=== 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: 9bf2b2015f7fa359f4ee539b0939531368083923 ################# 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/01 job id: 45235451 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/02 job id: 45235452 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/03 job id: 45235453 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/04 job id: 45235456 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/05 job id: 45235459 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/06 job id: 45235460 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/07 job id: 45235461 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/08 job id: 45235462 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/09 job id: 45235463 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/10 job id: 45235464 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/11 job id: 45235466 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/12 job id: 45235470 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/13 job id: 45235471 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/14 job id: 45235472 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/15 job id: 45235473 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/16 job id: 45235474 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/17 job id: 45235475 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/18 job id: 45235476 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/19 job id: 45235477 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/20 job id: 45235479 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/21 job id: 45235481 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/22 job id: 45235482 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/23 job id: 45235488 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/24 job id: 45235491 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/25 job id: 45235494 --- 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/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/26 job id: 45235495 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/27 job id: 45235498 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 0.022 0.036 135.038 135.039 farming_run 1 2.0 134.487 134.489 135.008 135.012 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463902E+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 231 1098811. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.026 116.270 116.271 qs_energies 1 2.0 0.000 0.000 115.976 115.977 mp2_main 1 3.0 0.000 0.000 113.900 113.901 mp2_gpw_main 1 4.0 0.020 0.026 112.917 112.919 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.781 94.201 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.451 55.871 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.767 47.144 get_2c_integrals 1 6.0 0.000 0.001 37.614 38.156 integrate_v_rspace 273 8.0 0.436 0.449 25.120 30.170 pw_transfer 6555 10.6 0.371 0.388 27.435 28.312 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.063 26.832 grid_integrate_task_list 273 9.0 20.930 26.450 20.930 26.450 fft_wrap_pw1pw2_100 2178 12.4 1.227 1.423 23.587 24.355 compute_2c_integrals 1 7.0 0.002 0.003 19.797 19.798 compute_2c_integrals_loop_lm 1 8.0 0.017 0.018 18.997 19.529 mp2_eri_2c_integrate_gpw 1 9.0 2.366 2.426 18.979 19.512 rpa_ri_compute_en 1 5.0 0.011 0.012 19.027 19.210 cp_fm_cholesky_decompose 12 8.2 17.733 18.313 17.733 18.313 cholesky_decomp 1 7.0 0.000 0.000 16.658 17.233 fft3d_s 5443 13.4 16.152 16.681 16.175 16.703 ao_to_mo_and_store_B_mult_1 272 7.0 10.854 15.561 10.854 15.561 calculate_wavefunction 272 8.0 5.467 5.626 12.569 13.193 rpa_num_int 1 6.0 0.001 0.009 10.505 10.505 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.447 10.454 calc_mat_Q 8 8.0 0.000 0.000 9.326 9.412 contract_S_to_Q 8 9.0 0.000 0.000 8.748 8.836 calc_potential_gpw 544 9.5 0.005 0.007 8.246 8.691 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.266 8.612 potential_pw2rs 545 10.0 0.107 0.109 7.750 8.471 parallel_gemm_fm 14 9.1 0.000 0.000 8.316 8.409 parallel_gemm_fm_cosma 14 10.1 8.316 8.409 8.316 8.409 create_integ_mat 1 6.0 0.022 0.027 7.996 7.996 collocate_single_gaussian 272 10.0 0.040 0.041 7.514 7.784 array2fm 1 7.0 0.000 0.000 6.797 7.386 pw_scatter_s 2720 13.7 4.392 4.601 4.392 4.601 pw_gather_s 2722 13.2 3.880 4.220 3.880 4.220 array2fm_buffer_send 1 8.0 3.020 3.166 3.020 3.166 pw_poisson_solve 545 10.5 1.123 1.170 2.205 2.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.919024, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.042 401.843 401.844 farming_run 1 2.0 401.061 401.065 401.790 401.791 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.226924E+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 744 385880. MP_Allreduce 1941 22272. MP_Sync 37 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 0.017 0.052 211.021 211.022 qs_energies 1 2.0 0.000 0.000 210.731 210.740 scf_env_do_scf 1 3.0 0.000 0.000 106.891 106.891 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.819 105.826 rebuild_ks_matrix 4 6.0 0.000 0.000 105.818 105.825 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 105.818 105.825 hfx_ks_matrix 4 8.0 0.001 0.001 105.375 105.378 integrate_four_center 4 9.0 0.144 0.451 105.374 105.377 mp2_main 1 3.0 0.000 0.000 103.498 103.508 mp2_gpw_main 1 4.0 0.035 0.055 102.142 102.153 integrate_four_center_main 4 10.0 0.101 0.579 96.682 99.640 integrate_four_center_bin 266 11.0 96.581 99.599 96.581 99.599 init_scf_loop 1 4.0 0.000 0.000 92.391 92.391 mp2_ri_gpw_compute_in 1 5.0 0.066 0.072 74.970 76.030 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.431 55.487 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.219 47.184 integrate_v_rspace 95 8.0 0.397 0.564 28.640 33.416 pw_transfer 2240 10.6 0.144 0.166 29.975 30.322 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.986 29.408 grid_integrate_task_list 95 9.0 23.878 28.878 23.878 28.878 mp2_ri_gpw_compute_en 1 5.0 0.054 0.064 27.007 28.735 ao_to_mo_and_store_B_mult_1 91 7.0 10.526 28.271 10.526 28.271 fft_wrap_pw1pw2_100 730 12.4 1.277 1.445 26.657 27.155 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.885 25.294 25.303 get_2c_integrals 1 6.0 0.000 0.000 20.365 20.476 compute_2c_integrals 1 7.0 0.003 0.006 19.331 19.350 compute_2c_integrals_loop_lm 1 8.0 0.003 0.010 18.872 19.188 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.862 18.869 19.187 fft3d_s 1823 13.4 18.436 18.834 18.449 18.847 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.498 14.498 calculate_wavefunction 91 8.0 2.026 2.058 9.751 10.018 potential_pw2rs 186 10.0 0.034 0.036 8.678 9.251 mp2_ri_gpw_compute_en_expansio 172 7.0 0.553 0.589 8.794 9.128 mp2_ri_gpw_compute_en_comm 22 7.0 0.480 0.495 8.288 8.899 local_gemm 172 8.0 8.240 8.583 8.240 8.583 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.240 8.578 collocate_single_gaussian 91 10.0 0.017 0.021 7.893 8.147 calc_potential_gpw 182 9.5 0.002 0.002 7.917 8.111 mp_sendrecv_dm3 2068 8.0 6.373 6.965 6.373 6.965 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.433 6.334 6.433 pw_gather_s 912 13.2 4.907 5.387 4.907 5.387 mp_sync 37 10.5 3.342 5.377 3.342 5.377 pw_scatter_s 910 13.7 3.928 4.300 3.928 4.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.136554, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 452.747264E+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 272. MP_Sync 530 MP_Alltoall 2083 589622. 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 0.020 0.083 58.751 58.752 qs_mol_dyn_low 1 2.0 0.005 0.018 58.386 58.408 qs_forces 11 3.9 0.017 0.026 58.157 58.166 qs_energies 11 4.9 0.001 0.001 56.423 56.442 scf_env_do_scf 11 5.9 0.001 0.001 49.893 49.894 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 47.213 47.213 qs_scf_new_mos 108 7.5 0.000 0.001 34.673 34.944 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.673 34.944 dbcsr_multiply_generic 2286 12.5 0.104 0.109 33.434 33.906 ot_scf_mini 108 9.5 0.002 0.002 33.029 33.200 velocity_verlet 10 3.0 0.001 0.002 28.612 28.616 multiply_cannon 2286 13.5 0.192 0.204 25.767 27.564 multiply_cannon_loop 2286 14.5 1.460 1.555 24.598 26.287 ot_mini 108 10.5 0.001 0.001 19.431 19.679 qs_ot_get_derivative 108 11.5 0.001 0.001 16.405 16.580 mp_waitall_1 245248 16.5 8.778 15.690 8.778 15.690 multiply_cannon_metrocomm3 54864 15.5 0.068 0.076 5.566 12.967 multiply_cannon_multrec 54864 15.5 4.280 6.597 7.522 10.933 rebuild_ks_matrix 119 8.3 0.000 0.000 9.346 9.506 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 9.346 9.506 qs_ot_get_p 119 10.4 0.001 0.001 9.017 9.278 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.215 8.343 multiply_cannon_sync_h2d 54864 15.5 5.950 6.721 5.950 6.721 mp_sum_l 7207 12.9 4.529 6.132 4.529 6.132 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.926 5.962 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.529 5.934 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.668 5.767 calculate_rho_elec 119 8.7 0.012 0.017 5.667 5.767 sum_up_and_integrate 119 10.3 0.012 0.015 5.651 5.659 integrate_v_rspace 119 11.3 0.003 0.003 5.639 5.647 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.512 5.610 init_scf_run 11 5.9 0.000 0.001 5.164 5.164 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.164 5.164 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.160 5.162 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.785 4.786 cp_fm_redistribute_end 50 14.0 2.431 4.735 2.449 4.741 cp_fm_diag_elpa_base 50 14.0 2.278 4.583 2.287 4.598 dbcsr_mm_accdrv_process 76910 16.1 1.130 1.828 3.163 4.547 rs_pw_transfer 974 11.9 0.011 0.013 4.065 4.263 density_rs2pw 119 9.7 0.004 0.005 3.508 3.618 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.951 3.170 apply_single 119 13.6 0.000 0.000 2.951 3.169 pw_transfer 1439 11.6 0.053 0.059 2.948 3.108 calculate_dm_sparse 119 9.5 0.000 0.001 2.940 3.055 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.871 3.036 potential_pw2rs 119 12.3 0.004 0.004 2.881 2.971 make_m2s 4572 13.5 0.054 0.057 2.826 2.967 multiply_cannon_metrocomm1 54864 15.5 0.054 0.060 1.695 2.965 make_images 4572 14.5 0.133 0.139 2.744 2.883 fft3d_ps 1201 14.6 0.370 0.481 2.646 2.813 ot_diis_step 108 11.5 0.006 0.006 2.761 2.762 init_scf_loop 11 6.9 0.000 0.000 2.648 2.649 jit_kernel_multiply 13 15.8 1.973 2.642 1.973 2.642 calculate_first_density_matrix 1 7.0 0.007 0.011 2.613 2.620 wfi_extrapolate 11 7.9 0.001 0.001 2.475 2.475 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.434 2.437 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.295 2.373 fft_wrap_pw1pw2_140 487 13.2 0.077 0.092 2.203 2.369 acc_transpose_blocks 54864 15.5 0.226 0.255 1.692 2.171 grid_integrate_task_list 119 12.3 2.003 2.112 2.003 2.112 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.049 2.099 mp_alltoall_d11v 2130 13.8 1.879 2.084 1.879 2.084 mp_sum_d 4135 12.0 1.350 1.984 1.350 1.984 mp_alltoall_z22v 1201 16.6 1.538 1.677 1.538 1.677 mp_waitany 12084 13.8 1.451 1.643 1.451 1.643 make_images_sizes 4572 15.5 0.004 0.005 1.332 1.639 mp_alltoall_i44 4572 16.5 1.328 1.635 1.328 1.635 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.475 1.505 prepare_preconditioner 11 7.9 0.000 0.000 1.387 1.411 make_preconditioner 11 8.9 0.000 0.000 1.387 1.411 grid_collocate_task_list 119 9.7 1.294 1.382 1.294 1.382 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.322 1.355 make_images_data 4572 15.5 0.038 0.043 0.807 1.227 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.182 1.214 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.752000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 487.452672E+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 54 MP_Alltoall 2060 1627150. 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 0.035 0.039 39.530 39.531 qs_mol_dyn_low 1 2.0 0.003 0.003 39.130 39.138 qs_forces 11 3.9 0.003 0.003 39.061 39.063 qs_energies 11 4.9 0.008 0.015 37.370 37.373 scf_env_do_scf 11 5.9 0.001 0.001 31.718 31.718 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.253 29.254 dbcsr_multiply_generic 2286 12.5 0.111 0.124 21.623 21.977 qs_scf_new_mos 108 7.5 0.001 0.001 19.905 20.145 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.905 20.144 ot_scf_mini 108 9.5 0.002 0.003 19.016 19.195 velocity_verlet 10 3.0 0.001 0.001 18.279 18.281 multiply_cannon 2286 13.5 0.207 0.216 16.599 18.192 multiply_cannon_loop 2286 14.5 0.894 0.969 15.526 16.847 ot_mini 108 10.5 0.001 0.001 11.729 11.969 mp_waitall_1 200699 16.5 5.697 10.992 5.697 10.992 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.187 9.692 qs_ot_get_derivative 108 11.5 0.001 0.001 9.282 9.464 multiply_cannon_multrec 27432 15.5 1.992 4.289 6.124 9.001 rebuild_ks_matrix 119 8.3 0.000 0.000 7.284 7.424 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.283 7.423 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.439 6.565 dbcsr_mm_accdrv_process 47894 16.0 3.080 5.233 4.063 5.985 qs_ot_get_p 119 10.4 0.001 0.001 4.423 4.655 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.608 4.461 init_scf_run 11 5.9 0.000 0.001 4.401 4.402 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.401 4.402 sum_up_and_integrate 119 10.3 0.025 0.028 4.263 4.272 integrate_v_rspace 119 11.3 0.002 0.003 4.239 4.