=== This is the CP2K Performance-Test === Updating e5cb9aef0..a5f7dffd4 Fast-forward src/response_solver.F | 20 ++--- src/rpa_im_time_force_methods.F | 24 ++++-- .../regtest-rpa-cubic-scaling-2/H2O_added_mos.inp | 86 ++++++++++++++++++++++ tests/QS/regtest-rpa-cubic-scaling-2/TEST_FILES | 1 + 4 files changed, 111 insertions(+), 20 deletions(-) create mode 100644 tests/QS/regtest-rpa-cubic-scaling-2/H2O_added_mos.inp Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2 ################# 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/01 job id: 45216655 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/02 job id: 45216656 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/03 job id: 45216657 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/04 job id: 45216661 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/05 job id: 45216662 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/06 job id: 45216663 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/07 job id: 45216665 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/08 job id: 45216666 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/09 job id: 45216667 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/10 job id: 45216668 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/11 job id: 45216669 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/12 job id: 45216670 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/13 job id: 45216671 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/14 job id: 45216672 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/15 job id: 45216673 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/16 job id: 45216674 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/17 job id: 45216675 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/18 job id: 45216676 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/19 job id: 45216677 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/20 job id: 45216678 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/21 job id: 45216679 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/22 job id: 45216680 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/23 job id: 45216681 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/24 job id: 45216683 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/25 job id: 45216684 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/26 job id: 45216685 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/27 job id: 45216686 --- 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.035 139.870 139.870 farming_run 1 2.0 139.290 139.293 139.822 139.825 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.448509E+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 0.014 0.029 118.412 118.412 qs_energies 1 2.0 0.000 0.000 118.110 118.112 mp2_main 1 3.0 0.000 0.000 115.393 115.396 mp2_gpw_main 1 4.0 0.028 0.035 114.040 114.043 mp2_ri_gpw_compute_in 1 5.0 0.180 0.205 94.097 94.536 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.522 55.963 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.163 41.838 47.218 get_2c_integrals 1 6.0 0.007 0.009 37.775 38.410 integrate_v_rspace 273 8.0 0.437 0.447 25.201 30.310 pw_transfer 6555 10.6 0.367 0.383 27.467 28.081 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.118 26.696 grid_integrate_task_list 273 9.0 20.998 26.588 20.998 26.588 fft_wrap_pw1pw2_100 2178 12.4 1.190 1.403 23.629 24.214 rpa_ri_compute_en 1 5.0 0.002 0.010 19.826 20.143 compute_2c_integrals 1 7.0 0.003 0.003 19.836 19.858 compute_2c_integrals_loop_lm 1 8.0 0.022 0.027 18.963 19.430 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.451 18.941 19.403 cp_fm_cholesky_decompose 12 8.2 18.085 18.682 18.085 18.682 cholesky_decomp 1 7.0 0.000 0.000 16.779 17.391 fft3d_s 5443 13.4 16.192 16.641 16.214 16.662 ao_to_mo_and_store_B_mult_1 272 7.0 10.859 15.590 10.859 15.590 calculate_wavefunction 272 8.0 5.429 5.573 12.559 13.191 rpa_num_int 1 6.0 0.000 0.005 10.970 10.970 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.708 10.920 calc_mat_Q 8 8.0 0.000 0.000 9.375 9.472 contract_S_to_Q 8 9.0 0.000 0.000 8.797 8.894 calc_potential_gpw 544 9.5 0.005 0.006 8.260 8.641 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.234 8.527 parallel_gemm_fm 14 9.1 0.000 0.000 8.388 8.470 parallel_gemm_fm_cosma 14 10.1 8.388 8.470 8.388 8.470 potential_pw2rs 545 10.0 0.107 0.109 7.721 8.415 create_integ_mat 1 6.0 0.023 0.028 8.316 8.317 collocate_single_gaussian 272 10.0 0.039 0.042 7.491 7.747 array2fm 1 7.0 0.000 0.000 6.786 7.407 pw_scatter_s 2720 13.7 4.414 4.637 4.414 4.637 pw_gather_s 2722 13.2 3.900 4.273 3.900 4.273 array2fm_buffer_send 1 8.0 2.939 3.185 2.939 3.185 scf_env_do_scf 1 3.0 0.000 0.000 2.423 2.424 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.042526, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 0.036 0.043 403.728 403.729 farming_run 1 2.0 402.888 402.899 403.656 403.658 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.229476E+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 17938732. 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 0.020 0.041 211.021 211.022 qs_energies 1 2.0 0.000 0.000 210.710 210.715 scf_env_do_scf 1 3.0 0.000 0.000 107.476 107.477 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.263 106.270 rebuild_ks_matrix 4 6.0 0.000 0.000 106.261 106.269 qs_ks_build_kohn_sham_matrix 4 7.0 0.065 0.072 106.261 106.269 hfx_ks_matrix 4 8.0 0.001 0.001 105.804 105.807 integrate_four_center 4 9.0 0.144 0.458 105.803 105.807 mp2_main 1 3.0 0.000 0.000 102.869 102.875 mp2_gpw_main 1 4.0 0.032 0.042 101.343 101.351 integrate_four_center_main 4 10.0 0.129 0.601 96.707 99.954 integrate_four_center_bin 270 11.0 96.578 99.866 96.578 99.866 init_scf_loop 1 4.0 0.000 0.000 92.795 92.795 mp2_ri_gpw_compute_in 1 5.0 0.064 0.066 74.818 75.855 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.440 55.475 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.158 42.120 47.197 integrate_v_rspace 95 8.0 0.400 0.571 28.562 33.455 pw_transfer 2240 10.6 0.146 0.161 29.899 30.285 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.933 29.271 grid_integrate_task_list 95 9.0 23.783 28.870 23.783 28.870 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 26.373 28.031 fft_wrap_pw1pw2_100 730 12.4 1.239 1.359 26.615 27.017 ao_to_mo_and_store_B_mult_1 91 7.0 10.637 26.899 10.637 26.899 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.825 1.873 24.685 24.695 get_2c_integrals 1 6.0 0.000 0.001 20.243 20.314 compute_2c_integrals 1 7.0 0.003 0.004 19.208 19.211 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.746 19.037 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.834 18.745 19.036 fft3d_s 1823 13.4 18.399 18.924 18.412 18.937 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.677 14.677 calculate_wavefunction 91 8.0 2.017 2.046 9.737 9.956 potential_pw2rs 186 10.0 0.033 0.035 8.672 9.295 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.586 8.789 9.155 local_gemm 172 8.0 8.234 8.606 8.234 8.606 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.179 8.556 mp2_ri_gpw_compute_en_comm 22 7.0 0.524 0.547 7.694 8.179 collocate_single_gaussian 91 10.0 0.017 0.022 7.828 8.093 calc_potential_gpw 182 9.5 0.002 0.002 7.878 8.063 mp_sync 37 10.5 3.773 7.348 3.773 7.348 mp2_ri_gpw_compute_en_ener 172 7.0 6.332 6.406 6.332 6.406 mp_sendrecv_dm3 2068 8.0 5.662 6.146 5.662 6.146 pw_gather_s 912 13.2 4.939 5.351 4.939 5.351 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.335598, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1517.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.850240E+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 0.033 0.111 60.072 60.073 qs_mol_dyn_low 1 2.0 0.003 0.005 59.673 59.681 qs_forces 11 3.9 0.015 0.081 59.555 59.556 qs_energies 11 4.9 0.006 0.017 57.836 57.858 scf_env_do_scf 11 5.9 0.001 0.001 51.134 51.134 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 48.437 48.438 qs_scf_new_mos 108 7.5 0.000 0.001 35.935 36.230 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.935 36.230 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.964 35.432 ot_scf_mini 108 9.5 0.002 0.004 34.239 34.416 velocity_verlet 10 3.0 0.001 0.002 29.136 29.146 multiply_cannon 2286 13.5 0.183 0.189 26.680 28.555 multiply_cannon_loop 2286 14.5 1.470 1.555 25.595 27.404 ot_mini 108 10.5 0.001 0.002 20.283 20.530 qs_ot_get_derivative 108 11.5 0.001 0.001 17.347 17.517 mp_waitall_1 245248 16.5 9.226 15.517 9.226 15.517 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 6.024 13.158 multiply_cannon_multrec 54864 15.5 4.261 6.560 7.911 11.480 qs_ot_get_p 119 10.4 0.001 0.001 9.098 9.382 rebuild_ks_matrix 119 8.3 0.000 0.000 9.235 9.380 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 9.234 9.380 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.161 8.296 mp_sum_l 7207 12.9 5.496 7.319 5.496 7.319 multiply_cannon_sync_h2d 54864 15.5 5.923 7.233 5.923 7.233 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.060 6.536 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.921 5.963 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.657 5.750 qs_rho_update_rho_low 119 7.7 0.001 0.002 5.626 5.723 calculate_rho_elec 119 8.7 0.012 0.018 5.625 5.722 sum_up_and_integrate 119 10.3 0.012 0.016 5.597 5.604 integrate_v_rspace 119 11.3 0.003 0.003 5.585 5.594 init_scf_run 11 5.9 0.000 0.001 5.290 5.290 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.290 5.290 cp_dbcsr_syevd 50 12.0 0.003 0.003 5.075 5.075 dbcsr_mm_accdrv_process 76910 16.1 1.145 1.827 3.571 5.014 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.712 4.713 cp_fm_redistribute_end 50 14.0 2.395 4.663 2.414 4.671 cp_fm_diag_elpa_base 50 14.0 2.245 4.515 2.253 4.526 rs_pw_transfer 974 11.9 0.011 0.012 3.949 4.088 density_rs2pw 119 9.7 0.004 0.005 3.477 3.576 calculate_dm_sparse 119 9.5 0.000 0.001 3.078 3.238 multiply_cannon_metrocomm1 54864 15.5 0.053 0.057 1.828 3.201 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.954 3.167 apply_single 119 13.6 0.000 0.000 2.953 3.167 pw_transfer 1439 11.6 0.053 0.058 2.978 3.141 jit_kernel_multiply 13 15.8 2.366 3.100 2.366 3.100 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.902 3.069 potential_pw2rs 119 12.3 0.004 0.004 2.829 2.895 fft3d_ps 1201 14.6 0.370 0.484 2.677 2.845 calculate_first_density_matrix 1 7.0 0.000 0.001 2.754 2.757 ot_diis_step 108 11.5 0.006 0.009 2.721 2.721 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.589 2.688 init_scf_loop 11 6.9 0.000 0.000 2.660 2.660 make_m2s 4572 13.5 0.054 0.056 2.513 2.616 make_images 4572 14.5 0.133 0.138 2.430 2.534 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.497 2.499 wfi_extrapolate 11 7.9 0.001 0.003 2.450 2.451 fft_wrap_pw1pw2_140 487 13.2 0.078 0.092 2.229 2.397 acc_transpose_blocks 54864 15.5 0.228 0.257 1.731 2.215 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.109 2.186 grid_integrate_task_list 119 12.3 2.006 2.132 2.006 2.132 mp_alltoall_d11v 2130 13.8 1.879 2.043 1.879 2.043 mp_sum_d 4135 12.0 1.301 1.908 1.301 1.908 mp_alltoall_z22v 1201 16.6 1.564 1.705 1.564 1.705 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.584 1.613 mp_waitany 12084 13.8 1.361 1.548 1.361 1.548 make_images_sizes 4572 15.5 0.004 0.005 1.152 1.441 prepare_preconditioner 11 7.9 0.000 0.000 1.406 1.439 make_preconditioner 11 8.9 0.000 0.000 1.406 1.439 mp_alltoall_i44 4572 16.5 1.148 1.437 1.148 1.437 grid_collocate_task_list 119 9.7 1.290 1.377 1.290 1.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.312 1.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.073000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 489.480192E+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 1030362. 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 0.025 0.042 46.738 46.740 qs_mol_dyn_low 1 2.0 0.003 0.003 46.436 46.476 qs_forces 11 3.9 0.003 0.010 46.372 46.373 qs_energies 11 4.9 0.002 0.003 44.517 44.522 scf_env_do_scf 11 5.9 0.001 0.002 38.641 38.641 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 35.353 35.353 dbcsr_multiply_generic 2286 12.5 0.101 0.104 23.623 24.069 qs_scf_new_mos 108 7.5 0.001 0.001 23.565 23.785 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.564 23.784 ot_scf_mini 108 9.5 0.002 0.003 22.585 22.756 velocity_verlet 10 3.0 0.001 0.002 21.833 21.834 multiply_cannon 2286 13.5 0.207 0.216 17.023 19.651 multiply_cannon_loop 2286 14.5 0.896 0.971 15.658 17.922 ot_mini 108 10.5 0.001 0.001 13.577 13.821 mp_waitall_1 200699 16.5 7.184 12.415 7.184 12.415 qs_ot_get_derivative 108 11.5 0.001 0.001 10.962 11.135 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.477 10.050 multiply_cannon_multrec 27432 15.5 1.954 4.337 6.054 9.096 rebuild_ks_matrix 119 8.3 0.000 0.000 8.709 8.919 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 8.708 8.918 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.754 7.946 qs_ot_get_p 119 10.4 0.001 0.001 5.923 6.162 dbcsr_mm_accdrv_process 47894 16.0 3.069 5.185 4.031 5.957 sum_up_and_integrate 119 10.3 0.024 0.028 5.578 5.586 integrate_v_rspace 119 11.3 0.003 0.003 5.554 5.563 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.983 5.355 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.275 5.294 calculate_rho_elec 119 8.7 0.021 0.024 5.274 5.294 rs_pw_transfer 974 11.9 0.010 0.011 4.733 5.160 mp_sum_l 7207 12.9 3.026 5.021 3.026 5.021 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.632 4.746 apply_single 119 13.6 0.000 0.000 3.632 4.746 init_scf_run 11 5.9 0.000 0.001 4.574 4.574 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.574 4.574 qs_ot_p2m_diag 50 11.0 0.009 0.013 4.003 4.021 density_rs2pw 119 9.7 0.004 0.004 3.551 3.933 cp_dbcsr_syevd 50 12.0 0.004 0.004 3.526 3.527 make_m2s 4572 13.5 0.052 0.054 3.128 3.338 potential_pw2rs 119 12.3 0.006 0.006 3.234 3.288 make_images 4572 14.5 0.200 0.237 3.041 3.249 init_scf_loop 11 6.9 0.000 0.000 3.247 3.248 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.953 2.954 cp_fm_redistribute_end 50 14.0 1.484 2.886 1.498 2.894 multiply_cannon_sync_h2d 27432 15.5 2.113 2.825 2.113 2.825 cp_fm_diag_elpa_base 50 14.0 1.350 2.729 1.385 2.793 pw_transfer 1439 11.6 0.065 0.072 2.606 2.701 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.575 2.664 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.514 2.612 calculate_first_density_matrix 1 7.0 0.001 0.002 2.574 2.575 ot_diis_step 108 11.5 0.011 0.011 2.515 2.516 calculate_dm_sparse 119 9.5 0.000 0.001 2.246 2.346 fft3d_ps 1201 14.6 0.517 0.569 2.218 2.