=== 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: 239d071ebeded0f727ee8686fa1a72666d665905 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/01 job id: 44779868 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/02 job id: 44779869 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/03 job id: 44779870 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/04 job id: 44779873 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/05 job id: 44779874 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/06 job id: 44779875 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/07 job id: 44779876 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/08 job id: 44779877 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/09 job id: 44779878 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/10 job id: 44779879 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/11 job id: 44779880 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/12 job id: 44779883 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/13 job id: 44779885 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/14 job id: 44779886 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/15 job id: 44779887 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/16 job id: 44779888 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/17 job id: 44779889 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/18 job id: 44779890 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/19 job id: 44779892 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/20 job id: 44779893 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/21 job id: 44779894 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/22 job id: 44779896 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/23 job id: 44779897 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/24 job id: 44779898 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/25 job id: 44779899 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/26 job id: 44779900 --- 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/239d071ebeded0f727ee8686fa1a72666d665905_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 0.036 133.571 133.571 farming_run 1 2.0 133.014 133.021 133.544 133.547 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463009E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.024 115.391 115.391 qs_energies 1 2.0 0.000 0.000 115.189 115.190 mp2_main 1 3.0 0.000 0.000 113.288 113.289 mp2_gpw_main 1 4.0 0.020 0.026 112.470 112.472 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.798 93.948 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.324 55.473 mp2_eri_3c_integrate_gpw 272 7.0 0.155 0.168 41.638 46.619 get_2c_integrals 1 6.0 0.000 0.000 37.645 38.302 integrate_v_rspace 273 8.0 0.441 0.453 25.025 29.644 pw_transfer 6555 10.6 0.369 0.391 27.396 27.813 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 26.028 26.275 grid_integrate_task_list 273 9.0 20.853 25.913 20.853 25.913 fft_wrap_pw1pw2_100 2178 12.4 1.242 1.405 23.544 23.806 compute_2c_integrals 1 7.0 0.003 0.003 19.601 19.601 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.957 19.332 mp2_eri_2c_integrate_gpw 1 9.0 2.345 2.418 18.954 19.330 cp_fm_cholesky_decompose 12 8.2 17.985 18.669 17.985 18.669 rpa_ri_compute_en 1 5.0 0.004 0.005 18.565 18.638 cholesky_decomp 1 7.0 0.000 0.000 16.888 17.567 fft3d_s 5443 13.4 16.097 16.257 16.119 16.279 ao_to_mo_and_store_B_mult_1 272 7.0 10.843 15.564 10.843 15.564 calculate_wavefunction 272 8.0 5.444 5.521 12.544 13.218 rpa_num_int 1 6.0 0.000 0.001 10.506 10.515 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.483 10.494 calc_mat_Q 8 8.0 0.000 0.000 9.338 9.429 contract_S_to_Q 8 9.0 0.000 0.000 8.759 8.848 calc_potential_gpw 544 9.5 0.006 0.006 8.279 8.649 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.273 8.527 parallel_gemm_fm 14 9.1 0.000 0.000 8.348 8.411 parallel_gemm_fm_cosma 14 10.1 8.348 8.411 8.348 8.411 potential_pw2rs 545 10.0 0.108 0.110 7.690 8.298 create_integ_mat 1 6.0 0.014 0.026 7.811 7.820 collocate_single_gaussian 272 10.0 0.039 0.042 7.518 7.758 array2fm 1 7.0 0.000 0.000 6.794 7.264 pw_scatter_s 2720 13.7 4.408 4.593 4.408 4.593 pw_gather_s 2722 13.2 3.867 4.366 3.867 4.366 array2fm_buffer_send 1 8.0 2.946 3.181 2.946 3.181 pw_poisson_solve 545 10.5 1.125 1.169 2.203 2.400 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.467571, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2734.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 0.026 0.036 396.075 396.076 farming_run 1 2.0 395.350 395.355 396.043 396.044 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227428E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.050 209.763 209.764 qs_energies 1 2.0 0.000 0.000 209.517 209.539 scf_env_do_scf 1 3.0 0.000 0.000 106.649 106.649 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.761 105.770 rebuild_ks_matrix 4 6.0 0.000 0.000 105.760 105.768 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.068 105.760 105.768 hfx_ks_matrix 4 8.0 0.001 0.001 105.379 105.382 integrate_four_center 4 9.0 0.144 0.460 105.378 105.381 mp2_main 1 3.0 0.000 0.000 102.583 102.605 mp2_gpw_main 1 4.0 0.034 0.063 101.759 101.784 integrate_four_center_main 4 10.0 0.089 0.624 97.373 99.566 integrate_four_center_bin 264 11.0 97.284 99.565 97.284 99.565 init_scf_loop 1 4.0 0.000 0.000 92.464 92.465 mp2_ri_gpw_compute_in 1 5.0 0.064 0.072 75.015 76.125 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.451 55.557 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.194 47.120 integrate_v_rspace 95 8.0 0.401 0.574 28.528 33.264 pw_transfer 2240 10.6 0.143 0.178 29.932 30.516 ao_to_mo_and_store_B_mult_1 91 7.0 10.568 29.537 10.568 29.537 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.940 29.521 grid_integrate_task_list 95 9.0 23.821 28.768 23.821 28.768 mp2_ri_gpw_compute_en 1 5.0 0.058 0.087 26.593 28.354 fft_wrap_pw1pw2_100 730 12.4 1.295 1.552 26.652 27.212 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.905 24.825 24.834 get_2c_integrals 1 6.0 0.000 0.000 20.478 20.502 compute_2c_integrals 1 7.0 0.005 0.025 19.454 19.465 compute_2c_integrals_loop_lm 1 8.0 0.004 0.021 18.907 19.312 mp2_eri_2c_integrate_gpw 1 9.0 1.726 1.945 18.903 19.310 fft3d_s 1823 13.4 18.395 18.767 18.409 18.780 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.183 14.183 calculate_wavefunction 91 8.0 2.055 2.088 9.777 10.036 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.583 8.721 9.395 potential_pw2rs 186 10.0 0.034 0.036 8.617 9.333 local_gemm 172 8.0 8.166 8.824 8.166 8.824 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.269 8.706 mp2_ri_gpw_compute_en_comm 22 7.0 0.489 0.516 7.872 8.338 calc_potential_gpw 182 9.5 0.002 0.003 7.949 8.184 collocate_single_gaussian 91 10.0 0.017 0.022 7.904 8.164 mp_sendrecv_dm3 2068 8.0 5.940 6.461 5.940 6.461 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.434 6.355 6.434 pw_gather_s 912 13.2 4.901 5.624 4.901 5.624 mp_sync 38 10.4 2.666 4.932 2.666 4.932 pw_scatter_s 910 13.7 3.954 4.205 3.954 4.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.750390, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.993600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578578. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.040 50.771 50.772 qs_mol_dyn_low 1 2.0 0.003 0.003 50.539 50.549 qs_forces 11 3.9 0.002 0.002 50.476 50.479 qs_energies 11 4.9 0.001 0.002 49.055 49.067 scf_env_do_scf 11 5.9 0.001 0.005 43.292 43.292 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 41.301 41.301 dbcsr_multiply_generic 2286 12.5 0.094 0.097 32.708 33.108 qs_scf_new_mos 108 7.5 0.000 0.001 31.417 31.698 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.416 31.698 ot_scf_mini 108 9.5 0.002 0.002 29.762 29.968 multiply_cannon 2286 13.5 0.190 0.201 25.650 27.171 multiply_cannon_loop 2286 14.5 1.468 1.555 24.982 26.543 velocity_verlet 10 3.0 0.001 0.001 24.884 24.885 ot_mini 108 10.5 0.001 0.001 18.879 19.144 qs_ot_get_derivative 108 11.5 0.001 0.001 15.930 16.139 mp_waitall_1 245248 16.5 8.111 13.963 8.111 13.963 multiply_cannon_metrocomm3 54864 15.5 0.069 0.073 5.841 12.566 multiply_cannon_multrec 54864 15.5 4.321 6.708 7.553 11.013 rebuild_ks_matrix 119 8.3 0.000 0.000 7.753 7.885 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.753 7.885 multiply_cannon_sync_h2d 54864 15.5 5.999 7.369 5.999 7.369 mp_sum_l 7207 12.9 5.213 6.975 5.213 6.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.843 6.968 qs_ot_get_p 119 10.4 0.001 0.001 6.384 6.666 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.402 5.849 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.341 5.452 init_scf_run 11 5.9 0.000 0.001 4.558 4.558 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.558 4.558 dbcsr_mm_accdrv_process 76910 16.1 1.123 1.826 3.154 4.437 sum_up_and_integrate 119 10.3 0.012 0.015 4.408 4.414 integrate_v_rspace 119 11.3 0.002 0.002 4.396 4.403 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.035 4.143 calculate_rho_elec 119 8.7 0.011 0.017 4.034 4.143 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.704 3.754 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.876 3.067 apply_single 119 13.6 0.000 0.000 2.876 3.067 calculate_dm_sparse 119 9.5 0.000 0.001 2.895 3.043 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.713 2.875 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.842 2.843 ot_diis_step 108 11.5 0.006 0.006 2.686 2.686 rs_pw_transfer 974 11.9 0.011 0.013 2.596 2.679 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.646 2.646 cp_fm_redistribute_end 50 14.0 2.409 2.626 2.414 2.627 jit_kernel_multiply 13 15.8 1.970 2.608 1.970 2.608 cp_fm_diag_elpa_base 50 14.0 0.210 2.543 0.211 2.556 calculate_first_density_matrix 1 7.0 0.000 0.000 2.487 2.492 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.313 2.394 density_rs2pw 119 9.7 0.004 0.004 2.106 2.206 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.192 2.193 acc_transpose_blocks 54864 15.5 0.223 0.249 1.687 2.134 grid_integrate_task_list 119 12.3 2.021 2.095 2.021 2.095 wfi_extrapolate 11 7.9 0.001 0.001 2.014 2.015 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.960 1.999 init_scf_loop 11 6.9 0.000 0.000 1.975 1.976 mp_sum_d 4127 12.0 1.293 1.920 1.293 1.920 potential_pw2rs 119 12.3 0.004 0.004 1.809 1.816 pw_transfer 1439 11.6 0.053 0.058 1.707 1.777 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.631 1.703 make_m2s 4572 13.5 0.054 0.056 1.530 1.573 make_images 4572 14.5 0.133 0.138 1.448 1.490 fft3d_ps 1201 14.6 0.366 0.469 1.400 1.461 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.415 grid_collocate_task_list 119 9.7 1.288 1.361 1.288 1.361 fft_wrap_pw1pw2_140 487 13.2 0.083 0.097 1.237 1.314 mp_alltoall_d11v 2130 13.8 1.143 1.305 1.143 1.305 mp_waitany 12084 13.8 1.180 1.285 1.180 1.285 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.714 1.091 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=50.772000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.818182, yerr=1.028519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 487.575552E+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 114897. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 38.507 38.508 qs_mol_dyn_low 1 2.0 0.003 0.003 38.281 38.289 qs_forces 11 3.9 0.002 0.002 38.206 38.207 qs_energies 11 4.9 0.001 0.002 36.531 36.534 scf_env_do_scf 11 5.9 0.001 0.001 31.293 31.294 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.832 28.833 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.467 21.830 qs_scf_new_mos 108 7.5 0.001 0.001 19.838 20.080 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.837 20.079 ot_scf_mini 108 9.5 0.002 0.003 18.957 19.133 multiply_cannon 2286 13.5 0.208 0.214 16.580 18.072 velocity_verlet 10 3.0 0.001 0.002 17.992 17.993 multiply_cannon_loop 2286 14.5 0.899 0.975 15.495 16.902 ot_mini 108 10.5 0.001 0.001 11.784 12.023 mp_waitall_1 200699 16.5 5.604 10.925 5.604 10.925 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.229 9.611 qs_ot_get_derivative 108 11.5 0.001 0.001 9.324 9.502 multiply_cannon_multrec 27432 15.5 1.995 4.271 6.099 8.897 rebuild_ks_matrix 119 8.3 0.000 0.000 7.104 7.245 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.104 7.245 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.276 6.404 dbcsr_mm_accdrv_process 47894 16.0 3.110 5.208 4.035 5.989 qs_ot_get_p 119 10.4 0.001 0.001 4.303 4.539 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.647 4.502 sum_up_and_integrate 119 10.3 0.024 0.027 4.123 4.129 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.049 4.124 apply_single 119 13.6 0.000 0.000 3.049 4.124 integrate_v_rspace 119 11.3 0.002 0.002 4.099 4.106 init_scf_run 11 5.9 0.000 0.001 4.047 4.048 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.047 4.047 mp_sum_l 7207 12.9 2.033 3.994 2.033 3.994 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.681 3.707 calculate_rho_elec 119 8.7 0.021 0.024 3.680 3.707 rs_pw_transfer 974 11.9 0.010 0.011 2.492 2.942 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.888 2.907 multiply_cannon_sync_h2d 27432 15.5 2.210 2.826 2.210 2.826 make_m2s 4572 13.5 0.052 0.054 2.402 2.616 calculate_first_density_matrix 1 7.0 0.000 0.000 2.580 2.582 make_images 4572 14.5 0.199 0.235 2.313 2.525 density_rs2pw 119 9.7 0.004 0.004 2.025 2.486 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.456 2.456 init_scf_loop 11 6.9 0.000 0.000 2.442 2.442 ot_diis_step 108 11.5 0.011 0.011 2.410 2.411 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.128 2.220 calculate_dm_sparse 119 9.5 0.000 0.001 2.129 2.202 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.133 2.133 cp_fm_redistribute_end 50 14.0 1.767 2.111 1.771 2.112 cp_fm_diag_elpa_base 50 14.0 0.327 2.000 0.339 2.061 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.042 2.044 jit_kernel_multiply 9 16.2 0.872 1.973 0.872 1.973 grid_integrate_task_list 119 12.3 1.834 1.925 1.834 1.925 pw_transfer 1439 11.6 0.064 0.071 1.864 1.894 potential_pw2rs 119 12.3 0.006 0.006 1.861 1.872 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.819 1.859 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.773 1.804 prepare_preconditioner 11 7.9 0.000 0.000 1.533 1.559 make_preconditioner 11 8.9 0.000 0.000 1.533 1.559 acc_transpose_blocks 27432 15.5 0.109 0.115 1.218 1.533 fft3d_ps 1201 14.6 0.512 0.566 1.474 1.