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.039 4.105 apply_single 119 13.6 0.000 0.000 3.039 4.105 mp_sum_l 7207 12.9 2.002 3.951 2.002 3.951 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.890 3.921 calculate_rho_elec 119 8.7 0.021 0.024 3.889 3.920 rs_pw_transfer 974 11.9 0.010 0.011 2.681 3.109 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.988 3.006 multiply_cannon_sync_h2d 27432 15.5 2.226 2.870 2.226 2.870 calculate_first_density_matrix 1 7.0 0.001 0.001 2.864 2.867 make_m2s 4572 13.5 0.052 0.054 2.562 2.791 make_images 4572 14.5 0.199 0.235 2.474 2.701 density_rs2pw 119 9.7 0.004 0.004 2.216 2.666 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.556 2.557 init_scf_loop 11 6.9 0.000 0.000 2.442 2.443 ot_diis_step 108 11.5 0.011 0.011 2.398 2.399 calculate_dm_sparse 119 9.5 0.000 0.001 2.245 2.322 jit_kernel_multiply 10 16.1 0.930 2.263 0.930 2.263 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.226 2.228 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.131 2.227 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.217 2.217 cp_fm_redistribute_end 50 14.0 1.124 2.189 1.127 2.192 cp_fm_diag_elpa_base 50 14.0 1.033 2.100 1.061 2.138 pw_transfer 1439 11.6 0.065 0.071 2.013 2.046 potential_pw2rs 119 12.3 0.006 0.006 1.983 1.994 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.922 1.957 grid_integrate_task_list 119 12.3 1.844 1.939 1.844 1.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.898 1.937 fft3d_ps 1201 14.6 0.516 0.567 1.623 1.654 make_images_data 4572 15.5 0.045 0.051 1.149 1.606 acc_transpose_blocks 27432 15.5 0.110 0.114 1.261 1.582 prepare_preconditioner 11 7.9 0.000 0.000 1.532 1.559 make_preconditioner 11 8.9 0.000 0.000 1.532 1.559 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.509 1.545 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.432 1.491 wfi_extrapolate 11 7.9 0.001 0.001 1.486 1.486 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.463 1.471 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.995 1.458 grid_collocate_task_list 119 9.7 1.226 1.364 1.226 1.364 mp_alltoall_d11v 2130 13.8 1.182 1.307 1.182 1.307 mp_allgather_i34 2286 14.5 0.518 1.273 0.518 1.273 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.139 1.188 mp_sum_d 4135 12.0 0.570 1.035 0.570 1.035 qs_energies_init_hamiltonians 11 5.9 0.016 0.022 0.972 0.975 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.529 0.961 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.728 0.961 mp_waitany 5720 13.7 0.508 0.945 0.508 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.902 0.915 mp_alltoall_z22v 1201 16.6 0.712 0.814 0.712 0.814 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.609 0.736 0.799 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.531000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.220096E+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 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1874908. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.030 0.081 39.236 39.238 qs_mol_dyn_low 1 2.0 0.003 0.006 38.857 38.866 qs_forces 11 3.9 0.010 0.060 38.650 38.684 qs_energies 11 4.9 0.013 0.073 36.908 36.911 scf_env_do_scf 11 5.9 0.002 0.002 31.213 31.214 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.800 27.801 velocity_verlet 10 3.0 0.001 0.002 18.564 18.570 dbcsr_multiply_generic 2286 12.5 0.095 0.099 17.809 17.903 qs_scf_new_mos 108 7.5 0.001 0.001 17.560 17.580 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.559 17.580 ot_scf_mini 108 9.5 0.003 0.004 16.755 16.775 multiply_cannon 2286 13.5 0.196 0.203 13.411 14.227 multiply_cannon_loop 2286 14.5 0.631 0.660 12.422 13.326 ot_mini 108 10.5 0.001 0.002 9.782 9.808 qs_ot_get_derivative 108 11.5 0.001 0.002 8.173 8.193 rebuild_ks_matrix 119 8.3 0.000 0.000 7.621 7.641 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 7.621 7.641 multiply_cannon_multrec 18288 15.5 1.924 2.833 6.734 7.072 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.768 6.786 dbcsr_mm_accdrv_process 38222 16.0 4.082 5.302 4.726 5.532 mp_waitall_1 158411 16.6 3.965 5.157 3.965 5.157 sum_up_and_integrate 119 10.3 0.030 0.033 5.043 5.049 integrate_v_rspace 119 11.3 0.003 0.003 5.012 5.017 qs_ot_get_p 119 10.4 0.001 0.001 4.619 4.636 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.586 4.593 calculate_rho_elec 119 8.7 0.031 0.032 4.585 4.592 init_scf_run 11 5.9 0.000 0.001 4.232 4.232 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.232 4.232 rs_pw_transfer 974 11.9 0.009 0.011 3.863 4.067 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.038 3.650 init_scf_loop 11 6.9 0.002 0.010 3.382 3.384 density_rs2pw 119 9.7 0.004 0.004 3.050 3.313 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.108 3.116 make_m2s 4572 13.5 0.045 0.046 2.665 2.803 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.790 2.790 potential_pw2rs 119 12.3 0.007 0.008 2.729 2.758 make_images 4572 14.5 0.190 0.202 2.579 2.717 calculate_first_density_matrix 1 7.0 0.002 0.004 2.670 2.672 pw_transfer 1439 11.6 0.066 0.071 2.546 2.583 multiply_cannon_metrocomm3 18288 15.5 0.043 0.045 1.529 2.533 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.453 2.487 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.099 2.443 apply_single 119 13.6 0.000 0.000 2.099 2.442 prepare_preconditioner 11 7.9 0.000 0.000 2.415 2.419 make_preconditioner 11 8.9 0.000 0.002 2.415 2.419 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.394 2.404 cp_fm_diag_elpa_base 50 14.0 2.362 2.381 2.392 2.402 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.232 2.333 fft3d_ps 1201 14.6 0.531 0.550 2.135 2.173 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.914 1.923 calculate_dm_sparse 119 9.5 0.000 0.001 1.897 1.908 grid_integrate_task_list 119 12.3 1.775 1.907 1.775 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.872 1.874 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.719 1.751 mp_sum_l 7207 12.9 1.402 1.737 1.402 1.737 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.599 1.614 ot_diis_step 108 11.5 0.012 0.014 1.575 1.576 multiply_cannon_sync_h2d 18288 15.5 1.344 1.567 1.344 1.567 wfi_extrapolate 11 7.9 0.001 0.002 1.486 1.486 make_images_data 4572 15.5 0.045 0.048 1.184 1.455 mp_alltoall_d11v 2130 13.8 1.077 1.396 1.077 1.396 grid_collocate_task_list 119 9.7 1.213 1.346 1.213 1.346 hybrid_alltoall_any 4725 16.4 0.056 0.116 1.035 1.307 jit_kernel_multiply 7 16.1 0.593 1.281 0.593 1.281 mp_alltoall_z22v 1201 16.6 1.133 1.262 1.133 1.262 acc_transpose_blocks 18288 15.5 0.076 0.079 1.231 1.252 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.199 1.204 make_images_sizes 4572 15.5 0.005 0.005 0.792 1.114 mp_alltoall_i44 4572 16.5 0.787 1.109 0.787 1.109 cp_fm_cholesky_invert 11 10.9 1.090 1.095 1.090 1.095 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.054 1.056 mp_sendrecv_dv 11067 12.7 1.005 1.041 1.005 1.041 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.999 1.018 mp_waitany 9880 13.7 0.654 0.937 0.654 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.930 0.935 rs_pw_transfer_PW2RS_50 119 14.3 0.402 0.411 0.835 0.911 mp_allgather_i34 2286 14.5 0.452 0.897 0.452 0.897 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.136 0.620 0.884 dbcsr_complete_redistribute 329 12.2 0.095 0.102 0.727 0.847 parallel_gemm_fm 81 9.0 0.000 0.000 0.839 0.842 parallel_gemm_fm_cosma 81 10.0 0.839 0.842 0.839 0.842 rs_pw_transfer_RS2PW_50 119 11.7 0.241 0.247 0.768 0.821 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.788 0.799 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.555 0.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.238000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 555.610112E+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 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.013 0.030 41.356 41.357 qs_mol_dyn_low 1 2.0 0.003 0.004 41.173 41.180 qs_forces 11 3.9 0.002 0.002 41.082 41.085 qs_energies 11 4.9 0.001 0.001 39.211 39.220 scf_env_do_scf 11 5.9 0.025 0.032 33.719 33.721 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 29.359 29.360 velocity_verlet 10 3.0 0.001 0.002 21.076 21.080 dbcsr_multiply_generic 2286 12.5 0.098 0.101 19.746 19.854 qs_scf_new_mos 108 7.5 0.001 0.001 18.979 19.041 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.979 19.040 ot_scf_mini 108 9.5 0.002 0.003 17.966 18.023 multiply_cannon 2286 13.5 0.227 0.265 14.883 15.351 multiply_cannon_loop 2286 14.5 0.934 0.967 13.705 14.136 ot_mini 108 10.5 0.001 0.001 10.548 10.618 multiply_cannon_multrec 27432 15.5 2.340 3.020 8.605 8.964 qs_ot_get_derivative 108 11.5 0.001 0.001 8.661 8.716 rebuild_ks_matrix 119 8.3 0.000 0.000 7.737 7.790 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.737 7.789 dbcsr_mm_accdrv_process 47916 15.9 5.234 7.214 6.171 7.367 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.878 6.923 sum_up_and_integrate 119 10.3 0.035 0.038 4.769 4.776 integrate_v_rspace 119 11.3 0.003 0.003 4.734 4.740 qs_ot_get_p 119 10.4 0.001 0.001 4.612 4.688 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.647 4.676 calculate_rho_elec 119 8.7 0.040 0.046 4.646 4.676 init_scf_loop 11 6.9 0.000 0.000 4.294 4.320 init_scf_run 11 5.9 0.000 0.001 4.048 4.048 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.048 4.048 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.165 3.714 mp_waitall_1 137007 16.6 3.112 3.653 3.112 3.653 rs_pw_transfer 974 11.9 0.009 0.009 3.353 3.581 prepare_preconditioner 11 7.9 0.000 0.000 3.281 3.289 make_preconditioner 11 8.9 0.000 0.000 3.281 3.289 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.854 3.199 make_m2s 4572 13.5 0.055 0.056 3.071 3.173 density_rs2pw 119 9.7 0.004 0.004 2.977 3.160 make_images 4572 14.5 0.270 0.330 2.963 3.071 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.975 2.986 pw_transfer 1439 11.6 0.066 0.072 2.655 2.704 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.562 2.615 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.609 2.610 potential_pw2rs 119 12.3 0.008 0.010 2.429 2.450 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.244 2.391 apply_single 119 13.6 0.000 0.000 2.244 2.391 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.306 2.334 calculate_first_density_matrix 1 7.0 0.008 0.021 2.328 2.330 fft3d_ps 1201 14.6 0.569 0.615 2.238 2.284 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.216 2.225 cp_fm_diag_elpa_base 50 14.0 2.172 2.189 2.212 2.222 calculate_dm_sparse 119 9.5 0.000 0.000 2.126 2.180 fft_wrap_pw1pw2_140 487 13.2 0.086 0.095 1.954 2.011 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.919 1.920 grid_integrate_task_list 119 12.3 1.825 1.907 1.825 1.907 ot_diis_step 108 11.5 0.012 0.012 1.844 1.845 jit_kernel_multiply 10 15.8 0.876 1.818 0.876 1.818 mp_sum_l 7207 12.9 1.371 1.697 1.371 1.697 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.677 1.693 wfi_extrapolate 11 7.9 0.001 0.001 1.658 1.658 make_images_data 4572 15.5 0.045 0.048 1.302 1.513 acc_transpose_blocks 27432 15.5 0.113 0.115 1.470 1.504 multiply_cannon_metrocomm3 27432 15.5 0.038 0.038 0.853 1.394 hybrid_alltoall_any 4725 16.4 0.062 0.152 1.080 1.347 grid_collocate_task_list 119 9.7 1.222 1.339 1.222 1.339 dbcsr_complete_redistribute 329 12.2 0.118 0.146 1.038 1.325 mp_alltoall_d11v 2130 13.8 1.143 1.307 1.143 1.307 mp_alltoall_z22v 1201 16.6 1.254 1.305 1.254 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.262 cp_fm_upper_to_full 72 14.2 0.863 1.229 0.863 1.229 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.204 1.218 cp_fm_cholesky_invert 11 10.9 1.161 1.165 1.161 1.165 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.105 1.108 make_images_sizes 4572 15.5 0.005 0.005 0.748 1.034 multiply_cannon_sync_h2d 27432 15.5 0.970 1.030 0.970 1.030 mp_alltoall_i44 4572 16.5 0.743 1.029 0.743 1.029 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.697 0.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.952 0.957 mp_sendrecv_dv 8211 12.7 0.889 0.915 0.889 0.915 yz_to_x 606 15.1 0.189 0.223 0.874 0.895 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.843 0.872 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.866 multiply_cannon_metrocomm1 27432 15.5 0.034 0.036 0.359 0.