cannon_loop 2055 14.4 0.939 0.961 29.703 30.740 ot_mini 99 10.5 0.001 0.001 18.169 18.320 multiply_cannon_multrec 24660 15.4 7.622 9.213 13.826 15.317 rebuild_ks_matrix 110 8.3 0.000 0.001 13.357 13.507 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.356 13.507 qs_ot_get_derivative 99 11.5 0.001 0.001 12.355 12.506 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.784 11.919 mp_waitall_1 186928 16.3 7.634 9.948 7.634 9.948 multiply_cannon_sync_h2d 24660 15.4 7.050 7.921 7.050 7.921 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.166 7.920 init_scf_loop 11 6.9 0.001 0.005 7.590 7.591 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.519 7.309 apply_single 110 13.6 0.000 0.001 6.519 7.309 sum_up_and_integrate 110 10.3 0.052 0.060 6.340 6.351 dbcsr_mm_accdrv_process 52282 16.1 4.670 5.569 6.045 6.330 qs_ot_get_p 110 10.4 0.001 0.001 6.147 6.310 integrate_v_rspace 110 11.3 0.002 0.002 6.288 6.299 init_scf_run 11 5.9 0.000 0.001 6.034 6.035 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.034 6.035 ot_diis_step 99 11.5 0.010 0.010 5.765 5.765 prepare_preconditioner 11 7.9 0.000 0.000 5.597 5.618 make_preconditioner 11 8.9 0.000 0.002 5.597 5.618 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.526 5.536 calculate_rho_elec 110 8.6 0.039 0.047 5.526 5.535 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.674 5.403 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.160 5.324 make_m2s 4110 13.4 0.056 0.059 4.027 4.463 make_images 4110 14.4 0.401 0.444 3.919 4.352 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.237 4.257 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.805 3.805 wfi_extrapolate 11 7.9 0.001 0.001 3.465 3.465 grid_integrate_task_list 110 12.3 3.143 3.308 3.143 3.308 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.178 3.273 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.266 3.268 cp_fm_redistribute_end 48 14.0 2.434 3.239 2.436 3.239 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.232 3.233 pw_transfer 1331 11.6 0.065 0.070 3.081 3.213 cp_fm_diag_elpa_base 48 14.0 0.768 3.116 0.799 3.202 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.975 3.105 density_rs2pw 110 9.6 0.004 0.004 2.883 3.077 rs_pw_transfer 902 11.9 0.012 0.013 2.834 3.040 calculate_dm_sparse 110 9.5 0.001 0.001 2.941 2.971 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.875 2.932 make_images_data 4110 15.4 0.047 0.051 2.200 2.639 fft_wrap_pw1pw2_140 451 13.1 0.200 0.219 2.494 2.627 cp_fm_cholesky_invert 11 10.9 2.605 2.612 2.605 2.612 hybrid_alltoall_any 4261 16.3 0.102 0.441 1.916 2.601 calculate_first_density_matrix 1 7.0 0.000 0.002 2.486 2.487 mp_sum_l 6514 12.8 1.885 2.482 1.885 2.482 fft3d_ps 1111 14.6 1.075 1.274 2.332 2.450 grid_collocate_task_list 110 9.6 2.119 2.297 2.119 2.297 potential_pw2rs 110 12.3 0.008 0.009 2.278 2.290 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.947 1.968 mp_alltoall_d11v 2046 13.8 1.637 1.859 1.637 1.859 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.838 1.839 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.769 1.771 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.738 1.748 jit_kernel_multiply 9 16.3 1.018 1.726 1.018 1.726 mp_irecv_dv 57340 16.2 0.644 1.651 0.644 1.651 acc_transpose_blocks 24660 15.4 0.107 0.110 1.526 1.549 mp_allgather_i34 2055 14.4 0.602 1.546 0.602 1.546 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.530 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.475 dbcsr_complete_redistribute 325 12.2 0.242 0.295 1.161 1.426 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=67.642000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.181818, yerr=6.899515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.148800E+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 0.015 0.031 59.359 59.361 qs_mol_dyn_low 1 2.0 0.003 0.003 59.093 59.103 qs_forces 11 3.9 0.002 0.002 59.020 59.021 qs_energies 11 4.9 0.001 0.002 55.881 55.884 scf_env_do_scf 11 5.9 0.001 0.001 48.050 48.050 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.420 39.420 velocity_verlet 10 3.0 0.002 0.002 32.181 32.183 dbcsr_multiply_generic 2055 12.4 0.107 0.109 28.285 28.501 qs_scf_new_mos 99 7.5 0.001 0.001 25.103 25.192 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.102 25.191 ot_scf_mini 99 9.5 0.002 0.003 23.894 24.002 multiply_cannon 2055 13.4 0.214 0.227 22.047 23.362 multiply_cannon_loop 2055 14.4 0.625 0.641 20.782 21.769 ot_mini 99 10.5 0.001 0.001 13.453 13.562 rebuild_ks_matrix 110 8.3 0.000 0.000 12.085 12.192 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.085 12.191 multiply_cannon_multrec 16440 15.4 4.031 5.248 9.777 10.869 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.668 10.761 mp_waitall_1 146766 16.3 6.896 9.951 6.896 9.951 qs_ot_get_derivative 99 11.5 0.001 0.001 9.059 9.172 init_scf_loop 11 6.9 0.000 0.000 8.595 8.596 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.102 6.954 prepare_preconditioner 11 7.9 0.000 0.000 6.831 6.852 make_preconditioner 11 8.9 0.000 0.000 6.831 6.852 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.177 6.520 sum_up_and_integrate 110 10.3 0.060 0.060 6.364 6.384 integrate_v_rspace 110 11.3 0.002 0.003 6.304 6.324 dbcsr_mm_accdrv_process 34862 16.1 4.714 5.301 5.600 5.786 qs_ot_get_p 110 10.4 0.001 0.001 5.617 5.733 init_scf_run 11 5.9 0.000 0.001 5.475 5.475 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.475 5.475 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.405 5.415 calculate_rho_elec 110 8.6 0.058 0.058 5.404 5.415 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.919 5.341 apply_single 110 13.6 0.000 0.000 4.919 5.341 make_m2s 4110 13.4 0.050 0.052 4.134 4.535 make_images 4110 14.4 0.395 0.516 4.019 4.420 ot_diis_step 99 11.5 0.011 0.011 4.364 4.365 multiply_cannon_sync_h2d 16440 15.4 3.689 4.152 3.689 4.152 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.000 4.005 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.106 3.746 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.657 3.657 grid_integrate_task_list 110 12.3 3.187 3.377 3.187 3.377 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.115 3.116 cp_fm_redistribute_end 48 14.0 1.936 3.086 1.938 3.086 pw_transfer 1331 11.6 0.064 0.071 3.040 3.052 cp_fm_diag_elpa_base 48 14.0 1.084 2.948 1.143 3.051 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.934 2.948 density_rs2pw 110 9.6 0.004 0.004 2.720 2.877 make_images_data 4110 15.4 0.044 0.048 2.372 2.857 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.089 2.828 rs_pw_transfer 902 11.9 0.010 0.012 2.671 2.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.806 2.807 cp_fm_cholesky_invert 11 10.9 2.744 2.750 2.744 2.750 calculate_dm_sparse 110 9.5 0.001 0.001 2.524 2.549 fft_wrap_pw1pw2_140 451 13.1 0.210 0.216 2.482 2.495 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.441 2.481 calculate_first_density_matrix 1 7.0 0.000 0.001 2.438 2.439 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.328 2.385 grid_collocate_task_list 110 9.6 2.173 2.337 2.173 2.337 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.848 2.306 fft3d_ps 1111 14.6 1.062 1.071 2.246 2.261 potential_pw2rs 110 12.3 0.010 0.010 2.219 2.229 mp_irecv_dv 48980 15.7 0.777 2.183 0.777 2.183 mp_sum_l 6514 12.8 1.487 2.060 1.487 2.060 mp_alltoall_d11v 2046 13.8 1.750 1.991 1.750 1.991 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.929 1.929 dbcsr_complete_redistribute 325 12.2 0.341 0.406 1.455 1.912 cp_fm_upper_to_full 70 13.6 1.371 1.825 1.371 1.825 cp_fm_cholesky_decompose 22 10.9 1.629 1.650 1.629 1.650 mp_allgather_i34 2055 14.4 0.583 1.648 0.583 1.648 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.629 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.495 1.505 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.349 1.474 jit_kernel_multiply 8 16.5 0.492 1.464 0.492 1.464 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.013 1.454 mp_waitany 17072 13.8 1.154 