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.191 2.194 prepare_preconditioner 11 7.9 0.000 0.000 2.139 2.166 make_preconditioner 11 8.9 0.000 0.000 2.139 2.166 acc_transpose_blocks 27432 15.5 0.109 0.112 1.246 2.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.017 2.076 jit_kernel_multiply 9 16.1 0.911 1.978 0.911 1.978 grid_integrate_task_list 119 12.3 1.823 1.955 1.823 1.955 wfi_extrapolate 11 7.9 0.001 0.001 1.926 1.926 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.827 1.900 mp_alltoall_d11v 2130 13.8 1.640 1.825 1.640 1.825 fft_wrap_pw1pw2_140 487 13.2 0.077 0.085 1.711 1.812 make_images_data 4572 15.5 0.045 0.051 1.297 1.810 mp_allgather_i34 2286 14.5 0.810 1.591 0.810 1.591 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.127 1.588 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.716 1.483 mp_sendrecv_dv 16779 12.7 1.373 1.443 1.373 1.443 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.409 1.417 make_images_sizes 4572 15.5 0.005 0.005 1.025 1.383 mp_alltoall_i44 4572 16.5 1.020 1.378 1.020 1.378 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.317 1.378 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.605 1.152 1.372 grid_collocate_task_list 119 9.7 1.230 1.358 1.230 1.358 mp_alltoall_z22v 1201 16.6 1.128 1.286 1.128 1.286 jit_kernel_transpose 5 15.5 0.536 1.218 0.536 1.218 mp_sum_d 4135 12.0 0.724 1.198 0.724 1.198 rs_pw_transfer_RS2PW_50 119 11.7 0.272 0.288 0.936 1.066 dbcsr_complete_redistribute 329 12.2 0.080 0.101 0.983 1.063 cp_fm_cholesky_invert 11 10.9 1.048 1.055 1.048 1.055 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.018 1.036 mp_waitany 5720 13.7 0.584 0.986 0.584 0.986 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.958 0.959 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.594 0.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=46.740000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=1.372697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 525.107200E+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 1607838. 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 0.018 0.033 34.244 34.245 qs_mol_dyn_low 1 2.0 0.003 0.019 33.999 34.007 qs_forces 11 3.9 0.003 0.003 33.935 33.936 qs_energies 11 4.9 0.003 0.017 32.318 32.320 scf_env_do_scf 11 5.9 0.001 0.002 27.541 27.542 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.714 24.714 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.861 17.036 velocity_verlet 10 3.0 0.001 0.002 16.209 16.212 qs_scf_new_mos 108 7.5 0.001 0.001 15.921 15.936 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.921 15.935 ot_scf_mini 108 9.5 0.002 0.003 15.159 15.176 multiply_cannon 2286 13.5 0.194 0.199 13.131 13.849 multiply_cannon_loop 2286 14.5 0.629 0.657 12.278 12.957 ot_mini 108 10.5 0.001 0.001 9.144 9.165 qs_ot_get_derivative 108 11.5 0.001 0.001 7.573 7.587 multiply_cannon_multrec 18288 15.5 1.904 2.850 6.700 7.047 rebuild_ks_matrix 119 8.3 0.000 0.000 6.758 6.790 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.757 6.790 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.987 6.015 dbcsr_mm_accdrv_process 38222 16.0 4.075 5.377 4.713 5.488 sum_up_and_integrate 119 10.3 0.031 0.032 4.253 4.257 integrate_v_rspace 119 11.3 0.003 0.003 4.221 4.230 mp_waitall_1 158411 16.6 3.030 4.108 3.030 4.108 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.784 3.794 calculate_rho_elec 119 8.7 0.031 0.032 3.783 3.793 qs_ot_get_p 119 10.4 0.001 0.001 3.752 3.780 init_scf_run 11 5.9 0.000 0.001 3.561 3.561 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.561 3.561 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.810 3.434 rs_pw_transfer 974 11.9 0.009 0.010 2.600 2.857 init_scf_loop 11 6.9 0.001 0.008 2.804 2.805 density_rs2pw 119 9.7 0.004 0.004 2.283 2.560 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.547 2.555 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.515 2.452 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.068 2.400 apply_single 119 13.6 0.000 0.000 2.067 2.400 cp_dbcsr_syevd 50 12.0 0.004 0.004 2.265 2.265 calculate_first_density_matrix 1 7.0 0.001 0.002 2.243 2.245 pw_transfer 1439 11.6 0.065 0.071 2.215 2.232 make_m2s 4572 13.5 0.044 0.045 2.008 2.207 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.121 2.142 make_images 4572 14.5 0.190 0.201 1.923 2.121 potential_pw2rs 119 12.3 0.007 0.008 1.992 2.004 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.948 1.957 cp_fm_diag_elpa_base 50 14.0 1.921 1.935 1.945 1.954 prepare_preconditioner 11 7.9 0.000 0.000 1.947 1.950 make_preconditioner 11 8.9 0.000 0.001 1.947 1.950 grid_integrate_task_list 119 12.3 1.801 1.908 1.801 1.908 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.785 1.871 calculate_dm_sparse 119 9.5 0.000 0.001 1.855 1.866 jit_kernel_multiply 10 16.4 0.588 1.860 0.588 1.860 fft3d_ps 1201 14.6 0.530 0.550 1.800 1.823 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.729 1.737 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.677 1.679 mp_sum_l 7207 12.9 1.338 1.677 1.338 1.677 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.630 1.646 multiply_cannon_sync_h2d 18288 15.5 1.356 1.580 1.356 1.580 ot_diis_step 108 11.5 0.011 0.011 1.535 1.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.510 1.522 grid_collocate_task_list 119 9.7 1.213 1.359 1.213 1.359 wfi_extrapolate 11 7.9 0.001 0.001 1.249 1.249 acc_transpose_blocks 18288 15.5 0.076 0.078 1.222 1.244 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.193 make_images_data 4572 15.5 0.045 0.048 0.921 1.136 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.455 1.075 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.796 1.018 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 0.964 0.965 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.944 0.963 mp_alltoall_z22v 1201 16.6 0.848 0.956 0.848 0.956 mp_alltoall_d11v 2130 13.8 0.788 0.929 0.788 0.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.846 0.849 mp_waitany 9880 13.7 0.563 0.838 0.563 0.838 cp_fm_cholesky_invert 11 10.9 0.827 0.833 0.827 0.833 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.127 0.544 0.801 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.784 0.798 mp_allgather_i34 2286 14.5 0.317 0.728 0.317 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.656 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.245000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.727273, yerr=2.733327 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 552.923136E+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 36.862 36.862 qs_mol_dyn_low 1 2.0 0.003 0.003 36.656 36.662 qs_forces 11 3.9 0.002 0.003 36.590 36.590 qs_energies 11 4.9 0.002 0.002 34.840 34.844 scf_env_do_scf 11 5.9 0.003 0.005 29.756 29.758 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.172 26.173 dbcsr_multiply_generic 2286 12.5 0.098 0.101 18.663 18.824 velocity_verlet 10 3.0 0.002 0.002 18.413 18.416 qs_scf_new_mos 108 7.5 0.001 0.001 17.187 17.244 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.186 17.243 ot_scf_mini 108 9.5 0.002 0.003 16.232 16.281 multiply_cannon 2286 13.5 0.225 0.267 14.802 15.361 multiply_cannon_loop 2286 14.5 0.934 0.963 13.792 14.168 ot_mini 108 10.5 0.001 0.001 9.809 9.878 multiply_cannon_multrec 27432 15.5 2.355 3.035 8.704 9.021 qs_ot_get_derivative 108 11.5 0.001 0.001 7.959 8.012 dbcsr_mm_accdrv_process 47916 15.9 5.356 7.066 6.256 7.447 rebuild_ks_matrix 119 8.3 0.000 0.000 6.871 6.922 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.871 6.922 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.112 6.157 sum_up_and_integrate 119 10.3 0.036 0.038 4.054 4.063 integrate_v_rspace 119 11.3 0.003 0.003 4.018 4.028 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.828 3.856 calculate_rho_elec 119 8.7 0.040 0.046 3.827 3.856 qs_ot_get_p 119 10.4 0.001 0.001 3.763 3.848 init_scf_run 11 5.9 0.000 0.001 3.744 3.744 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.744 3.744 init_scf_loop 11 6.9 0.000 0.000 3.557 3.560 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.762 3.208 prepare_preconditioner 11 7.9 0.000 0.000 2.692 2.700 make_preconditioner 11 8.9 0.000 0.000 2.692 2.700 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.292 2.623 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.113 2.577 apply_single 119 13.6 0.000 0.000 2.112 2.576 mp_waitall_1 137007 16.6 2.235 2.549 2.235 2.549 rs_pw_transfer 974 11.9 0.009 0.009 2.297 2.493 make_m2s 4572 13.5 0.054 0.056 2.341 2.481 density_rs2pw 119 9.7 0.004 0.004 2.239 2.432 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.406 2.415 make_images 4572 14.5 0.271 0.331 2.234 2.372 calculate_first_density_matrix 1 7.0 0.000 0.001 2.313 2.314 pw_transfer 1439 11.6 0.066 0.072 2.266 2.310 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.173 2.221 calculate_dm_sparse 119 9.5 0.000 0.000 2.083 2.132 cp_dbcsr_syevd 50 12.0 0.004 0.005 2.069 2.069 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.031 2.057 grid_integrate_task_list 119 12.3 1.815 1.929 1.815 1.929 fft3d_ps 1201 14.6 0.565 0.620 1.847 1.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.868 1.869 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.756 1.809 ot_diis_step 108 11.5 0.012 0.012 1.806 1.806 potential_pw2rs 119 12.3 0.009 0.010 1.787 1.795 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.770 1.781 cp_fm_diag_elpa_base 50 14.0 1.733 1.751 1.767 1.777 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.683 1.699 mp_sum_l 7207 12.9 1.094 1.604 1.094 1.604 jit_kernel_multiply 8 15.9 0.840 1.581 0.840 1.581 acc_transpose_blocks 27432 15.5 0.113 0.116 1.468 1.493 wfi_extrapolate 11 7.9 0.001 0.001 1.379 1.379 grid_collocate_task_list 119 9.7 1.221 1.343 1.221 1.343 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.923 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.252 1.261 cp_fm_upper_to_full 72 14.2 0.827 1.179 0.827 1.179 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.134 1.151 dbcsr_complete_redistribute 329 12.2 0.117 0.148 0.824 1.107 multiply_cannon_sync_h2d 27432 15.5 0.995 1.094 0.995 1.094 make_images_data 4572 15.5 0.045 0.049 0.938 1.093 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.072 1.073 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.809 0.998 mp_alltoall_d11v 2130 13.8 0.809 0.980 0.809 0.980 mp_alltoall_z22v 1201 16.6 0.866 0.908 0.866 0.908 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.599 0.872 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.844 0.862 cp_fm_cholesky_invert 11 10.9 0.858 0.861 0.858 0.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.855 mp_alltoall_i22 627 13.8 0.442 0.747 0.442 0.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.862000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.090909, yerr=2.065058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 599.982080E+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 0.012 0.029 30.056 30.058 qs_mol_dyn_low 1 2.0 0.003 0.003 29.889 29.897 qs_forces 11 3.9 0.002 0.003 29.827 29.827 qs_energies 11 4.9 0.001 0.002 28.092 28.094 scf_env_do_scf 11 5.9 0.004 0.005 23.267 23.268 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 20.595 20.596 velocity_verlet 10 3.0 0.001 0.001 15.058 15.061 dbcsr_multiply_generic 2286 12.5 0.091 0.096 12.827 12.914 qs_scf_new_mos 108 7.5 0.001 0.001 12.118 12.145 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.117 12.145 ot_scf_mini 108 9.5 0.002 0.002 11.421 11.448 multiply_cannon 2286 13.5 0.233 0.241 10.068 10.425 multiply_cannon_loop 2286 14.5 0.327 0.336 9.096 9.311 ot_mini 108 10.5 0.001 0.001 6.272 6.308 rebuild_ks_matrix 119 8.3 0.000 0.000 6.242 6.262 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.242 6.262 multiply_cannon_multrec 9144 15.5 1.701 1.928 6.018 6.254 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.577 5.596 qs_ot_get_derivative 108 11.5 0.001 0.001 4.908 4.937 dbcsr_mm_accdrv_process 12550 15.8 3.079 4.230 4.215 4.294 sum_up_and_integrate 119 10.3 0.038 0.042 3.873 3.878 integrate_v_rspace 119 11.3 0.003 0.003 3.835 3.840 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.806 3.815 calculate_rho_elec 119 8.7 0.060 0.061 3.805 3.814 init_scf_run 11 5.9 0.000 0.001 3.376 3.376 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.375 3.376 qs_ot_get_p 119 10.4 0.001 0.001 3.292 3.346 init_scf_loop 11 6.9 0.000 0.000 2.646 2.648 pw_transfer 1439 11.6 0.066 0.069 2.313 2.323 mp_waitall_1 115863 16.7 1.827 2.298 1.827 2.298 density_rs2pw 119 9.7 0.004 0.004 2.103 2.246 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.236 2.239 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.218 2.231 calculate_first_density_matrix 1 7.0 0.001 0.001 2.164 2.166 make_m2s 4572 13.5 0.035 0.036 1.895 2.092 rs_pw_transfer 974 11.9 0.008 0.008 1.883 2.031 cp_dbcsr_syevd 50 12.0 0.004 0.004 2.024 2.025 make_images 4572 14.5 0.267 0.299 1.806 2.003 grid_integrate_task_list 119 12.3 1.854 1.967 1.854 1.967 jit_kernel_multiply 10 15.7 1.098 1.923 1.098 1.923 fft3d_ps 1201 14.6 0.567 0.577 1.878 1.891 prepare_preconditioner 11 7.9 0.000 0.000 1.877 1.881 make_preconditioner 11 8.9 0.000 0.000 1.877 1.881 calculate_dm_sparse 119 9.5 0.000 0.000 1.818 1.838 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.790 1.