=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 0e722df91b89847b24d3e2fa39485a0bdfbfba12 ################# 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 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.12.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/01 job id: 44177744 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/02 job id: 44177745 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/03 job id: 44177746 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/04 job id: 44177747 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/05 job id: 44177748 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/06 job id: 44177749 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/07 job id: 44177750 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/08 job id: 44177751 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/09 job id: 44177752 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/10 job id: 44177753 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/11 job id: 44177754 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/12 job id: 44177755 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/13 job id: 44177756 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/14 job id: 44177757 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/15 job id: 44177758 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/16 job id: 44177759 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/17 job id: 44177760 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/18 job id: 44177761 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/19 job id: 44177762 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/20 job id: 44177763 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/21 job id: 44177764 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/22 job id: 44177765 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/23 job id: 44177766 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/24 job id: 44177767 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/25 job id: 44177768 --- 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/26 job id: 44177769 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 135.062 135.063 farming_run 1 2.0 134.391 134.392 135.035 135.038 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.454871E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 116.304 116.304 qs_energies 1 2.0 0.000 0.000 115.949 115.950 mp2_main 1 3.0 0.000 0.000 114.021 114.022 mp2_gpw_main 1 4.0 0.020 0.026 113.144 113.145 mp2_ri_gpw_compute_in 1 5.0 0.174 0.210 94.208 94.508 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.608 55.909 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.954 47.210 get_2c_integrals 1 6.0 0.001 0.001 37.428 38.389 integrate_v_rspace 273 8.0 0.436 0.452 25.289 30.355 pw_transfer 6555 10.6 0.374 0.387 27.460 28.289 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.137 26.782 grid_integrate_task_list 273 9.0 21.073 26.651 21.073 26.651 fft_wrap_pw1pw2_100 2178 12.4 1.166 1.399 23.660 24.296 compute_2c_integrals 1 7.0 0.003 0.003 19.660 19.661 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.879 19.392 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.430 18.876 19.388 rpa_ri_compute_en 1 5.0 0.000 0.000 18.831 19.139 cp_fm_cholesky_decompose 12 8.2 17.728 18.661 17.728 18.661 cholesky_decomp 1 7.0 0.000 0.000 16.617 17.546 fft3d_s 5443 13.4 16.238 16.672 16.260 16.692 ao_to_mo_and_store_B_mult_1 272 7.0 10.840 15.587 10.840 15.587 calculate_wavefunction 272 8.0 5.441 5.594 12.589 13.130 rpa_num_int 1 6.0 0.000 0.007 10.577 10.577 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.504 10.522 calc_mat_Q 8 8.0 0.000 0.000 9.348 9.454 contract_S_to_Q 8 9.0 0.000 0.000 8.768 8.869 calc_potential_gpw 544 9.5 0.005 0.006 8.254 8.669 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.214 8.568 potential_pw2rs 545 10.0 0.107 0.109 7.713 8.497 parallel_gemm_fm 14 9.1 0.000 0.000 8.359 8.426 parallel_gemm_fm_cosma 14 10.1 8.359 8.426 8.359 8.426 create_integ_mat 1 6.0 0.006 0.008 7.860 7.860 collocate_single_gaussian 272 10.0 0.041 0.044 7.449 7.726 array2fm 1 7.0 0.000 0.000 6.909 7.306 pw_scatter_s 2720 13.7 4.422 4.645 4.422 4.645 pw_gather_s 2722 13.2 3.900 4.240 3.900 4.240 array2fm_buffer_send 1 8.0 3.018 3.181 3.018 3.181 pw_poisson_solve 545 10.5 1.121 1.175 2.151 2.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.144552, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 0.027 0.040 397.275 397.276 farming_run 1 2.0 396.403 396.410 397.237 397.239 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227272E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.044 210.271 210.271 qs_energies 1 2.0 0.000 0.000 210.032 210.042 scf_env_do_scf 1 3.0 0.000 0.000 106.887 106.887 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.998 106.008 rebuild_ks_matrix 4 6.0 0.000 0.000 105.997 106.007 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.066 105.997 106.007 hfx_ks_matrix 4 8.0 0.001 0.001 105.611 105.617 integrate_four_center 4 9.0 0.144 0.454 105.611 105.616 mp2_main 1 3.0 0.000 0.000 102.851 102.860 mp2_gpw_main 1 4.0 0.041 0.092 102.011 102.019 integrate_four_center_main 4 10.0 0.091 0.473 96.942 100.235 integrate_four_center_bin 261 11.0 96.851 100.197 96.851 100.197 init_scf_loop 1 4.0 0.000 0.000 92.361 92.362 mp2_ri_gpw_compute_in 1 5.0 0.070 0.118 75.142 76.258 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.025 54.697 55.823 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.158 42.229 47.422 integrate_v_rspace 95 8.0 0.397 0.566 28.593 33.577 pw_transfer 2240 10.6 0.145 0.160 29.932 30.423 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.944 29.440 grid_integrate_task_list 95 9.0 23.880 29.064 23.880 29.064 mp2_ri_gpw_compute_en 1 5.0 0.058 0.079 26.712 28.523 ao_to_mo_and_store_B_mult_1 91 7.0 10.768 28.174 10.768 28.174 fft_wrap_pw1pw2_100 730 12.4 1.267 1.479 26.661 27.158 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.827 1.895 24.926 24.937 get_2c_integrals 1 6.0 0.002 0.009 20.358 20.386 compute_2c_integrals 1 7.0 0.004 0.015 19.340 19.357 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.849 19.183 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.887 18.846 19.175 fft3d_s 1823 13.4 18.415 18.816 18.428 18.829 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.524 14.524 calculate_wavefunction 91 8.0 2.020 2.049 9.750 9.971 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.581 8.838 9.378 potential_pw2rs 186 10.0 0.033 0.035 8.626 9.215 local_gemm 172 8.0 8.282 8.806 8.282 8.806 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.520 7.882 8.526 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.010 8.227 8.518 calc_potential_gpw 182 9.5 0.002 0.003 7.914 8.224 collocate_single_gaussian 91 10.0 0.017 0.022 7.876 8.100 mp_sendrecv_dm3 2068 8.0 5.909 6.559 5.909 6.559 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.414 6.334 6.414 mp_sync 38 10.4 3.320 5.906 3.320 5.906 pw_gather_s 912 13.2 4.937 5.393 4.937 5.393 pw_scatter_s 910 13.7 3.933 4.463 3.933 4.463 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.006821, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 451.584000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578578. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.085 52.907 52.911 qs_mol_dyn_low 1 2.0 0.005 0.014 52.454 52.467 qs_forces 11 3.9 0.002 0.004 51.911 51.915 qs_energies 11 4.9 0.003 0.006 50.430 50.456 scf_env_do_scf 11 5.9 0.001 0.002 44.387 44.388 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.302 42.302 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.446 33.846 qs_scf_new_mos 108 7.5 0.000 0.001 32.118 32.407 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.117 32.406 ot_scf_mini 108 9.5 0.002 0.003 30.506 30.707 multiply_cannon 2286 13.5 0.187 0.195 26.069 27.635 multiply_cannon_loop 2286 14.5 1.489 1.557 25.378 26.979 velocity_verlet 10 3.0 0.001 0.002 26.063 26.070 ot_mini 108 10.5 0.001 0.001 19.534 19.789 qs_ot_get_derivative 108 11.5 0.001 0.001 16.669 16.844 mp_waitall_1 245248 16.5 8.553 14.979 8.553 14.979 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.005 13.535 multiply_cannon_multrec 54864 15.5 4.210 6.477 7.638 11.138 rebuild_ks_matrix 119 8.3 0.000 0.000 7.978 8.121 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.977 8.121 multiply_cannon_sync_h2d 54864 15.5 5.931 7.558 5.931 7.558 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.039 7.169 mp_sum_l 7207 12.9 5.455 7.133 5.455 7.133 qs_ot_get_p 119 10.4 0.001 0.001 6.410 6.736 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.493 5.983 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.509 5.617 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.833 3.350 4.814 init_scf_run 11 5.9 0.000 0.001 4.763 4.763 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.762 4.763 sum_up_and_integrate 119 10.3 0.012 0.014 4.585 4.592 integrate_v_rspace 119 11.3 0.002 0.002 4.573 4.582 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.186 4.269 calculate_rho_elec 119 8.7 0.011 0.017 4.186 4.269 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.683 3.751 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.856 3.106 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.868 3.063 apply_single 119 13.6 0.000 0.000 2.868 3.063 calculate_dm_sparse 119 9.5 0.000 0.001 2.858 3.008 jit_kernel_multiply 13 15.8 2.102 2.892 2.102 2.892 rs_pw_transfer 974 11.9 0.012 0.013 2.772 2.892 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.849 2.849 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.639 2.639 ot_diis_step 108 11.5 0.006 0.006 2.623 2.623 cp_fm_redistribute_end 50 14.0 2.403 2.618 2.408 2.620 calculate_first_density_matrix 1 7.0 0.003 0.027 2.589 2.592 cp_fm_diag_elpa_base 50 14.0 0.210 2.529 0.211 2.540 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.363 2.429 density_rs2pw 119 9.7 0.004 0.005 2.237 2.360 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.269 2.271 acc_transpose_blocks 54864 15.5 0.231 0.249 1.761 2.198 grid_integrate_task_list 119 12.3 2.063 2.152 2.063 2.152 wfi_extrapolate 11 7.9 0.001 0.001 2.115 2.116 init_scf_loop 11 6.9 0.001 0.004 2.068 2.069 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.954 2.013 potential_pw2rs 119 12.3 0.004 0.004 1.900 1.913 mp_sum_d 4125 12.0 1.296 1.902 1.296 1.902 pw_transfer 1439 11.6 0.053 0.058 1.822 1.894 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.745 1.821 make_m2s 4572 13.5 0.054 0.057 1.633 1.674 make_images 4572 14.5 0.133 0.139 1.550 1.591 fft3d_ps 1201 14.6 0.375 0.480 1.515 1.582 mp_waitany 12084 13.8 1.268 1.440 1.268 1.440 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.410 1.437 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.352 1.429 mp_alltoall_d11v 2130 13.8 1.231 1.392 1.231 1.392 grid_collocate_task_list 119 9.7 1.290 1.382 1.290 1.382 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.713 1.124 acc_transpose_blocks_kernels 54864 16.5 0.252 0.384 0.774 1.075 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.911000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 489.328640E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1030216. 