1.322 1.154 1.322 acc_transpose_blocks 16440 15.4 0.072 0.076 1.231 1.268 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.259 1.267 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.361000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.818182, yerr=9.113608 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 739.995648E+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 0.014 0.034 64.357 64.358 qs_mol_dyn_low 1 2.0 0.003 0.003 64.077 64.090 qs_forces 11 3.9 0.002 0.002 64.004 64.005 qs_energies 11 4.9 0.001 0.001 60.670 60.675 scf_env_do_scf 11 5.9 0.000 0.001 52.193 52.196 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.403 40.403 velocity_verlet 10 3.0 0.002 0.002 36.427 36.429 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.776 29.978 qs_scf_new_mos 99 7.5 0.001 0.001 26.345 26.453 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.345 26.452 ot_scf_mini 99 9.5 0.003 0.003 24.737 24.836 multiply_cannon 2055 13.4 0.245 0.258 22.847 23.954 multiply_cannon_loop 2055 14.4 0.889 0.908 21.404 21.952 ot_mini 99 10.5 0.001 0.001 14.110 14.231 multiply_cannon_multrec 24660 15.4 4.226 6.946 13.006 14.119 rebuild_ks_matrix 110 8.3 0.000 0.000 11.715 11.808 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.714 11.807 init_scf_loop 11 6.9 0.000 0.000 11.744 11.745 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.372 10.456 qs_ot_get_derivative 99 11.5 0.001 0.001 9.977 10.081 prepare_preconditioner 11 7.9 0.000 0.000 10.048 10.063 make_preconditioner 11 8.9 0.000 0.000 10.048 10.063 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.242 9.740 dbcsr_mm_accdrv_process 52304 16.0 7.103 8.835 8.632 9.622 sum_up_and_integrate 110 10.3 0.068 0.071 6.203 6.217 mp_waitall_1 126806 16.4 4.344 6.182 4.344 6.182 integrate_v_rspace 110 11.3 0.003 0.003 6.135 6.150 qs_ot_get_p 110 10.4 0.001 0.001 5.651 5.781 init_scf_run 11 5.9 0.000 0.001 5.778 5.778 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.778 5.778 make_m2s 4110 13.4 0.059 0.061 5.341 5.708 make_images 4110 14.4 0.580 0.701 5.201 5.565 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.406 5.415 calculate_rho_elec 110 8.6 0.077 0.081 5.406 5.414 cp_fm_upper_to_full 70 13.8 3.358 4.780 3.358 4.780 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.014 4.101 apply_single 110 13.6 0.000 0.000 4.014 4.101 ot_diis_step 99 11.5 0.011 0.012 4.099 4.100 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.860 3.874 dbcsr_complete_redistribute 325 12.2 0.418 0.464 2.699 3.871 grid_integrate_task_list 110 12.3 3.273 3.438 3.273 3.438 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.381 3.382 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.317 3.364 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.154 3.326 multiply_cannon_sync_h2d 24660 15.4 3.161 3.302 3.161 3.302 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.446 3.135 calculate_dm_sparse 110 9.5 0.001 0.001 3.069 3.100 pw_transfer 1331 11.6 0.064 0.072 3.052 3.077 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.008 3.069 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.231 3.051 make_images_data 4110 15.4 0.046 0.050 2.682 3.036 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.004 3.006 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.947 2.975 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.797 2.946 wfi_extrapolate 11 7.9 0.001 0.001 2.933 2.933 mp_alltoall_i22 605 13.7 1.680 2.914 1.680 2.914 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.824 2.827 density_rs2pw 110 9.6 0.004 0.004 2.633 2.806 cp_fm_redistribute_end 48 14.0 1.407 2.793 1.408 2.794 cp_fm_diag_elpa_base 48 14.0 1.302 2.657 1.383 2.770 calculate_first_density_matrix 1 7.0 0.000 0.000 2.762 2.763 cp_fm_cholesky_invert 11 10.9 2.739 2.747 2.739 2.747 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.651 2.685 rs_pw_transfer 902 11.9 0.010 0.011 2.359 2.549 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 2.490 2.522 grid_collocate_task_list 110 9.6 2.222 2.342 2.222 2.342 jit_kernel_multiply 11 15.6 1.195 2.335 1.195 2.335 fft3d_ps 1111 14.6 1.057 1.091 2.249 2.268 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.208 2.209 potential_pw2rs 110 12.3 0.013 0.013 2.035 2.043 mp_alltoall_d11v 2046 13.8 1.653 1.874 1.653 1.874 cp_fm_cholesky_decompose 22 10.9 1.693 1.738 1.693 1.738 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.689 1.721 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.700 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.698 acc_transpose_blocks 24660 15.4 0.104 0.106 1.623 1.662 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.598 mp_allgather_i34 2055 14.4 0.542 1.580 0.542 1.580 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.847 1.452 mp_sum_l 6514 12.8 0.900 1.443 0.900 1.443 mp_irecv_dv 62702 16.1 0.748 1.376 0.748 1.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.358000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.090909, yerr=13.296989 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 830.423040E+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 0.030 0.072 53.964 53.981 qs_mol_dyn_low 1 2.0 0.003 0.004 53.706 53.716 qs_forces 11 3.9 0.002 0.002 53.631 53.632 qs_energies 11 4.9 0.001 0.001 50.049 50.054 scf_env_do_scf 11 5.9 0.000 0.001 42.022 42.022 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.318 34.318 velocity_verlet 10 3.0 0.002 0.002 30.239 30.243 dbcsr_multiply_generic 2055 12.4 0.104 0.105 22.478 22.660 qs_scf_new_mos 99 7.5 0.001 0.001 20.472 20.528 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.471 20.528 ot_scf_mini 99 9.5 0.002 0.002 19.245 19.273 multiply_cannon 2055 13.4 0.248 0.265 17.187 18.322 multiply_cannon_loop 2055 14.4 0.324 0.337 15.854 16.175 rebuild_ks_matrix 110 8.3 0.000 0.000 11.263 11.287 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.263 11.286 ot_mini 99 10.5 0.001 0.001 10.360 10.377 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.035 10.057 multiply_cannon_multrec 8220 15.4 3.201 4.479 7.442 8.389 init_scf_loop 11 6.9 0.000 0.000 7.657 7.659 mp_waitall_1 106626 16.5 5.874 7.577 5.874 7.577 qs_ot_get_derivative 99 11.5 0.001 0.001 6.579 6.604 sum_up_and_integrate 110 10.3 0.079 0.081 6.106 6.118 prepare_preconditioner 11 7.9 0.000 0.000 6.037 6.043 make_preconditioner 11 8.9 0.000 0.000 6.037 6.043 integrate_v_rspace 110 11.3 0.003 0.003 6.026 6.038 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.619 5.696 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.497 5.515 calculate_rho_elec 110 8.6 0.115 0.115 5.496 5.514 qs_ot_get_p 110 10.4 0.001 0.001 5.070 5.101 dbcsr_mm_accdrv_process 17442 15.9 2.813 3.649 4.111 5.032 init_scf_run 11 5.9 0.000 0.001 4.937 4.937 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.937 4.937 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.958 4.285 make_m2s 4110 13.4 0.038 0.040 4.014 4.267 make_images 4110 14.4 0.641 0.700 3.885 4.138 ot_diis_step 99 11.5 0.012 0.012 3.758 3.758 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.679 3.714 apply_single 110 13.6 0.000 0.000 3.679 3.713 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.698 3.702 grid_integrate_task_list 110 12.3 3.364 3.523 3.364 3.523 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.396 3.396 pw_transfer 1331 11.6 0.064 0.070 3.171 3.183 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.065 3.080 multiply_cannon_sync_h2d 8220 15.4 2.920 3.047 2.920 3.047 cp_fm_cholesky_invert 11 10.9 2.926 2.931 2.926 2.931 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.862 2.862 cp_fm_redistribute_end 48 14.0 0.726 2.832 0.730 2.833 cp_fm_diag_elpa_base 48 14.0 1.918 2.637 2.095 2.798 density_rs2pw 110 9.6 0.004 0.004 2.586 2.696 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.648 2.649 wfi_extrapolate 11 7.9 0.001 0.001 2.637 2.637 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.620 2.636 make_images_data 4110 15.4 0.038 0.043 2.218 2.635 