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.763 1.790 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.691 1.707 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.672 1.681 cp_fm_diag_elpa_base 50 14.0 1.644 1.663 1.670 1.679 potential_pw2rs 119 12.3 0.010 0.011 1.562 1.565 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.524 1.526 grid_collocate_task_list 119 9.7 1.276 1.405 1.276 1.405 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.357 1.364 ot_diis_step 108 11.5 0.012 0.013 1.348 1.348 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.226 1.240 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.206 1.232 apply_single 119 13.6 0.000 0.000 1.206 1.231 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.222 1.223 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.217 1.222 wfi_extrapolate 11 7.9 0.001 0.001 1.157 1.157 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.870 1.136 make_images_data 4572 15.5 0.039 0.042 0.890 1.130 mp_alltoall_d11v 2130 13.8 0.889 1.008 0.889 1.008 cp_fm_cholesky_invert 11 10.9 0.967 0.971 0.967 0.971 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.924 acc_transpose_blocks 9144 15.5 0.038 0.038 0.908 0.916 mp_alltoall_z22v 1201 16.6 0.758 0.813 0.758 0.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.790 0.792 multiply_cannon_sync_h2d 9144 15.5 0.719 0.789 0.719 0.789 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.777 0.785 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.423 0.778 mp_allgather_i34 2286 14.5 0.304 0.753 0.304 0.753 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.439 0.724 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.676 0.724 yz_to_x 606 15.1 0.264 0.275 0.661 0.679 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.669 0.674 x_to_yz 595 16.2 0.276 0.289 0.637 0.645 mp_waitany 5200 13.7 0.474 0.612 0.474 0.612 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.058000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.181818, yerr=3.270201 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 779.108352E+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 0.020 0.040 43.088 43.089 qs_mol_dyn_low 1 2.0 0.003 0.003 42.874 42.882 qs_forces 11 3.9 0.002 0.002 42.805 42.806 qs_energies 11 4.9 0.002 0.002 40.771 40.774 scf_env_do_scf 11 5.9 0.001 0.002 34.923 34.923 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.707 26.708 velocity_verlet 10 3.0 0.002 0.002 24.164 24.170 dbcsr_multiply_generic 2286 12.5 0.098 0.100 17.857 17.893 qs_scf_new_mos 108 7.5 0.001 0.001 16.564 16.644 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.563 16.643 ot_scf_mini 108 9.5 0.002 0.002 15.460 15.546 multiply_cannon 2286 13.5 0.302 0.308 13.826 14.584 multiply_cannon_loop 2286 14.5 0.342 0.349 12.522 13.159 ot_mini 108 10.5 0.001 0.001 9.165 9.268 multiply_cannon_multrec 9144 15.5 3.404 4.818 8.592 8.715 init_scf_loop 11 6.9 0.000 0.000 8.184 8.185 rebuild_ks_matrix 119 8.3 0.000 0.000 7.546 7.680 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.546 7.679 qs_ot_get_derivative 108 11.5 0.001 0.001 7.119 7.203 prepare_preconditioner 11 7.9 0.000 0.000 7.183 7.196 make_preconditioner 11 8.9 0.000 0.000 7.183 7.196 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.648 7.065 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.828 6.948 dbcsr_mm_accdrv_process 12550 15.8 4.084 5.605 5.064 6.336 cp_fm_upper_to_full 72 14.2 3.278 4.750 3.278 4.750 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.470 4.525 calculate_rho_elec 119 8.7 0.118 0.122 4.469 4.524 sum_up_and_integrate 119 10.3 0.065 0.067 4.152 4.157 integrate_v_rspace 119 11.3 0.004 0.004 4.086 4.094 init_scf_run 11 5.9 0.000 0.001 3.791 3.791 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.791 3.791 qs_ot_get_p 119 10.4 0.001 0.001 3.482 3.607 mp_waitall_1 94719 16.7 2.742 3.512 2.742 3.512 dbcsr_complete_redistribute 329 12.2 0.286 0.292 2.117 3.002 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.575 2.949 pw_transfer 1439 11.6 0.069 0.069 2.927 2.932 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.828 2.833 make_m2s 4572 13.5 0.038 0.039 2.483 2.689 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.779 2.649 make_images 4572 14.5 0.354 0.389 2.361 2.566 fft3d_ps 1201 14.6 0.599 0.607 2.451 2.454 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.142 2.428 apply_single 119 13.6 0.000 0.000 2.142 2.428 mp_alltoall_i22 627 13.8 1.534 2.412 1.534 2.412 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.530 2.391 fft_wrap_pw1pw2_140 487 13.2 0.095 0.098 2.362 2.369 density_rs2pw 119 9.7 0.004 0.004 2.332 2.356 calculate_first_density_matrix 1 7.0 0.001 0.001 2.239 2.292 calculate_dm_sparse 119 9.5 0.000 0.000 2.228 2.274 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.500 2.188 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.123 2.124 grid_integrate_task_list 119 12.3 2.063 2.077 2.063 2.077 ot_diis_step 108 11.5 0.014 0.014 2.013 2.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.874 1.875 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.840 1.841 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.742 1.787 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.779 1.779 mp_sum_l 7207 12.9 1.005 1.713 1.005 1.713 rs_pw_transfer 974 11.9 0.009 0.010 1.586 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.586 1.589 cp_fm_cholesky_invert 11 10.9 1.557 1.560 1.557 1.560 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.254 1.540 potential_pw2rs 119 12.3 0.014 0.015 1.532 1.536 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.519 1.520 cp_fm_diag_elpa_base 50 14.0 1.367 1.424 1.517 1.517 grid_collocate_task_list 119 9.7 1.494 1.505 1.494 1.505 make_images_data 4572 15.5 0.043 0.046 1.231 1.498 wfi_extrapolate 11 7.9 0.001 0.001 1.430 1.430 jit_kernel_multiply 8 15.5 0.952 1.424 0.952 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.277 1.326 mp_alltoall_d11v 2130 13.8 1.179 1.241 1.179 1.241 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.160 1.181 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.114 1.125 multiply_cannon_sync_h2d 9144 15.5 1.045 1.048 1.045 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.033 yz_to_x 606 15.1 0.459 0.471 0.993 1.004 qs_create_task_list 11 7.9 0.000 0.001 0.934 0.948 generate_qs_task_list 11 8.9 0.368 0.387 0.934 0.948 mp_alltoall_z22v 1201 16.6 0.900 0.947 0.900 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.943 acc_transpose_blocks 9144 15.5 0.038 0.039 0.892 0.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.089000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=732.272727, yerr=16.766259 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.087104E+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 0.014 0.032 92.474 92.475 qs_mol_dyn_low 1 2.0 0.003 0.003 92.209 92.219 qs_forces 11 3.9 0.003 0.004 92.093 92.094 qs_energies 11 4.9 0.001 0.002 88.963 88.979 scf_env_do_scf 11 5.9 0.001 0.001 79.473 79.475 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 72.512 72.513 dbcsr_multiply_generic 2055 12.4 0.107 0.111 52.128 52.327 qs_scf_new_mos 99 7.5 0.000 0.001 51.765 51.937 qs_scf_loop_do_ot 99 8.5 0.000 0.001 51.764 51.936 ot_scf_mini 99 9.5 0.002 0.002 49.294 49.428 velocity_verlet 10 3.0 0.001 0.003 47.589 47.591 multiply_cannon 2055 13.4 0.180 0.187 42.273 43.133 multiply_cannon_loop 2055 14.4 1.501 1.543 40.952 41.846 ot_mini 99 10.5 0.001 0.001 27.490 27.605 qs_ot_get_derivative 99 11.5 0.001 0.001 20.605 20.739 multiply_cannon_multrec 49320 15.4 12.382 13.103 17.229 17.914 rebuild_ks_matrix 110 8.3 0.000 0.000 16.519 16.671 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 16.518 16.670 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.535 14.672 qs_ot_get_p 110 10.4 0.001 0.001 13.067 13.192 mp_waitall_1 220248 16.4 11.885 12.894 11.885 12.894 multiply_cannon_sync_h2d 49320 15.4 10.225 10.921 10.225 10.921 qs_ot_p2m_diag 48 11.0 0.012 0.019 9.327 9.360 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.510 8.662 calculate_rho_elec 110 8.6 0.021 0.026 8.509 8.662 sum_up_and_integrate 110 10.3 0.036 0.043 8.625 8.639 integrate_v_rspace 110 11.3 0.003 0.004 8.589 8.611 cp_dbcsr_syevd 48 12.0 0.003 0.003 8.312 8.313 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.527 8.167 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.367 7.897 apply_single 110 13.6 0.000 0.001 7.366 7.896 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.189 7.496 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.424 7.443 cp_fm_diag_elpa_base 48 14.0 7.399 7.421 7.421 7.441 init_scf_run 11 5.9 0.000 0.001 7.265 7.266 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.265 7.265 init_scf_loop 11 6.9 0.000 0.000 6.919 6.921 ot_diis_step 99 11.5 0.006 0.006 6.639 6.639 rs_pw_transfer 902 11.9 0.011 0.013 5.468 6.132 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.940 6.018 density_rs2pw 110 9.6 0.004 0.005 5.276 5.958 mp_sum_l 6514 12.8 4.427 5.424 4.427 5.424 pw_transfer 1331 11.6 0.055 0.065 5.155 5.309 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.066 5.224 make_m2s 4110 13.4 0.061 0.064 4.970 5.078 dbcsr_mm_accdrv_process 87628 16.1 1.910 1.991 4.723 5.010 make_images 4110 14.4 0.178 0.195 4.873 4.983 fft3d_ps 1111 14.6 0.777 0.893 4.601 4.754 prepare_preconditioner 11 7.9 0.000 0.000 4.558 4.575 make_preconditioner 11 8.9 0.000 0.000 4.558 4.575 wfi_extrapolate 11 7.9 0.001 0.001 4.524 4.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.298 4.356 fft_wrap_pw1pw2_140 451 13.1 0.166 0.187 4.070 4.229 potential_pw2rs 110 12.3 0.006 0.007 4.120 4.172 calculate_dm_sparse 110 9.5 0.001 0.001 3.866 3.966 multiply_cannon_metrocomm1 49320 15.4 0.061 0.065 2.649 3.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.646 3.650 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.399 3.467 grid_integrate_task_list 110 12.3 3.224 3.456 3.224 3.456 mp_alltoall_z22v 1111 16.6 2.983 3.340 2.983 3.340 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.191 3.247 mp_alltoall_d11v 2046 13.8 2.397 2.952 2.397 2.952 jit_kernel_multiply 13 15.9 2.538 2.644 2.538 2.644 calculate_first_density_matrix 1 7.0 0.000 0.001 2.631 2.636 mp_waitany 14300 13.8 1.933 2.606 1.933 2.606 cp_fm_cholesky_invert 11 10.9 2.519 2.525 2.519 2.525 make_images_data 4110 15.4 0.042 0.045 2.116 2.290 grid_collocate_task_list 110 9.6 2.082 2.283 2.082 2.283 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.128 2.154 acc_transpose_blocks 49320 15.4 0.210 0.223 2.061 2.135 hybrid_alltoall_any 4261 16.3 0.082 0.471 1.807 2.022 yz_to_x 341 15.4 0.211 0.244 1.789 2.012 mp_sum_d 3889 11.9 1.492 1.980 1.492 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=92.475000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.636364, yerr=2.496278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 589.856768E+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 0.012 0.030 76.017 76.018 qs_mol_dyn_low 1 2.0 0.003 0.003 75.787 75.797 qs_forces 11 3.9 0.003 0.003 75.713 75.714 qs_energies 11 4.9 0.002 0.002 72.204 72.207 scf_env_do_scf 11 5.9 0.001 0.002 63.287 63.290 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 54.251 54.252 velocity_verlet 10 3.0 0.001 0.002 39.778 39.780 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.792 38.955 qs_scf_new_mos 99 7.5 0.001 0.001 36.024 36.178 qs_scf_loop_do_ot 99 8.5 0.001 0.001 36.023 36.177 ot_scf_mini 99 9.5 0.003 0.003 34.331 34.477 multiply_cannon 2055 13.4 0.221 0.242 31.170 32.505 multiply_cannon_loop 2055 14.4 0.920 0.945 29.661 30.662 ot_mini 99 10.5 0.001 0.001 19.198 19.370 multiply_cannon_multrec 24660 15.4 7.647 9.630 13.820 15.954 rebuild_ks_matrix 110 8.3 0.000 0.000 14.978 15.118 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.020 14.978 15.118 qs_ot_get_derivative 99 11.5 0.001 0.001 13.358 13.512 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.200 13.319 mp_waitall_1 176588 16.5 8.531 11.143 8.531 11.143 init_scf_loop 11 6.9 0.000 0.000 8.988 8.989 qs_ot_get_p 110 10.4 0.001 0.001 8.364 8.558 multiply_cannon_sync_h2d 24660 15.4 7.008 8.059 7.008 8.059 sum_up_and_integrate 110 10.3 0.052 0.059 7.675 7.685 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.153 7.654 integrate_v_rspace 110 11.3 0.003 0.003 7.623 7.635 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.553 7.295 apply_single 110 13.6 0.000 0.001 6.553 7.295 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.264 7.270 calculate_rho_elec 110 8.6 0.040 0.048 7.263 7.270 prepare_preconditioner 11 7.9 0.000 0.000 6.795 6.812 make_preconditioner 11 8.9 0.000 0.000 6.795 6.812 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.349 6.522 init_scf_run 11 5.9 0.000 0.001 6.462 6.462 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.461 