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 0.017 0.033 39.255 39.256 qs_mol_dyn_low 1 2.0 0.003 0.003 39.005 39.013 qs_forces 11 3.9 0.002 0.005 38.933 38.934 qs_energies 11 4.9 0.002 0.007 37.238 37.240 scf_env_do_scf 11 5.9 0.000 0.001 31.838 31.838 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.407 29.408 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.827 22.192 qs_scf_new_mos 108 7.5 0.001 0.001 20.167 20.402 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.166 20.401 ot_scf_mini 108 9.5 0.002 0.003 19.282 19.455 multiply_cannon 2286 13.5 0.208 0.219 16.750 18.403 velocity_verlet 10 3.0 0.001 0.002 18.134 18.135 multiply_cannon_loop 2286 14.5 0.896 0.970 15.674 17.118 ot_mini 108 10.5 0.001 0.001 12.042 12.280 mp_waitall_1 200699 16.5 5.896 11.461 5.896 11.461 multiply_cannon_metrocomm3 27432 15.5 0.068 0.069 4.383 9.959 qs_ot_get_derivative 108 11.5 0.001 0.001 9.586 9.762 multiply_cannon_multrec 27432 15.5 1.984 4.574 6.066 9.072 rebuild_ks_matrix 119 8.3 0.000 0.000 7.246 7.389 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.246 7.389 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.402 6.531 dbcsr_mm_accdrv_process 47894 16.0 3.064 5.376 4.012 6.104 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.836 4.706 qs_ot_get_p 119 10.4 0.001 0.001 4.387 4.618 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.143 4.406 apply_single 119 13.6 0.000 0.000 3.142 4.406 sum_up_and_integrate 119 10.3 0.024 0.028 4.239 4.244 mp_sum_l 7207 12.9 2.106 4.227 2.106 4.227 integrate_v_rspace 119 11.3 0.002 0.003 4.215 4.221 init_scf_run 11 5.9 0.000 0.001 4.150 4.151 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.150 4.150 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.814 3.849 calculate_rho_elec 119 8.7 0.021 0.024 3.814 3.848 rs_pw_transfer 974 11.9 0.010 0.011 2.608 3.020 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.983 3.001 multiply_cannon_sync_h2d 27432 15.5 2.201 2.817 2.201 2.817 make_m2s 4572 13.5 0.052 0.054 2.520 2.746 make_images 4572 14.5 0.229 0.309 2.431 2.654 calculate_first_density_matrix 1 7.0 0.002 0.013 2.643 2.644 density_rs2pw 119 9.7 0.004 0.004 2.145 2.562 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.475 2.476 init_scf_loop 11 6.9 0.000 0.000 2.411 2.412 ot_diis_step 108 11.5 0.010 0.011 2.406 2.407 calculate_dm_sparse 119 9.5 0.000 0.001 2.145 2.220 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.112 2.200 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.153 2.153 cp_fm_redistribute_end 50 14.0 1.782 2.129 1.786 2.130 cp_fm_diag_elpa_base 50 14.0 0.329 2.018 0.342 2.073 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.038 2.040 pw_transfer 1439 11.6 0.065 0.071 1.993 2.026 potential_pw2rs 119 12.3 0.006 0.006 1.959 1.970 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.901 1.937 grid_integrate_task_list 119 12.3 1.839 1.925 1.839 1.925 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.841 1.881 jit_kernel_multiply 9 16.0 0.896 1.875 0.896 1.875 fft3d_ps 1201 14.6 0.513 0.566 1.603 1.633 acc_transpose_blocks 27432 15.5 0.108 0.112 1.252 1.594 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.499 1.535 prepare_preconditioner 11 7.9 0.000 0.000 1.503 1.531 make_preconditioner 11 8.9 0.000 0.000 1.503 1.531 make_images_data 4572 15.5 0.047 0.053 1.132 1.517 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.407 1.463 wfi_extrapolate 11 7.9 0.001 0.001 1.458 1.458 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.981 1.439 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.372 1.380 grid_collocate_task_list 119 9.7 1.228 1.363 1.228 1.363 mp_alltoall_d11v 2130 13.8 1.162 1.313 1.162 1.313 mp_allgather_i34 2286 14.5 0.518 1.282 0.518 1.282 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.148 1.194 mp_sum_d 4125 12.0 0.566 1.023 0.566 1.023 mp_waitany 5720 13.7 0.522 0.974 0.522 0.974 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.722 0.973 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.146 0.547 0.951 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.937 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.916 mp_alltoall_z22v 1201 16.6 0.699 0.803 0.699 0.803 make_images_sizes 4572 15.5 0.005 0.005 0.512 0.796 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.753 0.795 mp_alltoall_i44 4572 16.5 0.507 0.792 0.507 0.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.256000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.000000, yerr=2.132007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.755520E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607838. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.134 32.472 32.474 qs_mol_dyn_low 1 2.0 0.003 0.004 31.892 31.900 qs_forces 11 3.9 0.003 0.005 31.823 31.826 qs_energies 11 4.9 0.002 0.005 30.237 30.241 scf_env_do_scf 11 5.9 0.001 0.003 25.572 25.572 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.992 22.992 dbcsr_multiply_generic 2286 12.5 0.093 0.096 16.208 16.300 velocity_verlet 10 3.0 0.001 0.002 15.273 15.275 qs_scf_new_mos 108 7.5 0.001 0.001 14.740 14.761 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.740 14.760 ot_scf_mini 108 9.5 0.002 0.003 14.020 14.038 multiply_cannon 2286 13.5 0.196 0.206 12.937 13.655 multiply_cannon_loop 2286 14.5 0.632 0.659 12.148 12.880 ot_mini 108 10.5 0.001 0.001 8.663 8.678 qs_ot_get_derivative 108 11.5 0.001 0.001 7.165 7.184 multiply_cannon_multrec 18288 15.5 1.931 2.900 6.741 7.080 rebuild_ks_matrix 119 8.3 0.000 0.000 6.435 6.458 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.434 6.457 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.675 5.695 dbcsr_mm_accdrv_process 38222 16.0 3.973 5.260 4.725 5.553 sum_up_and_integrate 119 10.3 0.030 0.031 3.998 4.004 integrate_v_rspace 119 11.3 0.002 0.003 3.967 3.976 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.474 3.483 calculate_rho_elec 119 8.7 0.030 0.031 3.474 3.482 init_scf_run 11 5.9 0.000 0.001 3.470 3.471 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.470 3.470 mp_waitall_1 158411 16.6 2.548 3.432 2.548 3.432 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.650 3.293 qs_ot_get_p 119 10.4 0.001 0.001 3.197 3.216 rs_pw_transfer 974 11.9 0.009 0.010 2.304 2.604 init_scf_loop 11 6.9 0.001 0.004 2.560 2.560 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.367 apply_single 119 13.6 0.000 0.000 2.060 2.367 density_rs2pw 119 9.7 0.004 0.004 1.998 2.277 calculate_first_density_matrix 1 7.0 0.008 0.032 2.257 2.258 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.134 2.139 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.351 2.103 pw_transfer 1439 11.6 0.066 0.071 1.956 1.964 make_m2s 4572 13.5 0.044 0.045 1.766 1.924 grid_integrate_task_list 119 12.3 1.796 1.902 1.796 1.902 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.862 1.874 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.868 1.868 jit_kernel_multiply 10 15.9 0.700 1.862 0.700 1.862 make_images 4572 14.5 0.190 0.203 1.681 1.838 calculate_dm_sparse 119 9.5 0.000 0.000 1.804 1.815 potential_pw2rs 119 12.3 0.007 0.008 1.768 1.777 prepare_preconditioner 11 7.9 0.000 0.000 1.756 1.759 make_preconditioner 11 8.9 0.000 0.001 1.756 1.759 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.611 1.693 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.638 1.639 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.633 1.634 cp_fm_redistribute_end 50 14.0 1.217 1.611 1.218 1.611 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.569 1.579 cp_fm_diag_elpa_base 50 14.0 0.377 1.524 0.391 1.573 fft3d_ps 1201 14.6 0.524 0.545 1.541 1.552 multiply_cannon_sync_h2d 18288 15.5 1.396 1.519 1.396 1.519 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.498 1.503 fft_wrap_pw1pw2_140 487 13.2 0.091 0.097 1.493 1.503 mp_sum_l 7207 12.9 1.152 1.489 1.152 1.489 ot_diis_step 108 11.5 0.011 0.011 1.476 1.476 grid_collocate_task_list 119 9.7 1.213 1.363 1.213 1.363 acc_transpose_blocks 18288 15.5 0.075 0.076 1.225 1.250 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.165 1.165 make_images_data 4572 15.5 0.047 0.050 0.787 0.956 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.948 0.951 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.876 0.897 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.678 0.881 mp_waitany 9880 13.7 0.528 0.860 0.528 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.832 0.834 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.511 0.814 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.787 0.801 mp_alltoall_d11v 2130 13.8 0.666 0.779 0.666 0.779 cp_fm_cholesky_invert 11 10.9 0.736 0.740 0.736 0.740 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.726 mp_alltoall_z22v 1201 16.6 0.621 0.719 0.621 0.719 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.474000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.909091, yerr=2.574510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 556.933120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 35.738 35.739 qs_mol_dyn_low 1 2.0 0.003 0.003 35.530 35.538 qs_forces 11 3.9 0.001 0.002 35.377 35.378 qs_energies 11 4.9 0.001 0.001 33.655 33.661 scf_env_do_scf 11 5.9 0.000 0.001 28.456 28.458 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.015 25.016 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.404 18.519 velocity_verlet 10 3.0 0.002 0.002 17.900 17.906 qs_scf_new_mos 108 7.5 0.001 0.001 16.463 16.509 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.463 16.508 ot_scf_mini 108 9.5 0.002 0.003 15.530 15.577 multiply_cannon 2286 13.5 0.232 0.291 14.784 15.305 multiply_cannon_loop 2286 14.5 0.935 0.964 13.850 14.239 ot_mini 108 10.5 0.001 0.001 9.567 9.626 multiply_cannon_multrec 27432 15.5 2.322 2.979 8.687 9.222 qs_ot_get_derivative 108 11.5 0.001 0.001 7.755 7.800 dbcsr_mm_accdrv_process 47916 15.9 5.377 6.589 6.271 7.644 rebuild_ks_matrix 119 8.3 0.000 0.000 6.623 6.679 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.622 6.679 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.882 5.930 sum_up_and_integrate 119 10.3 0.035 0.038 3.849 3.857 init_scf_run 11 5.9 0.000 0.001 3.849 3.849 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.849 3.849 integrate_v_rspace 119 11.3 0.002 0.002 3.814 3.822 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.604 3.636 calculate_rho_elec 119 8.7 0.040 0.046 3.604 3.636 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.075 3.572 qs_ot_get_p 119 10.4 0.001 0.001 3.408 3.478 init_scf_loop 11 6.9 0.000 0.000 3.420 3.420 mp_waitall_1 137007 16.6 1.983 2.710 1.983 2.710 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.145 2.559 apply_single 119 13.6 0.000 0.000 2.145 2.559 prepare_preconditioner 11 7.9 0.000 0.000 2.537 2.545 make_preconditioner 11 8.9 0.000 0.000 2.537 2.545 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.145 2.471 calculate_first_density_matrix 1 7.0 0.000 0.000 2.463 2.465 make_m2s 4572 13.5 0.054 0.056 2.146 2.258 rs_pw_transfer 974 11.9 0.009 0.009 2.023 2.222 density_rs2pw 119 9.7 0.004 0.004 2.020 2.208 pw_transfer 1439 11.6 0.066 0.071 2.121 2.161 make_images 4572 14.5 0.272 0.333 2.039 2.148 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.104 2.113 jit_kernel_multiply 10 16.1 0.834 2.107 0.834 2.107 calculate_dm_sparse 119 9.5 0.000 0.000 2.022 2.085 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.028 2.072 acc_transpose_blocks 27432 15.5 0.112 0.114 1.564 1.979 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.901 1.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.901 1.902 grid_integrate_task_list 119 12.3 1.819 1.894 1.819 1.894 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.789 1.790 ot_diis_step 108 11.5 0.012 0.012 1.771 1.771 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.728 1.745 fft3d_ps 1201 14.6 0.558 0.609 1.698 1.733 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.673 1.724 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.916 1.661 potential_pw2rs 119 12.3 0.008 0.009 1.607 1.614 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.549 1.550 cp_fm_redistribute_end 50 14.0 1.025 1.524 1.026 1.525 mp_sum_l 7207 12.9 1.050 1.521 1.050 1.521 cp_fm_diag_elpa_base 50 14.0 0.477 1.462 0.496 1.500 acc_transpose_blocks_kernels 27432 16.5 0.267 0.277 0.940 1.344 wfi_extrapolate 11 7.9 0.001 0.001 1.335 1.335 grid_collocate_task_list 119 9.7 1.222 1.335 1.222 1.335 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.217 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.099 1.115 cp_fm_upper_to_full 72 13.5 0.796 1.106 0.796 1.106 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.099 1.100 jit_kernel_transpose 5 15.6 0.673 1.072 0.673 1.072 multiply_cannon_sync_h2d 27432 15.5 0.993 1.061 0.993 1.061 dbcsr_complete_redistribute 329 12.2 0.121 0.152 0.779 1.051 make_images_data 4572 15.5 0.046 0.050 0.834 0.951 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.715 0.870 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.867 mp_alltoall_d11v 2130 13.8 0.727 0.857 0.727 0.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.830 0.836 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.832 mp_alltoall_z22v 1201 16.6 0.724 0.757 0.724 0.757 cp_fm_cholesky_invert 11 10.9 0.754 0.757 0.754 0.757 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.241 0.738 mp_alltoall_i22 627 13.8 0.426 0.716 0.426 0.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.739000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=3.175318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 598.409216E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.050 28.480 28.482 qs_mol_dyn_low 1 2.0 0.003 0.003 28.153 28.160 qs_forces 11 3.9 0.002 0.002 28.096 28.096 qs_energies 11 4.9 0.001 0.001 26.378 26.381 scf_env_do_scf 11 5.9 0.000 0.001 21.440 21.440 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.937 18.938 velocity_verlet 10 3.0 0.002 0.002 14.448 14.451 dbcsr_multiply_generic 2286 12.5 0.092 0.095 11.980 12.054 qs_scf_new_mos 108 7.5 0.001 0.001 10.838 10.867 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.837 10.866 ot_scf_mini 108 9.5 0.002 0.002 10.167 10.195 multiply_cannon 2286 13.5 0.230 0.240 9.479 9.935 multiply_cannon_loop 2286 14.5 0.329 0.338 8.583 8.773 rebuild_ks_matrix 119 8.3 0.000 0.000 6.053 6.070 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.052 6.070 multiply_cannon_multrec 9144 15.5 1.584 1.813 5.734 5.975 ot_mini 108 10.5 0.001 0.001 5.582 5.616 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.396 5.412 qs_ot_get_derivative 108 11.5 0.001 0.001 4.287 4.315 dbcsr_mm_accdrv_process 12550 15.8 3.172 4.057 4.048 4.127 sum_up_and_integrate 119 10.3 0.038 0.041 3.727 3.732 integrate_v_rspace 119 11.3 0.002 0.003 3.689 3.695 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.601 3.609 calculate_rho_elec 119 8.7 0.059 0.061 3.601 3.608 init_scf_run 11 5.9 0.000 0.001 3.458 3.458 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.457 3.458 qs_ot_get_p 119 10.4 0.001 0.001 2.816 2.858 init_scf_loop 11 6.9 0.000 0.000 2.481 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.300 2.301 pw_transfer 1439 11.6 0.066 0.069 2.192 2.203 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.098 2.111 density_rs2pw 119 9.7 0.004 0.004 1.930 2.075 grid_integrate_task_list 119 12.3 1.848 1.944 1.848 1.944 mp_waitall_1 115863 16.7 1.420 1.900 1.420 1.900 make_m2s 4572 13.5 0.034 0.035 1.734 1.895 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.881 1.883 rs_pw_transfer 974 11.9 0.008 0.008 1.732 1.865 make_images 4572 14.5 0.269 0.303 1.644 1.805 fft3d_ps 1201 14.6 0.563 0.575 1.757 1.765 jit_kernel_multiply 10 15.8 0.838 1.746 0.838 1.746 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.731 1.746 prepare_preconditioner 11 7.9 0.000 0.000 1.731 1.735 make_preconditioner 11 8.9 0.000 0.000 1.730 1.735 calculate_dm_sparse 119 9.5 0.000 0.000 1.705 1.722 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.622 1.646 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.559 1.560 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.500 1.516 potential_pw2rs 119 12.3 0.010 0.010 1.451 1.456 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.402 1.408 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.394 1.394 grid_collocate_task_list 119 9.7 1.272 1.392 1.272 1.392 cp_fm_redistribute_end 50 14.0 0.696 1.372 0.696 1.373 cp_fm_diag_elpa_base 50 14.0 0.632 1.298 0.675 1.355 ot_diis_step 108 11.5 0.012 0.013 1.283 1.283 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.268 1.