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.357000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.000000, yerr=3.074824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 604.942336E+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 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 35.449 35.450 qs_mol_dyn_low 1 2.0 0.003 0.003 35.258 35.265 qs_forces 11 3.9 0.002 0.003 34.888 34.890 qs_energies 11 4.9 0.001 0.001 33.023 33.027 scf_env_do_scf 11 5.9 0.023 0.024 27.861 27.861 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.428 24.429 velocity_verlet 10 3.0 0.001 0.001 18.249 18.271 dbcsr_multiply_generic 2286 12.5 0.092 0.095 14.487 14.645 qs_scf_new_mos 108 7.5 0.001 0.001 14.572 14.614 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.572 14.613 ot_scf_mini 108 9.5 0.002 0.002 13.744 13.776 multiply_cannon 2286 13.5 0.233 0.238 10.672 11.328 multiply_cannon_loop 2286 14.5 0.329 0.343 9.365 9.754 ot_mini 108 10.5 0.001 0.001 7.323 7.372 rebuild_ks_matrix 119 8.3 0.000 0.000 7.192 7.240 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.192 7.239 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.436 6.479 multiply_cannon_multrec 9144 15.5 1.651 1.887 5.818 6.059 qs_ot_get_derivative 108 11.5 0.001 0.001 5.849 5.889 sum_up_and_integrate 119 10.3 0.038 0.042 4.602 4.607 integrate_v_rspace 119 11.3 0.003 0.003 4.564 4.570 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.465 4.480 calculate_rho_elec 119 8.7 0.060 0.061 4.464 4.480 qs_ot_get_p 119 10.4 0.001 0.001 4.276 4.333 dbcsr_mm_accdrv_process 12550 15.8 3.341 4.081 4.065 4.140 mp_waitall_1 115863 16.7 3.128 4.136 3.128 4.136 init_scf_run 11 5.9 0.000 0.001 3.613 3.613 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.613 3.613 init_scf_loop 11 6.9 0.000 0.000 3.378 3.383 make_m2s 4572 13.5 0.034 0.035 2.652 3.077 make_images 4572 14.5 0.267 0.301 2.562 2.987 rs_pw_transfer 974 11.9 0.008 0.008 2.749 2.883 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.877 2.879 density_rs2pw 119 9.7 0.004 0.004 2.672 2.818 pw_transfer 1439 11.6 0.066 0.069 2.686 2.698 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.607 2.608 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.592 2.606 prepare_preconditioner 11 7.9 0.000 0.000 2.496 2.502 make_preconditioner 11 8.9 0.000 0.000 2.496 2.502 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.346 2.394 fft3d_ps 1201 14.6 0.572 0.582 2.253 2.266 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.176 2.182 cp_fm_diag_elpa_base 50 14.0 2.141 2.160 2.172 2.179 potential_pw2rs 119 12.3 0.010 0.011 2.165 2.175 calculate_first_density_matrix 1 7.0 0.003 0.004 2.098 2.098 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 2.001 2.017 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.976 2.000 grid_integrate_task_list 119 12.3 1.926 1.979 1.926 1.979 calculate_dm_sparse 119 9.5 0.000 0.000 1.881 1.909 make_images_data 4572 15.5 0.039 0.042 1.331 1.862 hybrid_alltoall_any 4725 16.4 0.063 0.175 1.208 1.783 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.653 1.655 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.595 1.615 jit_kernel_multiply 9 16.0 0.686 1.584 0.686 1.584 grid_collocate_task_list 119 9.7 1.328 1.453 1.328 1.453 ot_diis_step 108 11.5 0.012 0.013 1.439 1.439 wfi_extrapolate 11 7.9 0.001 0.001 1.437 1.439 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.413 1.423 cp_fm_cholesky_invert 11 10.9 1.336 1.341 1.336 1.341 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.302 1.334 apply_single 119 13.6 0.000 0.000 1.302 1.334 mp_allgather_i34 2286 14.5 0.634 1.319 0.634 1.319 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.249 1.251 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.462 1.220 mp_alltoall_d11v 2130 13.8 1.115 1.186 1.115 1.186 mp_alltoall_z22v 1201 16.6 1.131 1.185 1.131 1.185 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.174 1.179 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.873 1.089 make_images_sizes 4572 15.5 0.005 0.005 0.659 1.018 mp_alltoall_i44 4572 16.5 0.654 1.013 0.654 1.013 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.962 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.923 acc_transpose_blocks 9144 15.5 0.038 0.039 0.903 0.922 mp_sum_l 7207 12.9 0.743 0.922 0.743 0.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.884 0.889 yz_to_x 606 15.1 0.261 0.272 0.844 0.860 dbcsr_complete_redistribute 329 12.2 0.163 0.170 0.793 0.842 x_to_yz 595 16.2 0.277 0.288 0.825 0.832 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.702 0.755 multiply_cannon_sync_h2d 9144 15.5 0.708 0.749 0.708 0.749 arnoldi_extremal 119 11.4 0.001 0.002 0.698 0.714 arnoldi_normal_ev 119 12.4 0.002 0.003 0.696 0.712 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=35.450000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.272727, yerr=4.494257 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 760.401920E+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 0.036 0.064 50.237 50.238 qs_mol_dyn_low 1 2.0 0.003 0.003 49.746 49.752 qs_forces 11 3.9 0.002 0.002 49.680 49.682 qs_energies 11 4.9 0.003 0.019 47.450 47.454 scf_env_do_scf 11 5.9 0.001 0.002 41.153 41.153 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 31.484 31.485 velocity_verlet 10 3.0 0.001 0.002 28.180 28.185 dbcsr_multiply_generic 2286 12.5 0.100 0.102 20.314 20.564 qs_scf_new_mos 108 7.5 0.001 0.001 19.700 19.796 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.699 19.795 ot_scf_mini 108 9.5 0.002 0.002 18.463 18.560 multiply_cannon 2286 13.5 0.306 0.316 14.858 16.209 multiply_cannon_loop 2286 14.5 0.344 0.352 13.296 14.360 ot_mini 108 10.5 0.001 0.001 10.610 10.739 init_scf_loop 11 6.9 0.000 0.000 9.627 9.630 multiply_cannon_multrec 9144 15.5 3.445 4.838 8.779 8.847 rebuild_ks_matrix 119 8.3 0.000 0.000 8.698 8.845 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 8.697 8.844 qs_ot_get_derivative 108 11.5 0.001 0.001 8.400 8.495 prepare_preconditioner 11 7.9 0.000 0.000 8.471 8.486 make_preconditioner 11 8.9 0.000 0.000 8.471 8.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.943 8.313 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.816 7.949 dbcsr_mm_accdrv_process 12550 15.8 4.250 5.734 5.208 6.485 mp_waitall_1 94719 16.7 4.247 5.679 4.247 5.679 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.239 5.246 calculate_rho_elec 119 8.7 0.118 0.121 5.239 5.245 sum_up_and_integrate 119 10.3 0.065 0.067 4.974 4.981 integrate_v_rspace 119 11.3 0.003 0.003 4.908 4.916 qs_ot_get_p 119 10.4 0.001 0.001 4.622 4.772 cp_fm_upper_to_full 72 14.2 3.245 4.665 3.245 4.665 init_scf_run 11 5.9 0.000 0.001 4.108 4.109 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.108 4.109 make_m2s 4572 13.5 0.038 0.038 3.443 3.754 make_images 4572 14.5 0.353 0.381 3.322 3.633 pw_transfer 1439 11.6 0.068 0.069 3.615 3.625 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.516 3.527 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.017 3.442 dbcsr_complete_redistribute 329 12.2 0.286 0.291 2.294 3.150 fft3d_ps 1201 14.6 0.600 0.607 3.139 3.148 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 2.021 3.051 density_rs2pw 119 9.7 0.004 0.004 3.013 3.045 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.923 2.928 fft_wrap_pw1pw2_140 487 13.2 0.095 0.098 2.846 2.859 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.852 2.713 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.351 2.594 apply_single 119 13.6 0.000 0.000 2.351 2.594 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.578 2.579 calculate_dm_sparse 119 9.5 0.000 0.000 2.515 2.536 cp_fm_cholesky_invert 11 10.9 2.526 2.532 2.526 2.532 mp_alltoall_i22 627 13.8 1.503 2.392 1.503 2.392 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.523 2.369 make_images_data 4572 15.5 0.043 0.045 1.863 2.369 rs_pw_transfer 974 11.9 0.009 0.010 2.325 2.364 calculate_first_density_matrix 1 7.0 0.002 0.002 2.300 2.301 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.813 2.277 mp_sum_l 7207 12.9 1.479 2.271 1.479 2.271 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.172 2.222 grid_integrate_task_list 119 12.3 2.149 2.177 2.149 2.177 ot_diis_step 108 11.5 0.014 0.014 2.157 2.157 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.151 2.151 cp_fm_diag_elpa_base 50 14.0 1.989 2.046 2.148 2.148 potential_pw2rs 119 12.3 0.014 0.014 2.131 2.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.942 1.943 qs_energies_init_hamiltonians 11 5.9 0.010 0.015 1.830 1.831 jit_kernel_multiply 9 15.5 0.930 1.756 0.930 1.756 wfi_extrapolate 11 7.9 0.001 0.001 1.729 1.729 mp_alltoall_d11v 2130 13.8 1.581 1.654 1.581 1.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.600 1.640 mp_alltoall_z22v 1201 16.6 1.590 1.623 1.590 1.623 grid_collocate_task_list 119 9.7 1.554 1.570 1.554 1.570 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.442 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.454 1.463 yz_to_x 606 15.1 0.458 0.468 1.349 1.357 mp_allgather_i34 2286 14.5 0.540 1.248 0.540 1.248 x_to_yz 595 16.2 0.480 0.491 1.178 1.183 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.150 1.169 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.099 1.113 multiply_cannon_sync_h2d 9144 15.5 1.048 1.051 1.048 1.051 make_images_sizes 4572 15.5 0.005 0.005 0.633 1.042 mp_alltoall_i44 4572 16.5 0.629 1.037 0.629 1.037 cp_fm_cholesky_decompose 22 10.9 1.023 1.031 1.023 1.031 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.988 1.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=50.238000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=709.636364, yerr=17.009477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 505.180160E+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 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.052 0.159 92.237 92.245 qs_mol_dyn_low 1 2.0 0.003 0.005 91.566 91.583 qs_forces 11 3.9 0.003 0.004 91.494 91.495 qs_energies 11 4.9 0.004 0.015 88.446 88.461 scf_env_do_scf 11 5.9 0.001 0.003 78.675 78.677 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 71.546 71.546 dbcsr_multiply_generic 2055 12.4 0.106 0.107 52.515 52.840 qs_scf_new_mos 99 7.5 0.000 0.001 51.506 51.602 qs_scf_loop_do_ot 99 8.5 0.001 0.001 51.505 51.601 ot_scf_mini 99 9.5 0.002 0.003 49.014 49.087 velocity_verlet 10 3.0 0.001 0.002 47.180 47.181 multiply_cannon 2055 13.4 0.184 0.188 42.993 43.888 multiply_cannon_loop 2055 14.4 1.494 1.523 41.663 42.484 ot_mini 99 10.5 0.001 0.001 27.787 27.860 qs_ot_get_derivative 99 11.5 0.001 0.001 21.042 21.123 multiply_cannon_multrec 49320 15.4 12.350 13.061 17.432 18.206 rebuild_ks_matrix 110 8.3 0.000 0.000 16.187 16.297 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 16.186 16.297 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.256 14.348 mp_waitall_1 220248 16.4 12.104 13.082 12.104 13.082 qs_ot_get_p 110 10.4 0.001 0.001 12.587 12.667 multiply_cannon_sync_h2d 49320 15.4 10.163 10.662 10.163 10.662 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.927 8.964 sum_up_and_integrate 110 10.3 0.036 0.043 8.365 8.378 integrate_v_rspace 110 11.3 0.003 0.004 8.328 8.351 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.114 8.240 calculate_rho_elec 110 8.6 0.021 0.026 8.113 8.240 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.624 8.171 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.950 7.951 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.216 7.715 apply_single 110 13.6 0.000 0.001 7.216 7.715 multiply_cannon_metrocomm3 49320 15.4 0.078 0.080 6.396 7.591 init_scf_run 11 5.9 0.000 0.001 7.459 7.459 scf_env_initial_rho_setup 11 6.9 0.000 0.002 7.459 7.459 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.116 7.136 cp_fm_diag_elpa_base 48 14.0 7.092 7.117 7.113 7.134 init_scf_loop 11 6.9 0.001 0.004 7.084 7.085 ot_diis_step 99 11.5 0.006 0.006 6.584 6.584 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.887 5.926 rs_pw_transfer 902 11.9 0.012 0.013 5.067 5.735 density_rs2pw 110 9.6 0.004 0.005 4.890 5.529 dbcsr_mm_accdrv_process 87628 16.1 1.907 2.007 4.957 5.377 make_m2s 4110 13.4 0.061 0.065 4.953 5.075 mp_sum_l 6514 12.8 4.205 5.003 4.205 5.003 make_images 4110 14.4 0.178 0.192 4.856 4.980 pw_transfer 1331 11.6 0.055 0.064 4.758 4.863 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.670 4.777 prepare_preconditioner 11 7.9 0.000 0.001 4.706 4.727 make_preconditioner 11 8.9 0.000 0.001 4.706 4.727 make_full_inverse_cholesky 11 9.9 0.000 0.001 4.485 4.542 wfi_extrapolate 11 7.9 0.001 0.001 4.532 4.532 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.913 4.407 fft3d_ps 1111 14.6 0.777 0.879 4.202 4.304 calculate_dm_sparse 110 9.5 0.000 0.001 3.879 3.964 fft_wrap_pw1pw2_140 451 13.1 0.167 0.187 3.759 3.872 potential_pw2rs 110 12.3 0.006 0.007 3.790 3.842 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.758 3.763 grid_integrate_task_list 110 12.3 3.236 3.430 3.236 3.430 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.308 3.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.318 3.358 jit_kernel_multiply 13 15.9 2.774 3.021 2.774 3.021 mp_alltoall_z22v 1111 16.6 2.600 2.906 2.600 2.906 calculate_first_density_matrix 1 7.0 0.001 0.003 2.818 2.826 mp_alltoall_d11v 2046 13.8 2.456 2.760 2.456 2.760 cp_fm_cholesky_invert 11 10.9 2.579 2.583 2.579 2.583 mp_waitany 14300 13.8 1.933 2.550 1.933 2.550 make_images_data 4110 15.4 0.042 0.045 2.078 2.327 grid_collocate_task_list 110 9.6 2.083 2.327 2.083 2.327 acc_transpose_blocks 49320 15.4 0.212 0.220 2.168 2.240 hybrid_alltoall_any 4261 16.3 0.082 0.474 1.779 2.137 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.078 2.093 mp_sum_d 3889 11.9 1.411 1.927 1.411 1.927 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=92.245000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.909091, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 587.018240E+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 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2462973. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.025 0.089 70.271 70.272 qs_mol_dyn_low 1 2.0 0.008 0.043 69.874 69.885 qs_forces 11 3.9 0.009 0.047 69.762 69.776 qs_energies 11 4.9 0.001 0.003 66.360 66.365 scf_env_do_scf 11 5.9 0.001 0.001 57.679 57.682 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.957 49.957 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.603 37.769 velocity_verlet 10 3.0 0.001 0.001 36.678 36.679 qs_scf_new_mos 99 7.5 0.001 0.001 33.125 33.264 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.124 33.264 multiply_cannon 2055 13.4 0.221 0.239 31.000 32.297 ot_scf_mini 99 9.5 0.003 0.004 31.462 31.586 multiply_cannon_loop 2055 14.4 0.921 0.939 29.683 30.644 ot_mini 99 10.5 0.001 0.002 18.160 18.288 multiply_cannon_multrec 24660 15.4 7.658 9.587 13.898 15.830 rebuild_ks_matrix 110 8.3 0.000 0.000 14.110 14.216 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.109 14.215 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.423 12.521 qs_ot_get_derivative 99 11.5 0.001 0.002 12.349 12.478 mp_waitall_1 176588 16.5 7.633 10.205 7.633 10.205 multiply_cannon_sync_h2d 24660 15.4 7.034 7.923 7.034 7.923 multiply_cannon_metrocomm3 24660 15.4 0.066 0.069 5.088 7.896 init_scf_loop 11 6.9 0.000 0.000 7.683 7.684 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.495 7.263 apply_single 110 13.6 0.000 0.001 6.495 7.262 sum_up_and_integrate 110 10.3 0.053 0.059 6.840 6.854 qs_ot_get_p 110 10.4 0.001 0.001 6.670 6.842 integrate_v_rspace 110 11.3 0.003 0.003 6.787 6.801 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.470 6.484 calculate_rho_elec 110 8.6 0.040 0.048 6.469 6.483 dbcsr_mm_accdrv_process 52282 16.1 4.609 5.448 6.079 6.379 init_scf_run 11 5.9 0.000 0.001 6.301 6.301 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.300 6.301 ot_diis_step 99 11.5 0.012 0.028 5.764 5.765 prepare_preconditioner 11 7.9 0.000 0.000 5.587 5.603 make_preconditioner 11 8.9 0.000 0.000 5.587 5.603 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.650 5.364 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.144 5.303 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.754 4.773 make_m2s 4110 13.4 0.057 0.059 4.222 4.711 make_images 4110 14.4 0.397 0.442 4.114 4.600 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.324 4.324 pw_transfer 1331 11.6 0.066 0.073 3.919 4.079 density_rs2pw 110 9.6 0.004 0.005 3.545 4.024 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.812 3.975 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.731 3.745 cp_fm_diag_elpa_base 48 14.0 3.681 3.700 3.728 3.743 rs_pw_transfer 902 11.9 0.012 0.014 3.175 3.695 wfi_extrapolate 11 7.9 0.001 0.001 3.587 3.588 fft_wrap_pw1pw2_140 451 13.1 0.202 0.220 3.214 3.378 grid_integrate_task_list 110 12.3 3.154 3.335 3.154 3.335 fft3d_ps 1111 14.6 1.106 1.324 3.162 3.