6.462 dbcsr_mm_accdrv_process 52282 16.1 4.738 5.540 6.013 6.308 qs_ot_p2m_diag 48 11.0 0.028 0.043 6.117 6.137 ot_diis_step 99 11.5 0.010 0.010 5.792 5.792 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.956 5.670 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.650 5.650 make_m2s 4110 13.4 0.057 0.060 4.968 5.480 make_images 4110 14.4 0.400 0.443 4.859 5.368 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.884 4.898 cp_fm_diag_elpa_base 48 14.0 4.825 4.846 4.881 4.893 density_rs2pw 110 9.6 0.004 0.004 4.365 4.860 rs_pw_transfer 902 11.9 0.012 0.014 4.209 4.741 pw_transfer 1331 11.6 0.066 0.073 4.519 4.701 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.412 4.598 fft3d_ps 1111 14.6 1.107 1.331 3.768 3.940 wfi_extrapolate 11 7.9 0.001 0.001 3.869 3.869 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.512 3.666 fft_wrap_pw1pw2_140 451 13.1 0.198 0.220 3.279 3.474 potential_pw2rs 110 12.3 0.008 0.009 3.406 3.452 grid_integrate_task_list 110 12.3 3.116 3.408 3.116 3.408 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.306 3.308 cp_fm_cholesky_invert 11 10.9 3.248 3.255 3.248 3.255 make_images_data 4110 15.4 0.046 0.050 2.678 3.236 hybrid_alltoall_any 4261 16.3 0.102 0.449 2.363 3.079 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.028 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.916 2.972 mp_sum_l 6514 12.8 2.031 2.796 2.031 2.796 mp_alltoall_d11v 2046 13.8 2.052 2.670 2.052 2.670 calculate_first_density_matrix 1 7.0 0.001 0.001 2.489 2.492 grid_collocate_task_list 110 9.6 2.054 2.474 2.054 2.474 mp_alltoall_z22v 1111 16.6 1.902 2.046 1.902 2.046 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.023 2.039 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.945 1.964 mp_waitany 10164 13.8 1.345 1.917 1.345 1.917 jit_kernel_multiply 10 16.4 0.920 1.904 0.920 1.904 mp_allgather_i34 2055 14.4 0.840 1.886 0.840 1.886 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.876 1.877 cp_fm_cholesky_decompose 22 10.9 1.822 1.834 1.822 1.834 multiply_cannon_metrocomm4 22605 15.4 0.076 0.082 0.785 1.744 dbcsr_complete_redistribute 325 12.2 0.234 0.295 1.333 1.638 mp_irecv_dv 57340 16.2 0.662 1.628 0.662 1.628 rs_gather_matrices 110 12.3 0.195 0.218 1.016 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.553 1.562 acc_transpose_blocks 24660 15.4 0.110 0.114 1.516 1.549 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.216 0.988 1.526 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=76.018000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.909091, yerr=7.525164 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.541440E+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 0.013 0.028 61.571 61.572 qs_mol_dyn_low 1 2.0 0.003 0.004 61.270 61.280 qs_forces 11 3.9 0.003 0.004 61.200 61.201 qs_energies 11 4.9 0.002 0.002 57.961 57.965 scf_env_do_scf 11 5.9 0.001 0.001 49.992 49.993 scf_env_do_scf_inner_loop 99 6.5 0.004 0.006 41.179 41.180 velocity_verlet 10 3.0 0.001 0.002 33.292 33.296 dbcsr_multiply_generic 2055 12.4 0.106 0.110 28.530 28.783 qs_scf_new_mos 99 7.5 0.001 0.001 25.616 25.704 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.615 25.703 ot_scf_mini 99 9.5 0.002 0.003 24.395 24.504 multiply_cannon 2055 13.4 0.209 0.217 22.025 23.198 multiply_cannon_loop 2055 14.4 0.613 0.627 20.843 21.886 ot_mini 99 10.5 0.001 0.001 13.643 13.751 rebuild_ks_matrix 110 8.3 0.000 0.000 12.727 12.872 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.726 12.872 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.222 11.353 multiply_cannon_multrec 16440 15.4 3.979 5.176 9.732 10.863 mp_waitall_1 139946 16.5 7.043 10.056 7.043 10.056 qs_ot_get_derivative 99 11.5 0.001 0.001 9.219 9.325 init_scf_loop 11 6.9 0.000 0.000 8.775 8.777 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.158 7.041 prepare_preconditioner 11 7.9 0.000 0.000 6.964 6.981 make_preconditioner 11 8.9 0.000 0.000 6.964 6.981 sum_up_and_integrate 110 10.3 0.060 0.062 6.791 6.807 integrate_v_rspace 110 11.3 0.003 0.003 6.730 6.748 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.285 6.657 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.233 6.248 calculate_rho_elec 110 8.6 0.059 0.059 6.233 6.248 qs_ot_get_p 110 10.4 0.001 0.001 5.918 6.038 dbcsr_mm_accdrv_process 34862 16.1 4.629 5.314 5.606 5.824 init_scf_run 11 5.9 0.000 0.001 5.567 5.568 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.567 5.567 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.948 5.365 apply_single 110 13.6 0.000 0.000 4.948 5.365 make_m2s 4110 13.4 0.050 0.052 4.370 4.716 make_images 4110 14.4 0.393 0.510 4.256 4.601 density_rs2pw 110 9.6 0.004 0.004 3.318 4.559 ot_diis_step 99 11.5 0.010 0.011 4.397 4.398 multiply_cannon_sync_h2d 16440 15.4 3.715 4.329 3.715 4.329 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.230 4.234 rs_pw_transfer 902 11.9 0.010 0.011 2.871 4.099 pw_transfer 1331 11.6 0.065 0.074 3.895 3.905 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.884 3.885 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.159 3.803 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.788 3.800 grid_integrate_task_list 110 12.3 3.192 3.400 3.192 3.400 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.274 3.282 cp_fm_diag_elpa_base 48 14.0 3.204 3.237 3.271 3.279 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.224 3.240 fft3d_ps 1111 14.6 1.092 1.105 3.086 3.096 wfi_extrapolate 11 7.9 0.001 0.001 3.051 3.052 make_images_data 4110 15.4 0.043 0.046 2.546 2.962 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.255 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.797 2.799 cp_fm_cholesky_invert 11 10.9 2.737 2.743 2.737 2.743 calculate_dm_sparse 110 9.5 0.001 0.001 2.517 2.550 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.461 2.509 potential_pw2rs 110 12.3 0.011 0.011 2.479 2.492 grid_collocate_task_list 110 9.6 2.083 2.480 2.083 2.480 mp_waitany 17072 13.8 1.205 2.475 1.205 2.475 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.392 2.450 calculate_first_density_matrix 1 7.0 0.000 0.001 2.424 2.426 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.861 2.289 mp_alltoall_d11v 2046 13.8 1.833 2.242 1.833 2.242 mp_irecv_dv 48980 15.7 0.791 2.166 0.791 2.166 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.182 0.893 2.106 mp_sum_l 6514 12.8 1.521 2.078 1.521 2.078 dbcsr_complete_redistribute 325 12.2 0.323 0.355 1.495 1.968 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.953 1.954 cp_fm_upper_to_full 70 14.2 1.413 1.792 1.413 1.792 cp_fm_cholesky_decompose 22 10.9 1.697 1.723 1.697 1.723 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.706 1.721 mp_allgather_i34 2055 14.4 0.510 1.640 0.510 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.481 1.495 jit_kernel_multiply 8 16.6 0.584 1.488 0.584 1.488 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.994 1.453 rs_gather_matrices 110 12.3 0.235 0.269 0.980 1.374 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.260 1.269 mp_alltoall_z22v 1111 16.6 1.227 1.257 1.227 1.257 acc_transpose_blocks 16440 15.4 0.073 0.074 1.227 1.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.572000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.363636, yerr=9.334907 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 739.381248E+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 0.025 0.054 66.286 66.289 qs_mol_dyn_low 1 2.0 0.003 0.004 65.945 65.958 qs_forces 11 3.9 0.004 0.011 65.875 65.876 qs_energies 11 4.9 0.002 0.008 62.424 62.428 scf_env_do_scf 11 5.9 0.001 0.001 53.678 53.682 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.861 41.863 velocity_verlet 10 3.0 0.001 0.002 37.314 37.319 dbcsr_multiply_generic 2055 12.4 0.113 0.119 29.655 29.946 qs_scf_new_mos 99 7.5 0.001 0.001 26.481 26.579 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.480 26.578 ot_scf_mini 99 9.5 0.003 0.003 24.868 24.964 multiply_cannon 2055 13.4 0.249 0.279 22.590 23.943 multiply_cannon_loop 2055 14.4 0.879 0.902 21.152 21.823 ot_mini 99 10.5 0.001 0.001 14.118 14.239 multiply_cannon_multrec 24660 15.4 4.245 6.852 12.733 13.991 rebuild_ks_matrix 110 8.3 0.000 0.000 12.441 12.536 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.440 12.536 init_scf_loop 11 6.9 0.001 0.004 11.776 11.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.989 11.073 qs_ot_get_derivative 99 11.5 0.001 0.001 9.994 10.095 prepare_preconditioner 11 7.9 0.000 0.000 9.915 9.931 make_preconditioner 11 8.9 0.000 0.001 9.915 9.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.113 9.588 dbcsr_mm_accdrv_process 52304 16.0 6.795 8.094 8.343 9.260 sum_up_and_integrate 110 10.3 0.068 0.072 6.642 6.658 integrate_v_rspace 110 11.3 0.003 0.004 6.573 6.588 mp_waitall_1 121746 16.5 4.498 6.328 4.498 6.328 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.283 6.291 calculate_rho_elec 110 8.6 0.078 0.082 6.283 6.291 init_scf_run 11 5.9 0.000 0.001 5.942 5.942 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.942 5.942 qs_ot_get_p 110 10.4 0.001 0.002 5.754 5.912 make_m2s 4110 13.4 0.060 0.062 5.496 5.861 make_images 4110 14.4 0.578 0.704 5.355 5.718 cp_fm_upper_to_full 70 14.2 3.378 4.893 3.378 4.893 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.996 4.091 apply_single 110 13.6 0.000 0.000 3.996 4.090 ot_diis_step 99 11.5 0.011 0.011 4.086 4.086 pw_transfer 1331 11.6 0.065 0.074 4.000 4.030 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.964 3.977 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.894 3.928 dbcsr_complete_redistribute 325 12.2 0.450 0.513 2.706 3.859 density_rs2pw 110 9.6 0.004 0.004 3.185 3.697 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.491 3.492 grid_integrate_task_list 110 12.3 3.264 3.443 3.264 3.443 fft_wrap_pw1pw2_140 451 13.1 0.203 0.213 3.358 3.394 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.202 3.340 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.258 3.305 multiply_cannon_sync_h2d 24660 15.4 3.185 3.298 3.185 3.298 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.522 3.245 hybrid_alltoall_any 4261 16.3 0.120 0.455 2.352 3.229 fft3d_ps 1111 14.6 1.097 1.135 3.186 3.213 make_images_data 4110 15.4 0.046 0.050 2.832 3.213 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.963 3.026 wfi_extrapolate 11 7.9 0.001 0.001 3.008 3.008 rs_pw_transfer 902 11.9 0.010 0.011 2.481 3.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.965 2.966 calculate_dm_sparse 110 9.5 0.001 0.001 2.929 2.960 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.938 2.948 cp_fm_diag_elpa_base 48 14.0 2.786 2.848 2.935 2.946 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.793 2.916 calculate_first_density_matrix 1 7.0 0.001 0.003 2.837 2.841 mp_alltoall_i22 605 13.7 1.649 2.814 1.649 2.814 cp_fm_cholesky_invert 11 10.9 2.675 2.683 2.675 2.683 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.552 2.588 grid_collocate_task_list 110 9.6 2.182 2.438 2.182 2.438 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.311 2.312 potential_pw2rs 110 12.3 0.013 0.013 2.281 2.288 jit_kernel_multiply 11 15.9 1.215 2.099 1.215 2.099 mp_alltoall_d11v 2046 13.8 1.800 2.046 1.800 2.046 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.688 1.724 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.697 1.709 cp_fm_cholesky_decompose 22 10.9 1.661 1.703 1.661 1.703 mp_allgather_i34 2055 14.4 0.529 1.659 0.529 1.659 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.851 1.600 mp_waitany 13376 13.8 1.073 1.585 1.073 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.569 acc_transpose_blocks 24660 15.4 0.107 0.110 1.505 1.528 mp_irecv_dv 62702 16.1 0.748 1.519 0.748 1.519 mp_sum_l 6514 12.8 0.897 1.450 0.897 1.450 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.259 1.366 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.176 0.828 1.359 mp_alltoall_z22v 1111 16.6 1.226 1.333 1.226 1.