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.158 2.559 grid_collocate_task_list 110 9.6 2.328 2.489 2.328 2.489 calculate_dm_sparse 110 9.5 0.001 0.001 2.452 2.489 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.464 2.465 fft3d_ps 1111 14.6 1.117 1.144 2.321 2.334 rs_pw_transfer 902 11.9 0.010 0.010 2.116 2.245 calculate_first_density_matrix 1 7.0 0.000 0.000 2.214 2.214 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.125 2.137 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.990 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.880 1.901 potential_pw2rs 110 12.3 0.015 0.016 1.876 1.884 mp_alltoall_d11v 2046 13.8 1.566 1.829 1.566 1.829 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.788 1.801 cp_fm_cholesky_decompose 22 10.9 1.672 1.696 1.672 1.696 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.489 1.609 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.559 1.562 dbcsr_complete_redistribute 325 12.2 0.561 0.595 1.428 1.521 mp_allgather_i34 2055 14.4 0.512 1.474 0.512 1.474 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.435 qs_create_task_list 11 7.9 0.000 0.000 1.220 1.321 generate_qs_task_list 11 8.9 0.378 0.447 1.220 1.320 jit_kernel_multiply 7 15.9 0.985 1.207 0.985 1.207 mp_waitany 9240 13.8 1.022 1.172 1.022 1.172 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.122 1.140 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.481 1.123 mp_irecv_dv 24056 15.7 0.458 1.084 0.458 1.084 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.981000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=784.909091, yerr=12.153971 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 1.379815E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.037 86.105 86.106 qs_mol_dyn_low 1 2.0 0.003 0.003 85.808 85.817 qs_forces 11 3.9 0.002 0.002 85.737 85.737 qs_energies 11 4.9 0.001 0.001 81.716 81.718 scf_env_do_scf 11 5.9 0.001 0.001 71.849 71.849 velocity_verlet 10 3.0 0.005 0.006 55.482 55.489 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.716 42.717 init_scf_loop 11 6.9 0.000 0.000 29.058 29.059 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.586 28.682 prepare_preconditioner 11 7.9 0.000 0.000 27.119 27.128 make_preconditioner 11 8.9 0.000 0.000 27.119 27.128 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.119 26.600 qs_scf_new_mos 99 7.5 0.001 0.001 26.398 26.453 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.397 26.452 ot_scf_mini 99 9.5 0.002 0.002 24.628 24.675 multiply_cannon 2055 13.4 0.344 0.374 21.437 22.307 multiply_cannon_loop 2055 14.4 0.343 0.347 19.572 19.943 cp_fm_upper_to_full 70 14.2 13.023 18.759 13.023 18.759 ot_mini 99 10.5 0.001 0.001 13.594 13.644 rebuild_ks_matrix 110 8.3 0.001 0.001 13.088 13.126 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.087 13.126 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.904 11.939 dbcsr_complete_redistribute 325 12.2 1.023 1.046 7.615 10.959 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.568 9.916 multiply_cannon_multrec 8220 15.4 4.347 4.550 9.553 9.626 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.984 9.304 qs_ot_get_derivative 99 11.5 0.001 0.001 9.000 9.042 mp_alltoall_i22 605 13.7 5.621 8.965 5.621 8.965 mp_waitall_1 87304 16.6 7.825 8.683 7.825 8.683 sum_up_and_integrate 110 10.3 0.151 0.153 6.579 6.595 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.447 6.481 calculate_rho_elec 110 8.6 0.227 0.227 6.447 6.481 integrate_v_rspace 110 11.3 0.004 0.004 6.428 6.446 make_m2s 4110 13.4 0.043 0.044 5.377 5.948 qs_ot_get_p 110 10.4 0.001 0.001 5.803 5.856 make_images 4110 14.4 0.880 0.924 5.190 5.760 init_scf_run 11 5.9 0.000 0.001 5.646 5.646 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.646 5.646 cp_fm_cholesky_invert 11 10.9 5.559 5.565 5.559 5.565 dbcsr_mm_accdrv_process 11614 15.7 3.312 3.663 5.063 5.286 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.730 5.179 apply_single 110 13.6 0.000 0.000 4.730 5.179 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.657 5.070 ot_diis_step 99 11.5 0.015 0.016 4.569 4.569 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.238 4.245 multiply_cannon_sync_h2d 8220 15.4 3.947 3.951 3.947 3.951 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.942 3.810 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.788 3.789 pw_transfer 1331 11.6 0.074 0.074 3.718 3.730 grid_integrate_task_list 110 12.3 3.655 3.714 3.655 3.714 make_images_data 4110 15.4 0.041 0.044 2.986 3.701 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.621 3.621 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.602 3.613 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.857 3.307 wfi_extrapolate 11 7.9 0.001 0.001 3.258 3.258 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.180 3.181 cp_fm_diag_elpa_base 48 14.0 2.629 2.832 3.179 3.179 calculate_dm_sparse 110 9.5 0.001 0.001 3.129 3.151 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.121 3.133 density_rs2pw 110 9.6 0.004 0.004 2.948 2.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.907 2.911 fft3d_ps 1111 14.6 1.265 1.282 2.812 2.824 grid_collocate_task_list 110 9.6 2.625 2.645 2.625 2.645 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.397 2.423 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.391 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.283 2.284 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.158 2.208 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.162 rs_pw_transfer 902 11.9 0.010 0.011 2.109 2.142 mp_alltoall_d11v 2046 13.8 2.013 2.081 2.013 2.081 cp_fm_cholesky_decompose 22 10.9 2.036 2.059 2.036 2.059 potential_pw2rs 110 12.3 0.021 0.021 2.036 2.041 qs_create_task_list 11 7.9 0.000 0.001 1.899 1.944 generate_qs_task_list 11 8.9 0.738 0.790 1.899 1.944 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.776 1.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.789 jit_kernel_multiply 10 15.3 1.547 1.776 1.547 1.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.106000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1225.000000, yerr=65.929715 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.895168E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955597216 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2753756. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.038 204.053 204.062 qs_mol_dyn_low 1 2.0 0.003 0.003 203.540 203.554 qs_forces 11 3.9 0.003 0.005 203.449 203.451 qs_energies 11 4.9 0.005 0.028 197.942 197.955 scf_env_do_scf 11 5.9 0.001 0.002 181.319 181.323 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.260 160.262 dbcsr_multiply_generic 2507 12.6 0.174 0.177 123.924 124.815 velocity_verlet 10 3.0 0.002 0.002 123.202 123.203 qs_scf_new_mos 117 7.6 0.001 0.001 121.534 121.739 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.533 121.738 ot_scf_mini 117 9.6 0.003 0.003 114.931 115.152 multiply_cannon 2507 13.6 0.239 0.247 100.646 102.725 multiply_cannon_loop 2507 14.6 2.128 2.197 98.424 100.590 ot_mini 117 10.6 0.001 0.001 65.429 65.674 multiply_cannon_multrec 60168 15.6 33.084 36.009 41.499 43.502 qs_ot_get_derivative 117 11.6 0.001 0.002 40.645 40.818 rebuild_ks_matrix 128 8.3 0.001 0.001 33.250 33.632 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.250 33.632 mp_waitall_1 291448 16.2 28.520 31.864 28.520 31.864 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.904 30.304 multiply_cannon_sync_h2d 60168 15.6 27.370 30.179 27.370 30.179 qs_ot_get_p 128 10.4 0.001 0.001 27.628 28.002 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.257 24.892 apply_single 128 13.6 0.001 0.001 24.257 24.892 ot_diis_step 117 11.6 0.007 0.008 24.421 24.422 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.931 21.014 init_scf_loop 11 6.9 0.000 0.000 20.980 20.981 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.788 18.940 multiply_cannon_metrocomm3 60168 15.6 0.111 0.119 15.608 18.432 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.263 18.264 