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.435 1.491 make_images_data 4572 15.5 0.045 0.051 1.101 1.491 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.391 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.419 1.419 grid_collocate_task_list 119 9.7 1.230 1.368 1.230 1.368 hybrid_alltoall_any 4725 16.4 0.050 0.112 0.954 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.355 1.362 mp_allgather_i34 2286 14.5 0.527 1.269 0.527 1.269 mp_alltoall_d11v 2130 13.8 1.122 1.243 1.122 1.243 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.156 1.207 mp_sum_d 4127 12.0 0.569 1.014 0.569 1.014 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.518 0.961 mp_waitany 5720 13.7 0.498 0.946 0.498 0.946 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.932 0.933 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.688 0.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.508000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.545455, yerr=1.437399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 523.415552E+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 847869. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.047 31.822 31.823 qs_mol_dyn_low 1 2.0 0.003 0.006 31.542 31.550 qs_forces 11 3.9 0.002 0.003 31.424 31.427 qs_energies 11 4.9 0.002 0.005 29.853 29.856 scf_env_do_scf 11 5.9 0.001 0.003 25.246 25.246 scf_env_do_scf_inner_loop 108 6.5 0.004 0.024 22.706 22.706 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.047 16.142 velocity_verlet 10 3.0 0.001 0.002 15.136 15.139 qs_scf_new_mos 108 7.5 0.001 0.001 14.531 14.556 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.531 14.555 ot_scf_mini 108 9.5 0.002 0.004 13.819 13.838 multiply_cannon 2286 13.5 0.196 0.203 12.887 13.702 multiply_cannon_loop 2286 14.5 0.632 0.670 12.126 12.975 ot_mini 108 10.5 0.001 0.001 8.557 8.577 qs_ot_get_derivative 108 11.5 0.001 0.001 7.057 7.077 multiply_cannon_multrec 18288 15.5 1.936 2.791 6.726 6.953 rebuild_ks_matrix 119 8.3 0.000 0.000 6.355 6.372 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.355 6.371 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.614 5.628 dbcsr_mm_accdrv_process 38222 16.0 4.163 5.446 4.706 5.532 sum_up_and_integrate 119 10.3 0.029 0.030 3.947 3.952 integrate_v_rspace 119 11.3 0.002 0.003 3.918 3.926 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.443 3.449 calculate_rho_elec 119 8.7 0.030 0.031 3.442 3.449 init_scf_run 11 5.9 0.000 0.001 3.404 3.404 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.404 3.404 mp_waitall_1 158411 16.6 2.463 3.388 2.463 3.388 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.638 3.280 qs_ot_get_p 119 10.4 0.001 0.001 3.124 3.147 init_scf_loop 11 6.9 0.001 0.003 2.523 2.523 rs_pw_transfer 974 11.9 0.009 0.010 2.271 2.516 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.062 2.370 apply_single 119 13.6 0.000 0.000 2.062 2.370 calculate_first_density_matrix 1 7.0 0.001 0.004 2.202 2.203 density_rs2pw 119 9.7 0.004 0.004 1.959 2.202 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.097 2.102 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.389 2.079 pw_transfer 1439 11.6 0.065 0.070 1.921 1.930 grid_integrate_task_list 119 12.3 1.803 1.894 1.803 1.894 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.828 1.841 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.839 1.839 make_m2s 4572 13.5 0.044 0.045 1.694 1.834 calculate_dm_sparse 119 9.5 0.000 0.000 1.796 1.806 make_images 4572 14.5 0.189 0.202 1.608 1.748 prepare_preconditioner 11 7.9 0.000 0.000 1.738 1.740 make_preconditioner 11 8.9 0.000 0.001 1.738 1.740 potential_pw2rs 119 12.3 0.007 0.008 1.717 1.721 make_full_inverse_cholesky 11 9.9 0.004 0.027 1.576 1.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.627 1.628 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.621 1.621 multiply_cannon_sync_h2d 18288 15.5 1.410 1.613 1.410 1.613 cp_fm_redistribute_end 50 14.0 1.208 1.601 1.209 1.601 cp_fm_diag_elpa_base 50 14.0 0.376 1.517 0.390 1.568 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.512 1.522 fft3d_ps 1201 14.6 0.522 0.540 1.508 1.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.493 1.499 fft_wrap_pw1pw2_140 487 13.2 0.091 0.094 1.473 1.483 ot_diis_step 108 11.5 0.012 0.019 1.482 1.482 mp_sum_l 7207 12.9 1.083 1.408 1.083 1.408 grid_collocate_task_list 119 9.7 1.215 1.366 1.215 1.366 acc_transpose_blocks 18288 15.5 0.076 0.077 1.222 1.246 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.179 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 jit_kernel_multiply 6 16.3 0.491 1.113 0.491 1.113 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.978 0.981 make_images_data 4572 15.5 0.045 0.049 0.749 0.903 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.861 0.885 mp_waitany 9880 13.7 0.551 0.824 0.551 0.824 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.653 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.812 0.814 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.784 0.796 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.125 0.538 0.783 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.738 mp_alltoall_d11v 2130 13.8 0.637 0.734 0.637 0.734 cp_fm_cholesky_invert 11 10.9 0.712 0.716 0.712 0.716 mp_alltoall_z22v 1201 16.6 0.594 0.702 0.594 0.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.823000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.545455, yerr=2.387986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 553.459712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 34.443 34.444 qs_mol_dyn_low 1 2.0 0.003 0.003 34.279 34.287 qs_forces 11 3.9 0.002 0.002 34.219 34.220 qs_energies 11 4.9 0.002 0.002 32.504 32.510 scf_env_do_scf 11 5.9 0.000 0.001 27.637 27.639 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.269 24.269 dbcsr_multiply_generic 2286 12.5 0.101 0.104 17.797 17.909 velocity_verlet 10 3.0 0.001 0.002 17.571 17.572 qs_scf_new_mos 108 7.5 0.001 0.001 15.846 15.899 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.846 15.898 ot_scf_mini 108 9.5 0.002 0.003 14.922 14.973 multiply_cannon 2286 13.5 0.228 0.266 14.276 14.634 multiply_cannon_loop 2286 14.5 0.930 0.960 13.371 13.804 ot_mini 108 10.5 0.001 0.001 9.123 9.190 multiply_cannon_multrec 27432 15.5 2.330 2.999 8.546 8.938 qs_ot_get_derivative 108 11.5 0.001 0.001 7.326 7.379 dbcsr_mm_accdrv_process 47916 15.9 5.314 6.925 6.123 7.318 rebuild_ks_matrix 119 8.3 0.000 0.000 6.558 6.614 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.558 6.613 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.823 5.872 sum_up_and_integrate 119 10.3 0.035 0.038 3.796 3.804 integrate_v_rspace 119 11.3 0.002 0.004 3.760 3.769 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.525 3.557 calculate_rho_elec 119 8.7 0.040 0.046 3.525 3.556 init_scf_run 11 5.9 0.000 0.001 3.553 3.553 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.552 3.553 init_scf_loop 11 6.9 0.000 0.000 3.349 3.349 qs_ot_get_p 119 10.4 0.001 0.001 3.230 3.306 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.656 3.114 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.088 2.547 apply_single 119 13.6 0.000 0.000 2.087 2.547 prepare_preconditioner 11 7.9 0.000 0.000 2.505 2.515 make_preconditioner 11 8.9 0.000 0.000 2.505 2.515 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.112 2.442 mp_waitall_1 137007 16.6 1.693 2.280 1.693 2.280 calculate_first_density_matrix 1 7.0 0.000 0.000 2.200 2.201 make_m2s 4572 13.5 0.054 0.056 2.095 2.201 rs_pw_transfer 974 11.9 0.009 0.009 1.959 2.146 density_rs2pw 119 9.7 0.004 0.004 1.955 2.137 pw_transfer 1439 11.6 0.065 0.071 2.066 2.105 make_images 4572 14.5 0.269 0.330 1.989 2.094 calculate_dm_sparse 119 9.5 0.000 0.000 2.009 2.062 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.973 2.017 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.986 1.995 grid_integrate_task_list 119 12.3 1.825 1.924 1.825 1.924 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.871 1.897 jit_kernel_multiply 10 15.8 0.748 1.828 0.748 1.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.777 1.778 ot_diis_step 108 11.5 0.012 0.012 1.758 1.758 fft3d_ps 1201 14.6 0.560 0.613 1.648 1.687 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.636 1.681 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.680 1.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.609 1.623 potential_pw2rs 119 12.3 0.008 0.010 1.559 1.562 mp_sum_l 7207 12.9 0.996 1.521 0.996 1.521 acc_transpose_blocks 27432 15.5 0.112 0.114 1.440 1.462 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.453 1.454 cp_fm_redistribute_end 50 14.0 0.962 1.430 0.963 1.431 cp_fm_diag_elpa_base 50 14.0 0.446 1.369 0.466 1.409 grid_collocate_task_list 119 9.7 1.220 1.355 1.220 1.355 wfi_extrapolate 11 7.9 0.001 0.001 1.304 1.305 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.729 1.285 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.201 1.211 cp_fm_upper_to_full 72 13.5 0.805 1.116 0.805 1.116 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.078 1.098 multiply_cannon_sync_h2d 27432 15.5 1.014 1.085 1.014 1.085 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.068 1.070 dbcsr_complete_redistribute 329 12.2 0.121 0.151 0.757 1.030 make_images_data 4572 15.5 0.045 0.050 0.805 0.919 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.872 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.682 0.863 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.820 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.830 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.557 0.825 mp_alltoall_d11v 2130 13.8 0.674 0.791 0.674 0.791 cp_fm_cholesky_invert 11 10.9 0.731 0.734 0.731 0.734 mp_alltoall_i22 627 13.8 0.425 0.723 0.425 0.723 mp_alltoall_z22v 1201 16.6 0.672 0.707 0.672 0.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.444000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.636364, yerr=2.672171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 598.777856E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.028 27.543 27.544 qs_mol_dyn_low 1 2.0 0.003 0.003 27.364 27.372 qs_forces 11 3.9 0.002 0.002 27.299 27.304 qs_energies 11 4.9 0.001 0.001 25.595 25.601 scf_env_do_scf 11 5.9 0.000 0.001 20.999 21.000 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.555 18.556 velocity_verlet 10 3.0 0.002 0.002 14.219 14.221 dbcsr_multiply_generic 2286 12.5 0.093 0.097 11.763 11.837 qs_scf_new_mos 108 7.5 0.001 0.001 10.565 10.594 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.565 10.593 ot_scf_mini 108 9.5 0.002 0.002 9.905 9.934 multiply_cannon 2286 13.5 0.228 0.238 9.348 9.789 multiply_cannon_loop 2286 14.5 0.328 0.339 8.476 8.653 rebuild_ks_matrix 119 8.3 0.000 0.000 5.984 6.002 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.983 6.002 multiply_cannon_multrec 9144 15.5 1.585 1.824 5.733 5.975 ot_mini 108 10.5 0.001 0.001 5.476 5.511 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.334 5.351 qs_ot_get_derivative 108 11.5 0.001 0.001 4.193 4.223 dbcsr_mm_accdrv_process 12550 15.8 3.091 3.763 4.048 4.127 sum_up_and_integrate 119 10.3 0.038 0.041 3.668 3.673 integrate_v_rspace 119 11.3 0.003 0.003 3.630 3.635 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.533 3.541 calculate_rho_elec 119 8.7 0.059 0.061 3.532 3.541 init_scf_run 11 5.9 0.000 0.001 3.170 3.170 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.169 3.169 qs_ot_get_p 119 10.4 0.001 0.001 2.687 2.734 init_scf_loop 11 6.9 0.000 0.000 2.425 2.426 pw_transfer 1439 11.6 0.065 0.069 2.153 2.162 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.059 2.070 calculate_first_density_matrix 1 7.0 0.000 0.000 2.049 2.049 density_rs2pw 119 9.7 0.004 0.004 1.876 2.016 grid_integrate_task_list 119 12.3 1.850 1.924 1.850 1.924 mp_waitall_1 115863 16.7 1.311 1.857 1.311 1.857 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.798 1.801 make_m2s 4572 13.5 0.034 0.035 1.665 1.798 rs_pw_transfer 974 11.9 0.008 0.008 1.642 1.763 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.719 1.730 fft3d_ps 1201 14.6 0.561 0.571 1.718 1.729 make_images 4572 14.5 0.266 0.302 1.576 1.708 calculate_dm_sparse 119 9.5 0.000 0.000 1.686 1.702 prepare_preconditioner 11 7.9 0.000 0.000 1.682 1.686 make_preconditioner 11 8.9 0.000 0.000 1.682 1.686 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.610 1.610 jit_kernel_multiply 9 15.9 0.919 1.603 0.919 1.603 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.574 1.599 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.503 1.504 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.468 1.485 potential_pw2rs 119 12.3 0.010 0.010 1.402 1.405 grid_collocate_task_list 119 9.7 1.271 1.390 1.271 1.