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.289000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.727273, yerr=10.694866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 821.043200E+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 0.015 0.039 58.455 58.463 qs_mol_dyn_low 1 2.0 0.003 0.012 58.204 58.213 qs_forces 11 3.9 0.020 0.021 58.131 58.131 qs_energies 11 4.9 0.002 0.002 54.379 54.383 scf_env_do_scf 11 5.9 0.001 0.002 45.651 45.652 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 37.364 37.365 velocity_verlet 10 3.0 0.005 0.005 32.604 32.612 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.589 23.758 qs_scf_new_mos 99 7.5 0.001 0.001 21.582 21.640 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.581 21.640 ot_scf_mini 99 9.5 0.002 0.002 20.298 20.334 multiply_cannon 2055 13.4 0.248 0.274 17.820 19.163 multiply_cannon_loop 2055 14.4 0.318 0.329 16.395 16.743 rebuild_ks_matrix 110 8.3 0.000 0.000 12.384 12.426 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.016 12.384 12.425 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.020 11.057 ot_mini 99 10.5 0.001 0.001 10.910 10.946 multiply_cannon_multrec 8220 15.4 3.252 4.917 7.642 8.867 init_scf_loop 11 6.9 0.000 0.000 8.236 8.237 mp_waitall_1 103326 16.6 6.479 8.101 6.479 8.101 qs_ot_get_derivative 99 11.5 0.001 0.001 6.968 7.006 sum_up_and_integrate 110 10.3 0.079 0.080 6.809 6.821 integrate_v_rspace 110 11.3 0.003 0.003 6.729 6.740 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.672 6.689 calculate_rho_elec 110 8.6 0.116 0.116 6.672 6.689 prepare_preconditioner 11 7.9 0.000 0.000 6.509 6.514 make_preconditioner 11 8.9 0.000 0.000 6.509 6.514 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.062 6.149 init_scf_run 11 5.9 0.000 0.001 5.516 5.516 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.516 5.516 qs_ot_get_p 110 10.4 0.001 0.001 5.356 5.407 dbcsr_mm_accdrv_process 17442 15.9 2.947 4.147 4.259 5.210 make_m2s 4110 13.4 0.039 0.040 4.402 4.634 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.149 4.619 make_images 4110 14.4 0.640 0.696 4.272 4.503 pw_transfer 1331 11.6 0.066 0.072 4.402 4.414 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.294 4.311 ot_diis_step 99 11.5 0.012 0.012 3.912 3.913 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.890 3.895 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.796 3.864 apply_single 110 13.6 0.000 0.000 3.796 3.863 density_rs2pw 110 9.6 0.004 0.004 3.367 3.744 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.676 3.696 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.573 3.573 fft3d_ps 1111 14.6 1.146 1.173 3.539 3.551 grid_integrate_task_list 110 12.3 3.356 3.544 3.356 3.544 cp_fm_cholesky_invert 11 10.9 3.266 3.270 3.266 3.270 multiply_cannon_sync_h2d 8220 15.4 2.904 3.061 2.904 3.061 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.994 3.002 cp_fm_diag_elpa_base 48 14.0 2.938 2.958 2.993 3.000 make_images_data 4110 15.4 0.038 0.044 2.525 2.931 wfi_extrapolate 11 7.9 0.001 0.001 2.868 2.868 hybrid_alltoall_any 4261 16.3 0.200 0.858 2.428 2.865 rs_pw_transfer 902 11.9 0.010 0.011 2.379 2.785 qs_energies_init_hamiltonians 11 5.9 0.018 0.021 2.732 2.733 calculate_dm_sparse 110 9.5 0.001 0.001 2.611 2.661 calculate_first_density_matrix 1 7.0 0.001 0.001 2.549 2.550 grid_collocate_task_list 110 9.6 2.280 2.538 2.280 2.538 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.534 2.536 potential_pw2rs 110 12.3 0.016 0.016 2.296 2.300 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.160 2.176 mp_alltoall_d11v 2046 13.8 1.805 2.118 1.805 2.118 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.054 2.072 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.981 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.916 1.934 mp_allgather_i34 2055 14.4 0.606 1.760 0.606 1.760 cp_fm_cholesky_decompose 22 10.9 1.748 1.759 1.748 1.759 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.702 1.706 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.553 1.679 dbcsr_complete_redistribute 325 12.2 0.566 0.586 1.535 1.628 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.550 mp_waitany 9240 13.8 1.116 1.533 1.116 1.533 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.902 1.419 qs_create_task_list 11 7.9 0.031 0.032 1.261 1.369 generate_qs_task_list 11 8.9 0.377 0.445 1.230 1.338 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.478 1.315 mp_alltoall_z22v 1111 16.6 1.283 1.315 1.283 1.315 rs_gather_matrices 110 12.3 0.327 0.375 0.984 1.288 mp_irecv_dv 24056 15.7 0.455 1.273 0.455 1.273 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.815 1.220 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.187 1.211 x_to_yz 550 16.1 0.572 0.581 1.193 1.199 yz_to_x 561 15.1 0.524 0.536 1.186 1.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.463000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=775.454545, yerr=10.030532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 1.335542E+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 0.020 0.045 90.534 90.535 qs_mol_dyn_low 1 2.0 0.003 0.004 90.162 90.172 qs_forces 11 3.9 0.003 0.003 90.071 90.072 qs_energies 11 4.9 0.002 0.002 85.826 85.828 scf_env_do_scf 11 5.9 0.002 0.002 75.657 75.658 velocity_verlet 10 3.0 0.002 0.002 57.732 57.741 scf_env_do_scf_inner_loop 99 6.5 0.005 0.008 46.316 46.318 dbcsr_multiply_generic 2055 12.4 0.118 0.121 29.238 29.320 init_scf_loop 11 6.9 0.000 0.000 29.264 29.267 qs_scf_new_mos 99 7.5 0.001 0.001 27.308 27.337 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.308 27.337 prepare_preconditioner 11 7.9 0.000 0.000 27.178 27.184 make_preconditioner 11 8.9 0.000 0.000 27.178 27.184 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.377 26.634 ot_scf_mini 99 9.5 0.002 0.002 25.506 25.529 multiply_cannon 2055 13.4 0.348 0.367 21.746 22.603 multiply_cannon_loop 2055 14.4 0.341 0.347 19.886 20.176 cp_fm_upper_to_full 70 14.2 12.868 18.376 12.868 18.376 rebuild_ks_matrix 110 8.3 0.001 0.001 14.624 14.676 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.016 14.623 14.676 ot_mini 99 10.5 0.001 0.001 14.017 14.044 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.241 13.289 dbcsr_complete_redistribute 325 12.2 1.018 1.045 7.422 10.768 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.367 9.702 multiply_cannon_multrec 8220 15.4 4.373 4.572 9.563 9.669 qs_ot_get_derivative 99 11.5 0.001 0.001 9.325 9.351 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.786 9.079 mp_waitall_1 84994 16.7 8.104 8.970 8.104 8.970 mp_alltoall_i22 605 13.7 5.426 8.765 5.426 8.765 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.063 8.096 calculate_rho_elec 110 8.6 0.228 0.228 8.063 8.096 sum_up_and_integrate 110 10.3 0.150 0.152 7.614 7.627 integrate_v_rspace 110 11.3 0.004 0.004 7.463 7.477 qs_ot_get_p 110 10.4 0.001 0.001 6.167 6.196 make_m2s 4110 13.4 0.043 0.044 5.614 6.174 make_images 4110 14.4 0.882 0.917 5.426 5.986 cp_fm_cholesky_invert 11 10.9 5.849 5.854 5.849 5.854 init_scf_run 11 5.9 0.000 0.001 5.811 5.811 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.811 5.811 pw_transfer 1331 11.6 0.075 0.076 5.571 5.576 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.454 5.459 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.971 5.329 dbcsr_mm_accdrv_process 11614 15.7 3.304 3.846 5.047 5.283 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.788 5.236 apply_single 110 13.6 0.000 0.000 4.788 5.236 fft_wrap_pw1pw2_140 451 13.1 0.222 0.225 4.717 4.726 ot_diis_step 99 11.5 0.015 0.016 4.661 4.662 fft3d_ps 1111 14.6 1.299 1.310 4.645 4.653 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.503 4.509 density_rs2pw 110 9.6 0.004 0.004 4.050 4.082 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.030 4.030 hybrid_alltoall_any 4261 16.3 0.258 0.553 3.136 3.961 multiply_cannon_sync_h2d 8220 15.4 3.949 3.959 3.949 3.959 make_images_data 4110 15.4 0.042 0.045 3.168 3.882 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.754 3.755 grid_integrate_task_list 110 12.3 3.675 3.732 3.675 3.732 wfi_extrapolate 11 7.9 0.001 0.001 3.425 3.425 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.950 3.403 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.384 3.384 cp_fm_diag_elpa_base 48 14.0 2.830 3.033 3.381 3.381 calculate_dm_sparse 110 9.5 0.001 0.001 3.171 3.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.940 2.946 grid_collocate_task_list 110 9.6 2.643 2.670 2.643 2.670 potential_pw2rs 110 12.3 0.021 0.022 2.661 2.667 rs_pw_transfer 902 11.9 0.011 0.011 2.504 2.542 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.512 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.401 2.421 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.266 2.308 calculate_first_density_matrix 1 7.0 0.001 0.001 2.262 2.263 mp_alltoall_d11v 2046 13.8 2.109 2.191 2.109 2.191 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.177 cp_fm_cholesky_decompose 22 10.9 2.150 2.166 2.150 2.166 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.989 1.995 qs_create_task_list 11 7.9 0.000 0.001 1.894 1.942 generate_qs_task_list 11 8.9 0.730 0.783 1.893 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.828 1.847 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.535000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1209.909091, yerr=53.022216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/15/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 1.094965E+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 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 628.850688E+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 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 2520945. 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 0.016 0.032 209.237 209.239 qs_mol_dyn_low 1 2.0 0.004 0.009 208.823 208.837 qs_forces 11 3.9 0.005 0.005 208.722 208.724 qs_energies 11 4.9 0.002 0.003 203.039 203.052 scf_env_do_scf 11 5.9 0.001 0.002 186.122 186.126 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 164.678 164.680 velocity_verlet 10 3.0 0.001 0.003 126.186 126.188 qs_scf_new_mos 117 7.6 0.001 0.001 123.196 123.478 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.195 123.477 dbcsr_multiply_generic 2507 12.6 0.177 0.181 122.822 123.431 ot_scf_mini 117 9.6 0.003 0.003 116.666 116.959 multiply_cannon 2507 13.6 0.241 0.250 100.539 101.758 multiply_cannon_loop 2507 14.6 1.994 2.101 98.219 99.661 ot_mini 117 10.6 0.001 0.001 64.178 64.466 multiply_cannon_multrec 60168 15.6 33.923 35.899 41.950 43.478 qs_ot_get_derivative 117 11.6 0.001 0.001 39.314 39.576 rebuild_ks_matrix 128 8.3 0.001 0.001 34.688 34.939 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 34.687 34.938 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.177 31.429 qs_ot_get_p 128 10.4 0.001 0.001 30.696 30.982 multiply_cannon_sync_h2d 60168 15.6 28.220 30.315 28.220 30.315 mp_waitall_1 267128 16.5 27.477 30.229 27.477 30.229 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.427 25.105 apply_single 128 13.6 0.001 0.001 24.427 25.104 ot_diis_step 117 11.6 0.008 0.008 24.667 24.669 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.120 24.196 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.701 21.702 init_scf_loop 11 6.9 0.000 0.000 21.358 21.360 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.547 18.582 cp_fm_diag_elpa_base 83 14.4 18.463 18.502 18.542 18.575 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 17.790 17.983 multiply_cannon_metrocomm3 60168 15.6 0.115 0.120 14.740 17.479 prepare_preconditioner 11 7.9 0.000 0.000 16.591 16.638 make_preconditioner 11 8.9 0.000 0.000 16.591 16.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.790 15.970 sum_up_and_integrate 128 10.3 0.089 0.109 14.925 14.940 integrate_v_rspace 128 11.3 0.004 0.005 14.835 14.854 make_m2s 5014 13.6 0.103 0.110 14.194 14.640 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.470 14.555 calculate_rho_elec 128 8.7 0.046 0.065 14.470 14.554 make_images 5014 14.6 0.401 0.420 14.013 14.469 init_scf_run 11 5.9 0.000 0.001 12.670 12.671 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.670 12.671 density_rs2pw 128 9.7 0.006 0.007 7.756 10.960 rs_pw_transfer 1046 11.9 0.016 0.019 6.362 9.581 cp_fm_cholesky_invert 11 10.9 9.541 9.548 9.541 9.548 wfi_extrapolate 11 7.9 0.001 0.001 9.309 9.310 pw_transfer 1547 11.6 0.077 0.099 8.549 8.833 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.342 8.634 calculate_dm_sparse 128 9.5 0.001 0.001 8.356 8.454 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.218 8.361 dbcsr_mm_accdrv_process 124484 16.2 2.879 3.218 7.562 8.182 mp_sum_l 7870 13.0 6.815 8.178 6.815 8.178 make_images_data 5014 15.6 0.065 0.071 7.005 7.913 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.656 7.765 grid_integrate_task_list 128 12.3 7.025 7.525 7.025 7.525 hybrid_alltoall_any 5200 16.5 0.292 2.270 6.125 7.464 fft_wrap_pw1pw2_140 523 13.2 0.442 0.511 7.094 7.455 fft3d_ps 1291 14.7 2.103 2.870 7.127 7.294 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 5.678 7.279 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.848 6.856 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.955 6.033 mp_waitany 16020 13.9 2.725 6.010 2.725 6.010 mp_alltoall_d11v 2415 14.1 4.545 5.813 4.545 5.813 grid_collocate_task_list 128 9.7 4.564 5.792 4.564 5.792 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.294 2.118 5.347 potential_pw2rs 128 12.3 0.010 0.012 5.215 5.259 cp_fm_cholesky_decompose 22 10.9 4.801 4.815 4.801 4.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.239000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.818182, yerr=6.630757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 827.707392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11058 964. MP_Sync 87 MP_Alltoall 1983 5621256. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.029 195.911 195.911 qs_mol_dyn_low 1 2.0 0.003 0.003 195.574 195.592 qs_forces 11 3.9 0.004 0.005 195.438 195.439 qs_energies 11 4.9 0.002 0.002 188.536 188.549 scf_env_do_scf 11 5.9 0.001 0.001 171.890 171.900 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 137.937 137.940 velocity_verlet 10 3.0 0.012 0.027 123.681 123.698 dbcsr_multiply_generic 2527 12.6 0.188 0.201 98.547 99.869 qs_scf_new_mos 118 7.6 0.001 0.001 96.634 97.148 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.633 97.147 ot_scf_mini 118 9.6 0.004 0.004 91.767 92.326 multiply_cannon 2527 13.6 0.481 0.536 77.991 83.061 multiply_cannon_loop 2527 14.6 1.255 1.293 74.813 77.760 ot_mini 118 10.6 0.001 0.001 50.681 51.200 mp_waitall_1 216438 16.6 25.632 39.180 25.632 39.180 multiply_cannon_multrec 30324 15.6 21.803 25.502 31.443 35.682 rebuild_ks_matrix 129 8.3 0.001 0.001 34.075 34.535 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 34.074 34.535 init_scf_loop 11 6.9 0.000 0.000 33.858 33.859 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.690 31.122 prepare_preconditioner 11 7.9 0.000 0.000 29.411 29.467 make_preconditioner 11 8.9 0.000 0.000 29.410 29.467 qs_ot_get_derivative 118 11.6 0.001 0.002 28.448 28.996 multiply_cannon_metrocomm3 30324 15.6 0.091 0.097 16.502 28.912 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.037 28.575 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.305 23.487 apply_single 129 13.6 0.001 0.001 22.304 23.486 qs_ot_get_p 129 10.4 0.001 0.001 22.820 23.474 ot_diis_step 118 11.6 0.015 0.015 22.058 22.060 multiply_cannon_sync_h2d 30324 15.6 19.076 21.268 19.076 21.268 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.835 17.878 cp_fm_cholesky_invert 11 10.9 17.191 17.204 17.191 17.204 cp_dbcsr_syevd 83 12.4 0.006 0.006 16.656 16.657 make_m2s 5054 13.6 0.090 0.095 14.581 15.941 make_images 5054 14.6 1.173 1.361 14.372 15.734 sum_up_and_integrate 129 10.3 0.117 0.132 15.332 15.365 integrate_v_rspace 129 11.3 0.004 0.004 15.215 15.249 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.635 14.664 calculate_rho_elec 129 8.7 0.089 0.107 14.634 14.663 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.376 13.407 cp_fm_diag_elpa_base 83 14.4 13.114 13.206 13.369 13.397 init_scf_run 11 5.9 0.000 0.001 11.812 11.813 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.812 11.813 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.118 11.504 multiply_cannon_metrocomm4 27797 15.6 0.099 0.114 3.883 10.838 make_images_data 5054 15.6 0.064 0.072 8.871 10.558 mp_irecv_dv 70031 16.3 3.686 10.442 3.686 10.442 density_rs2pw 129 9.7 0.006 0.006 7.920 10.348 hybrid_alltoall_any 5240 16.5 0.346 1.519 7.580 9.824 dbcsr_mm_accdrv_process 62734 16.2 4.443 5.338 9.108 9.708 pw_transfer 1559 11.6 0.088 0.107 9.422 9.516 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.194 9.295 rs_pw_transfer 1054 12.0 0.014 0.016 6.113 8.547 wfi_extrapolate 11 7.9 0.001 0.001 8.490 8.490 fft_wrap_pw1pw2_140 527 13.2 0.476 0.532 8.039 8.152 grid_integrate_task_list 129 12.3 7.224 7.721 7.224 7.721 fft3d_ps 1301 14.7 2.777 2.957 7.549 7.604 cp_fm_cholesky_decompose 22 10.9 7.202 7.289 7.202 7.289 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.424 7.144 calculate_dm_sparse 129 9.5 0.001 0.001 6.522 6.663 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.368 6.378 mp_sum_l 7930 13.1 4.183 6.307 4.183 6.307 grid_collocate_task_list 129 9.7 4.744 5.931 4.744 5.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.490 5.636 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.426 5.502 potential_pw2rs 129 12.3 0.015 0.018 5.429 5.459 mp_alltoall_d11v 2423 14.1 4.399 5.183 4.399 5.183 mp_waitany 11836 13.9 2.509 5.010 2.509 5.010 mp_allgather_i34 2527 14.6 1.746 4.795 1.746 4.795 rs_pw_transfer_RS2PW_140 140 11.5 0.355 0.391 2.095 4.519 dbcsr_complete_redistribute 395 12.7 0.770 0.865 3.316 4.177 mp_sum_d 4499 12.1 2.703 4.124 2.703 4.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.911000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.636364, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/17/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 2.928533E+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 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 953.327616E+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 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 10977 1007. 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 0.018 0.037 194.724 194.725 qs_mol_dyn_low 1 2.0 0.003 0.004 194.294 194.313 qs_forces 11 3.9 0.004 0.005 193.955 193.960 qs_energies 11 4.9 0.002 0.002 187.119 187.131 scf_env_do_scf 11 5.9 0.001 0.002 170.970 170.970 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.027 131.028 velocity_verlet 10 3.0 0.006 0.012 124.382 124.397 qs_scf_new_mos 117 7.6 0.001 0.001 90.307 90.790 qs_scf_loop_do_ot 117 8.6 0.001 0.001 90.306 90.789 ot_scf_mini 117 9.6 0.003 0.004 86.048 86.538 dbcsr_multiply_generic 2507 12.6 0.179 0.183 84.774 86.103 multiply_cannon 2507 13.6 0.499 0.517 63.271 69.055 multiply_cannon_loop 2507 14.6 0.852 0.875 59.728 63.396 ot_mini 117 10.6 0.001 0.001 45.232 45.749 init_scf_loop 11 6.9 0.000 0.000 39.834 39.836 mp_waitall_1 170520 16.6 27.682 39.201 27.682 39.201 prepare_preconditioner 11 7.9 0.000 0.000 35.611 35.667 make_preconditioner 11 8.9 0.000 0.000 35.611 35.667 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.160 34.620 rebuild_ks_matrix 128 8.3 0.001 0.001 32.276 32.857 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 32.276 32.857 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.106 29.641 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 16.722 28.132 qs_ot_get_p 128 10.4 0.001 0.001 25.803 26.341 qs_ot_get_derivative 117 11.6 0.002 0.002 25.125 25.633 multiply_cannon_multrec 20056 15.6 13.226 16.409 21.925 25.124 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.284 21.550 apply_single 128 13.6 0.001 0.001 20.284 21.550 qs_ot_p2m_diag 83 11.4 0.266 0.272 21.012 21.051 ot_diis_step 117 11.6 0.018 0.018 19.980 19.980 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.819 19.822 cp_fm_cholesky_invert 11 10.9 17.468 17.482 17.468 17.482 make_m2s 5014 13.6 0.081 0.085 15.635 16.726 make_images 5014 14.6 1.185 1.286 15.404 16.498 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.041 16.051 cp_fm_diag_elpa_base 83 14.4 15.584 15.755 16.036 16.046 sum_up_and_integrate 128 10.3 0.133 0.144 15.867 15.893 integrate_v_rspace 128 11.3 0.004 0.004 15.733 15.762 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.638 15.666 calculate_rho_elec 128 8.7 0.132 0.147 15.637 15.666 multiply_cannon_sync_h2d 20056 15.6 13.960 15.600 13.960 15.600 make_images_data 5014 15.6 0.060 0.067 9.598 11.233 init_scf_run 11 5.9 0.000 0.001 10.991 10.991 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.991 10.991 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.990 10.349 density_rs2pw 128 9.7 0.006 0.006 8.404 10.190 hybrid_alltoall_any 5200 16.5 0.433 1.990 8.362 10.160 pw_transfer 1547 11.6 0.087 0.106 9.864 9.997 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.638 9.777 multiply_cannon_metrocomm4 17549 15.6 0.063 0.074 3.545 9.575 mp_irecv_dv 50230 16.2 3.423 9.327 3.423 9.327 cp_fm_cholesky_decompose 22 10.9 8.906 8.955 8.906 8.955 dbcsr_mm_accdrv_process 41502 16.2 4.462 5.279 8.187 8.426 rs_pw_transfer 1046 11.9 0.014 0.015 6.620 8.415 fft_wrap_pw1pw2_140 523 13.2 0.473 0.518 8.221 8.368 fft3d_ps 1291 14.7 2.676 2.912 7.901 7.990 wfi_extrapolate 11 7.9 0.001 0.001 7.970 7.971 grid_integrate_task_list 128 12.3 7.273 7.787 7.273 7.787 cp_fm_upper_to_full 105 14.8 5.810 7.427 5.810 7.427 dbcsr_complete_redistribute 395 12.7 1.167 1.192 4.868 6.727 mp_sum_l 7870 13.0 4.068 5.978 4.068 5.978 calculate_dm_sparse 128 9.5 0.001 0.001 5.872 5.950 potential_pw2rs 128 12.3 0.021 0.023 5.846 5.873 grid_collocate_task_list 128 9.7 4.899 5.833 4.899 5.833 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.997 5.792 mp_alltoall_d11v 2415 14.1 4.729 5.683 4.729 5.683 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.583 5.589 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.534 5.388 mp_allgather_i34 2507 14.6 2.069 5.229 2.069 5.229 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.585 4.721 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.304 4.355 mp_waitany 11748 13.9 2.502 4.325 2.502 4.325 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.432 4.215 mp_alltoall_i22 716 14.1 2.079 4.085 2.079 4.085 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=194.725000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=904.181818, yerr=8.653094 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.128591E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58138. MP_Allreduce 11015 1090. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.075 190.388 190.389 qs_mol_dyn_low 1 2.0 0.003 0.004 189.902 189.914 qs_forces 11 3.9 0.005 0.007 189.774 189.781 qs_energies 11 4.9 0.003 0.009 182.542 182.549 scf_env_do_scf 11 5.9 0.001 0.002 165.720 165.726 velocity_verlet 10 3.0 0.002 0.002 125.496 125.500 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 118.716 118.717 qs_scf_new_mos 116 7.6 0.001 0.001 80.751 81.101 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.751 81.100 dbcsr_multiply_generic 2485 12.5 0.184 0.189 78.614 79.342 ot_scf_mini 116 9.6 0.003 0.004 76.284 76.597 multiply_cannon 2485 13.5 0.550 0.590 54.087 57.216 multiply_cannon_loop 2485 14.5 1.167 1.197 50.354 52.131 init_scf_loop 11 6.9 0.001 0.012 46.874 46.876 prepare_preconditioner 11 7.9 0.000 0.001 42.685 42.714 make_preconditioner 11 8.9 0.001 0.004 42.685 42.714 ot_mini 116 10.6 0.001 0.001 41.934 42.241 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.343 41.354 multiply_cannon_multrec 29820 15.5 14.138 19.111 25.966 30.532 rebuild_ks_matrix 127 8.3 0.001 0.001 30.163 30.470 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.025 30.162 30.469 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.153 27.422 mp_waitall_1 146592 16.7 17.195 27.355 17.195 27.355 qs_ot_get_derivative 116 11.6 0.001 0.002 22.529 22.842 make_m2s 4970 13.5 0.096 0.101 20.262 21.361 make_images 4970 14.5 1.929 2.213 19.959 21.060 qs_ot_get_p 127 10.4 0.001 0.001 20.742 21.034 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.847 19.345 apply_single 127 13.6 0.001 0.001 18.847 19.345 ot_diis_step 116 11.6 0.017 0.019 19.290 19.291 cp_fm_upper_to_full 104 14.8 11.536 16.947 11.536 16.947 qs_ot_p2m_diag 82 11.4 0.339 0.385 16.568 16.620 cp_fm_cholesky_invert 11 10.9 16.403 16.412 16.403 16.412 cp_dbcsr_syevd 82 12.4 0.006 0.006 15.295 15.296 sum_up_and_integrate 127 10.3 0.139 0.150 14.950 14.975 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.197 14.939 integrate_v_rspace 127 11.3 0.004 0.004 14.811 14.841 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.779 14.822 calculate_rho_elec 127 8.7 0.174 0.189 14.778 14.822 make_images_data 4970 15.5 0.063 0.068 10.981 12.948 dbcsr_complete_redistribute 393 12.7 1.515 1.632 8.957 12.748 multiply_cannon_sync_h2d 29820 15.5 11.572 12.550 11.572 12.550 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.130 12.142 cp_fm_diag_elpa_base 82 14.4 11.142 11.457 12.123 12.138 hybrid_alltoall_any 5155 16.4 0.522 2.190 9.738 12.100 dbcsr_mm_accdrv_process 61748 16.2 7.230 7.969 11.404 11.945 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.570 11.357 init_scf_run 11 5.9 0.000 0.001 10.816 10.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.816 10.817 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.319 10.017 pw_transfer 1535 11.6 0.088 0.107 9.670 9.742 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.295 9.545 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.442 9.521 mp_alltoall_i22 712 14.1 5.576 9.316 5.576 9.316 density_rs2pw 127 9.7 0.006 0.007 7.279 8.763 fft_wrap_pw1pw2_140 519 13.2 0.480 0.496 8.339 8.432 grid_integrate_task_list 127 12.3 7.413 7.971 7.413 7.971 cp_fm_cholesky_decompose 22 10.9 7.694 7.852 7.694 7.852 fft3d_ps 1281 14.7 2.756 2.829 7.678 7.731 wfi_extrapolate 11 7.9 0.001 0.001 7.704 7.704 multiply_cannon_metrocomm4 24850 15.5 0.083 0.094 2.714 6.524 rs_pw_transfer 1038 11.9 0.013 0.015 4.839 6.383 mp_irecv_dv 75445 16.2 2.564 6.255 2.564 6.255 calculate_dm_sparse 127 9.5 0.001 0.001 6.132 6.227 mp_alltoall_d11v 2401 14.1 4.919 5.728 4.919 5.728 grid_collocate_task_list 127 9.7 5.030 5.703 5.030 5.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.411 5.441 potential_pw2rs 127 12.3 0.023 0.024 4.791 4.817 qs_energies_init_hamiltonians 11 5.9 0.002 0.009 4.535 4.536 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.392 4.493 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.322 4.413 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.116 4.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.389000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1063.090909, yerr=14.228362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/19/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 5.865089E+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 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 1.513624E+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 5.532238E+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 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. 