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.310 3.311 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.175 3.237 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.964 3.004 calculate_dm_sparse 110 9.5 0.001 0.001 2.965 2.991 make_images_data 4110 15.4 0.046 0.049 2.360 2.855 hybrid_alltoall_any 4261 16.3 0.103 0.442 2.075 2.794 calculate_first_density_matrix 1 7.0 0.002 0.002 2.621 2.624 cp_fm_cholesky_invert 11 10.9 2.609 2.616 2.609 2.616 potential_pw2rs 110 12.3 0.008 0.008 2.586 2.605 jit_kernel_multiply 13 16.3 1.115 2.520 1.115 2.520 grid_collocate_task_list 110 9.6 2.056 2.467 2.056 2.467 mp_sum_l 6514 12.8 1.809 2.363 1.809 2.363 mp_alltoall_d11v 2046 13.8 1.879 2.328 1.879 2.328 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.943 1.967 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.869 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.851 1.863 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.787 1.751 mp_waitany 10164 13.8 1.219 1.722 1.219 1.722 mp_allgather_i34 2055 14.4 0.649 1.666 0.649 1.666 mp_irecv_dv 57340 16.2 0.662 1.633 0.662 1.633 acc_transpose_blocks 24660 15.4 0.112 0.114 1.526 1.552 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.537 1.550 mp_alltoall_z22v 1111 16.6 1.372 1.500 1.372 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.472 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.217 0.959 1.446 dbcsr_complete_redistribute 325 12.2 0.243 0.294 1.168 1.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.272000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.818182, yerr=6.191169 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 662.036480E+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 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 66.103 66.104 qs_mol_dyn_low 1 2.0 0.003 0.003 65.845 65.855 qs_forces 11 3.9 0.003 0.003 65.769 65.770 qs_energies 11 4.9 0.001 0.001 62.440 62.443 scf_env_do_scf 11 5.9 0.001 0.001 54.194 54.194 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.403 44.404 velocity_verlet 10 3.0 0.001 0.002 35.568 35.570 dbcsr_multiply_generic 2055 12.4 0.106 0.111 29.480 29.715 qs_scf_new_mos 99 7.5 0.001 0.001 27.842 27.941 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.841 27.940 ot_scf_mini 99 9.5 0.002 0.002 26.569 26.686 multiply_cannon 2055 13.4 0.211 0.225 22.329 23.680 multiply_cannon_loop 2055 14.4 0.613 0.628 20.967 21.981 ot_mini 99 10.5 0.001 0.001 14.406 14.526 rebuild_ks_matrix 110 8.3 0.000 0.000 13.317 13.435 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.316 13.434 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.731 11.836 multiply_cannon_multrec 16440 15.4 3.963 5.276 9.690 11.008 mp_waitall_1 139946 16.5 7.761 10.289 7.761 10.289 qs_ot_get_derivative 99 11.5 0.001 0.001 9.953 10.073 init_scf_loop 11 6.9 0.000 0.000 9.746 9.747 prepare_preconditioner 11 7.9 0.000 0.000 7.894 7.909 make_preconditioner 11 8.9 0.000 0.000 7.894 7.909 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.174 7.565 qs_ot_get_p 110 10.4 0.001 0.001 7.237 7.371 sum_up_and_integrate 110 10.3 0.061 0.062 7.318 7.332 integrate_v_rspace 110 11.3 0.003 0.003 7.257 7.270 multiply_cannon_metrocomm3 16440 15.4 0.041 0.043 4.191 6.857 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.795 6.811 calculate_rho_elec 110 8.6 0.059 0.059 6.794 6.810 init_scf_run 11 5.9 0.000 0.001 5.776 5.776 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.776 5.776 dbcsr_mm_accdrv_process 34862 16.1 4.688 5.173 5.580 5.702 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.989 5.412 apply_single 110 13.6 0.000 0.000 4.989 5.412 make_m2s 4110 13.4 0.050 0.051 4.891 5.344 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.262 5.269 make_images 4110 14.4 0.390 0.511 4.776 5.230 density_rs2pw 110 9.6 0.004 0.005 3.839 5.082 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.862 4.863 rs_pw_transfer 902 11.9 0.010 0.012 3.645 4.825 ot_diis_step 99 11.5 0.011 0.011 4.420 4.421 pw_transfer 1331 11.6 0.066 0.072 4.150 4.189 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.148 4.156 cp_fm_diag_elpa_base 48 14.0 4.066 4.109 4.146 4.154 multiply_cannon_sync_h2d 16440 15.4 3.685 4.096 3.685 4.096 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.043 4.084 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.297 3.951 grid_integrate_task_list 110 12.3 3.174 3.435 3.174 3.435 fft_wrap_pw1pw2_140 451 13.1 0.211 0.216 3.345 3.387 cp_fm_cholesky_invert 11 10.9 3.379 3.386 3.379 3.386 fft3d_ps 1111 14.6 1.091 1.104 3.345 3.385 make_images_data 4110 15.4 0.043 0.048 2.713 3.341 wfi_extrapolate 11 7.9 0.001 0.001 3.314 3.314 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.373 3.250 potential_pw2rs 110 12.3 0.011 0.011 2.947 2.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.827 2.829 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.726 2.790 mp_waitany 17072 13.8 1.316 2.614 1.316 2.614 calculate_dm_sparse 110 9.5 0.001 0.001 2.568 2.601 mp_alltoall_d11v 2046 13.8 2.097 2.537 2.097 2.537 grid_collocate_task_list 110 9.6 2.074 2.476 2.074 2.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.472 calculate_first_density_matrix 1 7.0 0.001 0.002 2.366 2.367 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.862 2.351 mp_sum_l 6514 12.8 1.647 2.240 1.647 2.240 mp_irecv_dv 48980 15.7 0.793 2.229 0.793 2.229 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.182 0.922 2.150 dbcsr_complete_redistribute 325 12.2 0.317 0.339 1.657 2.145 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.977 1.978 cp_fm_cholesky_decompose 22 10.9 1.955 1.975 1.955 1.975 cp_fm_upper_to_full 70 14.2 1.406 1.837 1.406 1.837 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.799 mp_allgather_i34 2055 14.4 0.689 1.746 0.689 1.746 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.090 1.559 rs_gather_matrices 110 12.3 0.234 0.264 1.051 1.543 mp_alltoall_z22v 1111 16.6 1.488 1.534 1.488 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.489 1.503 jit_kernel_multiply 8 16.5 0.501 1.467 0.501 1.467 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.354 1.467 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.332 1.342 make_images_sizes 4110 15.4 0.005 0.005 0.825 1.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=66.104000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=9.232533 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 733.388800E+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 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.040 0.088 72.071 72.074 qs_mol_dyn_low 1 2.0 0.003 0.005 71.623 71.633 qs_forces 11 3.9 0.003 0.004 71.555 71.556 qs_energies 11 4.9 0.003 0.009 68.029 68.033 scf_env_do_scf 11 5.9 0.001 0.002 59.277 59.280 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.986 45.988 velocity_verlet 10 3.0 0.001 0.001 40.364 40.366 dbcsr_multiply_generic 2055 12.4 0.114 0.119 30.633 30.822 qs_scf_new_mos 99 7.5 0.001 0.001 29.413 29.531 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.413 29.530 ot_scf_mini 99 9.5 0.003 0.004 27.726 27.827 multiply_cannon 2055 13.4 0.243 0.260 22.569 23.793 multiply_cannon_loop 2055 14.4 0.879 0.905 21.051 21.637 ot_mini 99 10.5 0.001 0.001 15.001 15.107 multiply_cannon_multrec 24660 15.4 4.225 6.882 12.638 13.898 init_scf_loop 11 6.9 0.001 0.003 13.240 13.242 rebuild_ks_matrix 110 8.3 0.000 0.000 13.081 13.179 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.080 13.178 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.568 11.657 prepare_preconditioner 11 7.9 0.000 0.000 11.329 11.343 make_preconditioner 11 8.9 0.000 0.001 11.329 11.342 make_full_inverse_cholesky 11 9.9 0.000 0.001 9.461 10.989 qs_ot_get_derivative 99 11.5 0.001 0.001 10.772 10.867 dbcsr_mm_accdrv_process 52304 16.0 6.857 8.097 8.266 9.221 qs_ot_get_p 110 10.4 0.001 0.001 7.574 7.714 mp_waitall_1 121746 16.5 5.227 7.221 5.227 7.221 sum_up_and_integrate 110 10.3 0.068 0.071 7.195 7.207 integrate_v_rspace 110 11.3 0.003 0.003 7.127 7.137 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.102 7.112 calculate_rho_elec 110 8.6 0.078 0.081 7.101 7.112 make_m2s 4110 13.4 0.060 0.062 6.296 6.648 make_images 4110 14.4 0.575 0.696 6.154 6.501 init_scf_run 11 5.9 0.000 0.001 5.884 5.885 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.884 5.884 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.450 5.465 cp_fm_upper_to_full 70 14.2 3.440 5.020 3.440 5.020 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.913 4.914 pw_transfer 1331 11.6 0.065 0.074 4.450 4.504 density_rs2pw 110 9.6 0.004 0.005 3.925 4.438 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.344 4.402 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.204 4.213 cp_fm_diag_elpa_base 48 14.0 4.030 4.099 4.201 4.211 ot_diis_step 99 11.5 0.011 0.012 4.186 4.186 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.087 4.155 apply_single 110 13.6 0.000 0.000 4.087 4.155 dbcsr_complete_redistribute 325 12.2 0.414 0.460 2.917 4.148 rs_pw_transfer 902 11.9 0.010 0.011 3.284 3.850 make_images_data 4110 15.4 0.046 0.050 3.225 3.747 fft3d_ps 1111 14.6 1.098 1.131 3.637 3.679 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 3.567 3.624 cp_fm_cholesky_invert 11 10.9 3.559 3.567 3.559 3.567 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.665 3.559 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.500 3.548 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.291 3.505 qs_ot_get_derivative_diag 47 12.0 0.003 0.017 3.422 3.481 grid_integrate_task_list 110 12.3 3.236 3.438 3.236 3.438 multiply_cannon_sync_h2d 24660 15.4 3.163 3.309 3.163 3.309 wfi_extrapolate 11 7.9 0.001 0.001 3.305 3.305 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.437 3.169 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.859 3.057 calculate_dm_sparse 110 9.5 0.001 0.001 2.979 3.017 mp_alltoall_i22 605 13.7 1.750 2.952 1.750 2.952 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.858 2.860 mp_alltoall_d11v 2046 13.8 2.136 2.818 2.136 2.818 potential_pw2rs 110 12.3 0.013 0.013 2.756 2.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.466 grid_collocate_task_list 110 9.6 2.181 2.458 2.181 2.458 calculate_first_density_matrix 1 7.0 0.001 0.003 2.455 2.457 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.309 2.310 cp_fm_cholesky_decompose 22 10.9 2.069 2.128 2.069 2.128 jit_kernel_multiply 10 15.9 1.073 1.854 1.073 1.854 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.794 1.828 mp_alltoall_z22v 1111 16.6 1.678 1.797 1.678 1.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.760 1.771 mp_waitany 13376 13.8 1.196 1.746 1.196 1.746 mp_sum_l 6514 12.8 1.125 1.707 1.125 1.707 mp_allgather_i34 2055 14.4 0.617 1.700 0.617 1.700 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.698 rs_gather_matrices 110 12.3 0.266 0.292 1.044 1.655 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.838 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.547 1.564 acc_transpose_blocks 24660 15.4 0.107 0.109 1.495 1.513 mp_irecv_dv 62702 16.1 0.738 1.484 0.738 1.484 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.074000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.363636, yerr=9.412494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 849.752064E+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 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.027 0.059 65.598 65.605 qs_mol_dyn_low 1 2.0 0.003 0.013 65.228 65.231 qs_forces 11 3.9 0.003 0.003 64.773 64.774 qs_energies 11 4.9 0.001 0.001 60.906 60.913 scf_env_do_scf 11 5.9 0.002 0.002 51.923 51.924 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.780 41.781 velocity_verlet 10 3.0 0.001 0.001 36.442 36.480 dbcsr_multiply_generic 2055 12.4 0.105 0.109 25.121 25.347 qs_scf_new_mos 99 7.5 0.001 0.001 24.875 24.969 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.874 24.969 ot_scf_mini 99 9.5 0.002 0.002 23.498 23.541 multiply_cannon 2055 13.4 0.246 0.260 18.433 20.135 multiply_cannon_loop 2055 14.4 0.319 0.331 16.999 17.575 rebuild_ks_matrix 110 8.3 0.000 0.000 13.037 13.112 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.037 13.111 ot_mini 99 10.5 0.001 0.001 11.985 12.021 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.588 11.657 init_scf_loop 11 6.9 0.000 0.000 10.075 10.076 mp_waitall_1 103326 16.6 7.569 9.089 7.569 9.089 multiply_cannon_multrec 8220 15.4 3.241 4.549 7.681 8.697 prepare_preconditioner 11 7.9 0.000 0.000 8.277 8.284 make_preconditioner 11 8.9 0.000 0.000 8.277 8.284 qs_ot_get_derivative 99 11.5 0.001 0.001 8.056 8.094 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.794 7.903 qs_ot_get_p 110 10.4 0.001 0.001 7.306 7.344 sum_up_and_integrate 110 10.3 0.080 0.082 7.317 7.328 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.265 7.282 