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.165 1.192 apply_single 119 13.6 0.000 0.000 1.165 1.192 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.132 1.146 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.129 1.134 wfi_extrapolate 11 7.9 0.001 0.001 1.106 1.106 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.801 1.032 make_images_data 4572 15.5 0.040 0.044 0.818 1.009 acc_transpose_blocks 9144 15.5 0.038 0.039 0.935 0.951 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.920 cp_fm_cholesky_invert 11 10.9 0.876 0.878 0.876 0.878 mp_alltoall_d11v 2130 13.8 0.779 0.872 0.779 0.872 multiply_cannon_sync_h2d 9144 15.5 0.716 0.802 0.716 0.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.779 0.781 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.701 0.774 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.723 0.731 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.694 0.707 mp_alltoall_z22v 1201 16.6 0.641 0.693 0.641 0.693 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.324 0.687 mp_allgather_i34 2286 14.5 0.230 0.653 0.230 0.653 mp_waitany 5200 13.7 0.481 0.624 0.481 0.624 yz_to_x 606 15.1 0.265 0.275 0.593 0.610 x_to_yz 595 16.2 0.276 0.289 0.589 0.595 rs_pw_transfer_RS2PW_140 130 11.5 0.102 0.109 0.450 0.591 jit_kernel_transpose 5 15.6 0.578 0.589 0.578 0.589 qs_create_task_list 11 7.9 0.001 0.001 0.545 0.571 generate_qs_task_list 11 8.9 0.190 0.213 0.544 0.570 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.482000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.000000, yerr=3.884702 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 739.512320E+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 0.050 0.099 42.150 42.161 qs_mol_dyn_low 1 2.0 0.003 0.003 41.746 41.807 qs_forces 11 3.9 0.001 0.002 41.680 41.680 qs_energies 11 4.9 0.001 0.001 39.689 39.693 scf_env_do_scf 11 5.9 0.001 0.001 33.789 33.789 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.887 25.889 velocity_verlet 10 3.0 0.002 0.002 23.606 23.612 dbcsr_multiply_generic 2286 12.5 0.101 0.103 17.591 17.765 qs_scf_new_mos 108 7.5 0.001 0.001 15.980 16.091 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.979 16.090 ot_scf_mini 108 9.5 0.002 0.002 14.897 15.011 multiply_cannon 2286 13.5 0.302 0.313 13.690 14.635 multiply_cannon_loop 2286 14.5 0.342 0.346 12.416 13.405 ot_mini 108 10.5 0.001 0.001 8.953 9.087 multiply_cannon_multrec 9144 15.5 3.519 5.046 8.645 8.732 init_scf_loop 11 6.9 0.000 0.000 7.875 7.878 rebuild_ks_matrix 119 8.3 0.000 0.000 7.392 7.555 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.391 7.554 qs_ot_get_derivative 108 11.5 0.001 0.001 6.917 7.029 prepare_preconditioner 11 7.9 0.000 0.000 6.889 6.904 make_preconditioner 11 8.9 0.000 0.000 6.889 6.904 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.687 6.834 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.446 6.779 dbcsr_mm_accdrv_process 12550 15.8 4.083 5.602 5.002 6.323 cp_fm_upper_to_full 72 14.2 3.180 4.559 3.180 4.559 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.359 4.411 calculate_rho_elec 119 8.7 0.118 0.121 4.358 4.410 sum_up_and_integrate 119 10.3 0.065 0.066 4.061 4.067 integrate_v_rspace 119 11.3 0.003 0.003 3.995 4.002 init_scf_run 11 5.9 0.000 0.001 3.836 3.836 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.835 3.836 mp_waitall_1 94719 16.7 2.397 3.427 2.397 3.427 qs_ot_get_p 119 10.4 0.001 0.001 3.198 3.357 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.499 2.898 pw_transfer 1439 11.6 0.069 0.070 2.841 2.849 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.989 2.825 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.741 2.750 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.673 2.508 make_m2s 4572 13.5 0.038 0.038 2.282 2.472 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.154 2.421 apply_single 119 13.6 0.000 0.000 2.154 2.421 calculate_first_density_matrix 1 7.0 0.000 0.000 2.323 2.378 fft3d_ps 1201 14.6 0.595 0.606 2.365 2.371 make_images 4572 14.5 0.351 0.382 2.161 2.350 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.308 2.317 mp_alltoall_i22 627 13.8 1.484 2.304 1.484 2.304 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.336 2.274 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.438 2.265 density_rs2pw 119 9.7 0.004 0.004 2.239 2.260 calculate_dm_sparse 119 9.5 0.000 0.000 2.194 2.242 grid_integrate_task_list 119 12.3 2.053 2.065 2.053 2.065 ot_diis_step 108 11.5 0.014 0.014 2.006 2.006 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.927 1.930 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.908 1.910 mp_sum_l 7207 12.9 1.103 1.878 1.103 1.878 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.799 1.800 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.684 1.742 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.656 1.657 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.594 1.611 rs_pw_transfer 974 11.9 0.009 0.010 1.483 1.502 grid_collocate_task_list 119 9.7 1.485 1.501 1.485 1.501 cp_fm_cholesky_invert 11 10.9 1.496 1.500 1.496 1.500 potential_pw2rs 119 12.3 0.014 0.014 1.470 1.472 jit_kernel_multiply 8 15.5 0.892 1.404 0.892 1.404 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.132 1.394 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.388 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.363 1.363 cp_fm_diag_elpa_base 50 14.0 1.220 1.271 1.362 1.362 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.309 make_images_data 4572 15.5 0.044 0.047 1.082 1.309 mp_alltoall_d11v 2130 13.8 1.115 1.169 1.115 1.169 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.145 1.168 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.134 1.154 multiply_cannon_sync_h2d 9144 15.5 1.045 1.049 1.045 1.049 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.024 qs_create_task_list 11 7.9 0.015 0.030 0.968 0.980 generate_qs_task_list 11 8.9 0.374 0.392 0.953 0.979 yz_to_x 606 15.1 0.460 0.471 0.950 0.960 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.934 acc_transpose_blocks 9144 15.5 0.037 0.038 0.899 0.908 mp_alltoall_z22v 1201 16.6 0.816 0.866 0.816 0.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.161000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.272727, yerr=11.985528 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 501.805056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.101 83.090 83.099 qs_mol_dyn_low 1 2.0 0.006 0.019 82.662 82.671 qs_forces 11 3.9 0.002 0.003 82.530 82.532 qs_energies 11 4.9 0.002 0.005 79.604 79.613 scf_env_do_scf 11 5.9 0.001 0.002 70.578 70.581 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.965 64.968 dbcsr_multiply_generic 2055 12.4 0.106 0.110 51.490 51.747 qs_scf_new_mos 99 7.5 0.000 0.001 47.497 47.627 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.496 47.626 ot_scf_mini 99 9.5 0.002 0.003 45.091 45.223 multiply_cannon 2055 13.4 0.181 0.188 42.153 42.997 multiply_cannon_loop 2055 14.4 1.536 1.575 41.225 42.098 velocity_verlet 10 3.0 0.001 0.002 41.755 41.758 ot_mini 99 10.5 0.001 0.001 27.150 27.259 qs_ot_get_derivative 99 11.5 0.001 0.001 20.355 20.468 multiply_cannon_multrec 49320 15.4 12.171 12.752 17.210 18.103 rebuild_ks_matrix 110 8.3 0.000 0.000 14.732 14.830 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.027 14.732 14.829 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.850 12.933 mp_waitall_1 220248 16.4 11.049 12.078 11.049 12.078 multiply_cannon_sync_h2d 49320 15.4 10.041 10.531 10.041 10.531 qs_ot_get_p 110 10.4 0.003 0.023 9.574 9.694 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 6.747 7.917 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.195 7.622 apply_single 110 13.6 0.000 0.001 7.195 7.621 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.872 7.398 sum_up_and_integrate 110 10.3 0.036 0.042 7.222 7.259 integrate_v_rspace 110 11.3 0.003 0.003 7.186 7.232 init_scf_run 11 5.9 0.000 0.001 6.874 6.874 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.874 6.874 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.689 6.856 calculate_rho_elec 110 8.6 0.021 0.026 6.688 6.856 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.556 6.604 ot_diis_step 99 11.5 0.005 0.006 6.521 6.521 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.316 6.346 mp_sum_l 6514 12.8 5.110 5.916 5.110 5.916 init_scf_loop 11 6.9 0.001 0.003 5.587 5.610 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.348 5.348 dbcsr_mm_accdrv_process 87628 16.1 2.067 2.170 4.917 5.239 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.891 4.892 cp_fm_redistribute_end 48 14.0 4.268 4.865 4.272 4.866 cp_fm_diag_elpa_base 48 14.0 0.588 4.742 0.592 4.763 rs_pw_transfer 902 11.9 0.012 0.013 3.701 4.238 density_rs2pw 110 9.6 0.010 0.048 3.469 4.038 wfi_extrapolate 11 7.9 0.001 0.001 4.035 4.035 make_m2s 4110 13.4 0.061 0.066 3.821 3.948 calculate_dm_sparse 110 9.5 0.000 0.001 3.788 3.916 make_images 4110 14.4 0.178 0.192 3.726 3.857 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.534 3.538 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.461 3.506 pw_transfer 1331 11.6 0.055 0.065 3.402 3.485 grid_integrate_task_list 110 12.3 3.261 3.421 3.261 3.421 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.314 3.400 prepare_preconditioner 11 7.9 0.000 0.000 3.367 3.390 make_preconditioner 11 8.9 0.000 0.001 3.367 3.390 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.171 3.230 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.185 3.217 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.148 3.186 fft_wrap_pw1pw2_140 451 13.1 0.172 0.193 2.845 2.935 fft3d_ps 1111 14.6 0.784 0.911 2.839 2.915 potential_pw2rs 110 12.3 0.006 0.007 2.727 2.771 calculate_first_density_matrix 1 7.0 0.001 0.003 2.695 2.703 jit_kernel_multiply 13 15.9 2.567 2.642 2.567 2.642 mp_alltoall_d11v 2046 13.8 2.090 2.524 2.090 2.524 mp_waitany 14300 13.8 1.858 2.420 1.858 2.420 grid_collocate_task_list 110 9.6 2.084 2.322 2.084 2.322 acc_transpose_blocks 49320 15.4 0.223 0.232 2.156 2.223 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.024 2.036 mp_sum_d 3879 11.9 1.474 1.918 1.474 1.918 make_images_data 4110 15.4 0.044 0.047 1.759 1.879 hybrid_alltoall_any 4261 16.3 0.082 0.485 1.527 1.812 cp_fm_cholesky_invert 11 10.9 1.808 1.812 1.808 1.812 mp_alltoall_z22v 1111 16.6 1.424 1.726 1.424 1.726 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.653 1.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.099000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.636364, yerr=2.384523 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.107776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1301187. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.056 70.751 70.752 qs_mol_dyn_low 1 2.0 0.011 0.014 70.090 70.099 qs_forces 11 3.9 0.008 0.041 69.994 69.995 qs_energies 11 4.9 0.001 0.003 66.619 66.624 scf_env_do_scf 11 5.9 0.000 0.001 57.484 57.488 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.511 49.511 dbcsr_multiply_generic 2055 12.4 0.114 0.120 38.037 38.271 velocity_verlet 10 3.0 0.001 0.002 36.750 36.752 qs_scf_new_mos 99 7.5 0.001 0.001 32.994 33.124 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.993 33.123 multiply_cannon 2055 13.4 0.223 0.248 31.232 32.365 ot_scf_mini 99 9.5 0.003 0.003 31.324 31.448 multiply_cannon_loop 2055 14.4 0.924 0.944 29.859 30.882 ot_mini 99 10.5 0.001 0.001 18.501 18.631 multiply_cannon_multrec 24660 15.4 7.660 9.566 13.919 15.647 rebuild_ks_matrix 110 8.3 0.000 0.000 13.928 14.058 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.927 14.058 qs_ot_get_derivative 99 11.5 0.001 0.001 12.667 12.796 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.272 12.390 mp_waitall_1 176588 16.5 7.635 10.757 7.635 10.757 multiply_cannon_sync_h2d 24660 15.4 6.995 8.500 6.995 8.500 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.177 7.957 init_scf_loop 11 6.9 0.000 0.001 7.933 7.934 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.590 7.242 apply_single 110 13.6 0.000 0.001 6.590 7.241 sum_up_and_integrate 110 10.3 0.053 0.059 6.718 6.729 integrate_v_rspace 110 11.3 0.002 0.003 6.665 6.675 init_scf_run 11 5.9 0.000 0.001 6.468 6.469 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.468 6.469 dbcsr_mm_accdrv_process 52282 16.1 4.775 5.806 6.099 6.442 qs_ot_get_p 110 10.4 0.001 0.001 6.214 6.387 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.290 6.297 calculate_rho_elec 110 8.6 0.039 0.047 6.289 6.296 prepare_preconditioner 11 7.9 0.000 0.000 5.841 5.859 make_preconditioner 11 8.9 0.000 0.000 5.841 5.859 ot_diis_step 99 11.5 0.010 0.010 5.776 5.776 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.437 5.598 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.676 5.441 make_m2s 4110 13.4 0.057 0.060 4.231 4.720 make_images 4110 14.4 0.399 0.444 4.122 4.607 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.284 4.306 pw_transfer 1331 11.6 0.066 0.074 3.849 3.996 density_rs2pw 110 9.6 0.004 0.005 3.435 3.920 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.741 3.892 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.842 3.843 rs_pw_transfer 902 11.9 0.012 0.014 3.198 3.725 wfi_extrapolate 11 7.9 0.001 0.001 3.574 3.574 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.401 3.403 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.308 3.379 grid_integrate_task_list 110 12.3 3.144 3.351 3.144 3.351 fft_wrap_pw1pw2_140 451 13.1 0.202 0.219 3.129 3.286 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.264 3.265 cp_fm_redistribute_end 48 14.0 2.440 3.236 2.442 3.236 fft3d_ps 1111 14.6 1.102 1.321 3.092 3.223 cp_fm_diag_elpa_base 48 14.0 0.760 3.079 0.791 3.176 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.034 3.088 calculate_dm_sparse 110 9.5 0.001 0.001 3.040 3.074 calculate_first_density_matrix 1 7.0 0.000 0.002 2.806 2.808 make_images_data 4110 15.4 0.048 0.052 2.302 2.801 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.035 2.783 cp_fm_cholesky_invert 11 10.9 2.710 2.717 2.710 2.717 mp_sum_l 6514 12.8 1.951 2.563 1.951 2.563 potential_pw2rs 110 12.3 0.008 0.008 2.513 2.527 grid_collocate_task_list 110 9.6 2.049 2.469 2.049 2.469 qs_energies_init_hamiltonians 11 5.9 0.002 0.011 2.095 2.097 mp_alltoall_d11v 2046 13.8 1.761 2.061 1.761 2.061 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.963 1.986 mp_waitany 10164 13.8 1.357 1.871 1.357 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.821 1.836 jit_kernel_multiply 9 16.4 0.976 1.722 0.976 1.722 mp_allgather_i34 2055 14.4 0.700 1.672 0.700 1.672 acc_transpose_blocks 24660 15.4 0.109 0.112 1.599 1.650 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.219 1.112 1.632 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.608 1.621 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.796 1.613 cp_fm_cholesky_decompose 22 10.9 1.492 1.509 1.492 1.509 mp_irecv_dv 57340 16.2 0.673 1.498 0.673 1.498 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.363 1.469 dbcsr_complete_redistribute 325 12.2 0.240 0.294 1.199 1.462 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.752000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.727273, yerr=7.921099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.349504E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 60.926 60.928 qs_mol_dyn_low 1 2.0 0.003 0.003 60.456 60.466 qs_forces 11 3.9 0.002 0.006 60.330 60.331 qs_energies 11 4.9 0.001 0.004 57.112 57.115 scf_env_do_scf 11 5.9 0.001 0.001 48.999 48.999 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.416 40.416 velocity_verlet 10 3.0 0.001 0.002 32.947 32.951 dbcsr_multiply_generic 2055 12.4 0.110 0.113 28.530 28.813 qs_scf_new_mos 99 7.5 0.001 0.001 25.183 25.300 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.183 25.299 ot_scf_mini 99 9.5 0.003 0.004 23.962 24.087 multiply_cannon 2055 13.4 0.211 0.225 22.080 23.211 multiply_cannon_loop 2055 14.4 0.613 0.630 20.909 22.000 ot_mini 99 10.5 0.001 0.001 13.690 13.820 rebuild_ks_matrix 110 8.3 0.000 0.000 12.505 12.652 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.505 12.652 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.025 11.157 multiply_cannon_multrec 16440 15.4 3.850 4.763 9.613 10.508 mp_waitall_1 139946 16.5 7.198 10.506 7.198 10.506 qs_ot_get_derivative 99 11.5 0.001 0.001 9.153 9.281 init_scf_loop 11 6.9 0.001 0.004 8.545 8.546 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.460 7.650 prepare_preconditioner 11 7.9 0.000 0.000 6.732 6.749 make_preconditioner 11 8.9 0.000 0.002 6.732 6.749 sum_up_and_integrate 110 10.3 0.060 0.063 6.633 6.650 integrate_v_rspace 110 11.3 0.003 0.004 6.573 6.589 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.085 6.446 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.070 6.080 calculate_rho_elec 110 8.6 0.058 0.059 6.069 6.079 dbcsr_mm_accdrv_process 34862 16.1 4.525 5.325 5.618 5.717 qs_ot_get_p 110 10.4 0.001 0.001 5.475 5.646 init_scf_run 11 5.9 0.000 0.001 5.594 5.594 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.593 5.594 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.037 5.486 apply_single 110 13.6 0.000 0.000 5.037 5.485 make_m2s 4110 13.4 0.050 0.052 4.271 4.606 ot_diis_step 99 11.5 0.010 0.011 4.506 4.506 make_images 4110 14.4 0.392 0.511 4.155 4.490 density_rs2pw 110 9.6 0.004 0.005 3.182 4.448 multiply_cannon_sync_h2d 16440 15.4 3.602 4.213 3.602 4.213 rs_pw_transfer 902 11.9 0.010 0.011 2.745 3.999 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.811 3.816 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.156 3.793 pw_transfer 1331 11.6 0.066 0.074 3.765 3.776 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.658 3.668 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.471 3.472 grid_integrate_task_list 110 12.3 3.154 3.398 3.154 3.398 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.155 3.167 wfi_extrapolate 11 7.9 0.001 0.001 3.000 3.000 fft3d_ps 1111 14.6 1.086 1.097 2.956 2.966 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.940 2.941 make_images_data 4110 15.4 0.045 0.049 2.490 2.928 cp_fm_redistribute_end 48 14.0 1.834 2.913 1.837 2.914 cp_fm_diag_elpa_base 48 14.0 1.014 2.762 1.073 2.870 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.215 2.854 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.803 2.805 cp_fm_cholesky_invert 11 10.9 2.676 2.682 2.676 2.682 calculate_dm_sparse 110 9.5 0.001 0.001 2.546 2.574 calculate_first_density_matrix 1 7.0 0.000 0.003 2.506 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.501 mp_waitany 17072 13.8 1.209 2.495 1.209 2.495 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.895 2.470 grid_collocate_task_list 110 9.6 2.078 2.437 2.078 2.437 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.364 2.433 potential_pw2rs 110 12.3 0.010 0.011 2.358 2.372 mp_irecv_dv 48980 15.7 0.825 2.348 0.825 2.348 mp_alltoall_d11v 2046 13.8 1.794 2.235 1.794 2.235 mp_sum_l 6514 12.8 1.566 2.163 1.566 2.163 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.903 2.150 qs_energies_init_hamiltonians 11 5.9 0.026 0.040 2.050 2.051 dbcsr_complete_redistribute 325 12.2 0.335 0.373 1.433 1.900 cp_fm_upper_to_full 70 13.6 1.394 1.887 1.394 1.887 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.681 1.698 mp_allgather_i34 2055 14.4 0.498 1.639 0.498 1.639 cp_fm_cholesky_decompose 22 10.9 1.564 1.582 1.564 1.582 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.507 1.521 jit_kernel_multiply 8 16.3 0.706 1.494 0.706 1.494 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.357 1.488 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.970 1.424 rs_gather_matrices 110 12.3 0.233 0.260 0.983 1.356 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.273 1.281 acc_transpose_blocks 16440 15.4 0.070 0.072 1.218 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.928000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.636364, yerr=9.227161 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 725.471232E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.084 66.337 66.338 qs_mol_dyn_low 1 2.0 0.003 0.004 65.185 66.062 qs_forces 11 3.9 0.002 0.002 65.012 65.013 qs_energies 11 4.9 0.001 0.002 61.588 61.591 scf_env_do_scf 11 5.9 0.000 0.001 53.265 53.268 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.591 41.592 velocity_verlet 10 3.0 0.001 0.002 37.193 37.198 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.431 29.653 qs_scf_new_mos 99 7.5 0.001 0.001 26.280 26.381 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.279 26.380 ot_scf_mini 99 9.5 0.003 0.003 24.662 24.763 multiply_cannon 2055 13.4 0.244 0.263 22.423 23.683 multiply_cannon_loop 2055 14.4 0.883 0.902 21.053 21.671 ot_mini 99 10.5 0.001 0.001 14.063 14.178 multiply_cannon_multrec 24660 15.4 4.249 6.764 12.720 13.990 rebuild_ks_matrix 110 8.3 0.000 0.000 12.367 12.487 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.366 12.486 init_scf_loop 11 6.9 0.000 0.000 11.633 11.633 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.924 11.028 qs_ot_get_derivative 99 11.5 0.001 0.001 9.917 10.019 prepare_preconditioner 11 7.9 0.000 0.000 9.881 9.898 make_preconditioner 11 8.9 0.000 0.000 9.881 9.898 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.100 9.549 dbcsr_mm_accdrv_process 52304 16.0 6.894 8.485 8.324 9.229 sum_up_and_integrate 110 10.3 0.068 0.071 6.626 6.640 integrate_v_rspace 110 11.3 0.002 0.003 6.558 6.570 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.246 6.255 calculate_rho_elec 110 8.6 0.077 0.081 6.246 6.254 mp_waitall_1 121746 16.5 4.315 6.114 4.315 6.114 qs_ot_get_p 110 10.4 0.001 0.001 5.605 5.739 make_m2s 4110 13.4 0.060 0.062 5.429 5.696 make_images 4110 14.4 0.578 0.701 5.288 5.550 init_scf_run 11 5.9 0.000 0.001 5.524 5.524 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.523 5.524 cp_fm_upper_to_full 70 13.8 3.262 4.639 3.262 4.639 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.027 4.112 apply_single 110 13.6 0.000 0.000 4.027 4.112 ot_diis_step 99 11.5 0.011 0.011 4.110 4.110 pw_transfer 1331 11.6 0.065 0.075 3.984 4.016 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.877 3.913 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.844 3.858 dbcsr_complete_redistribute 325 12.2 0.417 0.458 2.673 3.816 density_rs2pw 110 9.6 0.004 0.004 3.166 3.680 grid_integrate_task_list 110 12.3 3.258 3.424 3.258 3.424 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.371 3.371 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 3.328 3.363 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.281 3.335 multiply_cannon_sync_h2d 24660 15.4 3.175 3.296 3.175 3.296 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.132 3.261 fft3d_ps 1111 14.6 1.095 1.133 3.168 3.192 make_images_data 4110 15.4 0.048 0.052 2.752 3.070 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.993 3.047 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.318 3.026 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.387 3.002 wfi_extrapolate 11 7.9 0.001 0.001 2.996 2.996 calculate_dm_sparse 110 9.5 0.001 0.001 2.963 2.992 rs_pw_transfer 902 11.9 0.010 0.011 2.418 2.941 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.773 2.889 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.830 2.832 cp_fm_redistribute_end 48 14.0 1.415 2.802 1.417 2.