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.353 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.339 1.340 cp_fm_redistribute_end 50 14.0 0.670 1.321 0.670 1.322 cp_fm_diag_elpa_base 50 14.0 0.606 1.251 0.650 1.308 ot_diis_step 108 11.5 0.013 0.013 1.272 1.272 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.220 1.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.158 1.191 apply_single 119 13.6 0.000 0.000 1.157 1.190 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.118 1.124 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.099 1.113 wfi_extrapolate 11 7.9 0.001 0.001 1.073 1.073 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.761 0.960 make_images_data 4572 15.5 0.039 0.043 0.780 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.918 acc_transpose_blocks 9144 15.5 0.038 0.039 0.881 0.896 cp_fm_cholesky_invert 11 10.9 0.851 0.854 0.851 0.854 mp_alltoall_d11v 2130 13.8 0.718 0.814 0.718 0.814 multiply_cannon_sync_h2d 9144 15.5 0.711 0.808 0.711 0.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.772 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.672 0.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.712 0.721 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.318 0.679 mp_alltoall_z22v 1201 16.6 0.606 0.660 0.606 0.660 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.642 0.653 mp_allgather_i34 2286 14.5 0.207 0.616 0.207 0.616 yz_to_x 606 15.1 0.265 0.277 0.579 0.595 x_to_yz 595 16.2 0.276 0.287 0.568 0.575 qs_create_task_list 11 7.9 0.001 0.001 0.540 0.567 generate_qs_task_list 11 8.9 0.188 0.211 0.540 0.566 mp_waitany 5200 13.7 0.444 0.556 0.444 0.556 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.544000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.727273, yerr=4.633689 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 764.125184E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 40.924 40.925 qs_mol_dyn_low 1 2.0 0.003 0.003 40.678 40.686 qs_forces 11 3.9 0.007 0.011 40.600 40.609 qs_energies 11 4.9 0.001 0.001 38.586 38.603 scf_env_do_scf 11 5.9 0.001 0.001 32.977 32.977 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.124 25.125 velocity_verlet 10 3.0 0.002 0.002 23.208 23.214 dbcsr_multiply_generic 2286 12.5 0.102 0.104 17.001 17.180 qs_scf_new_mos 108 7.5 0.001 0.001 15.371 15.465 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.370 15.464 ot_scf_mini 108 9.5 0.002 0.002 14.309 14.405 multiply_cannon 2286 13.5 0.301 0.310 13.394 14.304 multiply_cannon_loop 2286 14.5 0.342 0.347 12.190 13.127 ot_mini 108 10.5 0.001 0.001 8.548 8.660 multiply_cannon_multrec 9144 15.5 3.369 4.754 8.482 8.570 init_scf_loop 11 6.9 0.000 0.000 7.827 7.828 rebuild_ks_matrix 119 8.3 0.000 0.001 7.288 7.431 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.287 7.430 prepare_preconditioner 11 7.9 0.000 0.000 6.861 6.875 make_preconditioner 11 8.9 0.000 0.000 6.861 6.875 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.375 6.753 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.587 6.717 qs_ot_get_derivative 108 11.5 0.001 0.001 6.561 6.658 dbcsr_mm_accdrv_process 12550 15.8 3.962 5.540 4.991 6.308 cp_fm_upper_to_full 72 14.2 3.248 4.679 3.248 4.679 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.232 4.237 calculate_rho_elec 119 8.7 0.118 0.121 4.231 4.236 sum_up_and_integrate 119 10.3 0.065 0.068 4.015 4.021 integrate_v_rspace 119 11.3 0.003 0.003 3.949 3.957 init_scf_run 11 5.9 0.000 0.001 3.582 3.582 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.581 3.582 qs_ot_get_p 119 10.4 0.001 0.001 3.095 3.227 mp_waitall_1 94719 16.7 2.184 3.120 2.184 3.120 dbcsr_complete_redistribute 329 12.2 0.290 0.296 1.997 2.875 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.411 2.818 pw_transfer 1439 11.6 0.068 0.069 2.780 2.786 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.681 2.687 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.702 2.561 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.139 2.427 apply_single 119 13.6 0.000 0.000 2.139 2.427 mp_alltoall_i22 627 13.8 1.516 2.404 1.516 2.404 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.481 2.333 fft3d_ps 1201 14.6 0.593 0.602 2.305 2.310 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.277 2.286 make_m2s 4572 13.5 0.038 0.038 2.103 2.238 density_rs2pw 119 9.7 0.004 0.004 2.169 2.191 calculate_first_density_matrix 1 7.0 0.000 0.000 2.168 2.170 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.284 2.146 calculate_dm_sparse 119 9.5 0.000 0.000 2.124 2.145 make_images 4572 14.5 0.353 0.382 1.982 2.116 grid_integrate_task_list 119 12.3 2.046 2.060 2.046 2.060 ot_diis_step 108 11.5 0.014 0.014 1.963 1.963 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.880 1.881 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.868 1.870 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.770 1.785 jit_kernel_multiply 10 15.5 1.002 1.733 1.002 1.733 mp_sum_l 7207 12.9 0.957 1.684 0.957 1.684 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.626 1.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.582 1.607 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.603 1.604 grid_collocate_task_list 119 9.7 1.481 1.503 1.481 1.503 rs_pw_transfer 974 11.9 0.009 0.010 1.431 1.460 potential_pw2rs 119 12.3 0.014 0.014 1.429 1.432 cp_fm_cholesky_invert 11 10.9 1.387 1.390 1.387 1.390 wfi_extrapolate 11 7.9 0.001 0.001 1.348 1.348 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.327 1.327 cp_fm_diag_elpa_base 50 14.0 1.182 1.236 1.325 1.325 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.207 1.216 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.019 1.210 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.107 1.126 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.100 1.113 make_images_data 4572 15.5 0.043 0.047 0.946 1.110 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.058 mp_alltoall_d11v 2130 13.8 1.021 1.053 1.021 1.053 multiply_cannon_sync_h2d 9144 15.5 1.044 1.047 1.044 1.047 qs_create_task_list 11 7.9 0.000 0.000 0.934 0.947 generate_qs_task_list 11 8.9 0.370 0.390 0.934 0.946 yz_to_x 606 15.1 0.461 0.473 0.921 0.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.912 0.925 acc_transpose_blocks 9144 15.5 0.038 0.039 0.889 0.894 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.925000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.363636, yerr=13.956248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 503.947264E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 84.392 84.393 qs_mol_dyn_low 1 2.0 0.003 0.003 84.151 84.162 qs_forces 11 3.9 0.002 0.002 82.286 82.288 qs_energies 11 4.9 0.001 0.002 79.449 79.468 scf_env_do_scf 11 5.9 0.000 0.001 70.816 70.818 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 65.467 65.467 dbcsr_multiply_generic 2055 12.4 0.108 0.110 52.027 52.385 qs_scf_new_mos 99 7.5 0.000 0.001 48.191 48.347 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.190 48.347 ot_scf_mini 99 9.5 0.002 0.002 45.782 45.905 velocity_verlet 10 3.0 0.001 0.002 43.905 43.924 multiply_cannon 2055 13.4 0.178 0.184 42.399 43.464 multiply_cannon_loop 2055 14.4 1.546 1.577 41.509 42.593 ot_mini 99 10.5 0.001 0.001 27.712 27.819 qs_ot_get_derivative 99 11.5 0.001 0.001 20.889 20.987 multiply_cannon_multrec 49320 15.4 12.147 12.882 17.166 18.028 rebuild_ks_matrix 110 8.3 0.000 0.001 14.491 14.610 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.491 14.609 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.711 12.826 mp_waitall_1 220248 16.4 11.329 12.799 11.329 12.799 multiply_cannon_sync_h2d 49320 15.4 9.968 10.554 9.968 10.554 qs_ot_get_p 110 10.4 0.001 0.001 9.498 9.631 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.993 8.621 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.198 7.753 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.163 7.601 apply_single 110 13.6 0.000 0.001 7.163 7.601 sum_up_and_integrate 110 10.3 0.036 0.042 7.032 7.047 integrate_v_rspace 110 11.3 0.003 0.003 6.996 7.019 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.660 6.746 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.566 6.704 calculate_rho_elec 110 8.6 0.021 0.026 6.565 6.704 init_scf_run 11 5.9 0.000 0.001 6.641 6.641 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.641 6.641 ot_diis_step 99 11.5 0.006 0.006 6.523 6.523 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.269 6.291 mp_sum_l 6514 12.8 5.386 6.066 5.386 6.066 init_scf_loop 11 6.9 0.000 0.000 5.322 5.323 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.301 5.301 dbcsr_mm_accdrv_process 87628 16.1 2.077 2.179 4.897 5.195 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.852 4.854 cp_fm_redistribute_end 48 14.0 4.238 4.832 4.241 4.832 cp_fm_diag_elpa_base 48 14.0 0.585 4.696 0.588 4.721 rs_pw_transfer 902 11.9 0.011 0.013 3.540 4.085 wfi_extrapolate 11 7.9 0.001 0.001 3.984 3.984 calculate_dm_sparse 110 9.5 0.000 0.001 3.779 3.892 make_m2s 4110 13.4 0.061 0.065 3.778 3.875 density_rs2pw 110 9.6 0.004 0.005 3.349 3.846 make_images 4110 14.4 0.177 0.190 3.683 3.782 multiply_cannon_metrocomm1 49320 15.4 0.059 0.064 2.593 3.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.488 3.492 grid_integrate_task_list 110 12.3 3.261 3.398 3.261 3.398 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.340 3.397 pw_transfer 1331 11.6 0.054 0.066 3.200 3.262 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.160 3.214 prepare_preconditioner 11 7.9 0.000 0.000 3.178 3.198 make_preconditioner 11 8.9 0.000 0.000 3.178 3.198 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.112 3.177 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.974 3.021 fft_wrap_pw1pw2_140 451 13.1 0.173 0.194 2.670 2.744 fft3d_ps 1111 14.6 0.777 0.873 2.637 2.690 mp_alltoall_d11v 2046 13.8 2.014 2.591 2.014 2.591 potential_pw2rs 110 12.3 0.006 0.007 2.569 2.586 calculate_first_density_matrix 1 7.0 0.000 0.000 2.563 2.574 jit_kernel_multiply 13 15.9 2.536 2.551 2.536 2.551 mp_waitany 14300 13.8 1.795 2.381 1.795 2.381 grid_collocate_task_list 110 9.6 2.083 2.303 2.083 2.303 acc_transpose_blocks 49320 15.4 0.225 0.236 2.169 2.258 mp_sum_d 3881 11.9 1.464 1.971 1.464 1.971 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.919 1.937 mp_sum_dm 438 4.9 1.865 1.900 1.865 1.900 make_images_data 4110 15.4 0.042 0.045 1.729 1.863 md_write_output 11 3.9 0.020 1.818 0.021 1.838 md_output 10 3.0 0.000 0.000 0.023 1.836 update_particle_set 20 4.0 0.000 0.000 1.814 1.833 hybrid_alltoall_any 4261 16.3 0.081 0.476 1.497 1.772 cp_fm_cholesky_invert 11 10.9 1.724 1.728 1.724 1.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.393000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=2.532434 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.120064E+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 2356267. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 69.717 69.719 qs_mol_dyn_low 1 2.0 0.003 0.005 69.116 69.456 qs_forces 11 3.9 0.002 0.002 69.034 69.037 qs_energies 11 4.9 0.001 0.002 65.670 65.676 scf_env_do_scf 11 5.9 0.000 0.001 57.029 57.032 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.281 49.281 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.624 37.777 velocity_verlet 10 3.0 0.001 0.002 36.620 36.621 qs_scf_new_mos 99 7.5 0.001 0.001 32.676 32.800 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.675 32.799 multiply_cannon 2055 13.4 0.220 0.247 30.957 32.198 ot_scf_mini 99 9.5 0.003 0.003 31.011 31.142 multiply_cannon_loop 2055 14.4 0.927 0.952 29.623 30.513 ot_mini 99 10.5 0.001 0.001 18.224 18.351 multiply_cannon_multrec 24660 15.4 7.639 9.552 13.798 15.672 rebuild_ks_matrix 110 8.3 0.000 0.000 13.958 14.056 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.958 14.055 qs_ot_get_derivative 99 11.5 0.001 0.001 12.414 12.539 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.291 12.380 mp_waitall_1 176588 16.5 7.606 10.292 7.606 10.292 multiply_cannon_sync_h2d 24660 15.4 7.038 8.235 7.038 8.235 multiply_cannon_metrocomm3 24660 15.4 0.070 0.071 5.133 7.947 init_scf_loop 11 6.9 0.000 0.000 7.711 7.711 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.497 7.231 apply_single 110 13.6 0.000 0.001 6.496 7.231 sum_up_and_integrate 110 10.3 0.051 0.057 6.722 6.736 integrate_v_rspace 110 11.3 0.002 0.003 6.671 6.686 dbcsr_mm_accdrv_process 52282 16.1 4.682 5.584 5.999 6.350 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.320 6.332 calculate_rho_elec 110 8.6 0.039 0.047 6.320 6.331 qs_ot_get_p 110 10.4 0.001 0.001 6.169 6.319 init_scf_run 11 5.9 0.000 0.001 6.257 6.258 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.257 6.258 ot_diis_step 99 11.5 0.010 0.010 5.760 5.760 prepare_preconditioner 11 7.9 0.000 0.000 5.636 5.656 make_preconditioner 11 8.9 0.000 0.000 5.636 5.656 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.679 5.405 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.224 5.379 make_m2s 4110 13.4 0.057 0.060 4.229 4.700 make_images 4110 14.4 0.396 0.440 4.120 4.587 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.235 4.255 pw_transfer 1331 11.6 0.066 0.072 3.840 3.988 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.734 3.885 density_rs2pw 110 9.6 0.004 0.004 3.427 3.882 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.801 3.802 wfi_extrapolate 11 7.9 0.001 0.001 3.539 3.539 rs_pw_transfer 902 11.9 0.012 0.013 3.021 3.495 grid_integrate_task_list 110 12.3 3.144 3.341 3.144 3.341 fft_wrap_pw1pw2_140 451 13.1 0.203 0.222 3.157 3.310 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.198 3.271 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.244 3.246 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.229 3.230 cp_fm_redistribute_end 48 14.0 2.426 3.217 2.427 3.217 fft3d_ps 1111 14.6 1.102 1.312 3.082 3.216 cp_fm_diag_elpa_base 48 14.0 0.756 3.074 0.787 3.159 calculate_dm_sparse 110 9.5 0.001 0.001 2.967 2.995 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.889 2.917 hybrid_alltoall_any 4261 16.3 0.102 0.438 2.060 2.847 make_images_data 4110 15.4 0.047 0.050 2.354 2.846 calculate_first_density_matrix 1 7.0 0.000 0.000 2.628 2.630 cp_fm_cholesky_invert 11 10.9 2.611 2.619 2.611 2.619 potential_pw2rs 110 12.3 0.008 0.009 2.493 2.507 grid_collocate_task_list 110 9.6 2.052 2.481 2.052 2.481 mp_sum_l 6514 12.8 1.812 2.402 1.812 2.402 mp_alltoall_d11v 2046 13.8 1.839 2.191 1.839 2.191 jit_kernel_multiply 11 16.4 0.966 2.112 0.966 2.112 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.955 1.977 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.874 1.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.831 1.847 mp_allgather_i34 2055 14.4 0.666 1.731 0.666 1.731 mp_waitany 10164 13.8 1.206 1.645 1.206 1.645 multiply_cannon_metrocomm4 22605 15.4 0.075 0.078 0.785 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.547 acc_transpose_blocks 24660 15.4 0.112 0.115 1.513 1.535 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.478 mp_irecv_dv 57340 16.2 0.660 1.465 0.660 1.465 dbcsr_complete_redistribute 325 12.2 0.243 0.303 1.175 1.441 mp_alltoall_z22v 1111 16.6 1.301 1.426 1.301 1.426 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.216 0.953 1.420 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.719000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.272727, yerr=7.544907 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 658.976768E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.049 59.761 59.764 qs_mol_dyn_low 1 2.0 0.004 0.013 59.330 59.341 qs_forces 11 3.9 0.002 0.002 59.236 59.240 qs_energies 11 4.9 0.001 0.002 56.038 56.044 scf_env_do_scf 11 5.9 0.000 0.001 48.125 48.125 scf_env_do_scf_inner_loop 99 6.5 0.006 0.017 39.695 39.695 velocity_verlet 10 3.0 0.001 0.002 32.485 32.487 dbcsr_multiply_generic 2055 12.4 0.109 0.111 28.124 28.386 qs_scf_new_mos 99 7.5 0.001 0.001 24.543 24.614 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.542 24.613 ot_scf_mini 99 9.5 0.002 0.003 23.329 23.436 multiply_cannon 2055 13.4 0.210 0.218 21.934 23.197 multiply_cannon_loop 2055 14.4 0.615 0.628 20.756 21.850 ot_mini 99 10.5 0.001 0.001 13.397 13.506 rebuild_ks_matrix 110 8.3 0.000 0.000 12.434 12.552 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.433 12.551 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.953 11.059 multiply_cannon_multrec 16440 15.4 4.023 5.144 9.759 10.909 mp_waitall_1 139946 16.5 6.704 9.274 6.704 9.274 qs_ot_get_derivative 99 11.5 0.001 0.001 8.969 9.079 init_scf_loop 11 6.9 0.000 0.000 8.395 8.396 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.112 6.820 prepare_preconditioner 11 7.9 0.000 0.000 6.638 6.655 make_preconditioner 11 8.9 0.000 0.000 6.638 6.655 sum_up_and_integrate 110 10.3 0.059 0.060 6.599 6.614 integrate_v_rspace 110 11.3 0.002 0.003 6.539 6.555 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.976 6.335 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.045 6.051 calculate_rho_elec 110 8.6 0.058 0.059 6.044 6.051 dbcsr_mm_accdrv_process 34862 16.1 4.681 5.277 5.589 5.809 init_scf_run 11 5.9 0.000 0.001 5.486 5.486 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.486 5.486 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.985 5.401 apply_single 110 13.6 0.000 0.000 4.985 5.401 qs_ot_get_p 110 10.4 0.001 0.001 5.186 5.312 make_m2s 4110 13.4 0.050 0.052 4.103 4.455 ot_diis_step 99 11.5 0.011 0.011 4.400 4.400 make_images 4110 14.4 0.390 0.509 3.988 4.339 density_rs2pw 110 9.6 0.004 0.005 3.139 4.256 multiply_cannon_sync_h2d 16440 15.4 3.726 4.227 3.726 4.227 rs_pw_transfer 902 11.9 0.010 0.011 2.664 3.770 pw_transfer 1331 11.6 0.065 0.073 3.748 3.755 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.078 3.742 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.642 3.649 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.606 3.611 grid_integrate_task_list 110 12.3 3.163 3.416 3.163 3.416 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.274 3.274 fft_wrap_pw1pw2_140 451 13.1 0.213 0.219 3.148 3.160 wfi_extrapolate 11 7.9 0.001 0.001 2.959 2.959 fft3d_ps 1111 14.6 1.088 1.101 2.941 2.952 make_images_data 4110 15.4 0.043 0.047 2.367 2.806 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.768 2.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.748 2.749 cp_fm_redistribute_end 48 14.0 1.728 2.745 1.731 2.745 cp_fm_diag_elpa_base 48 14.0 0.953 2.595 1.011 2.704 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.059 2.641 cp_fm_cholesky_invert 11 10.9 2.612 2.618 2.612 2.618 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.466 grid_collocate_task_list 110 9.6 2.082 2.457 2.082 2.457 calculate_first_density_matrix 1 7.0 0.000 0.000 2.442 2.442 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.851 2.363 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.286 2.352 potential_pw2rs 110 12.3 0.010 0.010 2.316 2.325 mp_waitany 17072 13.8 1.174 2.310 1.174 2.310 mp_irecv_dv 48980 15.7 0.782 2.239 0.782 2.239 mp_alltoall_d11v 2046 13.8 1.737 2.172 1.737 2.172 mp_sum_l 6514 12.8 1.483 2.096 1.483 2.096 qs_energies_init_hamiltonians 11 5.9 0.038 0.044 2.004 2.007 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.196 0.900 2.003 dbcsr_complete_redistribute 325 12.2 0.330 0.376 1.422 1.908 cp_fm_upper_to_full 70 13.6 1.399 1.852 1.399 1.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.679 1.694 cp_fm_cholesky_decompose 22 10.9 1.556 1.581 1.556 1.581 mp_allgather_i34 2055 14.4 0.506 1.542 0.506 1.542 jit_kernel_multiply 8 16.6 0.519 1.540 0.519 1.540 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.488 1.501 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.471 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.956 1.418 rs_gather_matrices 110 12.3 0.234 0.265 0.982 1.354 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.242 1.250 acc_transpose_blocks 16440 15.4 0.072 0.075 1.201 1.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.764000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.545455, yerr=9.038942 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 727.408640E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.040 65.219 65.221 qs_mol_dyn_low 1 2.0 0.003 0.006 64.822 64.833 qs_forces 11 3.9 0.004 0.008 64.750 64.752 qs_energies 11 4.9 0.003 0.007 61.317 61.322 scf_env_do_scf 11 5.9 0.003 0.015 53.041 53.044 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.399 41.402 velocity_verlet 10 3.0 0.002 0.003 37.040 37.043 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.211 29.437 qs_scf_new_mos 99 7.5 0.001 0.001 26.244 26.349 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.243 26.348 ot_scf_mini 99 9.5 0.003 0.004 24.631 24.720 multiply_cannon 2055 13.4 0.241 0.259 22.220 23.201 multiply_cannon_loop 2055 14.4 0.884 0.906 20.890 21.511 ot_mini 99 10.5 0.001 0.001 13.973 14.081 multiply_cannon_multrec 24660 15.4 4.212 6.825 12.607 13.616 rebuild_ks_matrix 110 8.3 0.000 0.000 12.252 12.356 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.252 12.355 init_scf_loop 11 6.9 0.001 0.003 11.596 11.597 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.817 10.908 qs_ot_get_derivative 99 11.5 0.001 0.001 9.811 9.903 prepare_preconditioner 11 7.9 0.000 0.000 9.849 9.867 make_preconditioner 11 8.9 0.000 0.003 9.849 9.867 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.075 9.542 dbcsr_mm_accdrv_process 52304 16.0 6.891 8.489 8.248 9.180 sum_up_and_integrate 110 10.3 0.067 0.070 6.524 6.537 integrate_v_rspace 110 11.3 0.003 0.003 6.457 6.470 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.179 6.192 calculate_rho_elec 110 8.6 0.077 0.081 6.178 6.191 mp_waitall_1 121746 16.5 4.195 6.039 4.195 6.039 qs_ot_get_p 110 10.4 0.001 0.001 5.687 5.822 make_m2s 4110 13.4 0.060 0.062 5.377 5.651 make_images 4110 14.4 0.573 0.692 5.238 5.509 init_scf_run 11 5.9 0.000 0.001 5.468 5.469 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.468 5.468 cp_fm_upper_to_full 70 13.8 3.306 4.702 3.306 4.702 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.023 4.132 apply_single 110 13.6 0.000 0.000 4.023 4.131 ot_diis_step 99 11.5 0.011 0.011 4.125 4.125 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.918 3.933 pw_transfer 1331 11.6 0.065 0.073 3.877 3.906 dbcsr_complete_redistribute 325 12.2 0.417 0.464 2.675 3.841 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.771 3.804 density_rs2pw 110 9.6 0.004 0.004 3.095 3.597 cp_dbcsr_syevd 48 12.0 0.004 0.005 3.439 3.440 grid_integrate_task_list 110 12.3 3.261 3.436 3.261 3.436 multiply_cannon_sync_h2d 24660 15.4 3.171 3.354 3.171 3.354 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.265 3.311 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.136 3.291 fft_wrap_pw1pw2_140 451 13.1 0.203 0.214 3.234 3.272 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.381 3.176 fft3d_ps 1111 14.6 1.090 1.133 3.062 3.085 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.009 3.058 make_images_data 4110 15.4 0.047 0.051 2.696 2.993 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.330 2.990 wfi_extrapolate 11 7.9 0.001 0.002 2.984 2.984 calculate_dm_sparse 110 9.5 0.001 0.001 2.899 2.931 rs_pw_transfer 902 11.9 0.010 0.011 2.418 2.922 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.765 2.899 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.831 2.833 mp_alltoall_i22 605 13.7 1.679 2.832 1.679 2.832 cp_fm_redistribute_end 48 14.0 1.417 2.805 1.419 2.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.771 2.778 cp_fm_diag_elpa_base 48 14.0 1.305 2.669 1.384 2.777 cp_fm_cholesky_invert 11 10.9 2.667 2.675 2.667 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.408 2.453 grid_collocate_task_list 110 9.6 2.179 2.428 2.179 2.428 calculate_first_density_matrix 1 7.0 0.006 0.021 2.392 2.395 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.297 2.299 potential_pw2rs 110 12.3 0.013 0.013 2.215 2.220 mp_alltoall_d11v 2046 13.8 1.770 2.050 1.770 2.050 jit_kernel_multiply 11 15.7 1.027 1.976 1.027 1.976 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.689 1.722 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.679 1.690 cp_fm_cholesky_decompose 22 10.9 1.637 1.685 1.637 1.685 mp_waitany 13376 13.8 1.084 1.542 1.084 1.542 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.527 1.541 mp_sum_l 6514 12.8 0.930 1.535 0.930 1.535 acc_transpose_blocks 24660 15.4 0.106 0.108 1.495 1.511 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.864 1.471 mp_irecv_dv 62702 16.1 0.764 1.394 0.764 1.394 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.239 1.351 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.180 0.828 1.336 mp_allgather_i34 2055 14.4 0.438 1.322 0.438 1.