prepare_preconditioner 11 7.9 0.000 0.000 16.408 16.453 make_preconditioner 11 8.9 0.000 0.000 16.408 16.453 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.573 15.741 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.389 15.395 cp_fm_redistribute_end 83 14.4 12.147 15.312 12.161 15.318 cp_fm_diag_elpa_base 83 14.4 3.109 15.025 3.142 15.138 make_m2s 5014 13.6 0.103 0.110 13.806 14.257 make_images 5014 14.6 0.403 0.422 13.627 14.088 sum_up_and_integrate 128 10.3 0.090 0.109 13.986 13.999 integrate_v_rspace 128 11.3 0.003 0.004 13.895 13.912 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.666 12.752 calculate_rho_elec 128 8.7 0.045 0.063 12.665 12.751 init_scf_run 11 5.9 0.000 0.001 12.309 12.309 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.308 12.309 cp_fm_cholesky_invert 11 10.9 9.428 9.436 9.428 9.436 mp_sum_l 7870 13.0 8.273 9.377 8.273 9.377 wfi_extrapolate 11 7.9 0.001 0.001 9.134 9.134 dbcsr_mm_accdrv_process 124484 16.2 3.341 3.506 7.973 8.535 calculate_dm_sparse 128 9.5 0.001 0.001 8.407 8.499 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.989 8.098 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.826 7.898 make_images_data 5014 15.6 0.069 0.075 6.831 7.735 multiply_cannon_metrocomm1 60168 15.6 0.088 0.092 5.878 7.704 grid_integrate_task_list 128 12.3 7.090 7.492 7.090 7.492 hybrid_alltoall_any 5200 16.5 0.291 2.269 5.973 7.314 pw_transfer 1547 11.6 0.074 0.108 6.796 7.066 density_rs2pw 128 9.7 0.006 0.007 6.458 6.977 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.593 6.835 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.656 6.664 rs_pw_transfer 1046 11.9 0.017 0.020 5.687 6.208 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.811 5.954 fft_wrap_pw1pw2_140 523 13.2 0.445 0.509 5.651 5.832 mp_alltoall_d11v 2415 14.1 4.305 5.715 4.305 5.715 fft3d_ps 1291 14.7 2.093 2.573 5.388 5.569 grid_collocate_task_list 128 9.7 4.724 5.176 4.724 5.176 cp_fm_cholesky_decompose 22 10.9 4.722 4.737 4.722 4.737 potential_pw2rs 128 12.3 0.010 0.011 4.467 4.495 mp_sum_d 4473 12.2 3.541 4.212 3.541 4.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.062000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.545455, yerr=7.831912 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 825.090048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 57094. MP_Allreduce 11272 945. MP_Sync 170 MP_Alltoall 1969 5754269. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 191.776 191.777 qs_mol_dyn_low 1 2.0 0.003 0.004 191.404 191.418 qs_forces 11 3.9 0.003 0.003 191.316 191.318 qs_energies 11 4.9 0.016 0.023 184.595 184.604 scf_env_do_scf 11 5.9 0.001 0.001 168.345 168.355 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.129 135.132 velocity_verlet 10 3.0 0.002 0.002 120.471 120.473 dbcsr_multiply_generic 2507 12.6 0.183 0.188 97.300 98.458 qs_scf_new_mos 117 7.6 0.001 0.001 96.425 96.844 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.424 96.843 ot_scf_mini 117 9.6 0.004 0.004 91.609 92.023 multiply_cannon 2507 13.6 0.488 0.553 77.390 81.895 multiply_cannon_loop 2507 14.6 1.267 1.308 74.025 76.450 ot_mini 117 10.6 0.001 0.001 49.868 50.287 mp_waitall_1 226760 16.4 24.494 37.778 24.494 37.778 multiply_cannon_multrec 30084 15.6 22.167 26.778 31.812 36.633 rebuild_ks_matrix 128 8.3 0.001 0.001 32.743 33.198 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.742 33.197 init_scf_loop 11 6.9 0.000 0.000 33.125 33.127 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.507 29.924 prepare_preconditioner 11 7.9 0.000 0.000 28.761 28.811 make_preconditioner 11 8.9 0.000 0.000 28.761 28.811 qs_ot_get_derivative 117 11.6 0.001 0.002 27.942 28.358 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.473 27.991 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.407 27.623 qs_ot_get_p 128 10.4 0.001 0.001 23.437 23.920 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.973 23.097 apply_single 128 13.6 0.001 0.001 21.973 23.097 multiply_cannon_sync_h2d 30084 15.6 19.425 21.832 19.425 21.832 ot_diis_step 117 11.6 0.014 0.015 21.758 21.760 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.615 18.651 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.454 17.455 cp_fm_cholesky_invert 11 10.9 16.862 16.875 16.862 16.875 make_m2s 5014 13.6 0.088 0.095 14.191 15.797 make_images 5014 14.6 1.166 1.345 13.983 15.589 sum_up_and_integrate 128 10.3 0.116 0.132 14.321 14.347 integrate_v_rspace 128 11.3 0.003 0.003 14.204 14.235 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.095 14.108 cp_fm_redistribute_end 83 14.4 8.228 14.018 8.239 14.019 cp_fm_diag_elpa_base 83 14.4 5.538 13.546 5.756 13.896 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.034 13.068 calculate_rho_elec 128 8.7 0.088 0.105 13.034 13.067 init_scf_run 11 5.9 0.000 0.001 11.454 11.455 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.454 11.455 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.048 11.359 make_images_data 5014 15.6 0.066 0.074 8.490 10.394 multiply_cannon_metrocomm4 27577 15.6 0.096 0.112 3.687 10.255 hybrid_alltoall_any 5200 16.5 0.342 1.508 7.333 9.926 mp_irecv_dv 69486 16.3 3.493 9.878 3.493 9.878 dbcsr_mm_accdrv_process 62242 16.2 4.583 5.241 9.102 9.701 wfi_extrapolate 11 7.9 0.001 0.001 8.385 8.385 pw_transfer 1547 11.6 0.085 0.099 7.730 7.808 grid_integrate_task_list 128 12.3 7.160 7.659 7.160 7.659 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.506 7.580 density_rs2pw 128 9.7 0.006 0.006 6.735 7.108 cp_fm_cholesky_decompose 22 10.9 7.021 7.095 7.021 7.095 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.235 7.023 calculate_dm_sparse 128 9.5 0.001 0.001 6.464 6.641 fft_wrap_pw1pw2_140 523 13.2 0.468 0.517 6.484 6.558 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.175 6.183 rs_pw_transfer 1046 11.9 0.015 0.016 5.628 6.090 fft3d_ps 1291 14.7 2.772 2.929 5.880 5.935 mp_sum_l 7870 13.0 3.939 5.682 3.939 5.682 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.365 5.431 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.275 5.423 grid_collocate_task_list 128 9.7 4.911 5.388 4.911 5.388 mp_allgather_i34 2507 14.6 1.919 4.957 1.919 4.957 mp_alltoall_d11v 2415 14.1 4.502 4.930 4.502 4.930 potential_pw2rs 128 12.3 0.014 0.016 4.834 4.850 mp_sum_d 4468 12.1 2.642 4.136 2.642 4.136 dbcsr_complete_redistribute 395 12.7 0.767 0.854 3.168 3.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.777000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.000000, yerr=2.296242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 972.935168E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 176.001 176.003 qs_mol_dyn_low 1 2.0 0.003 0.004 175.536 175.550 qs_forces 11 3.9 0.003 0.004 175.436 175.442 qs_energies 11 4.9 0.002 0.005 168.878 168.890 scf_env_do_scf 11 5.9 0.001 0.001 153.495 153.498 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.663 118.666 velocity_verlet 10 3.0 0.002 0.002 112.417 112.420 dbcsr_multiply_generic 2507 12.6 0.180 0.184 81.571 82.767 qs_scf_new_mos 117 7.6 0.001 0.001 82.212 82.550 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.211 82.549 ot_scf_mini 117 9.6 0.004 0.004 78.095 78.508 multiply_cannon 2507 13.6 0.501 0.521 61.749 65.896 multiply_cannon_loop 2507 14.6 0.867 0.893 58.535 61.334 ot_mini 117 10.6 0.001 0.001 42.645 43.088 init_scf_loop 11 6.9 0.000 0.002 34.728 34.737 mp_waitall_1 178456 16.5 25.010 34.608 25.010 34.608 prepare_preconditioner 11 7.9 0.000 0.000 30.711 30.766 make_preconditioner 11 8.9 0.000 0.002 30.711 30.766 rebuild_ks_matrix 128 8.3 0.001 0.001 30.083 30.543 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.083 30.542 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.402 29.714 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.101 27.511 multiply_cannon_multrec 20056 15.6 13.501 16.940 22.203 25.579 multiply_cannon_metrocomm3 20056 15.6 0.058 0.063 15.062 24.241 qs_ot_get_derivative 117 11.6 0.002 0.002 22.926 23.348 qs_ot_get_p 128 10.4 0.001 0.002 20.842 21.353 