=== 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: c94756ddc75c22fb829516d4d1ee9958f089b1c9 ################# 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 2.6.6, ELPA 2023.05.001, # HDF5 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 13.09.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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.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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/01 job id: 49135648 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/02 job id: 49135649 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/03 job id: 49135650 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/04 job id: 49135651 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/05 job id: 49135652 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/06 job id: 49135653 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/07 job id: 49135654 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/08 job id: 49135655 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/09 job id: 49135656 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/10 job id: 49135657 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/11 job id: 49135658 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/12 job id: 49135660 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/13 job id: 49135662 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/14 job id: 49135663 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/15 job id: 49135664 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/16 job id: 49135665 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/17 job id: 49135666 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/18 job id: 49135667 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/19 job id: 49135668 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/20 job id: 49135669 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/21 job id: 49135670 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/22 job id: 49135671 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/23 job id: 49135672 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/24 job id: 49135673 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/25 job id: 49135674 --- 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/26 job id: 49135675 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/27 job id: 49135676 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 424 8. 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 0.020 0.037 144.885 144.886 farming_run 1 2.0 144.293 144.295 144.837 144.842 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.498227E+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 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 119.102 119.103 qs_energies 1 2.0 0.000 0.000 118.813 118.814 mp2_main 1 3.0 0.000 0.000 115.073 115.075 mp2_gpw_main 1 4.0 0.030 0.041 113.474 113.476 mp2_ri_gpw_compute_in 1 5.0 0.180 0.202 92.560 92.876 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.664 54.987 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 41.055 46.117 get_2c_integrals 1 6.0 0.008 0.009 37.089 37.716 integrate_v_rspace 273 8.0 0.440 0.454 24.630 29.534 pw_transfer 6555 10.6 0.384 0.395 26.833 27.316 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.462 26.040 grid_integrate_task_list 273 9.0 20.515 25.897 20.515 25.897 fft_wrap_pw1pw2_100 2178 12.4 1.095 1.301 22.978 23.530 rpa_ri_compute_en 1 5.0 0.020 0.020 20.790 20.920 cp_fm_cholesky_decompose 12 8.2 18.396 19.332 18.396 19.332 compute_2c_integrals 1 7.0 0.002 0.003 19.294 19.294 compute_2c_integrals_loop_lm 