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.781 2.784 mp_alltoall_i22 605 13.7 1.630 2.780 1.630 2.780 cp_fm_diag_elpa_base 48 14.0 1.303 2.664 1.383 2.776 cp_fm_cholesky_invert 11 10.9 2.685 2.693 2.685 2.693 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.468 grid_collocate_task_list 110 9.6 2.181 2.444 2.181 2.444 calculate_first_density_matrix 1 7.0 0.000 0.000 2.434 2.437 qs_energies_init_hamiltonians 11 5.9 0.009 0.024 2.303 2.304 potential_pw2rs 110 12.3 0.012 0.013 2.256 2.265 mp_alltoall_d11v 2046 13.8 1.809 2.081 1.809 2.081 jit_kernel_multiply 10 15.6 1.101 1.894 1.101 1.894 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.694 1.726 cp_fm_cholesky_decompose 22 10.9 1.661 1.707 1.661 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.702 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.596 1.694 multiply_cannon_metrocomm4 20550 15.4 0.057 0.061 0.844 1.601 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.586 1.598 mp_allgather_i34 2055 14.4 0.472 1.557 0.472 1.557 mp_waitany 13376 13.8 1.051 1.551 1.051 1.551 acc_transpose_blocks 24660 15.4 0.105 0.108 1.518 1.537 mp_irecv_dv 62702 16.1 0.746 1.526 0.746 1.526 mp_sum_l 6514 12.8 0.891 1.449 0.891 1.449 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.235 1.334 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.338000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.181818, yerr=7.333834 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 837.758976E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.055 55.285 55.286 qs_mol_dyn_low 1 2.0 0.003 0.003 54.999 55.007 qs_forces 11 3.9 0.002 0.002 54.935 54.936 qs_energies 11 4.9 0.001 0.001 51.245 51.249 scf_env_do_scf 11 5.9 0.001 0.001 42.747 42.748 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.041 35.042 velocity_verlet 10 3.0 0.002 0.002 31.112 31.115 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.556 22.669 qs_scf_new_mos 99 7.5 0.001 0.001 19.909 19.955 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.908 19.955 ot_scf_mini 99 9.5 0.002 0.002 18.680 18.710 multiply_cannon 2055 13.4 0.244 0.251 17.205 18.444 multiply_cannon_loop 2055 14.4 0.320 0.332 15.882 16.107 rebuild_ks_matrix 110 8.3 0.000 0.000 11.964 11.985 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.964 11.984 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.626 10.645 ot_mini 99 10.5 0.001 0.001 10.106 10.132 multiply_cannon_multrec 8220 15.4 3.178 4.478 7.542 8.551 init_scf_loop 11 6.9 0.000 0.000 7.659 7.660 mp_waitall_1 103326 16.6 5.779 7.376 5.779 7.376 sum_up_and_integrate 110 10.3 0.079 0.081 6.567 6.577 integrate_v_rspace 110 11.3 0.003 0.003 6.487 6.497 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.337 6.351 calculate_rho_elec 110 8.6 0.115 0.116 6.336 6.350 qs_ot_get_derivative 99 11.5 0.001 0.001 6.316 6.346 prepare_preconditioner 11 7.9 0.000 0.000 6.000 6.005 make_preconditioner 11 8.9 0.000 0.000 6.000 6.005 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.590 5.658 init_scf_run 11 5.9 0.000 0.001 5.291 5.291 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.291 5.291 dbcsr_mm_accdrv_process 17442 15.9 2.801 3.785 4.236 5.129 qs_ot_get_p 110 10.4 0.001 0.001 4.705 4.734 make_m2s 4110 13.4 0.039 0.040 4.102 4.351 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.975 4.300 make_images 4110 14.4 0.638 0.690 3.971 4.220 pw_transfer 1331 11.6 0.066 0.071 4.118 4.125 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.011 4.022 ot_diis_step 99 11.5 0.012 0.012 3.771 3.771 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.691 3.730 apply_single 110 13.6 0.000 0.000 3.691 3.730 grid_integrate_task_list 110 12.3 3.355 3.535 3.355 3.535 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.479 3.493 density_rs2pw 110 9.6 0.004 0.004 3.105 3.480 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.403 3.407 fft3d_ps 1111 14.6 1.154 1.204 3.252 3.259 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.104 3.105 multiply_cannon_sync_h2d 8220 15.4 2.916 3.018 2.916 3.018 cp_fm_cholesky_invert 11 10.9 2.913 2.917 2.913 2.917 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.764 2.764 wfi_extrapolate 11 7.9 0.001 0.001 2.741 2.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.686 2.687 make_images_data 4110 15.4 0.040 0.045 2.303 2.684 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.247 2.630 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.606 2.607 rs_pw_transfer 902 11.9 0.010 0.011 2.228 2.600 cp_fm_redistribute_end 48 14.0 0.662 2.580 0.665 2.581 cp_fm_diag_elpa_base 48 14.0 1.735 2.392 1.909 2.555 grid_collocate_task_list 110 9.6 2.276 2.534 2.276 2.534 calculate_dm_sparse 110 9.5 0.001 0.001 2.447 2.491 calculate_first_density_matrix 1 7.0 0.000 0.000 2.455 2.456 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.341 2.348 potential_pw2rs 110 12.3 0.015 0.015 2.130 2.133 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.982 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.871 1.887 mp_alltoall_d11v 2046 13.8 1.570 1.849 1.570 1.849 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.771 1.787 jit_kernel_multiply 9 15.7 1.127 1.731 1.127 1.731 qs_env_update_s_mstruct 11 6.9 0.032 0.090 1.598 1.724 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.678 cp_fm_cholesky_decompose 22 10.9 1.643 1.666 1.643 1.666 mp_allgather_i34 2055 14.4 0.508 1.524 0.508 1.524 dbcsr_complete_redistribute 325 12.2 0.569 0.593 1.423 1.518 mp_waitany 9240 13.8 1.128 1.507 1.128 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.418 1.430 qs_create_task_list 11 7.9 0.001 0.001 1.228 1.334 generate_qs_task_list 11 8.9 0.382 0.451 1.227 1.333 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.840 1.211 rs_gather_matrices 110 12.3 0.324 0.367 0.915 1.186 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.130 1.162 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.485 1.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.286000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.545455, yerr=14.418193 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 1.356919E+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 0.039 0.078 87.291 87.292 qs_mol_dyn_low 1 2.0 0.003 0.003 86.912 86.922 qs_forces 11 3.9 0.002 0.003 86.839 86.839 qs_energies 11 4.9 0.001 0.001 82.661 82.663 scf_env_do_scf 11 5.9 0.001 0.001 72.456 72.456 velocity_verlet 10 3.0 0.016 0.017 55.762 55.768 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.230 44.231 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.403 28.524 init_scf_loop 11 6.9 0.000 0.000 28.150 28.152 prepare_preconditioner 11 7.9 0.000 0.000 26.112 26.119 make_preconditioner 11 8.9 0.000 0.000 26.112 26.119 qs_scf_new_mos 99 7.5 0.001 0.001 25.880 25.928 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.879 25.927 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.375 25.585 ot_scf_mini 99 9.5 0.002 0.002 24.102 24.135 multiply_cannon 2055 13.4 0.350 0.365 21.512 22.209 multiply_cannon_loop 2055 14.4 0.340 0.343 19.741 20.049 cp_fm_upper_to_full 70 14.2 12.608 18.059 12.608 18.059 rebuild_ks_matrix 110 8.3 0.000 0.001 14.195 14.239 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.194 14.239 ot_mini 99 10.5 0.001 0.001 13.482 13.517 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.863 12.903 dbcsr_complete_redistribute 325 12.2 1.017 1.043 7.289 10.437 multiply_cannon_multrec 8220 15.4 4.348 4.520 9.569 9.705 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.278 9.437 qs_ot_get_derivative 99 11.5 0.001 0.001 8.932 8.966 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.721 8.852 mp_alltoall_i22 605 13.7 5.365 8.560 5.365 8.560 mp_waitall_1 84994 16.7 7.439 8.326 7.439 8.326 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.763 7.797 calculate_rho_elec 110 8.6 0.227 0.227 7.762 7.797 sum_up_and_integrate 110 10.3 0.150 0.151 7.368 7.384 integrate_v_rspace 110 11.3 0.004 0.004 7.218 7.234 init_scf_run 11 5.9 0.000 0.001 5.919 5.919 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.919 5.919 make_m2s 4110 13.4 0.043 0.044 5.128 5.630 qs_ot_get_p 110 10.4 0.001 0.001 5.454 5.501 make_images 4110 14.4 0.881 0.934 4.940 5.441 dbcsr_mm_accdrv_process 11614 15.7 3.382 4.145 5.079 5.367 pw_transfer 1331 11.6 0.075 0.076 5.322 5.327 cp_fm_cholesky_invert 11 10.9 5.251 5.255 5.251 5.255 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.205 5.211 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.681 5.161 apply_single 110 13.6 0.000 0.000 4.681 5.161 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.803 5.066 fft_wrap_pw1pw2_140 451 13.1 0.222 0.225 4.536 4.546 ot_diis_step 99 11.5 0.015 0.016 4.525 4.526 fft3d_ps 1111 14.6 1.293 1.311 4.397 4.404 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.952 3.959 multiply_cannon_sync_h2d 8220 15.4 3.952 3.956 3.952 3.956 density_rs2pw 110 9.6 0.004 0.004 3.797 3.830 grid_integrate_task_list 110 12.3 3.655 3.713 3.655 3.713 qs_energies_init_hamiltonians 11 5.9 0.021 0.023 3.695 3.696 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.866 3.518 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.515 3.515 make_images_data 4110 15.4 0.044 0.047 2.765 3.412 wfi_extrapolate 11 7.9 0.001 0.001 3.334 3.334 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.836 3.316 calculate_dm_sparse 110 9.5 0.001 0.001 3.152 3.179 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.923 2.926 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.922 2.922 cp_fm_diag_elpa_base 48 14.0 2.392 2.580 2.920 2.921 grid_collocate_task_list 110 9.6 2.628 2.647 2.628 2.647 potential_pw2rs 110 12.3 0.021 0.021 2.507 2.509 calculate_first_density_matrix 1 7.0 0.000 0.000 2.461 2.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.414 2.448 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.357 2.375 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.208 2.273 rs_pw_transfer 902 11.9 0.011 0.011 2.238 2.272 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.148 mp_alltoall_d11v 2046 13.8 1.941 2.045 1.941 2.045 cp_fm_cholesky_decompose 22 10.9 1.988 2.008 1.988 2.008 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.938 1.942 qs_create_task_list 11 7.9 0.000 0.001 1.894 1.942 generate_qs_task_list 11 8.9 0.737 0.789 1.893 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.781 1.839 jit_kernel_multiply 10 15.3 1.496 1.793 1.496 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.292000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.363636, yerr=65.906646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.455744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2530532. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 204.916 204.918 qs_mol_dyn_low 1 2.0 0.009 0.033 204.311 204.326 qs_forces 11 3.9 0.003 0.005 204.167 204.187 qs_energies 11 4.9 0.002 0.005 198.597 198.626 scf_env_do_scf 11 5.9 0.001 0.001 181.311 181.315 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.603 160.606 dbcsr_multiply_generic 2507 12.6 0.193 0.198 123.547 124.184 velocity_verlet 10 3.0 0.001 0.002 123.022 123.023 qs_scf_new_mos 117 7.6 0.001 0.001 120.157 120.409 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.156 120.409 ot_scf_mini 117 9.6 0.003 0.003 113.609 113.857 multiply_cannon 2507 13.6 0.236 0.245 100.556 102.077 multiply_cannon_loop 2507 14.6 2.114 2.182 98.316 99.621 ot_mini 117 10.6 0.001 0.001 65.249 65.534 multiply_cannon_multrec 60168 15.6 33.037 35.135 41.807 43.343 qs_ot_get_derivative 117 11.6 0.001 0.001 40.491 40.710 rebuild_ks_matrix 128 8.3 0.001 0.001 33.896 34.081 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.033 33.895 34.080 mp_waitall_1 267128 16.5 27.707 31.058 27.707 31.058 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.470 30.652 multiply_cannon_sync_h2d 60168 15.6 27.312 28.722 27.312 28.722 qs_ot_get_p 128 10.4 0.001 0.001 26.767 27.031 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.242 24.799 apply_single 128 13.6 0.001 0.001 24.242 24.799 ot_diis_step 117 11.6 0.007 0.008 24.436 24.437 init_scf_loop 11 6.9 0.000 0.000 20.631 20.632 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.016 20.080 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.674 18.865 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.470 17.471 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 15.279 17.277 prepare_preconditioner 11 7.9 0.000 0.000 15.976 16.027 make_preconditioner 11 8.9 0.000 0.000 15.976 16.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.191 15.383 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.613 14.617 cp_fm_redistribute_end 83 14.4 11.553 14.539 11.564 14.541 sum_up_and_integrate 128 10.3 0.089 0.107 14.401 14.416 integrate_v_rspace 128 11.3 0.003 0.004 14.312 14.330 make_m2s 5014 13.6 0.105 0.113 13.979 14.325 cp_fm_diag_elpa_base 83 14.4 2.935 14.171 2.966 14.277 make_images 5014 14.6 0.403 0.421 13.798 14.153 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.029 14.144 calculate_rho_elec 128 8.7 0.045 0.064 14.029 14.143 init_scf_run 11 5.9 0.000 0.001 13.027 13.027 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.026 13.027 density_rs2pw 128 9.7 0.006 0.007 7.397 10.593 wfi_extrapolate 11 7.9 0.001 0.001 9.131 9.131 cp_fm_cholesky_invert 11 10.9 9.100 9.108 9.100 9.108 rs_pw_transfer 1046 11.9 0.017 0.019 5.895 9.089 mp_sum_l 7870 13.0 7.769 8.855 7.769 8.855 dbcsr_mm_accdrv_process 124484 16.2 3.344 3.481 8.332 8.854 calculate_dm_sparse 128 9.5 0.001 0.001 8.493 8.582 pw_transfer 1547 11.6 0.077 0.099 8.216 8.467 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.010 8.267 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.962 8.087 make_images_data 5014 15.6 0.070 0.076 6.864 7.821 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.688 7.782 grid_integrate_task_list 128 12.3 7.087 7.601 7.087 7.601 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 5.811 7.489 hybrid_alltoall_any 5200 16.5 0.290 2.254 6.009 7.435 fft_wrap_pw1pw2_140 523 13.2 0.444 0.508 6.927 7.216 fft3d_ps 1291 14.7 2.110 2.897 6.796 6.998 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.851 6.