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.893 20.884 apply_single 128 13.6 0.001 0.001 19.893 20.884 ot_diis_step 117 11.6 0.018 0.018 19.615 19.616 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.514 16.522 multiply_cannon_sync_h2d 20056 15.6 14.327 16.072 14.327 16.072 make_m2s 5014 13.6 0.080 0.086 14.617 15.673 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.508 15.509 make_images 5014 14.6 1.183 1.269 14.385 15.434 cp_fm_cholesky_invert 11 10.9 14.437 14.446 14.437 14.446 sum_up_and_integrate 128 10.3 0.134 0.144 14.175 14.207 integrate_v_rspace 128 11.3 0.003 0.004 14.041 14.076 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.979 13.046 calculate_rho_elec 128 8.7 0.132 0.146 12.978 13.045 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.441 12.442 cp_fm_redistribute_end 83 14.4 4.690 12.372 4.705 12.374 cp_fm_diag_elpa_base 83 14.4 7.243 11.797 7.646 12.262 make_images_data 5014 15.6 0.061 0.069 8.964 10.413 init_scf_run 11 5.9 0.000 0.001 10.352 10.353 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.352 10.353 hybrid_alltoall_any 5200 16.5 0.433 1.981 7.627 9.901 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.445 9.368 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.024 9.332 mp_irecv_dv 50230 16.2 3.322 9.125 3.322 9.125 dbcsr_mm_accdrv_process 41502 16.2 4.403 5.265 8.159 8.302 pw_transfer 1547 11.6 0.085 0.106 7.793 7.913 grid_integrate_task_list 128 12.3 7.320 7.796 7.320 7.796 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.569 7.694 wfi_extrapolate 11 7.9 0.001 0.001 7.373 7.374 cp_fm_cholesky_decompose 22 10.9 7.251 7.322 7.251 7.322 cp_fm_upper_to_full 105 14.5 5.641 7.315 5.641 7.315 fft_wrap_pw1pw2_140 523 13.2 0.477 0.527 6.656 6.782 density_rs2pw 128 9.7 0.006 0.006 6.438 6.680 dbcsr_complete_redistribute 395 12.7 1.183 1.219 4.489 6.208 fft3d_ps 1291 14.7 2.695 2.899 5.835 5.910 calculate_dm_sparse 128 9.5 0.001 0.001 5.713 5.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.445 5.452 grid_collocate_task_list 128 9.7 5.072 5.443 5.072 5.443 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.559 5.284 rs_pw_transfer 1046 11.9 0.014 0.015 4.993 5.233 mp_allgather_i34 2507 14.6 1.727 5.075 1.727 5.075 mp_alltoall_d11v 2415 14.1 4.193 5.044 4.193 5.044 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.266 4.981 mp_sum_l 7870 13.0 3.386 4.804 3.386 4.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.539 4.663 potential_pw2rs 128 12.3 0.020 0.023 4.527 4.559 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.016 4.054 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.289 3.930 mp_alltoall_i22 716 14.1 1.902 3.760 1.902 3.760 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.732 3.733 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.491 3.541 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.003000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.909091, yerr=22.182936 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.140634E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56895. MP_Allreduce 11306 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 83 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.041 189.945 189.946 qs_mol_dyn_low 1 2.0 0.003 0.004 189.539 189.555 qs_forces 11 3.9 0.003 0.004 189.421 189.433 qs_energies 11 4.9 0.002 0.005 182.393 182.405 scf_env_do_scf 11 5.9 0.001 0.001 165.786 165.801 velocity_verlet 10 3.0 0.002 0.002 126.181 126.184 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.311 118.312 qs_scf_new_mos 118 7.6 0.001 0.001 82.491 82.754 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.490 82.754 dbcsr_multiply_generic 2527 12.6 0.191 0.198 79.843 80.479 ot_scf_mini 118 9.6 0.003 0.004 77.946 78.218 multiply_cannon 2527 13.6 0.563 0.596 54.973 57.556 multiply_cannon_loop 2527 14.6 1.193 1.226 51.125 52.737 init_scf_loop 11 6.9 0.000 0.002 47.348 47.350 prepare_preconditioner 11 7.9 0.000 0.000 43.253 43.285 make_preconditioner 11 8.9 0.000 0.002 43.253 43.285 ot_mini 118 10.6 0.001 0.001 42.689 42.929 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.869 41.951 multiply_cannon_multrec 30324 15.6 14.375 19.561 26.356 31.013 rebuild_ks_matrix 129 8.3 0.001 0.001 29.198 29.461 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.198 29.460 mp_waitall_1 154994 16.5 17.405 26.956 17.405 26.956 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.401 26.638 qs_ot_get_derivative 118 11.6 0.002 0.003 22.846 23.108 make_m2s 5054 13.6 0.096 0.100 20.420 21.670 qs_ot_get_p 129 10.4 0.001 0.001 21.375 21.635 make_images 5054 14.6 1.983 2.303 20.112 21.362 ot_diis_step 118 11.6 0.018 0.019 19.712 19.713 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.241 19.701 apply_single 129 13.6 0.001 0.001 19.241 19.701 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.168 17.221 cp_fm_cholesky_invert 11 10.9 16.789 16.800 16.789 16.800 cp_fm_upper_to_full 105 14.7 11.227 16.626 11.227 16.626 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.870 15.871 multiply_cannon_metrocomm3 30324 15.6 0.047 0.049 6.217 15.284 sum_up_and_integrate 129 10.3 0.142 0.154 14.169 14.191 integrate_v_rspace 129 11.3 0.003 0.004 14.027 14.056 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.264 13.325 calculate_rho_elec 129 8.7 0.176 0.192 13.263 13.324 make_images_data 5054 15.6 0.064 0.068 10.952 12.998 dbcsr_complete_redistribute 395 12.7 1.517 1.631 9.026 12.890 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.710 12.712 cp_fm_redistribute_end 83 14.4 2.181 12.619 2.196 12.623 multiply_cannon_sync_h2d 30324 15.6 11.895 12.616 11.895 12.616 cp_fm_diag_elpa_base 83 14.4 9.800 11.998 10.393 12.484 dbcsr_mm_accdrv_process 62756 16.2 7.351 8.464 11.555 12.088 hybrid_alltoall_any 5240 16.5 0.530 2.223 9.793 11.854 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.615 11.437 init_scf_run 11 5.9 0.000 0.001 10.669 10.671 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.669 10.670 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.361 10.049 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.415 9.614 mp_alltoall_i22 718 14.1 5.601 9.343 5.601 9.343 pw_transfer 1559 11.6 0.086 0.099 7.924 7.998 cp_fm_cholesky_decompose 22 10.9 7.783 7.961 7.783 7.961 grid_integrate_task_list 129 12.3 7.599 7.880 7.599 7.880 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.697 7.777 wfi_extrapolate 11 7.9 0.001 0.001 7.613 7.613 multiply_cannon_metrocomm4 25270 15.6 0.075 0.085 2.715 6.948 fft_wrap_pw1pw2_140 527 13.2 0.486 0.495 6.775 6.865 density_rs2pw 129 9.7 0.006 0.006 6.355 6.701 mp_irecv_dv 76685 16.2 2.570 6.679 2.570 6.679 calculate_dm_sparse 129 9.5 0.001 0.001 6.212 6.309 fft3d_ps 1301 14.7 2.819 2.894 5.915 5.993 grid_collocate_task_list 129 9.7 5.237 5.619 5.237 5.619 mp_alltoall_d11v 2423 14.1 4.797 5.486 4.797 5.486 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.405 5.456 rs_pw_transfer 1054 12.0 0.014 0.015 4.739 5.151 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.462 4.547 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.443 4.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.332 4.423 potential_pw2rs 129 12.3 0.022 0.023 4.350 4.368 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.246 4.308 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.946000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1075.454545, yerr=19.147103 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/19/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 5.820057E+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 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 1.495773E+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 716108490000 ------------------------------------------------------------------------------- - - - 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 3989 58394. MP_Allreduce 10970 1176. MP_Sync 86 MP_Alltoall 1700 18828174. 