1 8.0 0.024 0.025 18.561 18.633 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.409 18.537 18.609 cholesky_decomp 1 7.0 0.000 0.000 16.638 17.289 fft3d_s 5443 13.4 16.104 16.438 16.127 16.462 ao_to_mo_and_store_B_mult_1 272 7.0 10.775 15.314 10.775 15.314 calculate_wavefunction 272 8.0 5.418 5.574 12.345 13.030 rpa_num_int 1 6.0 0.000 0.002 11.623 11.632 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 11.271 11.611 calc_mat_Q 8 8.0 0.000 0.000 9.534 9.670 contract_S_to_Q 8 9.0 0.000 0.000 8.954 9.090 create_integ_mat 1 6.0 0.014 0.027 8.750 8.760 parallel_gemm_fm 14 9.1 0.000 0.000 8.568 8.705 parallel_gemm_fm_cosma 14 10.1 8.568 8.705 8.568 8.705 calc_potential_gpw 544 9.5 0.005 0.006 8.238 8.672 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.047 8.252 potential_pw2rs 545 10.0 0.106 0.108 7.479 8.116 collocate_single_gaussian 272 10.0 0.040 0.042 7.305 7.475 array2fm 1 7.0 0.000 0.000 6.756 7.201 pw_scatter_s 2720 13.7 4.372 4.613 4.372 4.613 pw_gather_s 2722 13.2 3.446 3.953 3.446 3.953 scf_env_do_scf 1 3.0 0.000 0.000 3.363 3.363 init_scf_loop 1 4.0 0.000 0.000 3.167 3.167 array2fm_buffer_send 1 8.0 3.045 3.143 3.045 3.143 prepare_preconditioner 1 5.0 0.000 0.000 2.888 2.889 cp_fm_syevd 4 7.5 0.000 0.000 2.831 2.836 make_preconditioner 1 6.0 0.000 0.000 2.803 2.803 make_full_all 1 7.0 0.000 0.000 2.726 2.727 cp_fm_redistribute_end 4 8.5 1.281 2.554 1.281 2.554 cp_fm_syevd_base 4 8.5 1.270 2.541 1.270 2.541 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.474129, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 424 9. 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 0.024 0.036 425.285 425.287 farming_run 1 2.0 424.514 424.517 425.249 425.251 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.240162E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 216.971 216.972 qs_energies 1 2.0 0.000 0.000 216.728 216.734 scf_env_do_scf 1 3.0 0.000 0.000 114.820 114.821 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.834 113.839 rebuild_ks_matrix 4 6.0 0.000 0.000 113.832 113.838 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 113.832 113.838 hfx_ks_matrix 4 8.0 0.001 0.001 113.476 113.480 integrate_four_center 4 9.0 0.154 0.471 113.475 113.480 integrate_four_center_main 4 10.0 0.120 0.464 101.953 104.854 integrate_four_center_bin 265 11.0 101.833 104.838 101.833 104.838 mp2_main 1 3.0 0.000 0.000 101.610 101.616 mp2_gpw_main 1 4.0 0.055 0.080 100.705 100.715 init_scf_loop 1 4.0 0.000 0.000 96.791 96.792 mp2_ri_gpw_compute_in 1 5.0 0.106 0.149 73.938 74.954 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.786 54.809 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.157 41.656 46.514 integrate_v_rspace 95 8.0 0.397 0.561 28.130 32.889 pw_transfer 2240 10.6 0.148 0.166 29.468 30.089 ao_to_mo_and_store_B_mult_1 91 7.0 10.451 29.483 10.451 29.483 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.494 29.063 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 26.600 28.675 grid_integrate_task_list 95 9.0 23.542 28.506 23.542 28.506 fft_wrap_pw1pw2_100 730 12.4 1.245 1.494 26.233 26.754 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.893 24.902 24.913 get_2c_integrals 1 6.0 0.000 0.000 20.032 20.095 compute_2c_integrals 1 7.0 0.003 0.004 19.008 19.062 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.512 18.921 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.824 18.510 18.919 fft3d_s 1823 13.4 18.478 18.840 18.493 18.854 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.027 18.028 calculate_wavefunction 91 8.0 2.016 2.053 9.624 9.861 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.752 9.187 potential_pw2rs 186 