calculate_rho_elec 110 8.6 0.116 0.116 7.265 7.282 integrate_v_rspace 110 11.3 0.003 0.003 7.237 7.247 init_scf_run 11 5.9 0.000 0.001 5.721 5.721 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.721 5.721 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.495 5.504 make_m2s 4110 13.4 0.039 0.040 5.034 5.295 dbcsr_mm_accdrv_process 17442 15.9 2.784 3.763 4.311 5.283 make_images 4110 14.4 0.637 0.691 4.904 5.166 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.104 5.106 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.133 4.961 pw_transfer 1331 11.6 0.066 0.072 4.776 4.792 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.668 4.689 cp_fm_cholesky_invert 11 10.9 4.519 4.525 4.519 4.525 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.369 4.376 cp_fm_diag_elpa_base 48 14.0 4.286 4.335 4.366 4.373 density_rs2pw 110 9.6 0.004 0.004 3.848 4.231 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.907 3.932 fft3d_ps 1111 14.6 1.150 1.163 3.913 3.927 ot_diis_step 99 11.5 0.012 0.012 3.883 3.883 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.773 3.822 apply_single 110 13.6 0.000 0.000 3.773 3.822 grid_integrate_task_list 110 12.3 3.365 3.473 3.365 3.473 rs_pw_transfer 902 11.9 0.010 0.011 3.003 3.391 make_images_data 4110 15.4 0.038 0.043 2.856 3.340 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.721 3.258 wfi_extrapolate 11 7.9 0.001 0.001 3.158 3.158 multiply_cannon_sync_h2d 8220 15.4 2.889 2.983 2.889 2.983 calculate_dm_sparse 110 9.5 0.001 0.001 2.701 2.770 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.729 2.731 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.712 2.713 potential_pw2rs 110 12.3 0.016 0.016 2.696 2.707 grid_collocate_task_list 110 9.6 2.286 2.560 2.286 2.560 calculate_first_density_matrix 1 7.0 0.002 0.002 2.459 2.460 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.349 2.368 mp_alltoall_d11v 2046 13.8 2.176 2.339 2.176 2.339 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.303 2.325 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.303 2.324 cp_fm_cholesky_decompose 22 10.9 2.182 2.205 2.182 2.205 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.998 jit_kernel_multiply 10 15.8 1.214 1.920 1.214 1.920 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.447 1.901 dbcsr_complete_redistribute 325 12.2 0.546 0.565 1.726 1.852 mp_allgather_i34 2055 14.4 0.620 1.836 0.620 1.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.792 mp_waitany 9240 13.8 1.299 1.721 1.299 1.721 mp_alltoall_z22v 1111 16.6 1.654 1.677 1.654 1.677 qs_env_update_s_mstruct 11 6.9 0.003 0.008 1.553 1.673 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.557 1.572 make_images_sizes 4110 15.4 0.005 0.005 0.760 1.418 mp_alltoall_i44 4110 16.4 0.755 1.413 0.755 1.413 x_to_yz 550 16.1 0.572 0.581 1.396 1.404 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.353 1.372 yz_to_x 561 15.1 0.524 0.535 1.354 1.369 qs_create_task_list 11 7.9 0.001 0.001 1.238 1.340 generate_qs_task_list 11 8.9 0.375 0.442 1.237 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=65.605000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=800.000000, yerr=17.098644 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 1.354093E+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 0.030 0.061 101.179 101.188 qs_mol_dyn_low 1 2.0 0.003 0.003 100.734 100.737 qs_forces 11 3.9 0.003 0.004 100.251 100.252 qs_energies 11 4.9 0.001 0.001 95.808 95.816 scf_env_do_scf 11 5.9 0.001 0.002 84.961 84.961 velocity_verlet 10 3.0 0.001 0.002 63.778 63.833 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 52.742 52.743 dbcsr_multiply_generic 2055 12.4 0.120 0.124 32.060 32.160 init_scf_loop 11 6.9 0.000 0.000 32.130 32.132 qs_scf_new_mos 99 7.5 0.001 0.001 32.026 32.069 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.025 32.069 ot_scf_mini 99 9.5 0.002 0.002 30.082 30.118 prepare_preconditioner 11 7.9 0.000 0.000 29.859 29.870 make_preconditioner 11 8.9 0.000 0.000 29.859 29.870 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.908 29.248 multiply_cannon 2055 13.4 0.343 0.357 23.263 24.294 multiply_cannon_loop 2055 14.4 0.341 0.346 21.161 21.591 cp_fm_upper_to_full 70 14.2 12.924 18.547 12.924 18.547 ot_mini 99 10.5 0.001 0.001 15.975 16.015 rebuild_ks_matrix 110 8.3 0.001 0.001 15.763 15.826 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 15.762 15.825 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.218 14.273 mp_waitall_1 84994 16.7 10.323 11.478 10.323 11.478 dbcsr_complete_redistribute 325 12.2 1.048 1.080 7.838 11.224 qs_ot_get_derivative 99 11.5 0.001 0.001 10.993 11.029 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.654 10.039 multiply_cannon_multrec 8220 15.4 4.404 4.565 9.616 9.718 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.935 9.292 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.090 9.127 calculate_rho_elec 110 8.6 0.228 0.228 9.090 9.126 mp_alltoall_i22 605 13.7 5.550 8.906 5.550 8.906 sum_up_and_integrate 110 10.3 0.150 0.151 8.442 8.456 qs_ot_get_p 110 10.4 0.001 0.001 8.363 8.437 integrate_v_rspace 110 11.3 0.004 0.004 8.291 8.305 cp_fm_cholesky_invert 11 10.9 7.494 7.501 7.494 7.501 make_m2s 4110 13.4 0.043 0.044 6.593 7.296 make_images 4110 14.4 0.880 0.924 6.402 7.106 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 6.171 6.633 pw_transfer 1331 11.6 0.075 0.075 6.427 6.436 init_scf_run 11 5.9 0.000 0.001 6.417 6.417 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.416 6.417 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 6.310 6.319 qs_ot_p2m_diag 48 11.0 0.151 0.156 6.264 6.271 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.692 5.693 fft3d_ps 1111 14.6 1.302 1.322 5.503 5.512 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.024 5.470 apply_single 110 13.6 0.000 0.000 5.024 5.469 fft_wrap_pw1pw2_140 451 13.1 0.222 0.227 5.334 5.346 dbcsr_mm_accdrv_process 11614 15.7 3.151 3.663 5.068 5.305 density_rs2pw 110 9.6 0.004 0.004 4.904 4.954 ot_diis_step 99 11.5 0.016 0.016 4.932 4.933 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.872 4.873 cp_fm_diag_elpa_base 48 14.0 4.292 4.521 4.869 4.869 make_images_data 4110 15.4 0.041 0.044 3.770 4.826 hybrid_alltoall_any 4261 16.3 0.257 0.556 3.622 4.651 multiply_cannon_sync_h2d 8220 15.4 3.953 3.961 3.953 3.961 wfi_extrapolate 11 7.9 0.001 0.001 3.922 3.922 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.436 3.881 grid_integrate_task_list 110 12.3 3.689 3.741 3.689 3.741 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.712 3.714 calculate_dm_sparse 110 9.5 0.001 0.001 3.345 3.377 potential_pw2rs 110 12.3 0.021 0.022 3.373 3.376 rs_pw_transfer 902 11.9 0.011 0.011 3.280 3.301 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.099 3.116 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.101 3.104 cp_fm_cholesky_decompose 22 10.9 2.990 3.008 2.990 3.008 mp_alltoall_d11v 2046 13.8 2.645 2.746 2.645 2.746 grid_collocate_task_list 110 9.6 2.654 2.708 2.654 2.708 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.474 calculate_first_density_matrix 1 7.0 0.001 0.001 2.336 2.337 mp_alltoall_z22v 1111 16.6 2.317 2.334 2.317 2.334 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.231 2.286 yz_to_x 561 15.1 0.937 0.943 2.168 2.174 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.170 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.153 2.162 mp_allgather_i34 2055 14.4 0.876 2.078 0.876 2.078 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.021 2.055 x_to_yz 550 16.1 0.933 0.941 2.019 2.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=101.188000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1205.181818, yerr=58.691517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 631.230464E+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 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 228.992 228.993 qs_mol_dyn_low 1 2.0 0.003 0.004 228.593 228.608 qs_forces 11 3.9 0.005 0.006 228.501 228.501 qs_energies 11 4.9 0.002 0.002 222.649 222.661 scf_env_do_scf 11 5.9 0.001 0.001 205.232 205.236 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 180.480 180.482 velocity_verlet 10 3.0 0.001 0.001 136.769 136.770 qs_scf_new_mos 117 7.6 0.001 0.001 136.549 136.732 qs_scf_loop_do_ot 117 8.6 0.001 0.001 136.549 136.731 ot_scf_mini 117 9.6 0.003 0.003 129.922 130.144 dbcsr_multiply_generic 2507 12.6 0.179 0.186 125.648 126.431 multiply_cannon 2507 13.6 0.242 0.257 101.166 102.741 multiply_cannon_loop 2507 14.6 2.019 2.112 98.560 100.124 ot_mini 117 10.6 0.001 0.001 66.839 67.037 multiply_cannon_multrec 60168 15.6 33.417 35.759 41.494 43.377 qs_ot_get_derivative 117 11.6 0.001 0.001 42.068 42.244 qs_ot_get_p 128 10.4 0.001 0.001 41.026 41.290 rebuild_ks_matrix 128 8.3 0.001 0.001 35.823 36.117 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 35.823 36.116 qs_ot_p2m_diag 83 11.4 0.079 0.091 33.598 33.650 mp_waitall_1 267128 16.5 29.997 33.048 29.997 33.048 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.231 32.498 cp_dbcsr_syevd 83 12.4 0.004 0.005 30.978 30.981 multiply_cannon_sync_h2d 60168 15.6 27.601 29.618 27.601 29.618 cp_fm_diag_elpa 83 13.4 0.000 0.000 26.981 27.014 cp_fm_diag_elpa_base 83 14.4 26.858 26.922 26.975 27.009 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.373 25.181 apply_single 128 13.6 0.001 0.001 24.373 25.180 init_scf_loop 11 6.9 0.000 0.000 24.661 24.663 ot_diis_step 117 11.6 0.008 0.008 24.564 24.565 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.897 20.047 prepare_preconditioner 11 7.9 0.000 0.000 19.825 19.871 make_preconditioner 11 8.9 0.000 0.000 19.825 19.871 make_full_inverse_cholesky 11 9.9 0.000 0.000 18.995 19.220 multiply_cannon_metrocomm3 60168 15.6 0.114 0.120 15.420 17.571 sum_up_and_integrate 128 10.3 0.089 0.107 16.175 16.192 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.981 16.109 calculate_rho_elec 128 8.7 0.046 0.065 15.980 16.109 integrate_v_rspace 128 11.3 0.004 0.005 16.086 16.105 make_m2s 5014 13.6 0.104 0.112 15.273 15.662 make_images 5014 14.6 0.404 0.423 15.092 15.489 init_scf_run 11 5.9 0.000 0.001 13.031 13.031 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.031 13.031 density_rs2pw 128 9.7 0.006 0.008 9.342 12.606 rs_pw_transfer 1046 11.9 0.016 0.019 8.369 11.646 cp_fm_cholesky_invert 11 10.9 11.522 11.530 11.522 11.530 wfi_extrapolate 11 7.9 0.001 0.001 9.765 9.765 pw_transfer 1547 11.6 0.078 0.100 9.339 9.657 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.131 9.456 mp_sum_l 7870 13.0 7.940 9.041 7.940 9.041 calculate_dm_sparse 128 9.5 0.001 0.001 8.455 8.562 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.318 8.432 dbcsr_mm_accdrv_process 124484 16.2 3.022 3.279 7.630 8.262 make_images_data 5014 15.6 0.065 0.070 7.352 8.254 fft3d_ps 1291 14.7 2.102 2.850 7.926 8.164 multiply_cannon_metrocomm1 60168 15.6 0.091 0.096 6.260 8.113 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.956 8.024 grid_integrate_task_list 128 12.3 7.030 7.800 7.030 7.800 fft_wrap_pw1pw2_140 523 13.2 0.432 0.512 7.034 7.395 hybrid_alltoall_any 5200 16.5 0.293 2.265 6.345 7.336 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.752 6.761 potential_pw2rs 128 12.3 0.009 0.011 6.482 6.590 cp_fm_cholesky_decompose 22 10.9 6.033 6.048 6.033 6.048 mp_alltoall_d11v 2415 14.1 4.552 6.014 4.552 6.014 mp_waitany 16020 13.9 2.781 6.009 2.781 6.009 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.812 5.904 grid_collocate_task_list 128 9.7 4.561 5.775 4.561 5.775 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.296 2.142 5.438 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=228.993000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.090909, yerr=6.734640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 829.341696E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3994 58335. MP_Allreduce 10983 967. MP_Sync 87 MP_Alltoall 1969 5618844. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.057 212.576 212.585 qs_mol_dyn_low 1 2.0 0.003 0.004 212.117 212.134 qs_forces 11 3.9 0.005 0.005 211.568 211.569 qs_energies 11 4.9 0.001 0.002 204.567 204.576 scf_env_do_scf 11 5.9 0.001 0.002 187.555 187.566 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 148.946 148.949 velocity_verlet 10 3.0 0.001 0.002 133.329 133.344 qs_scf_new_mos 117 7.6 0.001 0.001 106.193 106.490 qs_scf_loop_do_ot 117 8.6 0.001 0.001 106.192 106.489 dbcsr_multiply_generic 2507 12.6 0.188 0.194 100.596 102.552 ot_scf_mini 117 9.6 0.004 0.004 101.260 101.636 multiply_cannon 2507 13.6 0.477 0.525 78.452 83.078 multiply_cannon_loop 2507 14.6 1.245 1.291 74.780 79.400 ot_mini 117 10.6 0.001 0.001 52.636 53.009 mp_waitall_1 214728 16.6 26.779 43.588 26.779 43.588 init_scf_loop 11 6.9 0.000 0.000 38.501 38.502 multiply_cannon_multrec 30084 15.6 21.803 26.399 31.444 36.443 rebuild_ks_matrix 128 8.3 0.001 0.001 34.831 35.329 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 34.830 35.328 prepare_preconditioner 11 7.9 0.000 0.000 33.961 34.021 make_preconditioner 11 8.9 0.000 0.000 33.961 34.020 multiply_cannon_metrocomm3 30084 15.6 0.090 0.098 16.665 33.516 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.602 33.184 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.395 31.841 qs_ot_get_derivative 117 11.6 0.001 0.002 30.377 30.746 qs_ot_get_p 128 10.4 0.001 0.001 29.916 30.479 qs_ot_p2m_diag 83 11.4 0.188 0.217 24.436 24.533 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.082 23.507 apply_single 128 13.6 0.001 0.001 22.081 23.507 cp_dbcsr_syevd 83 12.4 0.005 0.006 23.010 23.012 ot_diis_step 117 11.6 0.014 0.015 22.008 22.010 multiply_cannon_sync_h2d 30084 15.6 19.087 21.214 19.087 21.214 cp_fm_cholesky_invert 11 10.9 20.020 20.037 20.020 20.037 cp_fm_diag_elpa 83 13.4 0.000 0.001 19.179 19.209 cp_fm_diag_elpa_base 83 14.4 18.887 19.014 19.172 19.203 make_m2s 5014 13.6 0.089 0.093 15.278 16.715 make_images 5014 14.6 1.152 1.335 15.072 16.509 sum_up_and_integrate 128 10.3 0.117 0.134 16.044 16.073 integrate_v_rspace 128 11.3 0.004 0.004 15.927 15.959 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.576 15.636 calculate_rho_elec 128 8.7 0.088 0.106 15.575 15.635 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.230 12.506 init_scf_run 11 5.9 0.000 0.001 12.101 12.103 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.101 12.102 density_rs2pw 128 9.7 0.006 0.007 8.918 11.395 multiply_cannon_metrocomm4 27577 15.6 0.098 0.113 3.786 11.018 make_images_data 5014 15.6 0.063 0.071 9.025 10.745 mp_irecv_dv 69486 16.3 3.592 10.625 3.592 10.625 rs_pw_transfer 1046 11.9 0.014 0.016 7.730 10.269 hybrid_alltoall_any 5200 16.5 0.344 1.513 7.677 9.992 pw_transfer 1547 11.6 0.088 0.109 9.613 9.700 dbcsr_mm_accdrv_process 62242 16.2 4.560 5.408 9.113 9.678 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.385 9.471 cp_fm_cholesky_decompose 22 10.9 9.001 9.106 9.001 9.106 wfi_extrapolate 11 7.9 0.001 0.001 8.931 8.932 fft_wrap_pw1pw2_140 523 13.2 0.471 0.531 7.820 7.925 fft3d_ps 1291 14.7 2.752 2.926 7.747 7.791 grid_integrate_task_list 128 12.3 7.168 7.711 7.168 7.711 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.622 7.333 mp_sum_l 7870 13.0 5.108 6.959 5.108 6.959 calculate_dm_sparse 128 9.5 0.001 0.001 6.603 6.762 potential_pw2rs 128 12.3 0.015 0.018 6.290 6.