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 0.037 0.061 173.285 173.290 qs_mol_dyn_low 1 2.0 0.003 0.004 172.854 172.869 qs_forces 11 3.9 0.003 0.003 172.723 172.726 qs_energies 11 4.9 0.001 0.002 165.385 165.391 scf_env_do_scf 11 5.9 0.001 0.001 148.230 148.244 velocity_verlet 10 3.0 0.002 0.002 112.916 112.922 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 112.235 112.236 qs_scf_new_mos 116 7.6 0.001 0.001 76.556 76.660 qs_scf_loop_do_ot 116 8.6 0.001 0.001 76.556 76.659 dbcsr_multiply_generic 2485 12.5 0.179 0.185 74.161 74.701 ot_scf_mini 116 9.6 0.003 0.004 72.112 72.179 multiply_cannon 2485 13.5 0.582 0.630 54.025 58.770 multiply_cannon_loop 2485 14.5 0.446 0.461 49.447 50.457 ot_mini 116 10.6 0.001 0.001 39.728 39.787 init_scf_loop 11 6.9 0.000 0.000 35.841 35.843 mp_waitall_1 128490 16.6 26.305 33.513 26.305 33.513 prepare_preconditioner 11 7.9 0.000 0.000 31.993 32.026 make_preconditioner 11 8.9 0.000 0.000 31.993 32.026 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.838 30.101 rebuild_ks_matrix 127 8.3 0.001 0.001 28.528 28.622 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.527 28.621 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.029 26.111 multiply_cannon_multrec 9940 15.5 10.322 14.245 17.597 20.550 qs_ot_get_derivative 116 11.6 0.002 0.002 19.851 19.914 ot_diis_step 116 11.6 0.020 0.020 19.808 19.809 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.311 19.689 apply_single 127 13.6 0.001 0.001 19.311 19.688 qs_ot_get_p 127 10.4 0.001 0.001 19.482 19.567 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.037 19.247 make_m2s 4970 13.5 0.066 0.072 16.455 18.635 make_images 4970 14.5 2.303 2.667 16.151 18.324 cp_fm_cholesky_invert 11 10.9 18.191 18.197 18.191 18.197 qs_ot_p2m_diag 82 11.4 0.489 0.496 15.676 15.690 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.518 14.518 sum_up_and_integrate 127 10.3 0.178 0.188 14.069 14.114 integrate_v_rspace 127 11.3 0.004 0.004 13.890 13.944 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.275 13.322 calculate_rho_elec 127 8.7 0.256 0.266 13.275 13.321 make_images_data 4970 15.5 0.051 0.060 10.086 12.588 hybrid_alltoall_any 5155 16.4 0.837 3.777 9.857 12.143 multiply_cannon_sync_h2d 9940 15.5 11.559 12.100 11.559 12.100 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.422 11.425 cp_fm_diag_elpa_base 82 14.4 11.158 11.251 11.414 11.417 init_scf_run 11 5.9 0.000 0.001 10.326 10.326 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.326 10.326 cp_fm_cholesky_decompose 22 10.9 8.080 8.172 8.080 8.172 grid_integrate_task_list 127 12.3 7.703 8.062 7.703 8.062 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.963 8.009 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.653 7.973 pw_transfer 1535 11.6 0.083 0.089 7.930 7.948 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.710 7.730 mp_allgather_i34 2485 14.5 2.786 7.587 2.786 7.587 dbcsr_mm_accdrv_process 20590 16.1 2.546 3.432 6.908 7.522 wfi_extrapolate 11 7.9 0.001 0.001 7.365 7.365 fft_wrap_pw1pw2_140 519 13.2 0.496 0.516 6.739 6.766 density_rs2pw 127 9.7 0.005 0.006 6.027 6.248 calculate_dm_sparse 127 9.5 0.001 0.001 6.110 6.183 fft3d_ps 1281 14.7 2.702 2.769 5.890 5.918 dbcsr_complete_redistribute 393 12.7 2.142 2.200 5.353 5.781 grid_collocate_task_list 127 9.7 5.450 5.732 5.450 5.732 mp_alltoall_d11v 2401 14.1 4.929 5.704 4.929 5.704 qs_energies_init_hamiltonians 11 5.9 0.006 0.009 5.267 5.268 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.218 5.224 rs_pw_transfer 1038 11.9 0.013 0.013 4.087 4.358 potential_pw2rs 127 12.3 0.026 0.027 4.140 4.154 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.089 4.109 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.840 3.901 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.548 3.879 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.503 3.848 mp_irecv_dv 28618 15.9 1.805 3.834 1.805 3.834 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.676 3.709 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.607 3.690 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.492 3.506 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.290000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1393.454545, yerr=43.953206 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_performance_tests/20/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 11.696234E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.048292E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 59142. MP_Allreduce 11002 1514. MP_Sync 87 MP_Alltoall 1712 36974125. MP_ISendRecv 3584 218624. MP_Wait 11594 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.050 292.500 292.501 qs_mol_dyn_low 1 2.0 0.003 0.004 291.898 291.948 qs_forces 11 3.9 0.003 0.003 291.798 291.799 qs_energies 11 4.9 0.001 0.001 283.233 283.244 scf_env_do_scf 11 5.9 0.001 0.001 261.549 261.560 velocity_verlet 10 3.0 0.005 0.006 211.249 211.257 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.859 135.861 init_scf_loop 11 6.9 0.000 0.000 125.426 125.429 prepare_preconditioner 11 7.9 0.000 0.000 120.785 120.821 make_preconditioner 11 8.9 0.000 0.000 120.785 120.821 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.290 117.903 qs_scf_new_mos 117 7.6 0.001 0.001 94.089 94.261 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.088 94.261 ot_scf_mini 117 9.6 0.003 0.004 89.322 89.405 dbcsr_multiply_generic 2507 12.6 0.213 0.220 82.866 83.663 cp_fm_upper_to_full 105 14.8 53.051 75.472 53.051 75.472 multiply_cannon 2507 13.6 0.705 0.744 57.672 58.626 multiply_cannon_loop 2507 14.6 0.476 0.482 53.924 55.787 ot_mini 117 10.6 0.001 0.001 44.851 44.964 dbcsr_complete_redistribute 395 12.7 4.036 4.064 29.289 42.221 copy_fm_to_dbcsr 209 11.7 0.001 0.001 25.827 38.713 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.449 36.153 cp_fm_cholesky_invert 11 10.9 34.696 34.703 34.696 34.703 mp_alltoall_i22 716 14.1 21.291 34.435 21.291 34.435 rebuild_ks_matrix 128 8.3 0.001 0.001 33.269 33.414 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 33.268 33.413 mp_waitall_1 105466 16.7 27.513 32.144 27.513 32.144 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.016 31.150 qs_ot_get_p 128 10.4 0.001 0.001 29.052 29.156 qs_ot_p2m_diag 83 11.4 0.878 0.885 24.790 24.822 qs_ot_get_derivative 117 11.6 0.002 0.002 24.610 24.701 cp_dbcsr_syevd 83 12.4 0.005 0.006 23.010 23.012 make_m2s 5014 13.6 0.076 0.079 20.068 21.053 make_images 5014 14.6 3.780 3.954 19.590 20.581 ot_diis_step 117 11.6 0.022 0.024 20.194 20.195 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.640 19.642 cp_fm_diag_elpa_base 83 14.4 15.193 16.801 19.634 19.637 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.190 19.588 apply_single 128 13.6 0.001 0.001 19.190 19.588 multiply_cannon_metrocomm3 10028 15.6 0.023 0.023 17.785 19.341 multiply_cannon_multrec 10028 15.6 10.524 12.268 17.933 18.024 sum_up_and_integrate 128 10.3 0.321 0.323 15.712 15.806 multiply_cannon_sync_h2d 10028 15.6 15.649 15.681 15.649 15.681 integrate_v_rspace 128 11.3 0.004 0.004 15.391 15.486 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.974 14.997 calculate_rho_elec 128 8.7 0.483 0.484 14.973 14.996 hybrid_alltoall_any 5200 16.5 1.307 3.068 10.719 12.612 make_images_data 5014 15.6 0.059 0.064 10.709 12.480 init_scf_run 11 5.9 0.000 0.001 11.765 11.765 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.764 11.765 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.637 9.713 cp_fm_cholesky_decompose 22 10.9 9.335 9.373 9.335 9.373 dbcsr_mm_accdrv_process 20762 16.1 3.799 5.531 7.173 9.039 wfi_extrapolate 11 7.9 0.001 0.001 8.735 8.735 grid_integrate_task_list 128 12.3 8.546 8.710 8.546 8.710 pw_transfer 1547 11.6 0.091 0.092 8.342 8.353 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.108 8.119 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.921 7.922 fft_wrap_pw1pw2_140 523 13.2 0.540 0.546 7.125 7.140 mp_alltoall_d11v 2415 14.1 6.924 7.067 6.924 7.067 calculate_dm_sparse 128 9.5 0.001 0.001 6.622 6.746 grid_collocate_task_list 128 9.7 6.339 6.376 6.339 6.