10.0 0.033 0.034 8.398 9.029 local_gemm 172 8.0 8.196 8.624 8.196 8.624 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.086 8.536 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.517 7.906 8.421 calc_potential_gpw 182 9.5 0.002 0.002 7.865 8.116 collocate_single_gaussian 91 10.0 0.017 0.022 7.704 7.967 integrate_four_center_load 4 10.0 0.000 0.000 6.744 6.747 hfx_load_balance 1 11.0 0.000 0.000 6.744 6.747 mp_sync 37 10.5 3.498 6.583 3.498 6.583 mp_sendrecv_dm3 2068 8.0 5.928 6.439 5.928 6.439 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.429 6.355 6.429 pw_gather_s 912 13.2 4.495 5.057 4.495 5.057 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.701966, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 456.347648E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. 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 0.012 0.042 57.424 57.430 qs_mol_dyn_low 1 2.0 0.003 0.003 57.082 57.161 qs_forces 11 3.9 0.002 0.003 56.997 56.999 qs_energies 11 4.9 0.002 0.002 55.457 55.479 scf_env_do_scf 11 5.9 0.000 0.001 46.945 46.945 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.658 44.659 dbcsr_multiply_generic 2286 12.5 0.095 0.100 35.921 36.481 qs_scf_new_mos 108 7.5 0.000 0.001 34.064 34.382 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.064 34.382 ot_scf_mini 108 9.5 0.002 0.002 32.391 32.560 multiply_cannon 2286 13.5 0.192 0.199 28.662 30.470 multiply_cannon_loop 2286 14.5 1.754 1.838 27.859 29.681 velocity_verlet 10 3.0 0.004 0.004 27.065 27.065 ot_mini 108 10.5 0.001 0.001 19.449 19.709 mp_waitall_1 245248 16.5 9.831 16.851 9.831 16.851 qs_ot_get_derivative 108 11.5 0.001 0.001 16.462 16.625 multiply_cannon_metrocomm3 54864 15.5 0.075 0.081 6.602 14.096 multiply_cannon_multrec 54864 15.5 3.810 5.904 8.853 13.381 qs_ot_get_p 119 10.4 0.066 0.075 8.377 8.658 rebuild_ks_matrix 119 8.3 0.000 0.000 8.116 8.265 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.116 8.265 dbcsr_mm_accdrv_process 76910 16.1 1.689 2.683 4.958 7.625 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.166 7.303 init_scf_run 11 5.9 0.000 0.001 7.183 7.183 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.183 7.183 mp_sum_l 7287 12.8 4.978 6.949 4.978 6.949 multiply_cannon_sync_h2d 54864 15.5 5.180 6.827 5.180 6.827 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.903 6.371 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.346 5.432 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.143 5.270 jit_kernel_multiply 13 15.8 3.209 5.075 3.209 5.075 calculate_first_density_matrix 1 7.0 0.001 0.001 4.918 4.924 calculate_dm_sparse 119 9.5 0.000 0.001 4.603 4.764 multiply_cannon_metrocomm1 54864 15.5 0.057 0.061 2.401 4.732 sum_up_and_integrate 119 10.3 0.001 0.002 4.682 4.689 integrate_v_rspace 119 11.3 0.002 0.002 4.671 4.680 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.511 4.647 calculate_rho_elec 119 8.7 0.011 0.016 4.511 4.646 cp_dbcsr_syevd 50 12.0 0.003 0.004 4.568 4.633 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.309 4.310 cp_fm_redistribute_end 50 14.0 2.195 4.273 2.205 4.277 cp_fm_diag_elpa_base 50 14.0 2.063 4.163 2.069 4.170 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.108 3.157 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.933 3.151 apply_single 119 13.6 0.000 0.000 2.933 3.151 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.974 2.976 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.675 2.744 ot_diis_step 108 11.5 0.006 0.006 2.731 2.732 acc_transpose_blocks 54864 15.5 0.212 0.239 2.103 2.647 density_rs2pw 119 9.7 0.004 0.004 2.498 2.595 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.362 2.419 