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 0.018 0.036 189.642 189.643 qs_mol_dyn_low 1 2.0 0.003 0.004 189.244 189.256 qs_forces 11 3.9 0.004 0.005 189.137 189.140 qs_energies 11 4.9 0.002 0.003 181.413 181.420 scf_env_do_scf 11 5.9 0.005 0.006 163.594 163.599 scf_env_do_scf_inner_loop 117 6.6 0.004 0.010 124.139 124.141 velocity_verlet 10 3.0 0.002 0.002 123.305 123.309 qs_scf_new_mos 117 7.6 0.001 0.001 83.464 83.584 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.464 83.583 ot_scf_mini 117 9.6 0.003 0.003 78.943 79.057 dbcsr_multiply_generic 2507 12.6 0.180 0.189 76.566 77.364 multiply_cannon 2507 13.6 0.582 0.617 55.631 59.172 multiply_cannon_loop 2507 14.6 0.442 0.454 50.753 52.465 ot_mini 117 10.6 0.001 0.001 41.657 41.791 init_scf_loop 11 6.9 0.000 0.000 39.288 39.290 prepare_preconditioner 11 7.9 0.000 0.000 35.208 35.233 make_preconditioner 11 8.9 0.000 0.000 35.208 35.233 mp_waitall_1 125778 16.7 27.370 34.786 27.370 34.786 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.103 33.418 rebuild_ks_matrix 128 8.3 0.001 0.001 31.135 31.348 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.023 31.134 31.347 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.298 28.496 qs_ot_get_p 128 10.4 0.001 0.001 24.117 24.245 qs_ot_get_derivative 117 11.6 0.002 0.002 21.453 21.564 multiply_cannon_multrec 10028 15.6 10.437 15.186 17.935 21.077 cp_fm_cholesky_invert 11 10.9 20.626 20.633 20.626 20.633 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.836 20.518 ot_diis_step 117 11.6 0.020 0.022 20.098 20.098 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.540 20.051 apply_single 128 13.6 0.001 0.001 19.540 20.051 qs_ot_p2m_diag 83 11.4 0.496 0.501 19.931 19.948 make_m2s 5014 13.6 0.067 0.069 16.794 18.952 make_images 5014 14.6 2.281 2.751 16.492 18.649 cp_dbcsr_syevd 83 12.4 0.006 0.007 18.634 18.636 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.460 16.511 calculate_rho_elec 128 8.7 0.259 0.269 16.459 16.510 sum_up_and_integrate 128 10.3 0.180 0.191 16.046 16.091 integrate_v_rspace 128 11.3 0.004 0.004 15.866 15.919 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.083 15.091 cp_fm_diag_elpa_base 83 14.4 14.826 14.908 15.069 15.077 make_images_data 5014 15.6 0.052 0.060 10.134 12.526 hybrid_alltoall_any 5200 16.5 0.843 3.815 9.850 12.270 multiply_cannon_sync_h2d 10028 15.6 11.545 12.099 11.545 12.099 init_scf_run 11 5.9 0.000 0.001 10.913 10.914 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.913 10.914 pw_transfer 1547 11.6 0.086 0.094 10.577 10.629 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 10.353 10.411 cp_fm_cholesky_decompose 22 10.9 8.911 9.030 8.911 9.030 fft_wrap_pw1pw2_140 523 13.2 0.494 0.518 8.928 8.983 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.868 8.965 density_rs2pw 128 9.7 0.005 0.005 7.871 8.742 fft3d_ps 1291 14.7 2.716 2.784 8.515 8.550 grid_integrate_task_list 128 12.3 7.779 8.202 7.779 8.202 dbcsr_mm_accdrv_process 20762 16.1 2.721 3.537 7.129 7.816 wfi_extrapolate 11 7.9 0.001 0.001 7.816 7.816 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.793 7.622 mp_allgather_i34 2507 14.6 3.086 7.254 3.086 7.254 calculate_dm_sparse 128 9.5 0.001 0.001 6.252 6.343 grid_collocate_task_list 128 9.7 5.395 6.099 5.395 6.099 mp_alltoall_d11v 2415 14.1 5.529 6.053 5.529 6.053 rs_pw_transfer 1046 11.9 0.012 0.013 5.063 5.955 dbcsr_complete_redistribute 395 12.7 2.116 2.160 5.425 5.827 potential_pw2rs 128 12.3 0.027 0.028 5.302 5.327 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 5.320 5.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.304 5.310 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.860 4.650 mp_irecv_dv 28860 15.9 1.825 4.560 1.825 4.560 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.131 4.167 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.950 3.991 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.860 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.763 3.859 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.480 3.810 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.776 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=189.643000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1415.636364, yerr=55.181144 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.036778E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108623056 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954382. 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 0.031 0.064 298.906 298.916 qs_mol_dyn_low 1 2.0 0.003 0.003 298.376 298.383 qs_forces 11 3.9 0.005 0.005 298.267 298.273 qs_energies 11 4.9 0.002 0.002 288.899 288.913 scf_env_do_scf 11 5.9 0.002 0.002 266.578 266.593 velocity_verlet 10 3.0 0.002 0.002 215.916 215.924 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 139.694 139.697 init_scf_loop 11 6.9 0.000 0.000 126.626 126.632 prepare_preconditioner 11 7.9 0.000 0.000 121.463 121.494 make_preconditioner 11 8.9 0.000 0.000 121.463 121.493 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.342 118.629 qs_scf_new_mos 116 7.6 0.001 0.001 89.200 89.315 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.199 89.314 ot_scf_mini 116 9.6 0.004 0.004 84.493 84.539 dbcsr_multiply_generic 2485 12.5 0.209 0.220 82.224 82.795 cp_fm_upper_to_full 104 14.8 53.975 78.194 53.975 78.194 multiply_cannon 2485 13.5 0.711 0.762 57.990 58.516 multiply_cannon_loop 2485 14.5 0.463 0.467 54.427 55.734 dbcsr_complete_redistribute 393 12.7 3.987 4.032 30.792 45.031 ot_mini 116 10.6 0.001 0.001 43.940 43.983 copy_fm_to_dbcsr 208 11.6 0.002 0.002 27.393 41.606 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.075 39.130 rebuild_ks_matrix 127 8.3 0.001 0.001 38.094 38.145 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.021 38.094 38.144 mp_alltoall_i22 712 14.1 22.922 37.270 22.922 37.270 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.017 35.068 cp_fm_cholesky_invert 11 10.9 33.280 33.286 33.280 33.286 mp_waitall_1 102768 16.8 27.785 31.677 27.785 31.677 qs_ot_get_p 127 10.4 0.001 0.001 25.290 25.304 qs_ot_get_derivative 116 11.6 0.002 0.002 24.135 24.181 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.327 21.353 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.712 20.729 calculate_rho_elec 127 8.7 0.480 0.481 20.711 20.729 make_m2s 4970 13.5 0.076 0.078 19.640 20.535 make_images 4970 14.5 3.727 3.873 19.166 20.066 ot_diis_step 116 11.6 0.022 0.022 19.781 19.781 multiply_cannon_metrocomm3 9940 15.5 0.022 0.024 18.558 19.748 cp_dbcsr_syevd 82 12.4 0.007 0.007 19.637 19.638 sum_up_and_integrate 127 10.3 0.318 0.321 19.458 19.548 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.143 19.349 apply_single 127 13.6 0.001 0.001 19.143 19.348 integrate_v_rspace 127 11.3 0.004 0.004 19.139 19.227 multiply_cannon_multrec 9940 15.5 10.400 12.147 17.810 17.875 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.514 16.515 cp_fm_diag_elpa_base 82 14.4 12.043 13.747 16.510 16.511 multiply_cannon_sync_h2d 9940 15.5 15.537 15.547 15.537 15.547 hybrid_alltoall_any 5155 16.4 1.291 3.022 10.666 12.540 pw_transfer 1535 11.6 0.092 0.093 12.390 12.395 init_scf_run 11 5.9 0.000 0.001 12.224 12.224 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.224 12.224 make_images_data 4970 15.5 0.060 0.064 10.464 12.179 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.157 12.162 fft_wrap_pw1pw2_140 519 13.2 0.538 0.542 10.808 10.815 fft3d_ps 1281 14.7 2.726 2.737 10.207 10.212 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.258 9.298 wfi_extrapolate 11 7.9 0.001 0.001 9.112 9.112 dbcsr_mm_accdrv_process 20590 16.0 3.689 5.599 7.174 8.996 mp_alltoall_d11v 2401 14.1 8.076 8.952 8.076 8.952 cp_fm_cholesky_decompose 22 10.9 8.895 8.928 8.895 8.928 grid_integrate_task_list 127 12.3 8.519 8.696 8.519 8.696 density_rs2pw 127 9.7 0.005 0.005 8.350 8.467 qs_energies_init_hamiltonians 11 5.9 0.003 0.003 8.136 8.136 calculate_dm_sparse 127 9.5 0.001 0.001 6.547 6.636 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.300 6.350 grid_collocate_task_list 127 9.7 6.316 6.340 6.316 6.340 rs_scatter_matrices 138 9.7 3.585 4.548 6.030 6.259 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.075 6.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.916000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2732.727273, yerr=161.077672 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 1.262113E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.118 85.538 85.539 qs_energies 1 2.0 0.000 0.000 85.019 85.031 ls_scf 1 3.0 0.000 0.000 84.114 84.126 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.867 73.046 multiply_cannon 111 7.7 0.018 0.021 56.150 57.631 multiply_cannon_loop 111 8.7 0.206 0.219 52.712 54.338 ls_scf_main 1 4.0 0.000 0.000 52.566 52.577 density_matrix_trs4 2 5.0 0.002 0.003 46.932 47.010 ls_scf_init_scf 1 4.0 0.000 0.001 28.484 28.485 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.356 27.418 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.253 25.282 mp_waitall_1 11031 10.9 22.703 25.196 22.703 25.196 multiply_cannon_multrec 2664 9.7 8.137 8.853 15.541 17.197 multiply_cannon_sync_h2d 2664 9.7 13.594 15.929 13.594 15.929 make_m2s 222 7.7 0.008 0.010 13.064 13.615 make_images 222 8.7 0.099 0.110 13.042 13.593 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.988 12.893 make_images_data 222 9.7 0.004 0.005 7.605 8.154 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.428 7.964 dbcsr_mm_accdrv_process 4760 10.4 0.517 0.637 7.021 7.945 hybrid_alltoall_any 227 10.6 0.215 1.835 6.493 7.734 dbcsr_mm_accdrv_process_sort 4760 11.4 6.304 7.175 6.304 7.175 calculate_norms 4752 9.8 5.514 6.116 5.514 6.116 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.123 5.239 mp_sum_l 807 5.4 3.123 4.400 3.123 4.400 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.060 3.807 mp_irecv_dv 6231 10.9 2.044 3.780 2.044 3.780 make_images_sizes 222 9.7 0.000 0.000 0.775 3.494 mp_alltoall_i44 222 10.7 0.774 3.493 0.774 3.493 arnoldi_extremal 4 6.8 0.000 0.000 3.443 3.468 arnoldi_normal_ev 4 7.8 0.001 0.003 3.443 3.468 build_subspace 16 8.4 0.009 0.012 3.344 3.347 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.323 3.276 ls_scf_post 1 4.0 0.000 0.000 3.064 3.076 ls_scf_store_result 1 5.0 0.000 0.000 2.883 2.920 dbcsr_special_finalize 555 9.7 0.005 0.006 2.330 2.822 dbcsr_merge_single_wm 555 10.7 0.467 0.598 2.321 2.813 make_images_pack 222 9.7 2.208 2.627 2.209 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.382 2.614 dbcsr_sort_data 658 11.4 2.111 2.505 2.111 2.505 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.343 2.465 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.460 2.067 2.461 buffer_matrices_ensure_size 222 8.7 1.767 2.084 1.767 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.812 1.813 rebuild_ks_matrix 3 7.3 0.000 0.000 1.802 1.804 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.802 1.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.539000, 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/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 2.102510E+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 0.034 0.056 92.011 92.012 qs_energies 1 2.0 0.000 0.000 91.461 91.466 ls_scf 1 3.0 0.000 0.001 90.124 90.128 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.611 75.934 multiply_cannon 111 7.7 0.028 0.041 53.295 58.006 ls_scf_main 1 4.0 0.000 0.001 55.657 55.663 multiply_cannon_loop 111 8.7 0.116 0.124 49.997 53.910 density_matrix_trs4 2 5.0 0.002 0.007 49.852 50.068 ls_scf_init_scf 1 4.0 0.000 0.002 30.954 30.956 mp_waitall_1 9105 10.9 21.048 29.923 21.048 29.923 ls_scf_init_matrix_S 1 5.0 0.000 0.004 29.735 29.841 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.312 27.331 multiply_cannon_multrec 1332 9.7 13.050 16.135 22.348 26.368 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.829 22.007 make_m2s 222 7.7 0.006 0.008 15.100 15.886 make_images 222 8.7 1.581 1.948 15.070 15.857 dbcsr_mm_accdrv_process 4041 10.4 0.262 0.430 8.903 10.427 dbcsr_mm_accdrv_process_sort 4041 11.4 8.498 10.003 8.498 10.003 make_images_data 222 9.7 0.004 0.004 8.593 9.607 mp_sum_l 807 5.4 5.936 9.203 5.936 9.203 hybrid_alltoall_any 227 10.6 0.521 2.463 7.949 8.917 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.197 7.841 mp_irecv_dv 3311 11.0 3.178 7.783 3.178 7.783 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.642 7.412 calculate_norms 2376 9.8 6.014 6.710 6.014 6.710 multiply_cannon_sync_h2d 1332 9.7 4.788 6.124 4.788 6.124 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.076 5.298 arnoldi_extremal 4 6.8 0.000 0.000 5.103 5.122 arnoldi_normal_ev 4 7.8 0.001 0.004 5.103 5.122 build_subspace 16 8.4 0.014 0.021 4.840 4.843 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.270 3.544 ls_scf_post 1 4.0 0.000 0.001 3.512 3.517 ls_scf_store_result 1 5.0 0.000 0.000 3.227 3.335 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.225 2.737 3.227 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.250 2.976 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.639 2.729 mp_allgather_i34 111 8.7 0.999 2.406 0.999 2.406 make_images_pack 222 9.7 2.021 2.361 2.023 2.363 dbcsr_sort_data 436 11.2 1.823 2.067 1.823 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 rebuild_ks_matrix 3 7.3 0.000 0.000 1.885 1.887 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.885 1.887 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.012000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1751.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 2.700485E+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 0.055 0.077 94.462 94.464 qs_energies 1 2.0 0.000 0.000 93.805 93.808 ls_scf 1 3.0 0.000 0.001 92.393 92.399 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.950 77.270 ls_scf_main 1 4.0 0.000 0.002 57.752 57.758 multiply_cannon 111 7.7 0.042 0.081 52.977 56.443 multiply_cannon_loop 111 8.7 0.099 0.104 49.423 53.311 density_matrix_trs4 2 5.0 0.002 0.008 51.778 51.965 mp_waitall_1 7281 11.0 24.064 34.450 24.064 34.450 ls_scf_init_scf 1 4.0 0.000 0.002 31.050 31.052 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.845 29.904 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.395 27.404 multiply_cannon_multrec 888 9.7 12.577 15.248 21.222 24.460 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.045 23.647 make_m2s 222 7.7 0.006 0.007 17.186 18.488 make_images 222 8.7 1.964 2.274 17.147 18.448 hybrid_alltoall_any 227 10.6 0.620 2.841 9.494 10.893 make_images_data 222 9.7 0.003 0.004 9.850 10.844 dbcsr_mm_accdrv_process 3754 10.4 0.268 0.457 8.151 9.416 dbcsr_mm_accdrv_process_sort 3754 11.4 7.772 8.959 7.772 8.959 mp_sum_l 807 5.4 5.250 8.959 5.250 8.959 multiply_cannon_sync_h2d 888 9.7 6.043 7.328 6.043 7.328 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.061 7.137 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.436 7.056 mp_irecv_dv 2335 11.1 2.422 7.016 2.422 7.016 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.791 6.875 arnoldi_extremal 4 6.8 0.000 0.000 5.130 5.144 arnoldi_normal_ev 4 7.8 0.001 0.005 5.130 5.144 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.918 5.106 build_subspace 16 8.4 0.014 0.020 4.823 4.831 calculate_norms 1584 9.8 4.267 4.741 4.267 4.741 mp_allgather_i34 111 8.7 1.391 3.880 1.391 3.880 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.417 3.751 ls_scf_post 1 4.0 0.001 0.003 3.591 3.597 dbcsr_matrix_vector_mult_local 304 10.0 3.009 3.581 3.011 3.583 ls_scf_store_result 1 5.0 0.000 0.000 3.310 3.419 ls_scf_dm_to_ks 2 5.0 0.001 0.013 2.867 2.966 make_images_sizes 222 9.7 0.000 0.000 1.046 2.208 mp_alltoall_i44 222 10.7 1.046 2.208 1.046 2.208 dbcsr_sort_data 325 11.1 1.849 2.146 1.849 2.146 make_images_pack 222 9.7 1.812 2.120 1.815 2.122 dbcsr_data_release 9322 10.9 1.332 1.982 1.332 1.982 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.950 1.951 rebuild_ks_matrix 3 7.3 0.000 0.000 1.930 1.932 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.930 1.932 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.464000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2195.