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.962 6.040 mp_waitany 16020 13.9 2.670 5.897 2.670 5.897 grid_collocate_task_list 128 9.7 4.556 5.770 4.556 5.770 mp_alltoall_d11v 2415 14.1 4.237 5.401 4.237 5.401 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.297 2.151 5.386 potential_pw2rs 128 12.3 0.009 0.011 4.774 4.798 cp_fm_cholesky_decompose 22 10.9 4.666 4.681 4.666 4.681 mp_sum_d 4455 12.2 3.563 4.297 3.563 4.297 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.918000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.545455, yerr=6.692786 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 834.170880E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57228. MP_Allreduce 11247 947. MP_Sync 168 MP_Alltoall 1955 5616370. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.146 0.220 192.492 192.495 qs_mol_dyn_low 1 2.0 0.004 0.005 191.529 191.543 qs_forces 11 3.9 0.004 0.007 191.445 191.446 qs_energies 11 4.9 0.004 0.008 184.597 184.610 scf_env_do_scf 11 5.9 0.002 0.003 167.847 167.857 scf_env_do_scf_inner_loop 116 6.6 0.052 0.102 134.682 134.685 velocity_verlet 10 3.0 0.002 0.003 119.940 119.941 dbcsr_multiply_generic 2485 12.5 0.187 0.216 96.901 98.042 qs_scf_new_mos 116 7.6 0.001 0.001 94.123 94.545 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.122 94.544 ot_scf_mini 116 9.6 0.004 0.005 89.359 89.827 multiply_cannon 2485 13.5 0.485 0.554 77.009 81.089 multiply_cannon_loop 2485 14.5 1.241 1.282 73.700 76.059 ot_mini 116 10.6 0.001 0.001 49.391 49.848 mp_waitall_1 212858 16.6 24.274 37.519 24.274 37.519 multiply_cannon_multrec 29820 15.5 21.986 26.345 31.642 36.475 rebuild_ks_matrix 127 8.3 0.001 0.001 33.689 34.135 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.031 33.688 34.134 init_scf_loop 11 6.9 0.002 0.006 33.067 33.069 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.285 30.703 prepare_preconditioner 11 7.9 0.000 0.000 28.494 28.553 make_preconditioner 11 8.9 0.001 0.002 28.494 28.553 qs_ot_get_derivative 116 11.6 0.001 0.002 27.713 28.175 multiply_cannon_metrocomm3 29820 15.5 0.093 0.099 15.408 27.955 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.164 27.680 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.806 22.903 apply_single 127 13.6 0.001 0.001 21.805 22.902 qs_ot_get_p 127 10.4 0.001 0.001 21.862 22.433 ot_diis_step 116 11.6 0.014 0.015 21.513 21.515 multiply_cannon_sync_h2d 29820 15.5 19.234 21.330 19.234 21.330 qs_ot_p2m_diag 82 11.4 0.185 0.213 17.069 17.102 cp_fm_cholesky_invert 11 10.9 16.614 16.627 16.614 16.627 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.928 15.930 make_m2s 4970 13.5 0.089 0.097 14.256 15.900 make_images 4970 14.5 1.146 1.335 14.048 15.692 sum_up_and_integrate 127 10.3 0.115 0.133 14.961 14.995 integrate_v_rspace 127 11.3 0.003 0.005 14.846 14.885 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.388 14.469 calculate_rho_elec 127 8.7 0.087 0.105 14.387 14.468 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.820 12.828 cp_fm_redistribute_end 82 14.4 7.484 12.749 7.495 12.750 cp_fm_diag_elpa_base 82 14.4 5.020 12.236 5.237 12.608 init_scf_run 11 5.9 0.000 0.001 11.869 11.871 scf_env_initial_rho_setup 11 6.9 0.008 0.022 11.869 11.871 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.767 11.092 make_images_data 4970 15.5 0.066 0.075 8.660 10.635 multiply_cannon_metrocomm4 27335 15.5 0.094 0.108 3.711 10.453 density_rs2pw 127 9.7 0.006 0.006 7.712 10.180 hybrid_alltoall_any 5155 16.4 0.341 1.502 7.347 10.100 mp_irecv_dv 68888 16.3 3.520 10.072 3.520 10.072 dbcsr_mm_accdrv_process 61726 16.2 4.764 5.405 9.115 9.752 pw_transfer 1535 11.6 0.087 0.107 9.139 9.211 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.914 8.986 wfi_extrapolate 11 7.9 0.001 0.001 8.484 8.484 rs_pw_transfer 1038 11.9 0.014 0.016 5.965 8.479 fft_wrap_pw1pw2_140 519 13.2 0.467 0.517 7.817 7.905 grid_integrate_task_list 127 12.3 7.123 7.470 7.123 7.470 fft3d_ps 1281 14.7 2.730 2.895 7.296 7.338 cp_fm_cholesky_decompose 22 10.9 6.980 7.059 6.980 7.059 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.258 7.024 calculate_dm_sparse 127 9.5 0.001 0.001 6.481 6.610 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.231 6.240 grid_collocate_task_list 127 9.7 4.668 5.832 4.668 5.832 mp_sum_l 7804 13.0 3.908 5.717 3.908 5.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.327 5.445 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.247 5.312 potential_pw2rs 127 12.3 0.015 0.017 5.250 5.277 mp_waitany 11660 13.9 2.517 5.087 2.517 5.087 mp_allgather_i34 2485 14.5 1.880 5.036 1.880 5.036 mp_alltoall_d11v 2401 14.1 4.253 4.908 4.253 4.908 rs_pw_transfer_RS2PW_140 138 11.5 0.349 0.375 2.094 4.589 dbcsr_complete_redistribute 393 12.7 0.783 0.895 3.251 4.100 mp_sum_d 4441 12.1 2.672 4.094 2.672 4.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.495000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.363636, yerr=4.772511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 939.601920E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57277. MP_Allreduce 11236 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.082 0.147 178.644 178.647 qs_mol_dyn_low 1 2.0 0.018 0.039 177.827 177.841 qs_forces 11 3.9 0.011 0.025 177.713 177.727 qs_energies 11 4.9 0.013 0.022 171.013 171.041 scf_env_do_scf 11 5.9 0.005 0.012 153.718 153.719 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 118.642 118.643 velocity_verlet 10 3.0 0.003 0.003 112.971 112.973 dbcsr_multiply_generic 2485 12.5 0.181 0.186 82.394 83.565 qs_scf_new_mos 116 7.6 0.001 0.001 80.256 80.586 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.255 80.585 ot_scf_mini 116 9.6 0.004 0.005 76.138 76.532 multiply_cannon 2485 13.5 0.497 0.520 62.684 67.011 multiply_cannon_loop 2485 14.5 0.844 0.874 59.467 62.174 ot_mini 116 10.6 0.001 0.001 42.253 42.623 mp_waitall_1 169034 16.6 26.304 35.550 26.304 35.550 init_scf_loop 11 6.9 0.006 0.024 34.967 34.969 rebuild_ks_matrix 127 8.3 0.001 0.001 30.885 31.361 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.025 30.884 31.360 prepare_preconditioner 11 7.9 0.000 0.000 30.848 30.896 make_preconditioner 11 8.9 0.000 0.001 30.848 30.896 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.485 29.914 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.807 28.238 multiply_cannon_multrec 19880 15.5 13.316 16.294 22.040 25.013 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 16.186 24.956 qs_ot_get_derivative 116 11.6 0.001 0.002 22.534 22.916 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.834 20.953 apply_single 127 13.6 0.001 0.001 19.834 20.953 qs_ot_get_p 127 10.4 0.001 0.001 19.276 19.722 ot_diis_step 116 11.6 0.018 0.022 19.619 19.619 multiply_cannon_sync_h2d 19880 15.5 14.033 16.059 14.033 16.059 make_m2s 4970 13.5 0.080 0.085 14.817 15.832 make_images 4970 14.5 1.162 1.252 14.585 15.600 qs_ot_p2m_diag 82 11.4 0.262 0.270 15.129 15.137 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.483 14.930 calculate_rho_elec 127 8.7 0.131 0.146 14.483 14.929 sum_up_and_integrate 127 10.3 0.132 0.144 14.733 14.768 integrate_v_rspace 127 11.3 0.004 0.005 14.601 14.634 cp_fm_cholesky_invert 11 10.9 14.305 14.314 14.305 14.314 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.132 14.133 init_scf_run 11 5.9 0.000 0.001 12.089 12.090 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.089 12.090 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.998 10.999 cp_fm_redistribute_end 82 14.4 4.157 10.933 4.172 10.936 cp_fm_diag_elpa_base 82 14.4 6.336 10.341 6.743 10.816 make_images_data 4970 15.5 0.063 0.071 9.210 10.641 hybrid_alltoall_any 5155 16.4 0.432 1.975 8.057 9.912 density_rs2pw 127 9.7 0.006 0.006 7.362 9.488 dbcsr_mm_accdrv_process 41158 16.2 4.369 6.062 8.180 9.385 pw_transfer 1535 11.6 0.086 0.109 9.188 9.300 multiply_cannon_metrocomm4 17395 15.5 0.062 0.071 3.486 9.250 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.965 9.081 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.772 9.048 mp_irecv_dv 49801 16.2 3.364 9.003 3.364 9.003 fft_wrap_pw1pw2_140 519 13.2 0.474 0.512 7.931 8.057 grid_integrate_task_list 127 12.3 7.199 7.719 7.199 7.719 cp_fm_upper_to_full 104 14.5 5.857 7.529 5.857 7.529 wfi_extrapolate 11 7.9 0.001 0.001 7.474 7.474 rs_pw_transfer 1038 11.9 0.014 0.014 5.269 7.358 fft3d_ps 1281 14.7 2.643 2.864 7.243 7.322 cp_fm_cholesky_decompose 22 10.9 7.270 7.299 7.270 7.299 calculate_dm_sparse 127 9.5 0.001 0.001 6.931 7.065 dbcsr_complete_redistribute 393 12.7 1.175 1.208 4.638 6.421 grid_collocate_task_list 127 9.7 4.858 5.700 4.858 5.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.504 5.508 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.598 5.263 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.441 5.222 mp_alltoall_d11v 2401 14.1 4.444 5.104 4.444 5.104 potential_pw2rs 127 12.3 0.020 0.022 4.880 4.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.599 4.761 mp_allgather_i34 2485 14.5 1.750 4.712 1.750 4.712 mp_sum_l 7804 13.0 3.114 4.632 3.114 4.632 mp_waitany 11660 13.9 2.416 4.550 2.416 4.550 calculate_first_density_matrix 1 7.0 0.016 0.041 4.412 4.432 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.343 4.108 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.345 1.962 4.053 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.993 4.028 mp_alltoall_i22 712 14.1 1.947 3.906 1.947 3.906 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.848 3.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.594 3.638 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.647000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.545455, yerr=9.866045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.159713E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4071 57384. MP_Allreduce 11216 1069. MP_Sync 168 MP_Alltoall 1700 12496391. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.134 187.306 187.308 qs_mol_dyn_low 1 2.0 0.006 0.020 186.630 186.644 qs_forces 11 3.9 0.009 0.034 186.483 186.489 qs_energies 11 4.9 0.002 0.004 179.262 179.272 scf_env_do_scf 11 5.9 0.010 0.025 162.222 162.229 velocity_verlet 10 3.0 0.003 0.004 123.234 123.237 scf_env_do_scf_inner_loop 116 6.6 0.015 0.042 115.787 115.791 dbcsr_multiply_generic 2485 12.5 0.189 0.193 78.623 79.178 qs_scf_new_mos 116 7.6 0.001 0.001 78.497 78.799 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.496 78.798 ot_scf_mini 116 9.6 0.004 0.005 74.076 74.348 multiply_cannon 2485 13.5 0.552 0.591 54.436 57.181 multiply_cannon_loop 2485 14.5 1.170 1.197 50.798 52.188 init_scf_loop 11 6.9 0.001 0.003 46.295 46.295 prepare_preconditioner 11 7.9 0.000 0.000 42.151 42.184 make_preconditioner 11 8.9 0.000 0.002 42.151 42.184 ot_mini 116 10.6 0.002 0.008 41.873 42.145 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.954 40.878 multiply_cannon_multrec 29820 15.5 14.100 18.772 26.246 30.920 rebuild_ks_matrix 127 8.3 0.001 0.001 29.820 30.012 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.033 29.819 30.011 mp_waitall_1 146592 16.7 16.691 27.224 16.691 27.224 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.816 26.987 qs_ot_get_derivative 116 11.6 0.001 0.002 22.418 22.691 make_m2s 4970 13.5 0.095 0.099 19.812 20.866 make_images 4970 14.5 1.939 2.269 19.506 20.562 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.899 19.392 apply_single 127 13.6 0.001 0.001 18.898 19.391 ot_diis_step 116 11.6 0.020 0.037 19.337 19.340 qs_ot_get_p 127 10.4 0.001 0.001 18.671 18.960 cp_fm_cholesky_invert 11 10.9 16.349 16.358 16.349 16.358 cp_fm_upper_to_full 104 14.7 10.896 16.138 10.896 16.138 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.219 15.274 sum_up_and_integrate 127 10.3 0.139 0.152 14.722 14.750 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.637 14.688 integrate_v_rspace 127 11.3 0.004 0.005 14.583 14.615 qs_rho_update_rho_low 127 7.7 0.001 0.002 14.524 14.558 calculate_rho_elec 127 8.7 0.174 0.189 14.523 14.558 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.365 13.366 multiply_cannon_sync_h2d 29820 15.5 11.690 12.901 11.690 12.901 dbcsr_complete_redistribute 393 12.7 1.501 1.615 8.714 12.389 make_images_data 4970 15.5 0.066 0.070 10.539 12.303 dbcsr_mm_accdrv_process 61748 16.2 7.509 8.369 11.724 12.246 hybrid_alltoall_any 5155 16.4 0.523 2.198 9.547 11.444 init_scf_run 11 5.9 0.000 0.001 11.113 11.115 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.113 11.114 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.374 11.047 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.314 10.317 cp_fm_redistribute_end 82 14.4 1.784 10.237 1.799 10.243 cp_fm_diag_elpa_base 82 14.4 7.847 9.636 8.412 10.111 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.174 9.770 pw_transfer 1535 11.6 0.087 0.103 9.391 9.456 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.262 9.456 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.165 9.238 mp_alltoall_i22 712 14.1 5.499 9.135 5.499 9.135 density_rs2pw 127 9.7 0.006 0.006 7.058 8.462 fft_wrap_pw1pw2_140 519 13.2 0.481 0.490 8.120 8.206 grid_integrate_task_list 127 12.3 7.489 7.860 7.489 7.860 cp_fm_cholesky_decompose 22 10.9 7.605 7.714 7.605 7.714 wfi_extrapolate 11 7.9 0.001 0.001 7.633 7.633 fft3d_ps 1281 14.7 2.762 2.846 7.401 7.469 multiply_cannon_metrocomm4 24850 15.5 0.075 0.085 2.702 7.132 mp_irecv_dv 75445 16.2 2.560 6.876 2.560 6.876 calculate_dm_sparse 127 9.5 0.001 0.001 6.442 6.537 rs_pw_transfer 1038 11.9 0.013 0.014 4.711 6.188 grid_collocate_task_list 127 9.7 5.035 5.791 5.035 5.