init_scf_loop 11 6.9 0.000 0.000 2.262 2.263 wfi_extrapolate 11 7.9 0.001 0.001 2.191 2.191 grid_integrate_task_list 119 12.3 2.018 2.127 2.018 2.127 potential_pw2rs 119 12.3 0.004 0.004 2.015 2.035 make_m2s 4572 13.5 0.054 0.056 1.971 2.033 pw_transfer 1439 11.6 0.052 0.058 1.936 2.024 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.860 1.950 make_images 4572 14.5 0.134 0.139 1.887 1.948 mp_sum_d 4135 12.0 1.360 1.940 1.360 1.940 mp_alltoall_d11v 2130 13.8 1.478 1.694 1.478 1.694 transfer_rs2pw 487 10.6 0.005 0.006 1.560 1.666 fft3d_ps 1201 14.6 0.371 0.476 1.577 1.665 fft_wrap_pw1pw2_140 487 13.2 0.139 0.156 1.439 1.529 transfer_pw2rs 487 13.2 0.006 0.006 1.438 1.457 mp_waitany 12084 13.8 1.299 1.456 1.299 1.456 grid_collocate_task_list 119 9.7 1.354 1.414 1.354 1.414 acc_transpose_blocks_sync 164592 16.5 1.091 1.326 1.091 1.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.430000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=435.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 489.398272E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1032117. 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 0.010 0.030 44.233 44.234 qs_mol_dyn_low 1 2.0 0.003 0.003 43.981 43.991 qs_forces 11 3.9 0.002 0.003 43.915 43.916 qs_energies 11 4.9 0.002 0.002 42.181 42.187 scf_env_do_scf 11 5.9 0.000 0.001 34.356 34.356 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.688 31.688 dbcsr_multiply_generic 2286 12.5 0.103 0.106 25.632 26.086 qs_scf_new_mos 108 7.5 0.001 0.001 22.200 22.429 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.200 22.428 multiply_cannon 2286 13.5 0.209 0.217 19.855 21.745 ot_scf_mini 108 9.5 0.002 0.003 21.210 21.371 multiply_cannon_loop 2286 14.5 1.198 1.257 18.544 20.437 velocity_verlet 10 3.0 0.001 0.002 19.762 19.763 mp_waitall_1 200699 16.5 7.358 13.536 7.358 13.536 ot_mini 108 10.5 0.001 0.001 12.985 13.225 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 5.269 11.954 multiply_cannon_multrec 27432 15.5 1.817 4.199 7.430 11.150 qs_ot_get_derivative 108 11.5 0.001 0.001 10.527 10.692 dbcsr_mm_accdrv_process 47894 16.0 4.608 8.021 5.534 8.759 rebuild_ks_matrix 119 8.3 0.000 0.000 7.190 7.362 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 7.190 7.362 init_scf_run 11 5.9 0.000 0.001 6.584 6.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.584 6.584 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.339 6.492 qs_ot_get_p 119 10.4 0.002 0.003 5.170 5.386 calculate_first_density_matrix 1 7.0 0.001 0.001 4.942 4.945 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.001 4.837 mp_sum_l 7287 12.8 2.342 4.324 2.342 4.324 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.143 4.175 calculate_rho_elec 119 8.7 0.021 0.024 4.143 4.174 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.037 4.104 apply_single 119 13.6 0.000 0.000 3.037 4.104 sum_up_and_integrate 119 10.3 0.001 0.001 4.040 4.045 integrate_v_rspace 119 11.3 0.002 0.003 4.025 4.031 calculate_dm_sparse 119 9.5 0.000 0.001 3.925 4.007 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.330 3.357 make_m2s 4572 13.5 0.053 0.055 2.984 3.288 make_images 4572 14.5 0.206 0.247 2.895 3.203 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.063 3.072 jit_kernel_multiply 9 16.4 0.868 2.963 0.868 2.963 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.880 2.882 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.776 2.779 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.585 2.670 init_scf_loop 11 6.9 0.000 0.000 2.643 2.