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.292 6.300 grid_collocate_task_list 128 9.7 4.700 5.881 4.700 5.881 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.690 5.778 mp_alltoall_d11v 2415 14.1 4.387 5.480 4.387 5.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.320 5.445 mp_allgather_i34 2507 14.6 2.252 5.356 2.252 5.356 mp_waitany 11748 13.9 2.690 5.283 2.690 5.283 rs_pw_transfer_RS2PW_140 139 11.5 0.355 0.379 2.129 4.580 dbcsr_complete_redistribute 395 12.7 0.786 0.877 3.516 4.440 mp_sum_d 4466 12.1 3.015 4.368 3.015 4.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=212.585000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.090909, yerr=1.975051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 939.552768E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11002 1006. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 178.199 178.200 qs_mol_dyn_low 1 2.0 0.003 0.004 177.771 177.785 qs_forces 11 3.9 0.004 0.005 177.678 177.685 qs_energies 11 4.9 0.001 0.002 171.012 171.024 scf_env_do_scf 11 5.9 0.001 0.001 155.129 155.130 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.237 120.239 velocity_verlet 10 3.0 0.001 0.002 114.184 114.186 dbcsr_multiply_generic 2507 12.6 0.178 0.184 82.166 83.445 qs_scf_new_mos 117 7.6 0.001 0.001 81.030 81.352 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.029 81.351 ot_scf_mini 117 9.6 0.003 0.004 76.883 77.281 multiply_cannon 2507 13.6 0.502 0.524 62.029 66.752 multiply_cannon_loop 2507 14.6 0.850 0.877 58.906 61.943 ot_mini 117 10.6 0.001 0.001 42.787 43.183 mp_waitall_1 170520 16.6 25.510 35.107 25.510 35.107 init_scf_loop 11 6.9 0.000 0.000 34.790 34.792 rebuild_ks_matrix 128 8.3 0.001 0.001 31.400 31.906 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.400 31.905 prepare_preconditioner 11 7.9 0.000 0.000 30.689 30.732 make_preconditioner 11 8.9 0.000 0.000 30.689 30.732 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.325 29.685 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.302 28.757 multiply_cannon_multrec 20056 15.6 13.498 16.368 22.284 25.168 multiply_cannon_metrocomm3 20056 15.6 0.057 0.062 15.229 24.837 qs_ot_get_derivative 117 11.6 0.001 0.002 23.035 23.446 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.872 20.899 apply_single 128 13.6 0.001 0.001 19.872 20.899 qs_ot_get_p 128 10.4 0.001 0.001 19.404 19.891 ot_diis_step 117 11.6 0.018 0.019 19.653 19.653 multiply_cannon_sync_h2d 20056 15.6 14.287 16.092 14.287 16.092 make_m2s 5014 13.6 0.081 0.087 15.067 16.073 make_images 5014 14.6 1.183 1.292 14.835 15.839 sum_up_and_integrate 128 10.3 0.134 0.146 15.009 15.035 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.001 15.009 integrate_v_rspace 128 11.3 0.004 0.004 14.874 14.904 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.751 14.783 calculate_rho_elec 128 8.7 0.132 0.148 14.750 14.782 cp_fm_cholesky_invert 11 10.9 14.296 14.304 14.296 14.304 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.005 14.006 make_images_data 5014 15.6 0.059 0.068 9.401 10.882 init_scf_run 11 5.9 0.000 0.001 10.851 10.852 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.851 10.851 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.798 10.815 cp_fm_diag_elpa_base 83 14.4 10.385 10.552 10.794 10.811 hybrid_alltoall_any 5200 16.5 0.435 1.982 8.145 9.856 density_rs2pw 128 9.7 0.006 0.006 7.572 9.760 pw_transfer 1547 11.6 0.087 0.110 9.493 9.612 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.267 9.391 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.025 9.323 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.443 9.265 mp_irecv_dv 50230 16.2 3.320 9.018 3.320 9.018 dbcsr_mm_accdrv_process 41502 16.2 4.474 5.357 8.236 8.417 fft_wrap_pw1pw2_140 523 13.2 0.477 0.527 8.168 8.300 grid_integrate_task_list 128 12.3 7.237 7.883 7.237 7.883 fft3d_ps 1291 14.7 2.669 2.891 7.532 7.614 rs_pw_transfer 1046 11.9 0.014 0.014 5.340 7.545 wfi_extrapolate 11 7.9 0.001 0.001 7.506 7.506 cp_fm_upper_to_full 105 14.8 5.764 7.256 5.764 7.256 cp_fm_cholesky_decompose 22 10.9 7.207 7.248 7.207 7.248 dbcsr_complete_redistribute 395 12.7 1.168 1.192 4.654 6.432 calculate_dm_sparse 128 9.5 0.001 0.001 5.805 5.921 grid_collocate_task_list 128 9.7 4.893 5.801 4.893 5.801 mp_alltoall_d11v 2415 14.1 4.606 5.636 4.606 5.636 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.547 5.554 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.653 5.347 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.444 5.219 potential_pw2rs 128 12.3 0.021 0.023 5.026 5.048 mp_sum_l 7870 13.0 3.250 4.815 3.250 4.815 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.655 4.792 mp_allgather_i34 2507 14.6 1.651 4.688 1.651 4.688 mp_waitany 11748 13.9 2.334 4.589 2.334 4.589 rs_pw_transfer_RS2PW_140 139 11.5 0.333 0.356 1.920 4.123 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.344 4.053 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.979 4.013 mp_alltoall_i22 716 14.1 1.948 3.850 1.948 3.850 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 3.774 3.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.611 3.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.200000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.909091, yerr=8.106726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.141879E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58096. MP_Allreduce 11022 1090. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.069 205.774 205.775 qs_mol_dyn_low 1 2.0 0.003 0.004 205.318 205.331 qs_forces 11 3.9 0.004 0.005 205.175 205.183 qs_energies 11 4.9 0.002 0.008 197.871 197.879 scf_env_do_scf 11 5.9 0.012 0.017 180.587 180.594 velocity_verlet 10 3.0 0.002 0.002 135.184 135.189 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 130.195 130.196 qs_scf_new_mos 117 7.6 0.001 0.001 89.961 90.260 qs_scf_loop_do_ot 117 8.6 0.001 0.001 89.960 90.259 ot_scf_mini 117 9.6 0.003 0.004 85.230 85.508 dbcsr_multiply_generic 2507 12.6 0.186 0.193 82.965 83.680 multiply_cannon 2507 13.6 0.556 0.588 55.467 58.212 multiply_cannon_loop 2507 14.6 1.172 1.202 51.163 52.931 init_scf_loop 11 6.9 0.001 0.002 50.248 50.260 prepare_preconditioner 11 7.9 0.000 0.000 45.996 46.024 make_preconditioner 11 8.9 0.000 0.002 45.996 46.024 ot_mini 117 10.6 0.001 0.001 44.675 44.947 make_full_inverse_cholesky 11 9.9 0.000 0.002 39.531 44.650 rebuild_ks_matrix 128 8.3 0.001 0.001 31.299 31.570 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 31.299 31.570 multiply_cannon_multrec 30084 15.6 14.298 19.384 26.239 30.728 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.252 28.503 mp_waitall_1 147882 16.7 19.094 28.162 19.094 28.162 qs_ot_get_p 128 10.4 0.001 0.001 25.961 26.234 qs_ot_get_derivative 117 11.6 0.001 0.002 24.855 25.131 make_m2s 5014 13.6 0.096 0.099 22.941 23.936 make_images 5014 14.6 1.953 2.600 22.633 23.628 qs_ot_p2m_diag 83 11.4 0.343 0.390 21.209 21.268 cp_dbcsr_syevd 83 12.4 0.005 0.007 19.805 19.807 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.208 19.687 apply_single 128 13.6 0.001 0.001 19.208 19.687 ot_diis_step 117 11.6 0.017 0.018 19.677 19.680 cp_fm_cholesky_invert 11 10.9 18.367 18.376 18.367 18.376 cp_fm_upper_to_full 105 14.8 11.677 17.244 11.677 17.244 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.274 16.288 cp_fm_diag_elpa_base 83 14.4 15.189 15.520 16.266 16.279 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.208 16.245 calculate_rho_elec 128 8.7 0.175 0.190 16.208 16.244 sum_up_and_integrate 128 10.3 0.140 0.152 15.776 15.798 multiply_cannon_metrocomm3 30084 15.6 0.046 0.048 6.323 15.753 integrate_v_rspace 128 11.3 0.004 0.004 15.636 15.663 make_images_data 5014 15.6 0.062 0.067 12.138 14.055 dbcsr_complete_redistribute 395 12.7 1.503 1.638 9.249 13.234 multiply_cannon_sync_h2d 30084 15.6 11.695 12.630 11.695 12.630 hybrid_alltoall_any 5200 16.5 0.534 2.226 10.619 12.448 dbcsr_mm_accdrv_process 62264 16.2 7.413 8.363 11.517 12.019 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.767 11.725 init_scf_run 11 5.9 0.000 0.001 11.248 11.250 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.248 11.249 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.573 10.774 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.443 10.313 pw_transfer 1547 11.6 0.089 0.106 10.169 10.253 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 9.939 10.029 mp_alltoall_i22 716 14.1 5.729 9.779 5.729 9.779 density_rs2pw 128 9.7 0.006 0.007 8.138 9.709 cp_fm_cholesky_decompose 22 10.9 8.803 8.905 8.803 8.905 fft_wrap_pw1pw2_140 523 13.2 0.480 0.493 8.647 8.759 fft3d_ps 1291 14.7 2.770 2.859 8.166 8.239 grid_integrate_task_list 128 12.3 7.509 8.066 7.509 8.066 wfi_extrapolate 11 7.9 0.002 0.008 7.960 7.960 rs_pw_transfer 1046 11.9 0.013 0.014 5.917 7.569 mp_alltoall_d11v 2415 14.1 5.902 6.856 5.902 6.856 multiply_cannon_metrocomm4 25070 15.6 0.082 0.092 2.717 6.787 calculate_dm_sparse 128 9.5 0.001 0.001 6.448 6.525 mp_irecv_dv 76098 16.2 2.569 6.523 2.569 6.523 grid_collocate_task_list 128 9.7 5.068 5.847 5.068 5.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.511 5.531 potential_pw2rs 128 12.3 0.024 0.024 5.469 5.515 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.764 4.849 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.662 4.729 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 4.512 4.512 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.322 4.396 mp_allgather_i34 2507 14.6 2.409 4.379 2.409 4.379 make_images_sizes 5014 15.6 0.007 0.007 2.504 4.355 mp_alltoall_i44 5014 16.6 2.498 4.348 2.498 4.348 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=205.775000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.090909, yerr=17.588501 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/19/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 5.865088E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.498448E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.057 189.734 189.735 qs_mol_dyn_low 1 2.0 0.003 0.003 189.243 189.257 qs_forces 11 3.9 0.004 0.004 189.133 189.143 qs_energies 11 4.9 0.001 0.002 181.305 181.312 scf_env_do_scf 11 5.9 0.030 0.033 163.283 163.291 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 123.957 123.958 velocity_verlet 10 3.0 0.001 0.002 122.997 123.002 qs_scf_new_mos 117 7.6 0.001 0.001 82.923 83.049 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.922 83.048 ot_scf_mini 117 9.6 0.003 0.004 78.366 78.439 dbcsr_multiply_generic 2507 12.6 0.181 0.191 76.601 77.047 multiply_cannon 2507 13.6 0.586 0.634 55.515 60.483 multiply_cannon_loop 2507 14.6 0.442 0.459 50.636 51.393 ot_mini 117 10.6 0.001 0.001 41.441 41.523 init_scf_loop 11 6.9 0.000 0.000 39.140 39.157 prepare_preconditioner 11 7.9 0.000 0.000 35.016 35.055 make_preconditioner 11 8.9 0.000 0.000 35.016 35.055 mp_waitall_1 125778 16.7 27.655 34.735 27.655 34.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.776 33.106 rebuild_ks_matrix 128 8.3 0.001 0.001 31.502 31.616 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.018 31.501 31.615 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.558 28.661 qs_ot_get_p 128 10.4 0.001 0.001 23.790 23.884 qs_ot_get_derivative 117 11.6 0.001 0.002 21.439 21.507 multiply_cannon_multrec 10028 15.6 10.437 14.131 17.894 21.035 cp_fm_cholesky_invert 11 10.9 19.942 19.948 19.942 19.948 ot_diis_step 117 11.6 0.020 0.020 19.928 19.928 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.541 19.893 apply_single 128 13.6 0.001 0.001 19.541 19.892 qs_ot_p2m_diag 83 11.4 0.496 0.501 19.551 19.573 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.243 19.554 make_m2s 5014 13.6 0.067 0.072 17.200 19.485 make_images 5014 14.6 2.286 2.587 16.896 19.186 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.297 18.299 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.619 16.672 calculate_rho_elec 128 8.7 0.259 0.269 16.619 16.671 sum_up_and_integrate 128 10.3 0.182 0.191 16.291 16.334 integrate_v_rspace 128 11.3 0.004 0.005 16.109 16.158 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.764 14.772 cp_fm_diag_elpa_base 83 14.4 14.442 14.588 14.756 14.765 make_images_data 5014 15.6 0.051 0.059 10.366 13.036 hybrid_alltoall_any 5200 16.5 0.794 3.632 10.003 12.282 multiply_cannon_sync_h2d 10028 15.6 11.598 12.081 11.598 12.081 init_scf_run 11 5.9 0.000 0.001 11.054 11.054 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.054 11.054 pw_transfer 1547 11.6 0.086 0.095 10.820 10.883 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 10.596 10.668 cp_fm_cholesky_decompose 22 10.9 9.259 9.397 9.259 9.397 fft_wrap_pw1pw2_140 523 13.2 0.495 0.516 9.160 9.230 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.832 8.880 density_rs2pw 128 9.7 0.005 0.005 8.007 8.878 fft3d_ps 1291 14.7 2.717 2.792 8.761 8.804 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 5.265 8.222 grid_integrate_task_list 128 12.3 7.750 8.194 7.750 8.194 mp_allgather_i34 2507 14.6 3.090 8.041 3.090 8.041 wfi_extrapolate 11 7.9 0.001 0.001 7.848 7.848 dbcsr_mm_accdrv_process 20762 16.1 2.624 3.584 7.090 7.779 mp_alltoall_d11v 2415 14.1 5.798 6.685 5.798 6.685 calculate_dm_sparse 128 9.5 0.001 0.001 6.311 6.380 rs_pw_transfer 1046 11.9 0.012 0.013 5.200 6.147 grid_collocate_task_list 128 9.7 5.393 6.058 5.393 6.058 dbcsr_complete_redistribute 395 12.7 2.152 2.209 5.560 6.015 potential_pw2rs 128 12.3 0.027 0.028 5.455 5.491 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 5.343 5.345 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.287 5.294 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.863 4.297 mp_irecv_dv 28860 15.9 1.827 4.226 1.827 4.