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.399 5.450 mp_alltoall_d11v 2401 14.1 4.532 5.422 4.532 5.422 potential_pw2rs 127 12.3 0.022 0.023 4.672 4.682 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.490 4.493 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.349 4.456 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.372 4.453 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.117 4.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.308000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1084.727273, yerr=29.539160 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.494172E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58279. MP_Allreduce 10990 1175. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.064 174.993 175.001 qs_mol_dyn_low 1 2.0 0.003 0.003 174.485 174.496 qs_forces 11 3.9 0.003 0.003 174.378 174.383 qs_energies 11 4.9 0.002 0.002 166.811 166.819 scf_env_do_scf 11 5.9 0.001 0.001 149.437 149.440 velocity_verlet 10 3.0 0.011 0.011 113.614 113.618 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 113.146 113.147 qs_scf_new_mos 116 7.6 0.001 0.001 74.545 74.641 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.544 74.641 dbcsr_multiply_generic 2485 12.5 0.182 0.188 73.827 74.354 ot_scf_mini 116 9.6 0.003 0.004 70.171 70.251 multiply_cannon 2485 13.5 0.579 0.612 54.515 58.741 multiply_cannon_loop 2485 14.5 0.441 0.455 49.929 50.858 ot_mini 116 10.6 0.001 0.001 39.576 39.672 init_scf_loop 11 6.9 0.000 0.000 36.136 36.137 mp_waitall_1 124680 16.7 25.947 32.953 25.947 32.953 prepare_preconditioner 11 7.9 0.000 0.000 32.028 32.069 make_preconditioner 11 8.9 0.000 0.000 32.028 32.069 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.903 30.163 rebuild_ks_matrix 127 8.3 0.001 0.001 30.022 30.114 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 30.022 30.114 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.258 27.337 multiply_cannon_multrec 9940 15.5 10.377 16.098 17.744 22.011 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.855 20.255 qs_ot_get_derivative 116 11.6 0.001 0.002 19.727 19.802 ot_diis_step 116 11.6 0.019 0.020 19.778 19.780 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.370 19.672 apply_single 127 13.6 0.001 0.001 19.370 19.671 cp_fm_cholesky_invert 11 10.9 18.479 18.484 18.479 18.484 make_m2s 4970 13.5 0.064 0.068 15.768 18.059 qs_ot_get_p 127 10.4 0.001 0.001 17.775 17.924 make_images 4970 14.5 2.239 2.573 15.465 17.751 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.413 15.450 calculate_rho_elec 127 8.7 0.257 0.267 15.412 15.450 sum_up_and_integrate 127 10.3 0.180 0.190 15.099 15.175 integrate_v_rspace 127 11.3 0.004 0.004 14.918 14.999 qs_ot_p2m_diag 82 11.4 0.489 0.496 14.039 14.054 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.902 12.905 multiply_cannon_sync_h2d 9940 15.5 11.474 12.116 11.474 12.116 make_images_data 4970 15.5 0.054 0.062 9.582 12.050 hybrid_alltoall_any 5155 16.4 0.832 3.768 9.461 11.814 init_scf_run 11 5.9 0.000 0.001 10.584 10.584 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.584 10.584 pw_transfer 1535 11.6 0.085 0.091 9.955 10.023 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.856 9.867 cp_fm_diag_elpa_base 82 14.4 9.621 9.697 9.853 9.863 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.733 9.802 fft_wrap_pw1pw2_140 519 13.2 0.493 0.514 8.590 8.663 grid_integrate_task_list 127 12.3 7.719 8.153 7.719 8.153 cp_fm_cholesky_decompose 22 10.9 7.971 8.050 7.971 8.050 fft3d_ps 1281 14.7 2.688 2.779 7.903 7.950 density_rs2pw 127 9.7 0.005 0.006 7.075 7.943 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.865 7.926 dbcsr_mm_accdrv_process 20590 16.1 2.737 3.745 7.004 7.619 wfi_extrapolate 11 7.9 0.001 0.001 7.384 7.384 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.251 7.251 mp_allgather_i34 2485 14.5 2.807 7.173 2.807 7.173 calculate_dm_sparse 127 9.5 0.001 0.001 6.023 6.124 grid_collocate_task_list 127 9.7 5.351 6.055 5.351 6.055 mp_alltoall_d11v 2401 14.1 5.027 5.655 5.027 5.655 dbcsr_complete_redistribute 393 12.7 2.106 2.156 5.221 5.613 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.866 5.568 mp_irecv_dv 28618 15.9 1.830 5.487 1.830 5.487 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.292 5.292 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.135 5.141 rs_pw_transfer 1038 11.9 0.012 0.013 4.100 5.001 potential_pw2rs 127 12.3 0.026 0.026 4.617 4.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.078 4.108 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.539 3.852 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.713 3.755 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.382 3.712 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.667 3.679 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.573 3.642 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.517 3.556 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.001000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1387.363636, yerr=40.387380 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410026932224 0.0% 0.0% 100.0% flops 32 x 32 x 32 1930587799552 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957877121024 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444723269632 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019199483904 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019199483904 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.803070E+12 0.0% 0.0% 100.0% flops max/rank 11.607223E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705605184 0.0% 0.0% 100.0% number of processed stacks 1948912 0.0% 0.0% 100.0% average stack size 0.0 0.0 3440.7 marketing flops 143.514185E+12 ------------------------------------------------------------------------------- # multiplications 2488 max memory usage/rank 2.896638E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99520 MPI messages size (bytes): total size 1.127523E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.329611E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44888 34846277632 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716109087456 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4046 58520. MP_Allreduce 11112 1512. MP_Sync 87 MP_Alltoall 1700 36954293. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.101 298.620 298.627 qs_mol_dyn_low 1 2.0 0.003 0.003 297.829 297.843 qs_forces 11 3.9 0.003 0.003 297.726 297.731 qs_energies 11 4.9 0.001 0.001 288.330 288.344 scf_env_do_scf 11 5.9 0.001 0.001 265.614 265.626 velocity_verlet 10 3.0 0.015 0.016 213.585 213.593 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.444 141.446 init_scf_loop 11 6.9 0.000 0.000 123.912 123.914 prepare_preconditioner 11 7.9 0.000 0.000 118.631 118.663 make_preconditioner 11 8.9 0.000 0.000 118.631 118.663 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.039 115.803 qs_scf_new_mos 116 7.6 0.001 0.001 90.482 90.632 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.481 90.631 ot_scf_mini 116 9.6 0.004 0.004 85.731 85.786 dbcsr_multiply_generic 2488 12.5 0.212 0.219 83.013 83.633 cp_fm_upper_to_full 105 14.8 52.286 74.794 52.286 74.794 multiply_cannon 2488 13.5 0.703 0.769 58.703 59.346 multiply_cannon_loop 2488 14.5 0.465 0.478 55.037 56.578 ot_mini 116 10.6 0.001 0.001 44.282 44.349 dbcsr_complete_redistribute 395 12.7 3.970 4.048 29.282 42.446 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.860 38.910 rebuild_ks_matrix 127 8.3 0.001 0.001 38.548 38.634 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.548 38.634 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.548 36.469 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.474 35.568 mp_alltoall_i22 714 14.1 21.353 34.277 21.353 34.277 cp_fm_cholesky_invert 11 10.9 33.518 33.525 33.518 33.525 mp_waitall_1 102888 16.8 28.231 31.751 28.231 31.751 qs_ot_get_p 127 10.4 0.001 0.001 25.995 26.079 qs_ot_get_derivative 116 11.6 0.002 0.002 24.082 24.141 qs_ot_p2m_diag 83 11.4 0.879 0.885 21.913 21.941 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.777 20.790 calculate_rho_elec 127 8.7 0.479 0.480 20.776 20.789 make_m2s 4976 13.5 0.076 0.078 19.575 20.488 multiply_cannon_metrocomm3 9952 15.5 0.024 0.024 19.069 20.380 ot_diis_step 116 11.6 0.022 0.022 20.161 20.161 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.158 20.159 make_images 4976 14.5 3.748 3.867 19.099 20.010 sum_up_and_integrate 127 10.3 0.319 0.321 19.772 19.838 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.318 19.630 apply_single 127 13.6 0.001 0.001 19.317 19.630 integrate_v_rspace 127 11.3 0.004 0.004 19.452 19.519 multiply_cannon_multrec 9952 15.5 10.428 12.182 17.855 17.937 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.944 16.945 cp_fm_diag_elpa_base 83 14.4 12.670 14.173 16.940 16.940 multiply_cannon_sync_h2d 9952 15.5 15.538 15.548 15.538 15.548 init_scf_run 11 5.9 0.000 0.001 12.610 12.611 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.610 12.610 pw_transfer 1535 11.6 0.092 0.093 12.384 12.397 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.151 12.163 make_images_data 4976 15.5 0.062 0.067 10.367 12.114 hybrid_alltoall_any 5162 16.4 1.289 3.013 10.419 11.998 fft_wrap_pw1pw2_140 519 13.2 0.538 0.539 10.795 10.813 fft3d_ps 1281 14.7 2.724 2.734 10.200 10.212 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.547 9.602 mp_alltoall_d11v 2407 14.1 8.293 9.325 8.293 9.325 wfi_extrapolate 11 7.9 0.001 0.001 9.173 9.173 dbcsr_mm_accdrv_process 20602 16.0 3.791 5.551 7.192 8.977 cp_fm_cholesky_decompose 22 10.9 8.940 8.961 8.940 8.961 grid_integrate_task_list 127 12.3 8.522 8.693 8.522 8.693 density_rs2pw 127 9.7 0.005 0.005 8.394 8.540 qs_energies_init_hamiltonians 11 5.9 0.017 0.022 8.086 8.087 calculate_dm_sparse 127 9.5 0.001 0.001 6.642 6.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.455 6.535 grid_collocate_task_list 127 9.7 6.317 6.347 6.317 6.347 rs_scatter_matrices 138 9.7 3.632 4.509 6.041 6.270 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.082 6.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.627000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2597.727273, yerr=156.008211 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 1.261806E+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 0.078 0.156 86.406 86.407 qs_energies 1 2.0 0.002 0.012 85.561 85.585 ls_scf 1 3.0 0.003 0.022 84.612 84.619 dbcsr_multiply_generic 111 6.7 0.015 0.019 73.079 73.242 multiply_cannon 111 7.7 0.018 0.021 55.910 57.205 multiply_cannon_loop 111 8.7 0.209 0.223 52.485 53.873 ls_scf_main 1 4.0 0.001 0.009 52.445 52.450 density_matrix_trs4 2 5.0 0.005 0.070 46.842 46.930 ls_scf_init_scf 1 4.0 0.000 0.003 29.114 29.133 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.521 27.570 mp_waitall_1 11031 10.9 22.575 26.203 22.575 26.203 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.902 24.926 multiply_cannon_multrec 2664 9.7 8.186 8.910 15.503 17.147 multiply_cannon_sync_h2d 2664 9.7 13.586 15.525 13.586 15.525 make_m2s 222 7.7 0.008 0.010 13.141 13.694 make_images 222 8.7 0.100 0.109 13.119 13.673 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.661 12.677 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.569 8.357 make_images_data 222 9.7 0.004 0.005 7.670 8.276 hybrid_alltoall_any 227 10.6 0.215 1.841 6.542 8.274 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.607 6.935 7.989 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.106 6.209 7.106 calculate_norms 4752 9.8 5.501 6.071 5.501 6.071 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.542 5.647 mp_sum_l 807 5.4 3.517 4.877 3.517 4.877 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.079 3.761 mp_irecv_dv 6231 10.9 2.061 3.730 2.061 3.730 make_images_sizes 222 9.7 0.000 0.000 0.875 3.507 mp_alltoall_i44 222 10.7 0.874 3.507 0.874 3.507 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.255 3.263 arnoldi_extremal 4 6.8 0.000 0.000 3.214 3.252 arnoldi_normal_ev 4 7.8 0.001 0.003 3.214 3.252 build_subspace 16 8.4 0.009 0.012 3.113 3.115 ls_scf_post 1 4.0 0.000 0.001 3.050 3.058 ls_scf_store_result 1 5.0 0.000 0.000 2.873 2.915 dbcsr_special_finalize 555 9.7 0.005 0.006 2.243 2.722 dbcsr_merge_single_wm 555 10.7 0.455 0.598 2.236 2.714 make_images_pack 222 9.7 2.206 2.612 2.208 2.614 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.552 dbcsr_sort_data 658 11.4 2.039 2.462 2.039 2.462 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.453 2.073 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.308 2.394 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.036 2.059 rebuild_ks_matrix 3 7.3 0.000 0.000 2.026 2.049 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.037 2.026 2.049 buffer_matrices_ensure_size 222 8.7 1.750 2.025 1.750 2.025 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.407000, 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/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 2.118816E+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 0.063 0.080 91.797 91.801 qs_energies 1 2.0 0.000 0.001 90.976 90.982 ls_scf 1 3.0 0.000 0.001 