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.474 2.519 density_rs2pw 119 9.7 0.004 0.004 2.408 2.494 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.466 2.467 cp_fm_redistribute_end 50 14.0 1.249 2.429 1.255 2.434 ot_diis_step 108 11.5 0.011 0.011 2.384 2.385 cp_fm_diag_elpa_base 50 14.0 1.142 2.317 1.173 2.356 multiply_cannon_sync_h2d 27432 15.5 1.639 2.185 1.639 2.185 multiply_cannon_metrocomm1 27432 15.5 0.032 0.036 0.531 2.162 pw_transfer 1439 11.6 0.065 0.069 2.093 2.138 acc_transpose_blocks 27432 15.5 0.113 0.119 1.733 2.103 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.001 2.048 grid_integrate_task_list 119 12.3 1.836 1.914 1.836 1.914 make_images_data 4572 15.5 0.048 0.054 1.371 1.828 potential_pw2rs 119 12.3 0.006 0.006 1.748 1.760 prepare_preconditioner 11 7.9 0.000 0.000 1.715 1.742 make_preconditioner 11 8.9 0.000 0.000 1.715 1.742 fft3d_ps 1201 14.6 0.520 0.574 1.637 1.679 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.609 1.668 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.220 1.634 wfi_extrapolate 11 7.9 0.001 0.001 1.576 1.577 mp_alltoall_d11v 2130 13.8 1.406 1.542 1.406 1.542 transfer_rs2pw 487 10.6 0.005 0.005 1.426 1.535 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.483 1.525 grid_collocate_task_list 119 9.7 1.293 1.361 1.293 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.356 mp_allgather_i34 2286 14.5 0.733 1.177 0.733 1.177 transfer_pw2rs 487 13.2 0.005 0.005 1.134 1.146 mp_sum_d 4135 12.0 0.611 1.116 0.611 1.116 make_images_sizes 4572 15.5 0.005 0.005 0.748 1.060 mp_alltoall_i44 4572 16.5 0.743 1.055 0.743 1.055 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.775 1.012 acc_transpose_blocks_sync 82296 16.5 0.820 0.945 0.820 0.945 mp_alltoall_z22v 1201 16.6 0.842 0.939 0.842 0.939 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.929 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=44.234000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=1.671343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 522.326016E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.125 41.620 41.622 qs_mol_dyn_low 1 2.0 0.003 0.004 37.946 41.237 qs_forces 11 3.9 0.003 0.008 37.243 37.244 qs_energies 11 4.9 0.005 0.026 35.603 35.608 scf_env_do_scf 11 5.9 0.003 0.019 28.196 28.196 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.418 25.419 dbcsr_multiply_generic 2286 12.5 0.098 0.100 20.462 20.628 multiply_cannon 2286 13.5 0.195 0.199 16.553 17.335 velocity_verlet 10 3.0 0.001 0.002 17.171 17.189 qs_scf_new_mos 108 7.5 0.001 0.001 17.041 17.055 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.041 17.055 multiply_cannon_loop 2286 14.5 0.862 0.891 15.675 16.625 ot_scf_mini 108 9.5 0.002 0.002 16.218 16.231 ot_mini 108 10.5 0.001 0.001 9.902 9.921 multiply_cannon_multrec 18288 15.5 1.878 2.936 8.954 9.431 qs_ot_get_derivative 108 11.5 0.001 0.001 8.343 8.357 dbcsr_mm_accdrv_process 38222 16.0 6.926 8.148 6.983 8.213 rebuild_ks_matrix 119 8.3 0.000 0.000 6.355 6.388 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.355 6.388 init_scf_run 11 5.9 0.000 0.003 6.099 6.099 scf_env_initial_rho_setup 11 6.9 0.002 0.002 6.099 6.099 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.595 5.623 mp_waitall_1 158411 16.6 3.716 5.000 3.716 5.000 calculate_first_density_matrix 1 7.0 0.002 0.006 4.724 4.725 md_output 10 3.0 0.000 0.000 0.127 3.970 md_write_output 11 3.9 0.002 0.037 0.125 3.969 qs_ot_get_p 119 10.4 0.002 0.003 3.920 3.958 write_trajectory 44 4.9 0.001 0.009 0.123 3.932 write_particle_coordinates 11 5.9 0.123 3.923 0.123 3.923 sum_up_and_integrate 119 10.3 0.001 0.001 3.882 3.887 integrate_v_rspace 119 11.3 0.002 0.003 3.869 3.876 calculate_dm_sparse 119 9.5 0.000 0.001 3.765 3.774 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.179 3.740 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.687 3.695 calculate_rho_elec 119 8.7 0.031 0.032 3.687 3.695 mp_max_l 33 2.8 3.290 3.399 3.290 3.399 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.991 3.