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.221000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.272727, yerr=8.582126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 824.848384E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 54.516 54.518 qs_mol_dyn_low 1 2.0 0.003 0.003 54.283 54.293 qs_forces 11 3.9 0.002 0.002 54.215 54.216 qs_energies 11 4.9 0.001 0.002 50.538 50.543 scf_env_do_scf 11 5.9 0.001 0.001 42.380 42.382 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.740 34.741 velocity_verlet 10 3.0 0.002 0.002 30.806 30.809 dbcsr_multiply_generic 2055 12.4 0.106 0.109 22.140 22.297 qs_scf_new_mos 99 7.5 0.001 0.001 19.798 19.839 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.797 19.838 ot_scf_mini 99 9.5 0.002 0.002 18.577 18.606 multiply_cannon 2055 13.4 0.244 0.262 16.934 18.270 multiply_cannon_loop 2055 14.4 0.319 0.331 15.621 16.023 rebuild_ks_matrix 110 8.3 0.000 0.000 11.826 11.872 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.826 11.872 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.511 10.551 ot_mini 99 10.5 0.001 0.001 10.017 10.045 multiply_cannon_multrec 8220 15.4 3.189 4.522 7.391 8.363 init_scf_loop 11 6.9 0.000 0.000 7.585 7.586 mp_waitall_1 103326 16.6 5.558 7.288 5.558 7.288 sum_up_and_integrate 110 10.3 0.079 0.081 6.480 6.490 integrate_v_rspace 110 11.3 0.003 0.003 6.401 6.413 qs_ot_get_derivative 99 11.5 0.001 0.001 6.267 6.296 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.259 6.273 calculate_rho_elec 110 8.6 0.115 0.116 6.259 6.272 prepare_preconditioner 11 7.9 0.000 0.000 5.941 5.944 make_preconditioner 11 8.9 0.000 0.000 5.941 5.944 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.534 5.605 init_scf_run 11 5.9 0.000 0.001 5.037 5.037 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.037 5.037 dbcsr_mm_accdrv_process 17442 15.9 2.849 3.781 4.074 4.972 qs_ot_get_p 110 10.4 0.001 0.001 4.726 4.751 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.885 4.378 make_m2s 4110 13.4 0.039 0.040 3.962 4.164 make_images 4110 14.4 0.636 0.689 3.833 4.034 pw_transfer 1331 11.6 0.066 0.071 4.002 4.010 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.894 3.908 ot_diis_step 99 11.5 0.012 0.012 3.731 3.732 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.649 3.673 apply_single 110 13.6 0.000 0.000 3.649 3.673 grid_integrate_task_list 110 12.3 3.347 3.569 3.347 3.569 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.422 3.425 density_rs2pw 110 9.6 0.004 0.004 3.040 3.416 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.366 3.381 fft3d_ps 1111 14.6 1.145 1.167 3.138 3.149 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.122 3.123 multiply_cannon_sync_h2d 8220 15.4 2.923 3.032 2.923 3.032 cp_fm_cholesky_invert 11 10.9 2.869 2.872 2.869 2.872 wfi_extrapolate 11 7.9 0.001 0.001 2.699 2.699 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.678 2.679 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.618 2.618 cp_fm_redistribute_end 48 14.0 0.666 2.593 0.670 2.594 cp_fm_diag_elpa_base 48 14.0 1.743 2.402 1.917 2.566 rs_pw_transfer 902 11.9 0.010 0.010 2.137 2.544 grid_collocate_task_list 110 9.6 2.277 2.540 2.277 2.540 make_images_data 4110 15.4 0.038 0.044 2.193 2.535 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.138 2.501 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.483 2.484 calculate_dm_sparse 110 9.5 0.001 0.001 2.442 2.482 calculate_first_density_matrix 1 7.0 0.000 0.000 2.243 2.244 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.095 2.109 potential_pw2rs 110 12.3 0.015 0.016 2.067 2.073 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.998 mp_alltoall_d11v 2046 13.8 1.516 1.925 1.516 1.925 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.856 1.872 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.756 1.772 cp_fm_cholesky_decompose 22 10.9 1.667 1.695 1.667 1.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.648 1.652 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.509 1.639 dbcsr_complete_redistribute 325 12.2 0.582 0.630 1.437 1.528 mp_waitany 9240 13.8 1.044 1.455 1.044 1.455 mp_allgather_i34 2055 14.4 0.506 1.444 0.506 1.444 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.421 1.432 qs_create_task_list 11 7.9 0.000 0.001 1.220 1.328 generate_qs_task_list 11 8.9 0.378 0.446 1.219 1.327 rs_gather_matrices 110 12.3 0.325 0.370 0.903 1.297 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.164 0.784 1.184 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.108 1.131 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.484 1.129 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.518000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=773.818182, yerr=11.239688 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 1.354416E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 87.234 87.234 qs_mol_dyn_low 1 2.0 0.003 0.003 86.897 86.906 qs_forces 11 3.9 0.002 0.002 85.942 85.943 qs_energies 11 4.9 0.002 0.002 81.751 81.752 scf_env_do_scf 11 5.9 0.001 0.001 71.877 71.878 velocity_verlet 10 3.0 0.002 0.002 56.156 56.273 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.653 43.655 init_scf_loop 11 6.9 0.000 0.000 28.153 28.157 dbcsr_multiply_generic 2055 12.4 0.121 0.122 27.720 27.768 prepare_preconditioner 11 7.9 0.000 0.000 26.148 26.154 make_preconditioner 11 8.9 0.000 0.000 26.148 26.154 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.391 25.642 qs_scf_new_mos 99 7.5 0.001 0.001 25.433 25.461 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.432 25.460 ot_scf_mini 99 9.5 0.002 0.002 23.701 23.724 multiply_cannon 2055 13.4 0.340 0.354 20.968 21.815 multiply_cannon_loop 2055 14.4 0.341 0.343 19.224 19.609 cp_fm_upper_to_full 70 14.2 12.757 18.249 12.757 18.249 rebuild_ks_matrix 110 8.3 0.001 0.001 14.090 14.099 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.090 14.099 ot_mini 99 10.5 0.001 0.001 13.170 13.195 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.764 12.773 dbcsr_complete_redistribute 325 12.2 1.027 1.057 7.317 10.521 multiply_cannon_multrec 8220 15.4 4.352 4.566 9.502 9.603 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.296 9.495 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.742 8.919 qs_ot_get_derivative 99 11.5 0.001 0.001 8.757 8.778 mp_alltoall_i22 605 13.7 5.367 8.556 5.367 8.556 mp_waitall_1 84994 16.7 6.992 7.756 6.992 7.756 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.704 7.742 calculate_rho_elec 110 8.6 0.227 0.227 7.704 7.741 sum_up_and_integrate 110 10.3 0.149 0.151 7.359 7.371 integrate_v_rspace 110 11.3 0.004 0.004 7.210 7.221 init_scf_run 11 5.9 0.000 0.001 5.600 5.600 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.600 5.600 make_m2s 4110 13.4 0.043 0.044 5.063 5.546 qs_ot_get_p 110 10.4 0.001 0.001 5.480 5.504 make_images 4110 14.4 0.879 0.928 4.875 5.358 pw_transfer 1331 11.6 0.075 0.075 5.255 5.261 dbcsr_mm_accdrv_process 11614 15.7 3.285 3.611 5.008 5.248 cp_fm_cholesky_invert 11 10.9 5.146 5.150 5.146 5.150 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.138 5.145 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.544 4.992 apply_single 110 13.6 0.000 0.000 4.544 4.992 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.387 4.807 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.475 4.482 ot_diis_step 99 11.5 0.015 0.015 4.390 4.391 fft3d_ps 1111 14.6 1.289 1.295 4.333 4.343 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.013 4.021 multiply_cannon_sync_h2d 8220 15.4 3.948 3.955 3.948 3.955 density_rs2pw 110 9.6 0.004 0.004 3.756 3.798 grid_integrate_task_list 110 12.3 3.663 3.718 3.663 3.718 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.686 3.686 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.576 3.577 hybrid_alltoall_any 4261 16.3 0.256 0.553 2.762 3.423 make_images_data 4110 15.4 0.042 0.045 2.709 3.309 wfi_extrapolate 11 7.9 0.001 0.001 3.262 3.262 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.802 3.248 calculate_dm_sparse 110 9.5 0.001 0.001 3.099 3.119 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.974 2.974 cp_fm_diag_elpa_base 48 14.0 2.444 2.643 2.972 2.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.775 2.778 grid_collocate_task_list 110 9.6 2.629 2.650 2.629 2.650 potential_pw2rs 110 12.3 0.021 0.022 2.504 2.507 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.323 2.338 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.294 2.298 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.206 2.272 rs_pw_transfer 902 11.9 0.011 0.011 2.240 2.267 calculate_first_density_matrix 1 7.0 0.000 0.000 2.212 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.188 cp_fm_cholesky_decompose 22 10.9 1.975 1.997 1.975 1.997 mp_alltoall_d11v 2046 13.8 1.859 1.964 1.859 1.964 qs_create_task_list 11 7.9 0.000 0.001 1.891 1.935 generate_qs_task_list 11 8.9 0.733 0.788 1.891 1.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.929 1.930 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.737 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.234000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1222.181818, yerr=57.853763 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 632.127488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 203.839 203.840 qs_mol_dyn_low 1 2.0 0.003 0.003 203.475 203.490 qs_forces 11 3.9 0.003 0.003 203.386 203.388 qs_energies 11 4.9 0.017 0.021 197.750 197.764 scf_env_do_scf 11 5.9 0.001 0.001 181.327 181.330 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 160.885 160.887 dbcsr_multiply_generic 2507 12.6 0.180 0.183 124.204 124.992 velocity_verlet 10 3.0 0.001 0.002 123.064 123.065 qs_scf_new_mos 117 7.6 0.001 0.001 121.128 121.472 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.127 121.472 ot_scf_mini 117 9.6 0.003 0.003 114.624 114.948 multiply_cannon 2507 13.6 0.235 0.243 101.056 103.084 multiply_cannon_loop 2507 14.6 2.094 2.180 98.953 100.878 ot_mini 117 10.6 0.001 0.001 65.945 66.249 multiply_cannon_multrec 60168 15.6 33.362 35.523 41.622 43.705 qs_ot_get_derivative 117 11.6 0.001 0.001 40.938 41.261 rebuild_ks_matrix 128 8.3 0.001 0.001 33.585 33.853 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.584 33.852 mp_waitall_1 267128 16.5 28.245 31.870 28.245 31.870 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.185 30.425 multiply_cannon_sync_h2d 60168 15.6 27.344 29.510 27.344 29.510 qs_ot_get_p 128 10.4 0.001 0.001 26.839 27.104 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.435 25.217 apply_single 128 13.6 0.001 0.001 24.434 25.217 ot_diis_step 117 11.6 0.007 0.008 24.658 24.659 init_scf_loop 11 6.9 0.000 0.000 20.370 20.371 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.163 20.281 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.934 19.194 multiply_cannon_metrocomm3 60168 15.6 0.115 0.120 15.830 18.325 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.545 17.546 prepare_preconditioner 11 7.9 0.000 0.000 15.759 15.830 make_preconditioner 11 8.9 0.000 0.000 15.759 15.830 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.010 15.214 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.718 14.723 cp_fm_redistribute_end 83 14.4 11.637 14.650 11.651 14.654 cp_fm_diag_elpa_base 83 14.4 2.959 14.284 2.991 14.403 sum_up_and_integrate 128 10.3 0.090 0.107 14.121 14.134 integrate_v_rspace 128 11.3 0.003 0.004 14.031 14.045 make_m2s 5014 13.6 0.105 0.112 13.459 13.830 make_images 5014 14.6 0.401 0.420 13.277 13.654 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.494 13.593 calculate_rho_elec 128 8.7 0.045 0.064 13.494 13.593 init_scf_run 11 5.9 0.000 0.001 12.247 12.248 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.247 12.247 density_rs2pw 128 9.7 0.006 0.007 6.948 10.310 mp_sum_l 7870 13.0 8.412 9.792 8.412 9.792 wfi_extrapolate 11 7.9 0.001 0.001 9.053 9.053 rs_pw_transfer 1046 11.9 0.016 0.019 5.644 9.033 cp_fm_cholesky_invert 11 10.9 9.007 9.014 9.007 9.014 dbcsr_mm_accdrv_process 124484 16.2 3.233 3.454 7.817 8.442 calculate_dm_sparse 128 9.5 0.001 0.001 8.299 8.428 multiply_cannon_metrocomm1 60168 15.6 0.088 0.095 6.202 8.407 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.177 8.311 pw_transfer 1547 11.6 0.076 0.099 7.678 7.924 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.770 7.869 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.472 7.723 grid_integrate_task_list 128 12.3 7.034 7.529 7.034 7.529 make_images_data 5014 15.6 0.066 0.072 6.522 7.375 fft_wrap_pw1pw2_140 523 13.2 0.448 0.515 6.525 6.838 hybrid_alltoall_any 5200 16.5 0.291 2.258 5.705 6.