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.090 4.116 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.671 4.044 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.892 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.845 3.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.848 3.859 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.783 3.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=189.735000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1380.090909, yerr=37.424966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410026932224 0.0% 0.0% 100.0% flops 32 x 32 x 32 1930587799552 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957877121024 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 11444723269632 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019199483904 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019199483904 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 92.803070E+12 0.0% 0.0% 100.0% flops max/rank 11.607223E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705605184 0.0% 0.0% 100.0% number of processed stacks 1948912 0.0% 0.0% 100.0% average stack size 0.0 0.0 3440.7 marketing flops 143.514185E+12 ------------------------------------------------------------------------------- # multiplications 2488 max memory usage/rank 3.005850E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99520 MPI messages size (bytes): total size 1.127523E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.329611E+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 44888 34846277632 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716109087456 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4034 58688. MP_Allreduce 11082 1513. MP_Sync 87 MP_Alltoall 1700 36954293. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.069 321.449 321.454 qs_mol_dyn_low 1 2.0 0.003 0.003 320.853 320.876 qs_forces 11 3.9 0.004 0.004 320.728 320.730 qs_energies 11 4.9 0.006 0.011 310.979 310.993 scf_env_do_scf 11 5.9 0.002 0.002 287.497 287.512 velocity_verlet 10 3.0 0.002 0.002 229.215 229.221 scf_env_do_scf_inner_loop 116 6.6 0.007 0.012 154.985 154.987 init_scf_loop 11 6.9 0.000 0.000 132.243 132.246 prepare_preconditioner 11 7.9 0.000 0.000 126.853 126.876 make_preconditioner 11 8.9 0.000 0.000 126.853 126.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 102.574 123.926 qs_scf_new_mos 116 7.6 0.001 0.001 100.827 100.933 qs_scf_loop_do_ot 116 8.6 0.001 0.001 100.826 100.933 ot_scf_mini 116 9.6 0.003 0.004 95.999 96.069 dbcsr_multiply_generic 2488 12.5 0.210 0.221 87.128 87.475 cp_fm_upper_to_full 105 14.8 53.499 76.832 53.499 76.832 multiply_cannon 2488 13.5 0.714 0.747 60.202 61.726 multiply_cannon_loop 2488 14.5 0.463 0.469 56.200 57.162 ot_mini 116 10.6 0.001 0.001 47.690 47.778 dbcsr_complete_redistribute 395 12.7 3.996 4.067 30.405 43.247 rebuild_ks_matrix 127 8.3 0.001 0.001 40.691 40.751 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 40.690 40.750 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.765 39.595 cp_fm_cholesky_invert 11 10.9 38.084 38.090 38.084 38.090 qs_ks_update_qs_env 127 7.6 0.001 0.001 37.363 37.416 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.234 36.907 mp_waitall_1 102888 16.8 31.024 35.113 31.024 35.113 mp_alltoall_i22 714 14.1 22.032 34.978 22.032 34.978 qs_ot_get_p 127 10.4 0.001 0.001 32.450 32.573 qs_ot_p2m_diag 83 11.4 0.878 0.884 27.828 27.848 qs_ot_get_derivative 116 11.6 0.002 0.002 26.576 26.643 cp_dbcsr_syevd 83 12.4 0.006 0.006 25.996 25.998 make_m2s 4976 13.5 0.076 0.079 21.808 22.966 make_images 4976 14.5 3.738 3.954 21.334 22.495 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.276 22.277 cp_fm_diag_elpa_base 83 14.4 17.647 19.360 22.272 22.272 qs_rho_update_rho_low 127 7.7 0.001 0.001 22.222 22.247 calculate_rho_elec 127 8.7 0.480 0.481 22.221 22.247 sum_up_and_integrate 127 10.3 0.320 0.322 21.316 21.397 integrate_v_rspace 127 11.3 0.004 0.004 20.996 21.078 ot_diis_step 116 11.6 0.022 0.022 21.028 21.028 multiply_cannon_metrocomm3 9952 15.5 0.022 0.023 19.725 20.498 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.235 20.390 apply_single 127 13.6 0.001 0.001 20.235 20.390 multiply_cannon_multrec 9952 15.5 10.696 12.778 18.347 18.403 multiply_cannon_sync_h2d 9952 15.5 15.539 15.557 15.539 15.557 make_images_data 4976 15.5 0.059 0.063 12.168 14.579 hybrid_alltoall_any 5162 16.4 1.298 3.064 12.163 14.468 pw_transfer 1535 11.6 0.092 0.092 13.951 13.965 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 13.717 13.731 init_scf_run 11 5.9 0.000 0.001 13.180 13.180 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.179 13.180 fft_wrap_pw1pw2_140 519 13.2 0.539 0.545 11.984 12.002 fft3d_ps 1281 14.7 2.732 2.762 11.760 11.781 cp_fm_cholesky_decompose 22 10.9 10.868 10.906 10.868 10.906 mp_alltoall_d11v 2407 14.1 9.861 10.543 9.861 10.543 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.369 10.429 wfi_extrapolate 11 7.9 0.001 0.001 9.957 9.957 density_rs2pw 127 9.7 0.005 0.005 9.544 9.665 dbcsr_mm_accdrv_process 20602 16.0 4.010 6.206 7.411 9.461 grid_integrate_task_list 127 12.3 8.510 8.677 8.510 8.677 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.168 8.171 calculate_dm_sparse 127 9.5 0.001 0.001 6.703 6.784 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.684 6.770 rs_scatter_matrices 138 9.7 3.570 4.599 6.389 6.618 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.472 6.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=321.454000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2677.909091, yerr=160.844620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 1.262060E+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 7111212. 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 0.025 0.077 85.616 85.617 qs_energies 1 2.0 0.000 0.000 85.043 85.050 ls_scf 1 3.0 0.002 0.017 84.073 84.080 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.875 73.033 multiply_cannon 111 7.7 0.018 0.020 56.062 57.115 multiply_cannon_loop 111 8.7 0.200 0.219 52.533 53.826 ls_scf_main 1 4.0 0.000 0.010 52.396 52.410 density_matrix_trs4 2 5.0 0.004 0.080 46.820 46.922 ls_scf_init_scf 1 4.0 0.000 0.002 28.629 28.631 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.373 27.425 mp_waitall_1 11031 10.9 22.701 26.390 22.701 26.390 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.191 25.211 multiply_cannon_multrec 2664 9.7 8.148 8.728 15.552 17.110 multiply_cannon_sync_h2d 2664 9.7 13.528 15.613 13.528 15.613 make_m2s 222 7.7 0.008 0.011 13.145 13.675 make_images 222 8.7 0.100 0.108 13.123 13.654 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.854 12.721 make_images_data 222 9.7 0.004 0.005 7.726 8.246 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.454 8.036 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.628 7.025 8.007 hybrid_alltoall_any 227 10.6 0.216 1.831 6.626 7.865 dbcsr_mm_accdrv_process_sort 4760 11.4 6.311 7.213 6.311 7.213 calculate_norms 4752 9.8 5.509 6.133 5.509 6.133 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.126 5.256 mp_sum_l 807 5.4 3.163 4.714 3.163 4.714 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.060 3.961 mp_irecv_dv 6231 10.9 2.044 3.926 2.044 3.926 make_images_sizes 222 9.7 0.000 0.000 0.716 3.524 mp_alltoall_i44 222 10.7 0.716 3.524 0.716 3.524 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.290 3.456 arnoldi_extremal 4 6.8 0.000 0.000 3.227 3.253 arnoldi_normal_ev 4 7.8 0.001 0.003 3.227 3.253 build_subspace 16 8.4 0.009 0.012 3.129 3.131 ls_scf_post 1 4.0 0.000 0.000 3.045 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.863 2.908 dbcsr_special_finalize 555 9.7 0.005 0.006 2.348 2.792 dbcsr_merge_single_wm 555 10.7 0.485 0.613 2.340 2.782 make_images_pack 222 9.7 2.207 2.616 2.209 2.618 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.311 2.553 dbcsr_sort_data 658 11.4 2.112 2.493 2.112 2.493 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.455 2.065 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.293 2.375 buffer_matrices_ensure_size 222 8.7 1.812 2.196 1.812 2.196 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.855 1.857 rebuild_ks_matrix 3 7.3 0.000 0.000 1.846 1.847 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.016 1.846 1.847 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.617000, 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/9bf2b2015f7fa359f4ee539b0939531368083923_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 2.105053E+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 0.064 91.569 91.569 qs_energies 1 2.0 0.000 0.000 91.032 91.036 ls_scf 1 3.0 0.000 0.000 89.714 89.719 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.251 75.621 multiply_cannon 111 7.7 0.029 0.047 53.579 57.176 ls_scf_main 1 4.0 0.000 0.000 55.188 55.194 multiply_cannon_loop 111 8.7 0.115 0.122 50.327 53.528 density_matrix_trs4 2 5.0 0.002 0.003 49.402 49.589 ls_scf_init_scf 1 4.0 0.000 0.000 30.961 30.962 mp_waitall_1 9105 10.9 21.297 30.495 21.297 30.495 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.737 29.838 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.226 27.250 multiply_cannon_multrec 1332 9.7 13.055 16.601 22.371 27.044 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.096 21.170 make_m2s 222 7.7 0.006 0.008 15.095 15.765 make_images 222 8.7 1.578 1.968 15.065 15.737 dbcsr_mm_accdrv_process 4041 10.4 0.288 0.445 8.917 10.439 dbcsr_mm_accdrv_process_sort 4041 11.4 8.510 9.997 8.510 9.997 make_images_data 222 9.7 0.004 0.004 8.564 9.529 hybrid_alltoall_any 227 10.6 0.521 2.477 8.004 9.075 mp_sum_l 807 5.4 5.283 8.259 5.283 8.259 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.261 8.011 mp_irecv_dv 3311 11.0 3.241 7.960 3.241 7.960 calculate_norms 2376 9.8 5.974 6.724 5.974 6.724 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.050 6.457 multiply_cannon_sync_h2d 1332 9.7 4.791 5.742 4.791 5.742 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.087 5.293 arnoldi_extremal 4 6.8 0.000 0.000 4.910 4.936 arnoldi_normal_ev 4 7.8 0.001 0.005 4.910 4.936 build_subspace 16 8.4 0.014 0.021 4.647 4.652 ls_scf_post 1 4.0 0.000 0.000 3.565 3.568 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.248 3.485 ls_scf_store_result 1 5.0 0.000 0.000 3.260 3.386 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.232 2.738 3.234 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.253 2.828 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.643 2.718 make_images_pack 222 9.7 2.028 2.432 2.030 2.434 mp_allgather_i34 111 8.7 0.954 2.311 0.954 2.311 dbcsr_sort_data 436 11.2 1.808 2.046 1.808 2.046 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.922 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.909 1.911 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.909 1.911 dbcsr_data_new 4174 10.1 1.606 1.861 1.606 1.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.569000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 2.717602E+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 0.032 0.056 93.636 93.637 qs_energies 1 2.0 0.000 0.000 92.968 92.981 ls_scf 1 3.0 0.000 0.001 91.550 91.562 dbcsr_multiply_generic 111 6.7 0.016 0.018 75.594 75.872 ls_scf_main 1 4.0 0.000 0.002 57.177 57.181 multiply_cannon 111 7.7 0.041 0.097 52.315 56.674 multiply_cannon_loop 111 8.7 0.099 0.105 48.654 51.779 density_matrix_trs4 2 5.0 0.003 0.008 51.165 51.340 mp_waitall_1 7281 11.0 22.528 31.749 22.528 31.749 ls_scf_init_scf 1 4.0 0.000 0.002 30.820 30.822 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.604 29.687 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.048 27.062 multiply_cannon_multrec 888 9.7 12.703 15.127 21.341 24.391 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.476 21.486 make_m2s 222 7.7 0.006 0.007 16.747 18.011 make_images 222 8.7 1.965 2.296 16.708 17.973 hybrid_alltoall_any 227 10.6 0.621 2.857 8.823 10.519 make_images_data 222 9.7 0.003 0.004 9.217 10.443 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.414 8.163 9.406 dbcsr_mm_accdrv_process_sort 3754 11.4 7.791 8.991 7.791 8.991 mp_sum_l 807 5.4 4.992 8.427 4.992 8.427 multiply_cannon_sync_h2d 888 9.7 6.053 7.337 6.053 7.337 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.433 7.289 mp_irecv_dv 2335 11.1 2.419 7.244 2.419 7.244 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.809 6.614 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.457 6.118 arnoldi_extremal 4 6.8 0.000 0.000 5.562 5.583 arnoldi_normal_ev 4 7.8 0.001 0.005 5.562 5.583 build_subspace 16 8.4 0.014 0.020 5.252 5.259 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.950 5.177 calculate_norms 1584 9.8 4.281 4.610 4.281 4.610 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.661 3.967 mp_allgather_i34 111 8.7 1.497 3.775 1.497 3.775 dbcsr_matrix_vector_mult_local 304 10.0 3.005 3.602 3.007 3.604 ls_scf_post 1 4.0 0.002 0.014 3.553 3.568 ls_scf_store_result 1 5.0 0.000 0.000 3.277 3.375 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.878 2.963 make_images_sizes 222 9.7 0.000 0.000 1.284 2.303 mp_alltoall_i44 222 10.7 1.284 2.303 1.284 2.303 make_images_pack 222 9.7 1.816 2.133 1.819 2.136 dbcsr_sort_data 325 11.1 1.826 2.125 1.826 2.125 dbcsr_data_release 9322 10.9 1.312 1.945 1.312 1.945 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.928 1.930 rebuild_ks_matrix 3 7.3 0.000 0.000 1.910 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.910 1.