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.313 6.375 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.186 6.276 fft3d_ps 1291 14.7 2.762 2.777 6.148 6.159 density_rs2pw 128 9.7 0.005 0.005 6.076 6.110 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.501000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2690.272727, yerr=181.901617 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 1.260351E+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 2292695. 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 0.021 0.036 84.741 84.741 qs_energies 1 2.0 0.000 0.000 84.312 84.327 ls_scf 1 3.0 0.000 0.000 83.416 83.431 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.499 72.711 multiply_cannon 111 7.7 0.018 0.021 55.724 57.221 multiply_cannon_loop 111 8.7 0.214 0.228 52.338 54.089 ls_scf_main 1 4.0 0.000 0.000 52.271 52.271 density_matrix_trs4 2 5.0 0.002 0.003 46.890 46.997 ls_scf_init_scf 1 4.0 0.000 0.000 28.113 28.114 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.064 27.113 mp_waitall_1 11316 10.9 21.801 25.354 21.801 25.354 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.963 24.984 multiply_cannon_multrec 2664 9.7 8.216 9.044 15.588 17.397 multiply_cannon_sync_h2d 2664 9.7 14.013 15.627 14.013 15.627 make_m2s 222 7.7 0.008 0.011 13.103 13.587 make_images 222 8.7 0.099 0.110 13.081 13.566 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.130 12.694 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.345 8.403 make_images_data 222 9.7 0.004 0.005 7.649 8.209 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.620 6.988 7.924 hybrid_alltoall_any 227 10.6 0.215 1.830 6.573 7.922 dbcsr_mm_accdrv_process_sort 4760 11.4 6.276 7.170 6.276 7.170 calculate_norms 4752 9.8 5.601 6.249 5.601 6.249 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.939 5.136 mp_sum_l 807 5.4 3.143 4.835 3.143 4.835 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.312 3.691 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.060 3.452 mp_irecv_dv 6231 10.9 2.043 3.433 2.043 3.433 make_images_sizes 222 9.7 0.000 0.000 0.713 3.349 mp_alltoall_i44 222 10.7 0.712 3.348 0.712 3.348 arnoldi_extremal 4 6.8 0.000 0.000 3.281 3.309 arnoldi_normal_ev 4 7.8 0.001 0.004 3.281 3.309 build_subspace 16 8.4 0.009 0.012 3.176 3.178 ls_scf_post 1 4.0 0.000 0.000 3.032 3.047 ls_scf_store_result 1 5.0 0.000 0.000 2.831 2.869 dbcsr_special_finalize 555 9.7 0.005 0.006 2.387 2.822 dbcsr_merge_single_wm 555 10.7 0.456 0.578 2.378 2.814 make_images_pack 222 9.7 2.209 2.635 2.211 2.636 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.333 2.570 dbcsr_sort_data 658 11.4 2.180 2.556 2.180 2.556 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.451 2.066 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.205 2.294 buffer_matrices_ensure_size 222 8.7 1.737 2.045 1.737 2.045 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.724 1.725 rebuild_ks_matrix 3 7.3 0.000 0.000 1.715 1.716 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.714 1.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.741000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 2.103489E+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 0.031 0.060 90.839 90.840 qs_energies 1 2.0 0.000 0.001 90.332 90.338 ls_scf 1 3.0 0.000 0.000 89.019 89.024 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.148 75.533 multiply_cannon 111 7.7 0.028 0.040 53.072 57.137 ls_scf_main 1 4.0 0.000 0.000 54.735 54.740 multiply_cannon_loop 111 8.7 0.117 0.124 49.786 53.535 density_matrix_trs4 2 5.0 0.002 0.003 49.108 49.324 ls_scf_init_scf 1 4.0 0.000 0.000 30.762 30.764 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.622 29.723 mp_waitall_1 9246 10.9 20.957 29.141 20.957 29.141 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.222 27.232 multiply_cannon_multrec 1332 9.7 13.113 16.443 22.398 26.932 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.562 20.811 make_m2s 222 7.7 0.006 0.008 15.360 16.059 make_images 222 8.7 1.579 1.971 15.330 16.031 dbcsr_mm_accdrv_process 4041 10.4 0.271 0.476 8.885 10.473 dbcsr_mm_accdrv_process_sort 4041 11.4 8.466 10.015 8.466 10.015 make_images_data 222 9.7 0.004 0.004 8.804 9.734 hybrid_alltoall_any 227 10.6 0.522 2.484 8.183 9.295 mp_sum_l 807 5.4 5.414 8.567 5.414 8.567 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.234 7.694 mp_irecv_dv 3311 11.0 3.214 7.637 3.214 7.637 calculate_norms 2376 9.8 6.015 6.771 6.015 6.771 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.226 6.760 multiply_cannon_sync_h2d 1332 9.7 4.786 6.149 4.786 6.149 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.024 5.224 arnoldi_extremal 4 6.8 0.000 0.000 4.642 4.661 arnoldi_normal_ev 4 7.8 0.001 0.005 4.642 4.661 build_subspace 16 8.4 0.014 0.021 4.392 4.395 ls_scf_post 1 4.0 0.000 0.000 3.522 3.528 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.148 3.382 ls_scf_store_result 1 5.0 0.000 0.000 3.225 3.344 dbcsr_matrix_vector_mult_local 304 10.0 2.781 3.254 2.783 3.256 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.213 2.731 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.510 2.605 make_images_pack 222 9.7 2.025 2.416 2.027 2.419 mp_allgather_i34 111 8.7 0.979 2.401 0.979 2.401 dbcsr_sort_data 436 11.2 1.810 2.093 1.810 2.093 dbcsr_data_new 4174 10.1 1.616 1.821 1.616 1.821 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.816 1.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.840000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1759.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 2.717344E+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 0.051 0.132 93.126 93.127 qs_energies 1 2.0 0.000 0.000 92.424 92.440 ls_scf 1 3.0 0.000 0.001 91.033 91.050 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.681 75.931 ls_scf_main 1 4.0 0.001 0.013 56.805 56.809 multiply_cannon 111 7.7 0.037 0.071 52.366 56.684 multiply_cannon_loop 111 8.7 0.100 0.107 48.830 51.905 density_matrix_trs4 2 5.0 0.002 0.003 51.021 51.138 mp_waitall_1 7374 11.0 23.088 31.961 23.088 31.961 ls_scf_init_scf 1 4.0 0.000 0.002 30.675 30.677 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.263 29.332 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.877 26.890 multiply_cannon_multrec 888 9.7 12.564 15.100 21.177 24.239 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.855 21.892 make_m2s 222 7.7 0.006 0.007 16.874 18.121 make_images 222 8.7 1.984 2.306 16.835 18.082 hybrid_alltoall_any 227 10.6 0.623 2.874 9.102 10.614 make_images_data 222 9.7 0.003 0.004 9.454 10.456 dbcsr_mm_accdrv_process 3754 10.4 0.266 0.416 8.108 9.312 dbcsr_mm_accdrv_process_sort 3754 11.4 7.737 8.896 7.737 8.896 mp_sum_l 807 5.4 4.865 8.401 4.865 8.401 multiply_cannon_sync_h2d 888 9.7 6.087 7.114 6.087 7.114 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.398 6.876 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.441 6.845 mp_irecv_dv 2335 11.1 2.427 6.797 2.427 6.797 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.729 6.571 arnoldi_extremal 4 6.8 0.000 0.000 5.122 5.141 arnoldi_normal_ev 4 7.8 0.001 0.006 5.122 5.141 apply_matrix_preconditioner 6 5.3 0.001 0.002 4.851 5.041 build_subspace 16 8.4 0.014 0.021 4.821 4.827 calculate_norms 1584 9.8 4.258 4.590 4.258 4.590 mp_allgather_i34 111 8.7 1.373 3.802 1.373 3.802 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.462 3.786 dbcsr_matrix_vector_mult_local 304 10.0 3.065 3.610 3.067 3.612 ls_scf_post 1 4.0 0.002 0.019 3.553 3.570 ls_scf_store_result 1 5.0 0.000 0.000 3.293 3.376 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.697 2.800 dbcsr_sort_data 325 11.1 1.882 2.154 1.882 2.154 make_images_pack 222 9.7 1.815 2.140 1.817 2.142 make_images_sizes 222 9.7 0.000 0.000 1.099 2.079 mp_alltoall_i44 222 10.7 1.098 2.079 1.098 2.079 dbcsr_data_release 9322 10.9 1.352 1.932 1.352 1.932 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.915 dbcsr_finalize 304 7.8 0.026 0.032 1.627 1.904 rebuild_ks_matrix 3 7.3 0.000 0.000 1.895 1.896 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.895 1.