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.612 6.620 fft3d_ps 1291 14.7 2.103 2.890 6.253 6.409 mp_waitany 16020 13.9 2.658 6.031 2.658 6.031 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.788 5.885 grid_collocate_task_list 128 9.7 4.565 5.784 4.565 5.784 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.301 2.117 5.511 mp_alltoall_d11v 2415 14.1 3.939 5.163 3.939 5.163 cp_fm_cholesky_decompose 22 10.9 4.606 4.620 4.606 4.620 potential_pw2rs 128 12.3 0.009 0.011 4.576 4.594 mp_sum_d 4457 12.1 3.694 4.450 3.694 4.450 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.840000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.727273, yerr=7.212249 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444712984576 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 2.166474E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 827.138048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843365392 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57200. MP_Allreduce 11252 947. MP_Sync 168 MP_Alltoall 1955 5620276. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.028 189.191 189.192 qs_mol_dyn_low 1 2.0 0.003 0.003 188.590 188.871 qs_forces 11 3.9 0.003 0.003 188.494 188.498 qs_energies 11 4.9 0.001 0.002 181.675 181.686 scf_env_do_scf 11 5.9 0.001 0.002 165.429 165.439 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.799 132.801 velocity_verlet 10 3.0 0.001 0.002 118.555 118.556 dbcsr_multiply_generic 2485 12.5 0.187 0.193 96.593 97.772 qs_scf_new_mos 116 7.6 0.001 0.001 92.822 93.282 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.822 93.282 ot_scf_mini 116 9.6 0.004 0.004 88.070 88.490 multiply_cannon 2485 13.5 0.475 0.528 77.025 81.235 multiply_cannon_loop 2485 14.5 1.241 1.287 73.711 76.023 ot_mini 116 10.6 0.001 0.001 48.959 49.463 mp_waitall_1 212858 16.6 23.672 36.970 23.672 36.970 multiply_cannon_multrec 29820 15.5 22.209 26.886 31.706 36.590 rebuild_ks_matrix 127 8.3 0.001 0.001 33.357 33.806 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.356 33.806 init_scf_loop 11 6.9 0.000 0.000 32.544 32.547 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.996 30.409 prepare_preconditioner 11 7.9 0.000 0.000 28.099 28.159 make_preconditioner 11 8.9 0.000 0.000 28.099 28.159 qs_ot_get_derivative 116 11.6 0.001 0.002 27.284 27.729 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.799 27.305 multiply_cannon_metrocomm3 29820 15.5 0.094 0.099 15.225 27.283 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.770 22.890 apply_single 127 13.6 0.001 0.001 21.769 22.889 multiply_cannon_sync_h2d 29820 15.5 19.513 21.714 19.513 21.714 ot_diis_step 116 11.6 0.014 0.015 21.506 21.508 qs_ot_get_p 127 10.4 0.001 0.001 20.903 21.458 cp_fm_cholesky_invert 11 10.9 16.455 16.468 16.455 16.468 qs_ot_p2m_diag 82 11.4 0.184 0.213 16.234 16.269 make_m2s 4970 13.5 0.090 0.095 14.032 15.688 make_images 4970 14.5 1.140 1.327 13.821 15.481 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.117 15.118 sum_up_and_integrate 127 10.3 0.113 0.132 14.623 14.651 integrate_v_rspace 127 11.3 0.003 0.004 14.510 14.545 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.996 14.033 calculate_rho_elec 127 8.7 0.087 0.105 13.995 14.032 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.073 12.085 cp_fm_redistribute_end 82 14.4 7.050 12.016 7.059 12.016 cp_fm_diag_elpa_base 82 14.4 4.723 11.506 4.935 11.888 init_scf_run 11 5.9 0.000 0.001 11.471 11.472 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.471 11.472 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.656 10.956 multiply_cannon_metrocomm4 27335 15.5 0.097 0.111 3.605 10.391 make_images_data 4970 15.5 0.063 0.072 8.392 10.336 mp_irecv_dv 68888 16.3 3.412 10.008 3.412 10.008 density_rs2pw 127 9.7 0.006 0.007 7.436 9.739 hybrid_alltoall_any 5155 16.4 0.339 1.488 7.029 9.563 dbcsr_mm_accdrv_process 61726 16.2 4.405 5.210 8.958 9.542 pw_transfer 1535 11.6 0.086 0.113 8.883 8.954 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.659 8.734 wfi_extrapolate 11 7.9 0.001 0.001 8.394 8.394 rs_pw_transfer 1038 11.9 0.014 0.016 5.610 7.937 fft_wrap_pw1pw2_140 519 13.2 0.469 0.526 7.660 7.748 grid_integrate_task_list 127 12.3 7.108 7.568 7.108 7.568 fft3d_ps 1281 14.7 2.727 2.899 7.037 7.066 cp_fm_cholesky_decompose 22 10.9 6.822 6.900 6.822 6.900 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.121 6.830 calculate_dm_sparse 127 9.5 0.001 0.001 6.401 6.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.147 6.155 grid_collocate_task_list 127 9.7 4.671 5.833 4.671 5.833 mp_sum_l 7804 13.0 3.756 5.647 3.756 5.647 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.341 5.492 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.259 5.327 potential_pw2rs 127 12.3 0.014 0.016 4.979 4.995 mp_allgather_i34 2485 14.5 1.902 4.951 1.902 4.951 mp_alltoall_d11v 2401 14.1 3.980 4.805 3.980 4.805 mp_waitany 11660 13.9 2.417 4.788 2.417 4.788 rs_pw_transfer_RS2PW_140 138 11.5 0.351 0.379 2.038 4.341 mp_sum_d 4444 12.1 2.600 3.989 2.600 3.989 dbcsr_complete_redistribute 393 12.7 0.777 0.854 3.042 3.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.765 3.799 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.192000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.909091, yerr=2.874798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 933.199872E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57277. MP_Allreduce 11236 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.074 175.772 175.782 qs_mol_dyn_low 1 2.0 0.003 0.004 175.312 175.327 qs_forces 11 3.9 0.003 0.003 175.060 175.064 qs_energies 11 4.9 0.003 0.006 168.476 168.486 scf_env_do_scf 11 5.9 0.001 0.002 152.920 152.921 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 117.833 117.833 velocity_verlet 10 3.0 0.002 0.002 112.196 112.203 dbcsr_multiply_generic 2485 12.5 0.182 0.187 80.552 81.596 qs_scf_new_mos 116 7.6 0.001 0.001 80.065 80.395 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.064 80.394 ot_scf_mini 116 9.6 0.004 0.004 75.957 76.345 multiply_cannon 2485 13.5 0.491 0.524 60.980 64.975 multiply_cannon_loop 2485 14.5 0.847 0.879 58.029 60.534 ot_mini 116 10.6 0.001 0.001 42.075 42.457 init_scf_loop 11 6.9 0.001 0.004 34.989 34.991 mp_waitall_1 169034 16.6 24.518 33.166 24.518 33.166 prepare_preconditioner 11 7.9 0.000 0.001 30.963 31.013 make_preconditioner 11 8.9 0.000 0.001 30.963 31.013 rebuild_ks_matrix 127 8.3 0.001 0.001 30.402 30.864 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 30.401 30.863 make_full_inverse_cholesky 11 9.9 0.002 0.014 28.571 29.973 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.379 27.801 multiply_cannon_multrec 19880 15.5 13.541 16.631 22.089 25.133 multiply_cannon_metrocomm3 19880 15.5 0.058 0.062 14.889 23.599 qs_ot_get_derivative 116 11.6 0.001 0.002 22.571 22.958 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.614 20.577 apply_single 127 13.6 0.001 0.001 19.614 20.576 qs_ot_get_p 127 10.4 0.001 0.002 19.493 19.903 ot_diis_step 116 11.6 0.017 0.018 19.400 19.401 multiply_cannon_sync_h2d 19880 15.5 14.242 15.609 14.242 15.609 make_m2s 4970 13.5 0.081 0.086 14.613 15.408 qs_ot_p2m_diag 82 11.4 0.262 0.269 15.262 15.271 make_images 4970 14.5 1.165 1.250 14.380 15.177 sum_up_and_integrate 127 10.3 0.131 0.141 14.511 14.539 integrate_v_rspace 127 11.3 0.004 0.005 14.380 14.411 cp_fm_cholesky_invert 11 10.9 14.335 14.344 14.335 14.344 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.270 14.272 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.177 14.206 calculate_rho_elec 127 8.7 0.131 0.146 14.177 14.206 cp_fm_diag_elpa 82 13.4 0.001 0.002 11.145 11.145 cp_fm_redistribute_end 82 14.4 4.211 11.079 4.225 11.082 cp_fm_diag_elpa_base 82 14.4 6.437 10.508 6.838 10.956 init_scf_run 11 5.9 0.000 0.001 10.483 10.483 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.483 10.483 make_images_data 4970 15.5 0.060 0.069 9.003 10.241 hybrid_alltoall_any 5155 16.4 0.430 1.968 7.858 9.499 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.823 9.108 density_rs2pw 127 9.7 0.006 0.006 7.165 9.082 pw_transfer 1535 11.6 0.085 0.108 8.849 8.969 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.379 8.788 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.626 8.753 mp_irecv_dv 49801 16.2 3.258 8.548 3.258 8.548 dbcsr_mm_accdrv_process 41158 16.2 4.431 5.152 8.006 8.124 fft_wrap_pw1pw2_140 519 13.2 0.475 0.521 7.622 7.749 grid_integrate_task_list 127 12.3 7.246 7.749 7.246 7.749 cp_fm_upper_to_full 104 14.5 5.798 7.640 5.798 7.640 cp_fm_cholesky_decompose 22 10.9 7.395 7.459 7.395 7.459 wfi_extrapolate 11 7.9 0.001 0.001 7.425 7.425 rs_pw_transfer 1038 11.9 0.013 0.015 5.109 7.059 fft3d_ps 1281 14.7 2.646 2.867 6.903 7.004 dbcsr_complete_redistribute 393 12.7 1.178 1.206 4.602 6.378 calculate_dm_sparse 127 9.5 0.001 0.001 5.710 5.797 grid_collocate_task_list 127 9.7 4.871 5.764 4.871 5.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.384 5.389 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.544 5.231 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.440 5.216 mp_alltoall_d11v 2401 14.1 4.122 4.859 4.122 4.859 potential_pw2rs 127 12.3 0.020 0.022 4.732 4.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.469 4.605 mp_sum_l 7804 13.0 3.167 4.567 3.167 4.567 mp_allgather_i34 2485 14.5 1.494 4.319 1.494 4.319 mp_waitany 11660 13.9 2.303 4.188 2.303 4.188 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.372 4.132 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.973 4.010 mp_alltoall_i22 712 14.1 1.982 3.954 1.982 3.954 qs_energies_init_hamiltonians 11 5.9 0.047 0.064 3.813 3.815 rs_pw_transfer_RS2PW_140 138 11.5 0.332 0.352 1.902 3.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.530 3.573 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.782000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.636364, yerr=9.402831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.142067E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4123 56680. MP_Allreduce 11346 1064. MP_Sync 172 MP_Alltoall 1724 12509605. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 84 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 0.038 0.108 189.242 189.244 qs_mol_dyn_low 1 2.0 0.003 0.004 188.750 188.763 qs_forces 11 3.9 0.003 0.006 188.635 188.645 qs_energies 11 4.9 0.009 0.022 181.325 181.337 scf_env_do_scf 11 5.9 0.001 0.004 164.673 164.679 velocity_verlet 10 3.0 0.002 0.002 125.711 125.715 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.403 118.404 qs_scf_new_mos 118 7.6 0.001 0.001 80.200 80.565 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.200 80.564 dbcsr_multiply_generic 2529 12.6 0.193 0.197 79.651 80.321 ot_scf_mini 118 9.6 0.004 0.005 75.652 75.970 multiply_cannon 2529 13.6 0.560 0.596 54.972 57.267 multiply_cannon_loop 2529 14.6 1.189 1.215 51.190 52.671 init_scf_loop 11 6.9 0.000 0.001 46.146 46.148 ot_mini 118 10.6 0.001 0.001 42.404 42.731 prepare_preconditioner 11 7.9 0.000 0.000 41.994 42.015 make_preconditioner 11 8.9 0.000 0.002 41.994 42.015 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.604 40.600 multiply_cannon_multrec 30348 15.6 14.423 20.249 26.349 31.611 rebuild_ks_matrix 129 8.3 0.001 0.001 30.463 30.701 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 30.463 30.701 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.372 27.584 mp_waitall_1 149172 16.7 17.336 27.467 17.336 27.467 qs_ot_get_derivative 118 11.6 0.001 0.002 22.708 23.041 make_m2s 5058 13.6 0.098 0.102 20.391 21.382 make_images 5058 14.6 1.971 2.251 20.090 21.086 qs_ot_get_p 129 10.4 0.001 0.002 19.443 19.777 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.160 19.594 apply_single 129 13.6 0.001 0.001 19.160 19.594 ot_diis_step 118 11.6 0.018 0.019 19.573 19.575 cp_fm_upper_to_full 106 14.7 10.957 16.207 10.957 16.207 cp_fm_cholesky_invert 11 10.9 16.066 16.075 16.066 16.075 multiply_cannon_metrocomm3 30348 15.6 0.046 0.048 6.333 15.508 qs_ot_p2m_diag 84 11.4 0.346 0.394 15.264 15.315 sum_up_and_integrate 129 10.3 0.142 0.154 15.021 15.053 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.898 14.937 calculate_rho_elec 129 8.7 0.176 0.192 14.898 14.936 integrate_v_rspace 129 11.3 0.004 0.004 14.879 14.918 cp_dbcsr_syevd 84 12.4 0.006 0.010 13.976 13.977 dbcsr_complete_redistribute 397 12.7 1.519 1.628 8.942 12.644 make_images_data 5058 15.6 0.064 0.071 11.000 12.487 multiply_cannon_sync_h2d 30348 15.6 11.714 12.278 11.714 12.278 dbcsr_mm_accdrv_process 62780 16.2 7.409 8.369 11.497 11.999 hybrid_alltoall_any 5245 16.5 0.530 2.231 9.869 11.423 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.589 11.306 cp_fm_diag_elpa 84 13.4 0.001 0.001 10.814 10.818 cp_fm_redistribute_end 84 14.4 1.870 10.731 1.886 10.736 init_scf_run 11 5.9 0.000 0.001 10.726 10.727 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.725 10.727 cp_fm_diag_elpa_base 84 14.4 8.237 10.103 8.817 10.595 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.366 10.032 pw_transfer 1559 11.6 0.088 0.107 9.709 9.786 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.446 9.681 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.479 9.562 mp_alltoall_i22 720 14.1 5.548 9.232 5.548 9.232 density_rs2pw 129 9.7 0.006 0.006 7.287 8.708 fft_wrap_pw1pw2_140 527 13.2 0.488 0.496 8.422 8.527 grid_integrate_task_list 129 12.3 7.525 7.929 7.525 7.929 fft3d_ps 1301 14.7 2.800 2.887 7.687 7.750 wfi_extrapolate 11 7.9 0.005 0.018 7.631 7.631 cp_fm_cholesky_decompose 22 10.9 7.477 7.588 7.477 7.588 multiply_cannon_metrocomm4 25290 15.6 0.078 0.088 2.738 6.902 mp_irecv_dv 76751 16.2 2.591 6.633 2.591 6.633 calculate_dm_sparse 129 9.5 0.001 0.001 6.211 6.298 rs_pw_transfer 1054 12.0 0.013 0.014 4.773 6.270 grid_collocate_task_list 129 9.7 5.115 5.911 5.115 5.