89.607 89.612 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.599 75.990 multiply_cannon 111 7.7 0.029 0.093 53.260 57.076 ls_scf_main 1 4.0 0.000 0.001 55.396 55.404 multiply_cannon_loop 111 8.7 0.116 0.122 49.951 53.532 density_matrix_trs4 2 5.0 0.003 0.004 49.667 49.848 ls_scf_init_scf 1 4.0 0.000 0.001 30.694 30.700 mp_waitall_1 9105 10.9 21.421 30.369 21.421 30.369 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.459 29.570 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.053 27.065 multiply_cannon_multrec 1332 9.7 13.062 16.174 22.267 26.477 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.921 22.004 make_m2s 222 7.7 0.006 0.007 15.549 16.311 make_images 222 8.7 1.575 1.949 15.519 16.281 dbcsr_mm_accdrv_process 4041 10.4 0.269 0.447 8.804 10.309 make_images_data 222 9.7 0.004 0.004 8.950 10.040 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.888 8.399 9.888 hybrid_alltoall_any 227 10.6 0.521 2.450 8.286 9.508 mp_sum_l 807 5.4 5.495 8.607 5.495 8.607 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.224 7.649 mp_irecv_dv 3311 11.0 3.204 7.578 3.204 7.578 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.338 7.065 calculate_norms 2376 9.8 6.004 6.765 6.004 6.765 multiply_cannon_sync_h2d 1332 9.7 4.790 5.730 4.790 5.730 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.198 arnoldi_extremal 4 6.8 0.000 0.000 4.606 4.624 arnoldi_normal_ev 4 7.8 0.001 0.005 4.606 4.624 build_subspace 16 8.4 0.014 0.021 4.357 4.360 ls_scf_post 1 4.0 0.000 0.001 3.516 3.521 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.128 3.351 ls_scf_store_result 1 5.0 0.000 0.000 3.199 3.300 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.215 2.740 3.217 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.623 2.709 mp_allgather_i34 111 8.7 1.011 2.538 1.011 2.538 make_images_pack 222 9.7 2.032 2.416 2.034 2.418 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.180 2.316 dbcsr_sort_data 436 11.2 1.845 2.110 1.845 2.110 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.912 1.918 rebuild_ks_matrix 3 7.3 0.000 0.000 1.899 1.905 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.009 1.899 1.905 dbcsr_data_new 4174 10.1 1.607 1.852 1.607 1.852 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.801000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 2.763981E+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 0.091 0.114 94.733 94.734 qs_energies 1 2.0 0.000 0.001 93.790 93.797 ls_scf 1 3.0 0.000 0.001 92.364 92.370 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.962 77.304 ls_scf_main 1 4.0 0.000 0.001 57.763 57.768 multiply_cannon 111 7.7 0.042 0.100 53.143 57.442 multiply_cannon_loop 111 8.7 0.099 0.105 49.576 53.650 density_matrix_trs4 2 5.0 0.002 0.004 51.774 51.958 mp_waitall_1 7281 11.0 23.979 34.340 23.979 34.340 ls_scf_init_scf 1 4.0 0.001 0.003 31.031 31.035 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.766 29.847 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.046 27.063 multiply_cannon_multrec 888 9.7 12.638 15.970 21.334 24.640 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.125 23.264 make_m2s 222 7.7 0.006 0.007 16.962 18.288 make_images 222 8.7 1.969 2.300 16.924 18.248 hybrid_alltoall_any 227 10.6 0.620 2.885 9.304 10.871 make_images_data 222 9.7 0.003 0.004 9.699 10.754 mp_sum_l 807 5.4 5.315 9.348 5.315 9.348 dbcsr_mm_accdrv_process 3754 10.4 0.274 0.653 8.212 9.231 dbcsr_mm_accdrv_process_sort 3754 11.4 7.668 8.812 7.668 8.812 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.132 7.310 multiply_cannon_sync_h2d 888 9.7 6.026 7.248 6.026 7.248 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.778 6.894 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.435 6.859 mp_irecv_dv 2335 11.1 2.420 6.803 2.420 6.803 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.229 5.448 arnoldi_extremal 4 6.8 0.000 0.000 5.078 5.099 arnoldi_normal_ev 4 7.8 0.001 0.005 5.078 5.099 build_subspace 16 8.4 0.014 0.020 4.769 4.776 calculate_norms 1584 9.8 4.261 4.625 4.261 4.625 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.456 3.773 mp_allgather_i34 111 8.7 1.410 3.680 1.410 3.680 dbcsr_matrix_vector_mult_local 304 10.0 3.020 3.594 3.022 3.596 ls_scf_post 1 4.0 0.000 0.001 3.569 3.574 ls_scf_store_result 1 5.0 0.000 0.000 3.261 3.371 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.847 2.969 dbcsr_sort_data 325 11.1 1.907 2.173 1.907 2.173 make_images_pack 222 9.7 1.820 2.135 1.823 2.137 make_images_sizes 222 9.7 0.000 0.000 0.925 2.048 mp_alltoall_i44 222 10.7 0.924 2.047 0.924 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.949 1.952 rebuild_ks_matrix 3 7.3 0.000 0.000 1.931 1.934 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.931 1.934 dbcsr_data_release 9322 10.9 1.306 1.904 1.306 1.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.734000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2158.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 3.313046E+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 0.129 0.164 98.199 98.200 qs_energies 1 2.0 0.000 0.001 97.244 97.251 ls_scf 1 3.0 0.000 0.001 95.553 95.559 dbcsr_multiply_generic 111 6.7 0.017 0.021 79.078 79.348 ls_scf_main 1 4.0 0.000 0.001 59.063 59.067 multiply_cannon 111 7.7 0.058 0.124 52.087 56.555 density_matrix_trs4 2 5.0 0.003 0.020 52.927 53.053 multiply_cannon_loop 111 8.7 0.114 0.126 46.954 49.762 ls_scf_init_scf 1 4.0 0.001 0.003 33.228 33.230 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.860 31.936 mp_waitall_1 6369 11.0 23.208 30.368 23.208 30.368 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.878 28.892 multiply_cannon_multrec 1332 9.7 14.134 17.320 22.018 24.550 make_m2s 222 7.7 0.007 0.008 21.080 22.541 make_images 222 8.7 3.143 3.600 21.030 22.493 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.705 17.884 make_images_data 222 9.7 0.004 0.005 11.786 13.336 hybrid_alltoall_any 227 10.6 0.799 3.829 11.084 13.067 dbcsr_mm_accdrv_process 3641 10.4 0.203 0.400 7.519 9.464 dbcsr_mm_accdrv_process_sort 3641 11.4 7.106 8.603 7.106 8.603 mp_sum_l 807 5.4 4.369 7.740 4.369 7.740 multiply_cannon_sync_h2d 1332 9.7 5.500 6.339 5.500 6.339 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.246 6.133 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.064 5.929 mp_irecv_dv 3229 10.9 2.042 5.861 2.042 5.861 arnoldi_extremal 4 6.8 0.000 0.000 5.302 5.315 arnoldi_normal_ev 4 7.8 0.001 0.005 5.302 5.315 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.071 5.220 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.549 5.146 build_subspace 16 8.4 0.014 0.021 4.966 4.974 mp_allgather_i34 111 8.7 2.254 4.722 2.254 4.722 calculate_norms 2376 9.8 4.184 4.541 4.184 4.541 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.652 3.968 dbcsr_matrix_vector_mult_local 304 10.0 3.207 3.702 3.209 3.704 dbcsr_sort_data 658 11.4 3.050 3.505 3.050 3.505 ls_scf_post 1 4.0 0.001 0.020 3.262 3.269 dbcsr_special_finalize 555 9.7 0.006 0.007 2.792 3.227 dbcsr_merge_single_wm 555 10.7 0.535 0.662 2.784 3.218 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.042 3.116 ls_scf_store_result 1 5.0 0.000 0.000 2.996 3.064 dbcsr_data_release 10477 10.7 1.583 2.452 1.583 2.452 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.048 2.050 dbcsr_finalize 304 7.8 0.049 0.061 1.798 2.043 rebuild_ks_matrix 3 7.3 0.000 0.000 2.025 2.027 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.022 2.025 2.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.200000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 4.629971E+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 0.042 0.058 98.808 98.809 qs_energies 1 2.0 0.000 0.000 98.032 98.039 ls_scf 1 3.0 0.000 0.000 96.076 96.087 dbcsr_multiply_generic 111 6.7 0.017 0.019 77.487 77.765 ls_scf_main 1 4.0 0.000 0.000 61.801 61.802 multiply_cannon 111 7.7 0.091 0.131 54.671 59.693 density_matrix_trs4 2 5.0 0.002 0.003 54.675 54.743 multiply_cannon_loop 111 8.7 0.069 0.080 50.197 53.054 mp_waitall_1 5436 11.0 25.679 31.369 25.679 31.369 ls_scf_init_scf 1 4.0 0.000 0.000 30.658 30.663 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.301 29.336 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.153 27.163 multiply_cannon_multrec 444 9.7 14.096 16.242 21.125 23.848 make_m2s 222 7.7 0.004 0.005 17.478 19.923 make_images 222 8.7 3.723 4.418 17.416 19.864 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.956 16.435 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.763 15.574 make_images_data 222 9.7 0.003 0.004 9.763 12.087 hybrid_alltoall_any 227 10.6 0.791 3.784 9.481 11.888 multiply_cannon_sync_h2d 444 9.7 6.432 8.016 6.432 8.016 dbcsr_mm_accdrv_process 3003 10.4 0.198 0.388 6.735 7.862 dbcsr_mm_accdrv_process_sort 3003 11.4 6.410 7.517 6.410 7.517 mp_allgather_i34 111 8.7 2.710 7.017 2.710 7.017 mp_sum_l 807 5.4 3.369 6.583 3.369 6.583 arnoldi_extremal 4 6.8 0.000 0.000 5.852 5.865 arnoldi_normal_ev 4 7.8 0.011 0.027 5.852 5.865 build_subspace 16 8.4 0.015 0.020 5.434 5.443 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.464 5.165 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.717 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.190 4.388 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.730 4.224 mp_irecv_dv 1241 11.2 1.710 4.198 1.710 4.198 dbcsr_matrix_vector_mult_local 304 10.0 3.712 4.184 3.714 4.186 calculate_norms 792 9.8 3.550 3.748 3.550 3.748 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.550 3.644 ls_scf_post 1 4.0 0.000 0.000 3.617 3.623 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.445 make_images_sizes 222 9.7 0.000 0.000 1.064 3.215 mp_alltoall_i44 222 10.7 1.064 3.215 1.064 3.215 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.264 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.148 2.150 rebuild_ks_matrix 3 7.3 0.000 0.000 2.115 2.116 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.115 2.116 dbcsr_merge_all 275 8.9 0.474 0.520 2.048 2.090 dbcsr_data_release 10123 10.8 1.327 1.989 1.327 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.809000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0e722df91b89847b24d3e2fa39485a0bdfbfba12_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 8.806076E+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 0.089 0.111 108.936 108.937 qs_energies 1 2.0 0.000 0.000 107.451 107.467 ls_scf 1 3.0 0.000 0.000 104.500 104.515 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.741 77.913 ls_scf_main 1 4.0 0.000 0.000 66.398 66.398 density_matrix_trs4 2 5.0 0.002 0.003 57.094 57.170 multiply_cannon 111 7.7 0.112 0.214 49.734 51.386 multiply_cannon_loop 111 8.7 0.070 0.085 46.227 47.712 ls_scf_init_scf 1 4.0 0.000 0.000 34.443 34.444 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.769 32.791 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.960 29.976 mp_waitall_1 4527 11.1 22.024 26.491 22.024 26.491 make_m2s 222 7.7 0.005 0.005 23.972 24.990 make_images 222 8.7 4.597 5.011 23.866 24.882 multiply_cannon_multrec 444 9.7 17.902 18.493 22.544 23.141 make_images_data 222 9.7 0.003 0.003 13.206 15.666 hybrid_alltoall_any 227 10.6 1.661 3.619 12.997 15.621 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.161 11.225 multiply_cannon_sync_h2d 444 9.7 8.840 8.867 8.840 8.867 arnoldi_extremal 4 6.8 0.000 0.000 7.460 7.479 arnoldi_normal_ev 4 7.8 0.003 0.009 7.460 7.479 build_subspace 16 8.4 0.026 0.036 6.890 6.900 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.548 5.694 dbcsr_matrix_vector_mult_local 304 10.0 5.104 5.416 5.106 5.419 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.256 5.349 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.009 5.304 dbcsr_mm_accdrv_process 1814 10.4 0.260 0.398 4.469 4.640 dbcsr_mm_accdrv_process_sort 1814 11.4 4.130 4.263 4.130 4.263 ls_scf_post 1 4.0 0.000 0.000 3.659 3.675 mp_allgather_i34 111 8.7 1.158 3.630 1.158 3.630 make_images_sizes 222 9.7 0.000 0.000 1.424 3.482 mp_alltoall_i44 222 10.7 1.424 3.481 1.424 3.481 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.421 calculate_norms 792 9.8 3.248 3.272 3.248 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.101 3.179 dbcsr_merge_all 275 8.9 0.890 0.917 2.875 2.958 qs_energies_init_hamiltonians 1 3.0 0.017 0.033 2.920 2.920 dbcsr_complete_redistribute 5 7.6 1.435 1.470 2.784 2.896 dbcsr_data_release 12724 10.6 2.358 2.895 2.358 2.895 mp_sum_l 807 5.4 1.704 2.823 1.704 2.823 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.594 2.595 matrix_ls_to_qs 2 6.0 0.000 0.000 2.417 2.534 rebuild_ks_matrix 3 7.3 0.000 0.000 2.528 2.529 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.528 2.529 dbcsr_sort_data 325 11.1 2.445 2.517 2.445 2.517 dbcsr_new_transposed 4 7.5 0.243 0.251 2.295 2.312 dbcsr_frobenius_norm 74 6.6 2.058 2.138 2.209 2.260 dbcsr_add_d 103 6.2 0.000 0.000 2.147 2.206 dbcsr_add_anytype 103 7.2 0.860 0.894 2.147 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.937000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6814.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 0e722df91b89847b24d3e2fa39485a0bdfbfba12 Summary: empty Status: OK