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.637000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2191.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 3.390456E+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 0.042 0.056 97.799 97.802 qs_energies 1 2.0 0.000 0.000 97.044 97.050 ls_scf 1 3.0 0.002 0.014 95.383 95.388 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.689 78.892 ls_scf_main 1 4.0 0.001 0.011 59.085 59.088 multiply_cannon 111 7.7 0.053 0.155 51.864 55.969 density_matrix_trs4 2 5.0 0.002 0.004 52.836 52.977 multiply_cannon_loop 111 8.7 0.114 0.124 46.824 50.015 ls_scf_init_scf 1 4.0 0.000 0.004 33.041 33.044 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.659 31.717 mp_waitall_1 6369 11.0 22.985 30.071 22.985 30.071 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.192 29.207 multiply_cannon_multrec 1332 9.7 14.150 17.109 22.088 24.686 make_m2s 222 7.7 0.007 0.007 21.187 22.644 make_images 222 8.7 3.140 3.613 21.136 22.596 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.418 16.993 make_images_data 222 9.7 0.004 0.004 11.814 13.387 hybrid_alltoall_any 227 10.6 0.798 3.781 11.083 12.810 dbcsr_mm_accdrv_process 3641 10.4 0.222 0.423 7.585 9.106 dbcsr_mm_accdrv_process_sort 3641 11.4 7.207 8.688 7.207 8.688 mp_sum_l 807 5.4 4.083 7.108 4.083 7.108 multiply_cannon_sync_h2d 1332 9.7 5.539 6.142 5.539 6.142 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.059 6.132 mp_irecv_dv 3229 10.9 2.036 6.085 2.036 6.085 arnoldi_extremal 4 6.8 0.000 0.000 5.477 5.495 arnoldi_normal_ev 4 7.8 0.001 0.005 5.477 5.494 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.116 5.359 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.572 5.317 build_subspace 16 8.4 0.014 0.020 5.132 5.140 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.576 4.725 calculate_norms 2376 9.8 4.189 4.556 4.189 4.556 mp_allgather_i34 111 8.7 2.160 4.537 2.160 4.537 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.699 3.964 dbcsr_matrix_vector_mult_local 304 10.0 3.179 3.690 3.181 3.692 dbcsr_sort_data 658 11.4 3.062 3.459 3.062 3.459 ls_scf_post 1 4.0 0.000 0.000 3.255 3.260 dbcsr_special_finalize 555 9.7 0.006 0.007 2.806 3.249 dbcsr_merge_single_wm 555 10.7 0.534 0.669 2.797 3.241 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.090 3.141 ls_scf_store_result 1 5.0 0.000 0.000 3.000 3.059 dbcsr_data_release 10477 10.7 1.580 2.410 1.580 2.410 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.065 2.067 rebuild_ks_matrix 3 7.3 0.000 0.000 2.041 2.043 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.025 2.041 2.043 dbcsr_finalize 304 7.8 0.049 0.061 1.803 1.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.802000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 4.637102E+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 0.044 0.054 99.551 99.552 qs_energies 1 2.0 0.000 0.000 98.689 98.693 ls_scf 1 3.0 0.000 0.000 96.715 96.720 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.502 77.765 ls_scf_main 1 4.0 0.000 0.000 62.107 62.108 multiply_cannon 111 7.7 0.070 0.139 54.890 60.045 density_matrix_trs4 2 5.0 0.002 0.003 54.696 54.772 multiply_cannon_loop 111 8.7 0.070 0.093 50.326 52.517 mp_waitall_1 5436 11.0 26.002 31.971 26.002 31.971 ls_scf_init_scf 1 4.0 0.000 0.000 30.992 30.998 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.596 29.644 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.427 27.446 multiply_cannon_multrec 444 9.7 13.916 16.242 21.033 22.660 make_m2s 222 7.7 0.004 0.005 17.758 20.360 make_images 222 8.7 3.730 4.432 17.697 20.300 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.806 15.619 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.084 14.276 make_images_data 222 9.7 0.003 0.004 9.903 12.477 hybrid_alltoall_any 227 10.6 0.790 3.764 9.636 12.182 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.346 6.826 7.960 dbcsr_mm_accdrv_process_sort 3003 11.4 6.511 7.614 6.511 7.614 multiply_cannon_sync_h2d 444 9.7 6.503 7.248 6.503 7.248 mp_allgather_i34 111 8.7 2.804 7.083 2.804 7.083 arnoldi_extremal 4 6.8 0.000 0.000 6.140 6.163 arnoldi_normal_ev 4 7.8 0.002 0.007 6.140 6.162 build_subspace 16 8.4 0.015 0.019 5.741 5.751 mp_sum_l 807 5.4 2.909 5.402 2.909 5.402 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.650 4.839 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.291 4.434 dbcsr_matrix_vector_mult_local 304 10.0 3.651 4.135 3.653 4.137 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.962 3.966 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.682 3.859 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.763 3.844 mp_irecv_dv 1241 11.2 1.663 3.833 1.663 3.833 calculate_norms 792 9.8 3.581 3.726 3.581 3.726 ls_scf_post 1 4.0 0.000 0.000 3.616 3.620 make_images_sizes 222 9.7 0.000 0.000 1.195 3.543 mp_alltoall_i44 222 10.7 1.195 3.543 1.195 3.543 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.435 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.290 2.291 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.277 rebuild_ks_matrix 3 7.3 0.000 0.000 2.257 2.259 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.003 2.257 2.259 dbcsr_merge_all 275 8.9 0.474 0.533 2.046 2.110 dbcsr_data_release 10123 10.8 1.326 2.000 1.326 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.552000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3694.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_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 8.730055E+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 0.090 0.145 108.923 108.924 qs_energies 1 2.0 0.000 0.000 107.453 107.464 ls_scf 1 3.0 0.000 0.000 104.498 104.509 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.092 77.235 ls_scf_main 1 4.0 0.000 0.000 66.343 66.345 density_matrix_trs4 2 5.0 0.002 0.003 56.768 56.836 multiply_cannon 111 7.7 0.172 0.232 49.005 50.517 multiply_cannon_loop 111 8.7 0.068 0.071 45.684 46.608 ls_scf_init_scf 1 4.0 0.000 0.000 34.496 34.496 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.765 32.782 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.911 29.919 mp_waitall_1 4527 11.1 21.677 25.605 21.677 25.605 make_m2s 222 7.7 0.005 0.005 24.260 25.090 make_images 222 8.7 4.594 4.977 24.154 24.983 multiply_cannon_multrec 444 9.7 17.836 18.673 22.478 23.172 hybrid_alltoall_any 227 10.6 1.658 3.623 12.967 15.890 make_images_data 222 9.7 0.003 0.003 13.328 15.724 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.774 10.312 multiply_cannon_sync_h2d 444 9.7 8.846 8.891 8.846 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.738 7.750 arnoldi_normal_ev 4 7.8 0.003 0.009 7.738 7.750 build_subspace 16 8.4 0.025 0.036 7.177 7.186 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.647 5.759 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.460 5.555 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.097 5.355 dbcsr_matrix_vector_mult_local 304 10.0 4.949 5.234 4.951 5.236 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.320 4.483 4.614 dbcsr_mm_accdrv_process_sort 1814 11.4 4.184 4.319 4.184 4.319 make_images_sizes 222 9.7 0.000 0.000 1.586 3.985 mp_alltoall_i44 222 10.7 1.586 3.984 1.586 3.984 ls_scf_post 1 4.0 0.000 0.000 3.659 3.669 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.418 calculate_norms 792 9.8 3.243 3.282 3.243 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.136 dbcsr_complete_redistribute 5 7.6 1.433 1.478 2.814 2.960 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.924 2.924 dbcsr_merge_all 275 8.9 0.891 0.922 2.862 2.912 dbcsr_data_release 12724 10.6 2.322 2.865 2.322 2.865 mp_allgather_i34 111 8.7 0.957 2.851 0.957 2.851 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.668 2.670 matrix_ls_to_qs 2 6.0 0.000 0.000 2.481 2.645 rebuild_ks_matrix 3 7.3 0.000 0.000 2.602 2.604 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.003 2.602 2.604 dbcsr_sort_data 325 11.1 2.436 2.499 2.436 2.499 mp_sum_l 807 5.4 1.519 2.469 1.519 2.469 dbcsr_new_transposed 4 7.5 0.244 0.253 2.387 2.401 mp_alltoall_d11v 48 9.2 2.233 2.286 2.233 2.286 dbcsr_frobenius_norm 74 6.6 2.059 2.139 2.206 2.256 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.209 dbcsr_add_anytype 103 7.2 0.860 0.892 2.132 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.924000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6859.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9bf2b2015f7fa359f4ee539b0939531368083923_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.532608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 753995063. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.125 293.170 293.174 qs_mol_dyn_low 1 2.0 0.004 0.025 292.098 292.119 qs_forces 5 3.8 0.006 0.042 291.917 291.921 qs_energies 5 4.8 0.003 0.036 288.409 288.437 scf_env_do_scf 5 5.8 0.001 0.009 273.078 273.080 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 239.625 239.628 qs_scf_new_mos 105 7.6 0.000 0.001 188.374 188.592 qs_scf_loop_do_ot 105 8.6 0.001 0.001 188.374 188.592 ot_scf_mini 105 9.6 0.003 0.005 178.380 178.605 dbcsr_multiply_generic 1445 12.2 0.126 0.133 135.408 136.210 velocity_verlet 4 3.0 0.002 0.019 118.308 118.311 multiply_cannon 1445 13.2 0.277 0.293 114.676 117.761 multiply_cannon_loop 1445 14.2 2.338 2.477 112.228 114.317 qs_ot_get_p 112 10.4 0.001 0.001 86.180 86.521 qs_ot_p2m_diag 40 11.0 0.020 0.031 74.025 74.125 cp_dbcsr_syevd 40 12.0 0.002 0.002 70.747 70.749 cp_fm_syevd 40 13.0 0.000 0.000 64.361 64.520 ot_mini 105 10.6 0.001 0.003 61.819 61.969 cp_fm_redistribute_end 40 14.0 29.456 58.812 29.466 58.815 cp_fm_syevd_base 40 14.0 29.331 58.683 29.331 58.683 mp_waitall_1 488190 16.1 38.575 47.573 38.575 47.573 multiply_cannon_multrec 69360 15.2 31.535 37.003 39.272 45.341 qs_ot_get_derivative 55 11.6 0.001 0.001 40.152 40.316 multiply_cannon_metrocomm3 69360 15.2 0.189 0.204 26.353 37.022 multiply_cannon_sync_h2d 69360 15.2 30.157 34.579 30.157 34.579 rebuild_ks_matrix 110 8.4 0.000 0.000 34.262 34.475 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.082 34.262 34.475 init_scf_loop 7 6.6 0.000 0.004 33.404 33.407 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.458 31.647 prepare_preconditioner 7 7.6 0.000 0.000 27.773 27.815 make_preconditioner 7 8.6 0.000 0.007 27.773 27.815 qs_rho_update_rho_low 110 7.6 0.001 0.002 24.082 24.398 calculate_rho_elec 110 8.6 0.029 0.032 24.081 24.397 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.042 23.292 apply_single 62 13.6 0.000 0.000 23.042 23.292 rs_pw_transfer 690 11.5 0.011 0.013 19.866 21.256 ot_new_cg_direction 55 11.6 0.001 0.006 21.023 21.026 make_full_inverse_cholesky 7 9.6 0.000 0.001 19.431 19.501 density_rs2pw 110 9.6 0.005 0.006 17.690 19.134 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.159 15.441 pw_transfer 1645 12.4 0.078 0.097 14.608 14.924 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 14.471 14.792 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.449 14.575 mp_sum_l 4684 12.4 12.359 13.511 12.359 13.511 cp_fm_cholesky_invert 7 10.6 13.108 13.125 13.108 13.125 fft_wrap_pw1pw2_240 915 15.0 0.415 0.459 12.605 12.870 init_scf_run 5 5.8 0.000 0.001 12.408 12.409 scf_env_initial_rho_setup 5 6.8 0.000 0.002 12.408 12.409 sum_up_and_integrate 60 10.3 0.028 0.033 12.039 12.052 qs_vxc_create 110 10.4 0.002 0.004 11.988 12.042 integrate_v_rspace 60 11.3 0.002 0.004 12.010 12.026 calculate_dm_sparse 110 9.5 0.000 0.001 11.616 11.834 fft3d_pb 915 16.0 2.329 2.773 11.429 11.725 check_diag 80 13.5 8.587 8.923 11.074 11.297 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.660 10.738 multiply_cannon_metrocomm1 69360 15.2 0.093 0.100 4.884 9.292 make_m2s 2890 13.2 0.076 0.083 7.971 8.540 dbcsr_mm_accdrv_process 154766 15.8 3.931 4.179 7.610 8.469 make_images 2890 14.2 0.244 0.263 7.868 8.437 xc_rho_set_and_dset_create 110 12.4 0.084 0.109 8.114 8.385 xc_vxc_pw_create 60 11.3 0.042 0.068 8.118 8.173 make_full_single_inverse 7 9.6 0.001 0.003 8.000 8.042 potential_pw2rs 60 12.3 0.003 0.003 7.920 8.010 rs_pw_transfer_RS2PW_30 110 11.6 1.542 1.633 7.204 7.680 xc_pw_derive 510 13.4 0.005 0.006 7.522 7.604 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.446 7.460 mp_sendrecv_dv 168740 12.6 7.074 7.288 7.074 7.288 mp_alltoall_z22v 2340 17.7 6.581 7.079 6.581 7.079 acc_transpose_blocks 69360 15.2 0.352 0.378 6.400 7.040 calculate_first_density_matrix 1 7.0 0.002 0.004 6.989 7.008 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.865 6.936 mp_waitany 7680 13.5 4.598 5.998 4.598 5.998 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=293.174000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.200000, yerr=3.429286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9bf2b2015f7fa359f4ee539b0939531368083923 Summary: empty Status: OK