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.127000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2183.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 3.338191E+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 0.037 0.057 96.562 96.563 qs_energies 1 2.0 0.000 0.000 95.882 95.887 ls_scf 1 3.0 0.000 0.001 94.224 94.228 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.070 78.307 ls_scf_main 1 4.0 0.000 0.001 58.487 58.489 multiply_cannon 111 7.7 0.057 0.126 51.309 55.388 density_matrix_trs4 2 5.0 0.002 0.003 52.456 52.604 multiply_cannon_loop 111 8.7 0.115 0.133 46.215 49.353 ls_scf_init_scf 1 4.0 0.000 0.003 32.482 32.484 ls_scf_init_matrix_S 1 5.0 0.001 0.023 31.348 31.421 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.831 28.843 mp_waitall_1 6438 11.0 22.115 27.853 22.115 27.853 multiply_cannon_multrec 1332 9.7 14.247 17.131 22.150 24.840 make_m2s 222 7.7 0.007 0.009 20.917 22.359 make_images 222 8.7 3.148 3.604 20.867 22.311 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.798 16.336 make_images_data 222 9.7 0.004 0.005 11.567 13.214 hybrid_alltoall_any 227 10.6 0.798 3.784 10.860 12.546 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.411 7.541 9.085 dbcsr_mm_accdrv_process_sort 3641 11.4 7.166 8.666 7.166 8.666 mp_sum_l 807 5.4 4.262 7.751 4.262 7.751 multiply_cannon_sync_h2d 1332 9.7 5.490 6.318 5.490 6.318 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.243 6.052 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.084 6.006 mp_irecv_dv 3229 10.9 2.061 5.944 2.061 5.944 arnoldi_extremal 4 6.8 0.000 0.000 5.329 5.344 arnoldi_normal_ev 4 7.8 0.001 0.005 5.329 5.344 build_subspace 16 8.4 0.015 0.021 4.993 5.001 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.573 4.815 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.558 4.751 mp_allgather_i34 111 8.7 2.208 4.696 2.208 4.696 calculate_norms 2376 9.8 4.178 4.522 4.178 4.522 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.655 3.940 dbcsr_matrix_vector_mult_local 304 10.0 3.247 3.727 3.249 3.729 dbcsr_sort_data 658 11.4 3.057 3.510 3.057 3.510 ls_scf_post 1 4.0 0.000 0.001 3.254 3.259 dbcsr_special_finalize 555 9.7 0.006 0.007 2.801 3.253 dbcsr_merge_single_wm 555 10.7 0.536 0.667 2.792 3.245 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.057 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.888 2.957 dbcsr_data_release 10477 10.7 1.589 2.426 1.589 2.426 dbcsr_finalize 304 7.8 0.049 0.061 1.807 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.563000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 4.635156E+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 0.040 0.055 98.808 98.809 qs_energies 1 2.0 0.000 0.000 97.975 97.979 ls_scf 1 3.0 0.000 0.000 96.080 96.085 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.734 77.876 ls_scf_main 1 4.0 0.000 0.000 61.861 61.861 multiply_cannon 111 7.7 0.077 0.135 55.381 61.220 density_matrix_trs4 2 5.0 0.002 0.003 54.851 54.947 multiply_cannon_loop 111 8.7 0.069 0.078 50.738 52.763 mp_waitall_1 5481 11.0 26.213 31.948 26.213 31.948 ls_scf_init_scf 1 4.0 0.000 0.000 30.613 30.618 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.424 29.482 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.258 multiply_cannon_multrec 444 9.7 13.995 16.145 21.069 23.581 make_m2s 222 7.7 0.004 0.005 17.458 20.011 make_images 222 8.7 3.721 4.428 17.396 19.952 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.159 16.388 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.125 14.129 make_images_data 222 9.7 0.003 0.004 9.726 12.217 hybrid_alltoall_any 227 10.6 0.788 3.751 9.535 12.057 multiply_cannon_sync_h2d 444 9.7 6.521 8.192 6.521 8.192 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.337 6.783 7.977 dbcsr_mm_accdrv_process_sort 3003 11.4 6.468 7.627 6.468 7.627 mp_allgather_i34 111 8.7 2.871 7.079 2.871 7.079 arnoldi_extremal 4 6.8 0.000 0.000 5.892 5.901 arnoldi_normal_ev 4 7.8 0.002 0.005 5.892 5.901 build_subspace 16 8.4 0.015 0.020 5.495 5.505 mp_sum_l 807 5.4 2.943 5.234 2.943 5.234 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.600 4.713 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.680 4.606 mp_irecv_dv 1241 11.2 1.661 4.573 1.661 4.573 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.195 4.399 dbcsr_matrix_vector_mult_local 304 10.0 3.729 4.168 3.731 4.170 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.061 4.013 calculate_norms 792 9.8 3.548 3.708 3.548 3.708 ls_scf_post 1 4.0 0.000 0.000 3.606 3.611 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.408 3.510 ls_scf_store_result 1 5.0 0.000 0.000 3.389 3.431 make_images_sizes 222 9.7 0.000 0.000 1.094 3.285 mp_alltoall_i44 222 10.7 1.093 3.285 1.093 3.285 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.264 dbcsr_merge_all 275 8.9 0.474 0.537 2.042 2.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.809000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3622.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/672e1570b1b7b1d8caa570061018a37a7bb1d41b_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 8.753156E+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 0.078 0.096 106.768 106.768 qs_energies 1 2.0 0.000 0.000 105.387 105.402 ls_scf 1 3.0 0.000 0.000 102.508 102.522 dbcsr_multiply_generic 111 6.7 0.024 0.027 76.630 76.714 ls_scf_main 1 4.0 0.000 0.000 65.164 65.166 density_matrix_trs4 2 5.0 0.002 0.003 56.328 56.387 multiply_cannon 111 7.7 0.148 0.206 49.640 51.455 multiply_cannon_loop 111 8.7 0.068 0.070 46.309 47.268 ls_scf_init_scf 1 4.0 0.000 0.000 33.694 33.696 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.380 32.387 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.555 29.566 mp_waitall_1 4569 11.1 21.777 25.055 21.777 25.055 make_m2s 222 7.7 0.005 0.005 23.504 24.325 make_images 222 8.7 4.589 4.955 23.398 24.217 multiply_cannon_multrec 444 9.7 17.814 18.563 22.443 23.088 hybrid_alltoall_any 227 10.6 1.659 3.618 12.609 15.086 make_images_data 222 9.7 0.003 0.003 12.761 15.016 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.429 11.075 multiply_cannon_sync_h2d 444 9.7 8.844 8.881 8.844 8.881 arnoldi_extremal 4 6.8 0.000 0.000 7.451 7.465 arnoldi_normal_ev 4 7.8 0.003 0.009 7.451 7.465 build_subspace 16 8.4 0.026 0.037 6.894 6.907 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.532 5.680 dbcsr_matrix_vector_mult_local 304 10.0 5.104 5.375 5.107 5.378 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.011 5.247 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.796 4.886 dbcsr_mm_accdrv_process 1814 10.4 0.233 0.321 4.460 4.592 dbcsr_mm_accdrv_process_sort 1814 11.4 4.157 4.295 4.157 4.295 ls_scf_post 1 4.0 0.000 0.000 3.649 3.662 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.410 make_images_sizes 222 9.7 0.000 0.000 1.415 3.388 mp_alltoall_i44 222 10.7 1.415 3.388 1.415 3.388 calculate_norms 792 9.8 3.243 3.273 3.243 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.124 mp_allgather_i34 111 8.7 0.966 3.017 0.966 3.017 dbcsr_complete_redistribute 5 7.6 1.447 1.484 2.774 2.909 dbcsr_merge_all 275 8.9 0.891 0.915 2.864 2.904 dbcsr_data_release 12724 10.6 2.318 2.884 2.318 2.884 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.849 2.849 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.566 dbcsr_sort_data 325 11.1 2.437 2.495 2.437 2.495 dbcsr_new_transposed 4 7.5 0.244 0.252 2.308 2.326 dbcsr_frobenius_norm 74 6.6 2.053 2.130 2.200 2.240 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.207 dbcsr_add_anytype 103 7.2 0.860 0.892 2.133 2.206 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.177 2.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.768000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6963.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 672e1570b1b7b1d8caa570061018a37a7bb1d41b Summary: empty Status: OK