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.416 5.466 mp_alltoall_d11v 2429 14.1 4.848 5.417 4.848 5.417 potential_pw2rs 129 12.3 0.023 0.023 4.779 4.806 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.473 4.475 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.357 4.458 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.361 4.430 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.191 4.248 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.244000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1075.727273, yerr=17.236805 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.498030E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 176.046 176.048 qs_mol_dyn_low 1 2.0 0.003 0.003 175.659 175.672 qs_forces 11 3.9 0.010 0.010 175.274 175.278 qs_energies 11 4.9 0.001 0.002 167.669 167.676 scf_env_do_scf 11 5.9 0.001 0.001 150.363 150.368 velocity_verlet 10 3.0 0.002 0.002 115.776 115.798 scf_env_do_scf_inner_loop 118 6.6 0.006 0.059 114.800 114.801 qs_scf_new_mos 118 7.6 0.001 0.001 75.405 75.526 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.405 75.525 dbcsr_multiply_generic 2529 12.6 0.186 0.192 74.760 75.299 ot_scf_mini 118 9.6 0.004 0.004 70.943 71.001 multiply_cannon 2529 13.6 0.590 0.623 54.685 59.534 multiply_cannon_loop 2529 14.6 0.445 0.455 49.886 50.784 ot_mini 118 10.6 0.001 0.001 39.736 39.791 init_scf_loop 11 6.9 0.000 0.000 35.414 35.416 mp_waitall_1 126876 16.7 26.135 33.406 26.135 33.406 prepare_preconditioner 11 7.9 0.000 0.000 31.395 31.427 make_preconditioner 11 8.9 0.000 0.000 31.395 31.427 rebuild_ks_matrix 129 8.3 0.001 0.001 30.391 30.466 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.390 30.465 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.255 29.516 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.632 27.699 multiply_cannon_multrec 10116 15.6 10.444 14.204 17.750 21.036 ot_diis_step 118 11.6 0.020 0.021 20.098 20.098 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.588 19.976 apply_single 129 13.6 0.001 0.001 19.588 19.976 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 12.365 19.702 qs_ot_get_derivative 118 11.6 0.001 0.002 19.572 19.632 make_m2s 5058 13.6 0.067 0.073 16.457 18.933 make_images 5058 14.6 2.295 2.678 16.146 18.623 qs_ot_get_p 129 10.4 0.001 0.001 18.080 18.149 cp_fm_cholesky_invert 11 10.9 17.838 17.844 17.838 17.844 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.648 15.700 calculate_rho_elec 129 8.7 0.261 0.270 15.647 15.700 sum_up_and_integrate 129 10.3 0.182 0.191 15.261 15.311 integrate_v_rspace 129 11.3 0.004 0.004 15.079 15.135 qs_ot_p2m_diag 84 11.4 0.501 0.506 14.297 14.314 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.148 13.150 make_images_data 5058 15.6 0.052 0.061 10.125 12.856 hybrid_alltoall_any 5245 16.5 0.832 3.714 9.942 12.248 multiply_cannon_sync_h2d 10116 15.6 11.776 11.994 11.776 11.994 init_scf_run 11 5.9 0.000 0.001 10.454 10.454 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.454 10.454 pw_transfer 1559 11.6 0.086 0.092 10.075 10.116 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.081 10.091 cp_fm_diag_elpa_base 84 14.4 9.834 9.921 10.074 10.085 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.851 9.897 fft_wrap_pw1pw2_140 527 13.2 0.499 0.524 8.701 8.759 grid_integrate_task_list 129 12.3 7.769 8.306 7.769 8.306 mp_allgather_i34 2529 14.6 3.007 8.055 3.007 8.055 density_rs2pw 129 9.7 0.005 0.006 7.174 8.045 fft3d_ps 1301 14.7 2.726 2.826 7.999 8.017 cp_fm_cholesky_decompose 22 10.9 7.899 7.996 7.899 7.996 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 7.905 7.949 dbcsr_mm_accdrv_process 20934 16.1 2.619 3.442 6.932 7.541 wfi_extrapolate 11 7.9 0.001 0.001 7.513 7.513 multiply_cannon_metrocomm1 10116 15.6 0.029 0.029 4.401 7.427 calculate_dm_sparse 129 9.5 0.001 0.001 6.142 6.218 grid_collocate_task_list 129 9.7 5.438 6.123 5.438 6.123 mp_alltoall_d11v 2429 14.1 5.074 6.092 5.074 6.092 dbcsr_complete_redistribute 397 12.7 2.136 2.189 5.167 5.582 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.345 5.346 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.202 5.228 rs_pw_transfer 1054 12.0 0.013 0.013 4.142 5.053 potential_pw2rs 129 12.3 0.027 0.028 4.619 4.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.128 4.153 multiply_cannon_metrocomm4 7587 15.6 0.025 0.027 1.848 3.992 mp_irecv_dv 29102 15.9 1.812 3.922 1.812 3.922 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.873 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.392 3.724 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.702 3.722 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.669 3.678 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.533 3.635 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.048000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1394.363636, yerr=40.721388 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237301E+12 0.0% 0.0% 100.0% flops max/rank 11.786605E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806449600 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 2.886402E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58052. MP_Allreduce 11197 1509. MP_Sync 87 MP_Alltoall 1724 36993608. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.086 299.012 299.020 qs_mol_dyn_low 1 2.0 0.003 0.003 298.435 298.449 qs_forces 11 3.9 0.003 0.003 298.337 298.343 qs_energies 11 4.9 0.002 0.002 288.867 288.882 scf_env_do_scf 11 5.9 0.001 0.001 266.531 266.549 velocity_verlet 10 3.0 0.004 0.006 215.570 215.578 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 141.924 141.925 init_scf_loop 11 6.9 0.000 0.000 124.354 124.357 prepare_preconditioner 11 7.9 0.000 0.000 119.196 119.227 make_preconditioner 11 8.9 0.000 0.000 119.196 119.227 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.337 116.446 qs_scf_new_mos 118 7.6 0.001 0.001 90.423 90.521 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.422 90.520 ot_scf_mini 118 9.6 0.004 0.004 85.675 85.794 dbcsr_multiply_generic 2531 12.6 0.217 0.223 82.863 83.061 cp_fm_upper_to_full 105 14.8 52.915 75.852 52.915 75.852 multiply_cannon 2531 13.6 0.713 0.783 58.307 59.494 multiply_cannon_loop 2531 14.6 0.476 0.484 54.739 55.091 ot_mini 118 10.6 0.001 0.001 44.406 44.555 dbcsr_complete_redistribute 395 12.7 3.994 4.046 29.491 42.503 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.095 39.076 rebuild_ks_matrix 129 8.3 0.001 0.001 38.790 38.979 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 38.789 38.978 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.814 36.646 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.715 35.900 mp_alltoall_i22 718 14.1 21.690 34.854 21.690 34.854 cp_fm_cholesky_invert 11 10.9 33.362 33.369 33.362 33.369 mp_waitall_1 104660 16.8 27.483 30.250 27.483 30.250 qs_ot_get_p 129 10.4 0.001 0.001 25.901 26.106 qs_ot_get_derivative 118 11.6 0.002 0.002 24.199 24.320 qs_ot_p2m_diag 83 11.4 0.879 0.885 21.801 21.830 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.960 20.983 calculate_rho_elec 129 8.7 0.487 0.487 20.959 20.983 make_m2s 5062 13.6 0.078 0.081 19.635 20.478 ot_diis_step 118 11.6 0.022 0.023 20.152 20.153 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.063 20.065 sum_up_and_integrate 129 10.3 0.323 0.324 19.969 20.044 make_images 5062 14.6 3.791 3.904 19.150 19.997 integrate_v_rspace 129 11.3 0.004 0.004 19.645 19.722 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.281 19.540 apply_single 129 13.6 0.001 0.001 19.281 19.539 multiply_cannon_metrocomm3 10124 15.6 0.024 0.024 18.415 18.963 multiply_cannon_multrec 10124 15.6 10.546 12.257 17.992 18.059 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.852 16.854 cp_fm_diag_elpa_base 83 14.4 12.453 14.093 16.848 16.848 multiply_cannon_sync_h2d 10124 15.6 15.783 15.817 15.783 15.817 pw_transfer 1559 11.6 0.093 0.093 12.465 12.473 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.229 12.236 init_scf_run 11 5.9 0.000 0.001 12.195 12.197 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.195 12.197 hybrid_alltoall_any 5248 16.5 1.308 3.053 10.420 12.022 make_images_data 5062 15.6 0.061 0.065 10.305 11.977 fft_wrap_pw1pw2_140 527 13.2 0.545 0.549 10.886 10.898 fft3d_ps 1301 14.7 2.759 2.768 10.249 10.257 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.390 9.483 mp_alltoall_d11v 2423 14.1 8.234 9.134 8.234 9.134 wfi_extrapolate 11 7.9 0.001 0.001 9.072 9.072 dbcsr_mm_accdrv_process 20942 16.1 3.821 5.758 7.206 9.036 cp_fm_cholesky_decompose 22 10.9 8.882 8.953 8.882 8.953 grid_integrate_task_list 129 12.3 8.648 8.801 8.648 8.801 density_rs2pw 129 9.7 0.005 0.005 8.440 8.554 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.126 8.127 calculate_dm_sparse 129 9.5 0.001 0.001 6.609 6.723 grid_collocate_task_list 129 9.7 6.410 6.443 6.410 6.443 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.228 6.323 rs_scatter_matrices 140 9.7 3.660 4.565 6.067 6.299 copy_dbcsr_to_fm 186 11.8 0.004 0.005 6.034 6.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.020000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2583.909091, yerr=163.618433 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 1.261818E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 84.972 84.973 qs_energies 1 2.0 0.000 0.000 84.520 84.527 ls_scf 1 3.0 0.000 0.000 83.606 83.613 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.643 72.817 multiply_cannon 111 7.7 0.017 0.020 55.924 56.944 multiply_cannon_loop 111 8.7 0.207 0.224 52.406 53.867 ls_scf_main 1 4.0 0.000 0.000 52.333 52.333 density_matrix_trs4 2 5.0 0.002 0.003 46.838 46.919 ls_scf_init_scf 1 4.0 0.000 0.000 28.227 28.228 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.123 27.168 mp_waitall_1 11031 10.9 21.937 25.326 21.937 25.326 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.034 25.050 multiply_cannon_multrec 2664 9.7 8.205 8.945 15.535 17.199 multiply_cannon_sync_h2d 2664 9.7 13.991 16.031 13.991 16.031 make_m2s 222 7.7 0.008 0.011 13.078 13.721 make_images 222 8.7 0.099 0.110 13.056 13.698 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.335 12.941 make_images_data 222 9.7 0.004 0.006 7.598 8.250 hybrid_alltoall_any 227 10.6 0.215 1.834 6.464 8.012 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.606 6.950 7.930 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.330 7.752 dbcsr_mm_accdrv_process_sort 4760 11.4 6.235 7.133 6.235 7.133 calculate_norms 4752 9.8 5.558 6.206 5.558 6.206 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.930 5.081 mp_sum_l 807 5.4 3.104 4.613 3.104 4.613 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.312 3.506 make_images_sizes 222 9.7 0.000 0.000 0.820 3.488 mp_alltoall_i44 222 10.7 0.820 3.488 0.820 3.488 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.080 3.440 mp_irecv_dv 6231 10.9 2.061 3.416 2.061 3.416 arnoldi_extremal 4 6.8 0.000 0.000 3.220 3.244 arnoldi_normal_ev 4 7.8 0.001 0.002 3.220 3.244 build_subspace 16 8.4 0.009 0.012 3.131 3.134 ls_scf_post 1 4.0 0.000 0.000 3.047 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.305 2.794 dbcsr_merge_single_wm 555 10.7 0.470 0.606 2.298 2.786 make_images_pack 222 9.7 2.208 2.625 2.210 2.626 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.338 2.577 dbcsr_sort_data 658 11.4 2.085 2.502 2.085 2.502 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.459 2.073 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.333 2.445 buffer_matrices_ensure_size 222 8.7 1.787 2.122 1.787 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.791 1.792 rebuild_ks_matrix 3 7.3 0.000 0.000 1.782 1.783 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.782 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.973000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 2.116604E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.106 91.380 91.382 qs_energies 1 2.0 0.000 0.000 90.704 90.708 ls_scf 1 3.0 0.000 0.000 89.319 89.324 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.342 75.638 multiply_cannon 111 7.7 0.029 0.040 53.287 57.464 ls_scf_main 1 4.0 0.000 