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 3.322393E+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 0.049 0.106 97.200 97.201 qs_energies 1 2.0 0.000 0.000 96.488 96.495 ls_scf 1 3.0 0.000 0.001 94.823 94.829 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.352 78.571 ls_scf_main 1 4.0 0.000 0.003 58.969 58.974 multiply_cannon 111 7.7 0.068 0.184 51.677 56.139 density_matrix_trs4 2 5.0 0.002 0.004 52.727 52.850 multiply_cannon_loop 111 8.7 0.115 0.127 46.633 49.242 ls_scf_init_scf 1 4.0 0.001 0.005 32.575 32.576 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.315 31.391 mp_waitall_1 6369 11.0 22.918 29.238 22.918 29.238 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.827 28.840 multiply_cannon_multrec 1332 9.7 14.121 17.312 22.047 24.506 make_m2s 222 7.7 0.006 0.007 21.198 22.662 make_images 222 8.7 3.134 3.596 21.148 22.614 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.275 17.197 make_images_data 222 9.7 0.004 0.004 11.883 13.536 hybrid_alltoall_any 227 10.6 0.798 3.832 11.251 13.174 dbcsr_mm_accdrv_process 3641 10.4 0.178 0.225 7.568 9.100 dbcsr_mm_accdrv_process_sort 3641 11.4 7.197 8.682 7.197 8.682 mp_sum_l 807 5.4 3.911 7.272 3.911 7.272 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.093 6.184 mp_irecv_dv 3229 10.9 2.069 6.112 2.069 6.112 multiply_cannon_sync_h2d 1332 9.7 5.491 6.092 5.491 6.092 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.931 5.695 arnoldi_extremal 4 6.8 0.000 0.000 5.292 5.305 arnoldi_normal_ev 4 7.8 0.001 0.005 5.292 5.305 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.587 5.234 build_subspace 16 8.4 0.014 0.021 4.952 4.961 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.621 4.837 mp_allgather_i34 111 8.7 2.178 4.582 2.178 4.582 calculate_norms 2376 9.8 4.195 4.481 4.195 4.481 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.611 3.931 dbcsr_matrix_vector_mult_local 304 10.0 3.175 3.675 3.177 3.677 dbcsr_sort_data 658 11.4 3.136 3.485 3.136 3.485 ls_scf_post 1 4.0 0.000 0.001 3.280 3.287 dbcsr_special_finalize 555 9.7 0.006 0.007 2.872 3.174 dbcsr_merge_single_wm 555 10.7 0.537 0.657 2.864 3.165 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.093 3.133 ls_scf_store_result 1 5.0 0.000 0.000 3.012 3.081 dbcsr_data_release 10477 10.7 1.580 2.518 1.580 2.518 dbcsr_finalize 304 7.8 0.049 0.061 1.815 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.993 1.995 rebuild_ks_matrix 3 7.3 0.000 0.000 1.969 1.970 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.969 1.970 dbcsr_merge_all 275 8.9 0.316 0.365 1.696 1.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.201000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 4.636799E+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 0.046 0.056 99.687 99.687 qs_energies 1 2.0 0.000 0.000 98.902 98.908 ls_scf 1 3.0 0.000 0.000 96.947 96.958 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.132 78.360 ls_scf_main 1 4.0 0.000 0.000 62.355 62.355 multiply_cannon 111 7.7 0.103 0.171 55.592 61.101 density_matrix_trs4 2 5.0 0.002 0.003 55.009 55.117 multiply_cannon_loop 111 8.7 0.069 0.077 51.000 52.733 mp_waitall_1 5436 11.0 26.615 32.124 26.615 32.124 ls_scf_init_scf 1 4.0 0.000 0.001 30.971 30.976 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.589 29.625 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.403 27.416 multiply_cannon_multrec 444 9.7 14.015 16.494 21.131 23.088 make_m2s 222 7.7 0.004 0.005 17.693 20.232 make_images 222 8.7 3.722 4.426 17.632 20.172 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.184 16.346 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.279 15.548 make_images_data 222 9.7 0.003 0.004 9.952 12.464 hybrid_alltoall_any 227 10.6 0.788 3.760 9.805 12.275 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.342 6.821 7.968 multiply_cannon_sync_h2d 444 9.7 6.507 7.883 6.507 7.883 dbcsr_mm_accdrv_process_sort 3003 11.4 6.506 7.623 6.506 7.623 mp_allgather_i34 111 8.7 2.833 7.036 2.833 7.036 arnoldi_extremal 4 6.8 0.000 0.000 5.880 5.896 arnoldi_normal_ev 4 7.8 0.002 0.005 5.880 5.896 build_subspace 16 8.4 0.015 0.019 5.484 5.493 mp_sum_l 807 5.4 2.894 5.303 2.894 5.303 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.650 4.773 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.197 4.377 dbcsr_matrix_vector_mult_local 304 10.0 3.667 4.150 3.669 4.152 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.047 4.072 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.673 3.980 mp_irecv_dv 1241 11.2 1.654 3.953 1.654 3.953 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.682 3.784 calculate_norms 792 9.8 3.578 3.746 3.578 3.746 ls_scf_post 1 4.0 0.000 0.000 3.621 3.628 make_images_sizes 222 9.7 0.000 0.000 1.093 3.522 mp_alltoall_i44 222 10.7 1.093 3.522 1.093 3.522 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.449 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.275 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.235 2.236 rebuild_ks_matrix 3 7.3 0.000 0.000 2.201 2.203 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.201 2.203 dbcsr_merge_all 275 8.9 0.474 0.530 2.048 2.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.687000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3646.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 8.747356E+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 0.081 0.102 108.760 108.761 qs_energies 1 2.0 0.000 0.000 107.310 107.330 ls_scf 1 3.0 0.000 0.000 104.360 104.379 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.055 77.189 ls_scf_main 1 4.0 0.000 0.000 66.030 66.042 density_matrix_trs4 2 5.0 0.002 0.003 56.541 56.606 multiply_cannon 111 7.7 0.172 0.252 49.611 51.235 multiply_cannon_loop 111 8.7 0.068 0.071 46.322 47.243 ls_scf_init_scf 1 4.0 0.001 0.001 34.633 34.636 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.889 32.903 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.052 30.063 mp_waitall_1 4527 11.1 21.950 25.640 21.950 25.640 make_m2s 222 7.7 0.005 0.005 23.821 24.593 make_images 222 8.7 4.598 5.003 23.715 24.485 multiply_cannon_multrec 444 9.7 17.827 18.556 22.483 23.120 hybrid_alltoall_any 227 10.6 1.666 3.639 12.847 15.508 make_images_data 222 9.7 0.003 0.003 12.992 15.284 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.400 10.712 multiply_cannon_sync_h2d 444 9.7 8.852 8.899 8.852 8.899 arnoldi_extremal 4 6.8 0.000 0.000 7.665 7.679 arnoldi_normal_ev 4 7.8 0.003 0.009 7.665 7.679 build_subspace 16 8.4 0.025 0.036 7.103 7.115 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.617 5.757 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.416 5.505 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.305 dbcsr_matrix_vector_mult_local 304 10.0 4.982 5.287 4.984 5.290 dbcsr_mm_accdrv_process 1814 10.4 0.234 0.339 4.489 4.608 dbcsr_mm_accdrv_process_sort 1814 11.4 4.181 4.312 4.181 4.312 ls_scf_post 1 4.0 0.000 0.000 3.696 3.715 make_images_sizes 222 9.7 0.000 0.000 1.480 3.664 mp_alltoall_i44 222 10.7 1.479 3.664 1.479 3.664 ls_scf_store_result 1 5.0 0.000 0.000 3.441 3.451 calculate_norms 792 9.8 3.237 3.273 3.237 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.183 dbcsr_merge_all 275 8.9 0.890 0.918 2.879 2.963 dbcsr_complete_redistribute 5 7.6 1.437 1.477 2.829 2.935 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 2.920 2.921 mp_allgather_i34 111 8.7 0.930 2.864 0.930 2.864 dbcsr_data_release 12724 10.6 2.336 2.853 2.336 2.853 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.678 2.681 rebuild_ks_matrix 3 7.3 0.000 0.000 2.611 2.614 qs_ks_build_kohn_sham_matrix 3 8.3 0.019 0.024 2.611 2.614 matrix_ls_to_qs 2 6.0 0.000 0.000 2.487 2.600 dbcsr_sort_data 325 11.1 2.450 2.518 2.450 2.518 dbcsr_new_transposed 4 7.5 0.243 0.253 2.394 2.406 mp_alltoall_d11v 48 9.2 2.225 2.284 2.225 2.284 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.197 2.243 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.205 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.761000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6849.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2_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 586.227712E+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 751836128. 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 0.069 0.176 274.282 274.284 qs_mol_dyn_low 1 2.0 0.005 0.031 273.078 273.099 qs_forces 5 3.8 0.005 0.045 272.915 272.917 qs_energies 5 4.8 0.003 0.054 269.390 269.425 scf_env_do_scf 5 5.8 0.001 0.002 255.085 255.088 scf_env_do_scf_inner_loop 105 6.6 0.003 0.024 223.937 223.952 qs_scf_new_mos 105 7.6 0.000 0.001 172.942 173.208 qs_scf_loop_do_ot 105 8.6 0.001 0.001 172.941 173.208 ot_scf_mini 105 9.6 0.003 0.004 163.014 163.238 dbcsr_multiply_generic 1445 12.2 0.125 0.134 133.547 134.047 velocity_verlet 4 3.0 0.004 0.027 116.510 116.513 multiply_cannon 1445 13.2 0.275 0.287 113.157 115.805 multiply_cannon_loop 1445 14.2 2.326 2.470 110.874 112.588 qs_ot_get_p 112 10.4 0.001 0.001 71.663 71.984 ot_mini 105 10.6 0.001 0.002 61.130 61.279 qs_ot_p2m_diag 40 11.0 0.020 0.031 59.678 59.791 cp_dbcsr_syevd 40 12.0 0.002 0.002 56.401 56.402 cp_fm_syevd 40 13.0 0.000 0.000 50.656 50.802 cp_fm_redistribute_end 40 14.0 22.736 45.388 22.745 45.392 cp_fm_syevd_base 40 14.0 22.630 45.281 22.630 45.281 multiply_cannon_multrec 69360 15.2 31.883 37.174 39.428 45.072 mp_waitall_1 488190 16.1 37.196 44.888 37.196 44.888 qs_ot_get_derivative 55 11.6 0.001 0.001 39.394 39.554 multiply_cannon_sync_h2d 69360 15.2 30.730 36.324 30.730 36.324 multiply_cannon_metrocomm3 69360 15.2 0.191 0.201 25.388 34.917 rebuild_ks_matrix 110 8.4 0.000 0.000 34.205 34.364 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.015 34.205 34.364 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.391 31.532 init_scf_loop 7 6.6 0.000 0.005 31.089 31.091 prepare_preconditioner 7 7.6 0.000 0.000 25.493 25.520 make_preconditioner 7 8.6 0.000 0.006 25.493 25.520 qs_rho_update_rho_low 110 7.6 0.001 0.003 23.919 24.339 calculate_rho_elec 110 8.6 0.029 0.032 23.918 24.338 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.008 23.202 apply_single 62 13.6 0.000 0.000 23.008 23.202 rs_pw_transfer 690 11.5 0.010 0.012 19.719 21.316 ot_new_cg_direction 55 11.6 0.001 0.003 20.981 20.982 density_rs2pw 110 9.6 0.005 0.006 17.555 19.185 make_full_inverse_cholesky 7 9.6 0.000 0.000 17.769 17.832 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.952 15.151 pw_transfer 1645 12.4 0.078 0.100 14.629 14.884 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 14.491 14.742 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.055 14.193 mp_sum_l 4684 12.4 12.166 13.147 12.166 13.147 fft_wrap_pw1pw2_240 915 15.0 0.413 0.463 12.842 13.091 cp_fm_cholesky_invert 7 10.6 12.088 12.102 12.088 12.102 qs_vxc_create 110 10.4 0.002 0.008 12.069 12.100 fft3d_pb 915 16.0 2.317 2.667 11.689 11.937 sum_up_and_integrate 60 10.3 0.028 0.031 11.876 11.887 integrate_v_rspace 60 11.3 0.002 0.002 11.848 11.859 calculate_dm_sparse 110 9.5 0.000 0.001 11.590 11.747 init_scf_run 5 5.8 0.000 0.001 11.377 11.378 scf_env_initial_rho_setup 5 6.8 0.001 0.004 11.377 11.378 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.620 10.699 check_diag 80 13.5 8.586 8.858 10.341 10.490 xc_rho_set_and_dset_create 110 12.4 0.084 0.106 8.291 8.546 make_m2s 2890 13.2 0.076 0.083 7.829 8.408 dbcsr_mm_accdrv_process 154766 15.8 3.823 4.158 7.414 8.369 make_images 2890 14.2 0.244 0.261 7.726 8.307 xc_vxc_pw_create 60 11.3 0.042 0.068 8.190 8.221 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.436 8.002 rs_pw_transfer_RS2PW_30 110 11.6 1.563 1.663 6.963 7.838 potential_pw2rs 60 12.3 0.003 0.003 7.728 7.811 xc_pw_derive 510 13.4 0.005 0.005 7.698 7.759 make_full_single_inverse 7 9.6 0.001 0.007 7.380 7.407 mp_sendrecv_dv 168740 12.6 7.001 7.234 7.001 7.234 mp_alltoall_z22v 2340 17.7 6.804 7.179 6.804 7.179 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.452 6.468 mp_waitany 7680 13.5 4.560 6.227 4.560 6.227 acc_transpose_blocks 69360 15.2 0.340 0.377 5.754 6.139 calculate_first_density_matrix 1 7.0 0.000 0.010 6.104 6.118 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.952 6.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=274.284000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.800000, yerr=1.939072 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a5f7dffd4754b6b6f5ee111240fda035c8bbe8b2 Summary: empty Status: OK