0.000 55.132 55.137 multiply_cannon_loop 111 8.7 0.116 0.123 50.031 53.494 density_matrix_trs4 2 5.0 0.002 0.003 49.378 49.547 ls_scf_init_scf 1 4.0 0.000 0.000 30.699 30.701 mp_waitall_1 9105 10.9 21.068 30.580 21.068 30.580 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.502 29.599 multiply_cannon_multrec 1332 9.7 13.054 17.063 22.276 27.433 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.132 27.145 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.844 21.269 make_m2s 222 7.7 0.006 0.007 15.172 15.788 make_images 222 8.7 1.580 1.963 15.142 15.759 dbcsr_mm_accdrv_process 4041 10.4 0.284 0.449 8.826 10.374 dbcsr_mm_accdrv_process_sort 4041 11.4 8.418 9.945 8.418 9.945 make_images_data 222 9.7 0.004 0.004 8.625 9.420 hybrid_alltoall_any 227 10.6 0.523 2.493 8.072 8.938 mp_sum_l 807 5.4 5.630 8.922 5.630 8.922 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.236 7.927 mp_irecv_dv 3311 11.0 3.216 7.859 3.216 7.859 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.361 7.049 calculate_norms 2376 9.8 5.999 6.660 5.999 6.660 multiply_cannon_sync_h2d 1332 9.7 4.874 5.930 4.874 5.930 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.263 arnoldi_extremal 4 6.8 0.000 0.000 4.620 4.641 arnoldi_normal_ev 4 7.8 0.001 0.005 4.620 4.641 build_subspace 16 8.4 0.014 0.021 4.361 4.364 ls_scf_post 1 4.0 0.000 0.000 3.488 3.493 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.129 3.370 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.313 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.237 3.241 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.228 2.737 3.229 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.623 2.711 make_images_pack 222 9.7 2.031 2.388 2.033 2.390 mp_allgather_i34 111 8.7 0.963 2.279 0.963 2.279 dbcsr_sort_data 436 11.2 1.820 2.055 1.820 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 rebuild_ks_matrix 3 7.3 0.000 0.000 1.887 1.889 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.887 1.889 dbcsr_data_new 4174 10.1 1.612 1.852 1.612 1.852 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.382000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 2.704126E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.067 93.188 93.189 qs_energies 1 2.0 0.000 0.000 92.627 92.631 ls_scf 1 3.0 0.000 0.000 91.223 91.227 dbcsr_multiply_generic 111 6.7 0.015 0.018 75.972 76.271 ls_scf_main 1 4.0 0.000 0.000 57.146 57.152 multiply_cannon 111 7.7 0.041 0.080 52.504 56.723 multiply_cannon_loop 111 8.7 0.100 0.111 48.946 52.722 density_matrix_trs4 2 5.0 0.002 0.003 51.145 51.298 mp_waitall_1 7281 11.0 23.277 33.193 23.277 33.193 ls_scf_init_scf 1 4.0 0.000 0.000 30.488 30.490 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.302 29.372 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.902 26.913 multiply_cannon_multrec 888 9.7 12.611 15.399 21.156 24.552 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.768 22.678 make_m2s 222 7.7 0.006 0.006 16.855 18.092 make_images 222 8.7 1.965 2.289 16.817 18.052 hybrid_alltoall_any 227 10.6 0.621 2.870 9.159 10.571 make_images_data 222 9.7 0.003 0.004 9.518 10.430 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.417 8.064 9.298 dbcsr_mm_accdrv_process_sort 3754 11.4 7.694 8.881 7.694 8.881 mp_sum_l 807 5.4 5.055 8.402 5.055 8.402 multiply_cannon_sync_h2d 888 9.7 6.053 7.321 6.053 7.321 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.428 6.984 mp_irecv_dv 2335 11.1 2.413 6.942 2.413 6.942 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.803 6.487 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.616 6.057 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.973 5.224 arnoldi_extremal 4 6.8 0.000 0.000 5.090 5.103 arnoldi_normal_ev 4 7.8 0.001 0.005 5.090 5.103 build_subspace 16 8.4 0.014 0.020 4.785 4.792 calculate_norms 1584 9.8 4.316 4.644 4.316 4.644 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.438 3.774 mp_allgather_i34 111 8.7 1.409 3.719 1.409 3.719 ls_scf_post 1 4.0 0.000 0.000 3.589 3.595 dbcsr_matrix_vector_mult_local 304 10.0 3.005 3.584 3.007 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.313 3.411 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.842 2.931 make_images_pack 222 9.7 1.823 2.139 1.826 2.142 dbcsr_sort_data 325 11.1 1.905 2.127 1.905 2.127 dbcsr_data_release 9322 10.9 1.317 1.968 1.317 1.968 make_images_sizes 222 9.7 0.000 0.000 0.997 1.960 mp_alltoall_i44 222 10.7 0.997 1.960 0.997 1.960 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.911 1.913 rebuild_ks_matrix 3 7.3 0.000 0.000 1.893 1.895 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.893 1.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.189000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2181.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 3.381072E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.047 97.130 97.130 qs_energies 1 2.0 0.000 0.000 96.531 96.539 ls_scf 1 3.0 0.000 0.000 94.896 94.902 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.624 78.824 ls_scf_main 1 4.0 0.000 0.000 59.051 59.051 multiply_cannon 111 7.7 0.042 0.143 51.634 55.926 density_matrix_trs4 2 5.0 0.002 0.003 52.897 53.034 multiply_cannon_loop 111 8.7 0.115 0.129 46.597 50.284 ls_scf_init_scf 1 4.0 0.000 0.000 32.564 32.565 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.331 31.403 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.860 28.870 mp_waitall_1 6369 11.0 22.614 28.546 22.614 28.546 multiply_cannon_multrec 1332 9.7 14.209 17.642 22.078 24.926 make_m2s 222 7.7 0.006 0.007 20.945 22.345 make_images 222 8.7 3.138 3.607 20.895 22.297 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.274 17.259 make_images_data 222 9.7 0.004 0.004 11.656 13.235 hybrid_alltoall_any 227 10.6 0.797 3.817 10.981 12.810 dbcsr_mm_accdrv_process 3641 10.4 0.219 0.408 7.504 9.011 dbcsr_mm_accdrv_process_sort 3641 11.4 7.132 8.611 7.132 8.611 mp_sum_l 807 5.4 4.463 8.602 4.463 8.602 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.468 6.863 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.111 multiply_cannon_sync_h2d 1332 9.7 5.516 6.098 5.516 6.098 mp_irecv_dv 3229 10.9 2.059 6.039 2.059 6.039 arnoldi_extremal 4 6.8 0.000 0.000 5.212 5.224 arnoldi_normal_ev 4 7.8 0.001 0.005 5.212 5.224 build_subspace 16 8.4 0.014 0.021 4.869 4.877 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.559 4.728 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.512 4.620 mp_allgather_i34 111 8.7 2.172 4.540 2.172 4.540 calculate_norms 2376 9.8 4.200 4.539 4.200 4.539 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.585 3.878 dbcsr_matrix_vector_mult_local 304 10.0 3.178 3.667 3.180 3.669 dbcsr_sort_data 658 11.4 3.088 3.365 3.088 3.365 ls_scf_post 1 4.0 0.000 0.000 3.281 3.289 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.108 dbcsr_merge_single_wm 555 10.7 0.531 0.638 2.821 3.100 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.098 ls_scf_store_result 1 5.0 0.000 0.000 3.022 3.083 dbcsr_data_release 10477 10.7 1.583 2.563 1.583 2.563 dbcsr_finalize 304 7.8 0.049 0.061 1.799 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.975 1.976 rebuild_ks_matrix 3 7.3 0.000 0.000 1.952 1.954 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.952 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.130000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 4.630430E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.054 99.222 99.223 qs_energies 1 2.0 0.000 0.000 98.457 98.462 ls_scf 1 3.0 0.000 0.000 96.509 96.519 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.746 77.979 ls_scf_main 1 4.0 0.000 0.000 62.197 62.198 multiply_cannon 111 7.7 0.101 0.190 55.348 60.473 density_matrix_trs4 2 5.0 0.002 0.003 54.843 54.933 multiply_cannon_loop 111 8.7 0.070 0.079 50.767 52.071 mp_waitall_1 5436 11.0 26.580 33.018 26.580 33.018 ls_scf_init_scf 1 4.0 0.000 0.000 30.711 30.716 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.381 29.425 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.204 27.216 multiply_cannon_multrec 444 9.7 13.946 16.453 20.984 23.565 make_m2s 222 7.7 0.004 0.005 17.688 20.277 make_images 222 8.7 3.720 4.416 17.626 20.217 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.055 16.444 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.345 14.877 hybrid_alltoall_any 227 10.6 0.789 3.776 9.803 12.455 make_images_data 222 9.7 0.003 0.004 9.981 12.446 multiply_cannon_sync_h2d 444 9.7 6.573 8.245 6.573 8.245 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.342 6.739 7.861 dbcsr_mm_accdrv_process_sort 3003 11.4 6.424 7.520 6.424 7.520 mp_allgather_i34 111 8.7 2.828 7.062 2.828 7.062 arnoldi_extremal 4 6.8 0.000 0.000 5.850 5.865 arnoldi_normal_ev 4 7.8 0.001 0.005 5.850 5.865 build_subspace 16 8.4 0.015 0.020 5.460 5.473 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.629 4.761 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.195 4.379 mp_sum_l 807 5.4 2.755 4.378 2.755 4.378 dbcsr_matrix_vector_mult_local 304 10.0 3.674 4.145 3.676 4.147 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.637 4.007 mp_irecv_dv 1241 11.2 1.617 3.952 1.617 3.952 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.704 3.822 calculate_norms 792 9.8 3.539 3.646 3.539 3.646 ls_scf_post 1 4.0 0.000 0.000 3.601 3.606 make_images_sizes 222 9.7 0.000 0.000 1.061 3.518 mp_alltoall_i44 222 10.7 1.061 3.518 1.061 3.518 ls_scf_store_result 1 5.0 0.000 0.000 3.381 3.429 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.880 3.166 dbcsr_finalize 304 7.8 0.062 0.078 2.194 2.249 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.209 2.211 rebuild_ks_matrix 3 7.3 0.000 0.000 2.178 2.179 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.178 2.179 dbcsr_merge_all 275 8.9 0.472 0.525 2.046 2.108 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.223000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3626.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/239d071ebeded0f727ee8686fa1a72666d665905_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 8.771002E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.077 0.098 108.656 108.657 qs_energies 1 2.0 0.000 0.000 107.258 107.272 ls_scf 1 3.0 0.000 0.000 104.288 104.301 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.646 77.769 ls_scf_main 1 4.0 0.000 0.000 66.442 66.443 density_matrix_trs4 2 5.0 0.002 0.003 57.062 57.111 multiply_cannon 111 7.7 0.108 0.186 50.090 52.143 multiply_cannon_loop 111 8.7 0.068 0.069 46.603 47.509 ls_scf_init_scf 1 4.0 0.000 0.000 34.180 34.181 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.513 32.538 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.700 29.708 mp_waitall_1 4527 11.1 22.396 26.264 22.396 26.264 make_m2s 222 7.7 0.005 0.005 23.907 24.925 make_images 222 8.7 4.594 4.975 23.801 24.816 multiply_cannon_multrec 444 9.7 17.893 18.709 22.491 23.093 hybrid_alltoall_any 227 10.6 1.662 3.630 12.955 15.614 make_images_data 222 9.7 0.003 0.003 13.150 15.604 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.664 11.128 multiply_cannon_sync_h2d 444 9.7 8.846 8.887 8.846 8.887 arnoldi_extremal 4 6.8 0.000 0.000 7.327 7.336 arnoldi_normal_ev 4 7.8 0.002 0.009 7.327 7.336 build_subspace 16 8.4 0.026 0.037 6.769 6.785 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.448 5.587 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.349 5.436 dbcsr_matrix_vector_mult_local 304 10.0 5.013 5.318 5.016 5.321 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.998 5.254 dbcsr_mm_accdrv_process 1814 10.4 0.183 0.304 4.434 4.559 dbcsr_mm_accdrv_process_sort 1814 11.4 4.137 4.258 4.137 4.258 ls_scf_post 1 4.0 0.000 0.000 3.665 3.679 mp_allgather_i34 111 8.7 1.157 3.631 1.157 3.631 make_images_sizes 222 9.7 0.000 0.000 1.427 3.499 mp_alltoall_i44 222 10.7 1.427 3.499 1.427 3.499 ls_scf_store_result 1 5.0 0.000 0.000 3.412 3.425 calculate_norms 792 9.8 3.242 3.277 3.242 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.160 dbcsr_merge_all 275 8.9 0.889 0.914 2.863 2.941 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.940 2.940 dbcsr_complete_redistribute 5 7.6 1.455 1.491 2.781 2.895 dbcsr_data_release 12724 10.6 2.320 2.861 2.320 2.861 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.608 2.610 matrix_ls_to_qs 2 6.0 0.000 0.000 2.450 2.579 rebuild_ks_matrix 3 7.3 0.000 0.000 2.541 2.543 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.541 2.543 dbcsr_sort_data 325 11.1 2.438 2.500 2.438 2.500 dbcsr_new_transposed 4 7.5 0.242 0.251 2.299 2.316 dbcsr_frobenius_norm 74 6.6 2.056 2.137 2.204 2.245 dbcsr_add_d 103 6.2 0.000 0.000 2.130 2.202 dbcsr_add_anytype 103 7.2 0.859 0.892 2.130 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.657000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6885.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 239d071ebeded0f727ee8686fa1a72666d665905 Summary: empty Status: OK