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.046 3.054 init_scf_loop 11 6.9 0.001 0.002 2.737 2.738 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.584 2.596 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.011 2.394 apply_single 119 13.6 0.000 0.000 2.011 2.394 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.353 2.354 make_m2s 4572 13.5 0.045 0.046 2.155 2.351 density_rs2pw 119 9.7 0.004 0.004 2.236 2.347 make_images 4572 14.5 0.194 0.207 2.068 2.264 cp_dbcsr_syevd 50 12.0 0.004 0.004 2.259 2.260 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.162 2.171 pw_transfer 1439 11.6 0.065 0.069 2.077 2.092 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.984 2.000 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.966 1.971 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.954 1.965 cp_fm_diag_elpa_base 50 14.0 1.929 1.940 1.952 1.963 grid_integrate_task_list 119 12.3 1.805 1.925 1.805 1.925 prepare_preconditioner 11 7.9 0.000 0.000 1.917 1.919 make_preconditioner 11 8.9 0.000 0.001 1.916 1.919 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.770 1.857 mp_sum_l 7287 12.8 1.410 1.854 1.410 1.854 acc_transpose_blocks 18288 15.5 0.080 0.082 1.744 1.813 potential_pw2rs 119 12.3 0.007 0.009 1.650 1.662 fft3d_ps 1201 14.6 0.531 0.551 1.552 1.565 ot_diis_step 108 11.5 0.012 0.012 1.527 1.527 fft_wrap_pw1pw2_140 487 13.2 0.212 0.217 1.493 1.507 transfer_rs2pw 487 10.6 0.005 0.005 1.282 1.401 grid_collocate_task_list 119 9.7 1.233 1.315 1.233 1.315 wfi_extrapolate 11 7.9 0.001 0.001 1.310 1.310 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.595 1.239 make_images_data 4572 15.5 0.048 0.052 0.944 1.156 multiply_cannon_sync_h2d 18288 15.5 0.997 1.146 0.997 1.146 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.085 1.103 transfer_pw2rs 487 13.2 0.004 0.005 1.041 1.049 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.805 1.042 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.994 0.996 acc_transpose_blocks_kernels 18288 16.5 0.217 0.226 0.897 0.914 mp_alltoall_z22v 1201 16.6 0.831 0.901 0.831 0.901 mp_alltoall_d11v 2130 13.8 0.803 0.900 0.803 0.900 cp_fm_cholesky_invert 11 10.9 0.874 0.878 0.874 0.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.836 0.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=41.622000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=1.799908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 557.584384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 57.505 57.506 qs_mol_dyn_low 1 2.0 0.003 0.003 57.182 57.195 qs_forces 11 3.9 0.002 0.003 57.097 57.100 qs_energies 11 4.9 0.005 0.006 54.996 55.008 scf_env_do_scf 11 5.9 0.000 0.001 46.824 46.825 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 40.310 40.311 velocity_verlet 10 3.0 0.002 0.002 28.754 28.757 dbcsr_multiply_generic 2286 12.5 0.104 0.106 27.054 27.421 qs_scf_new_mos 108 7.5 0.001 0.001 26.588 26.674 qs_scf_loop_do_ot 108 8.5 0.001 0.001 26.587 26.673 ot_scf_mini 108 9.5 0.002 0.003 25.311 25.367 multiply_cannon 2286 13.5 0.220 0.225 19.550 20.597 multiply_cannon_loop 2286 14.5 1.521 1.623 17.799 18.928 ot_mini 108 10.5 0.001 0.001 14.280 14.363 qs_ot_get_derivative 108 11.5 0.001 0.001 11.997 12.057 multiply_cannon_multrec 27432 15.5 2.487 3.178 10.799 11.100 rebuild_ks_matrix 119 8.3 0.000 0.000 9.746 9.870 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 9.746 9.870 dbcsr_mm_accdrv_process 47916 15.9 7.712 9.496 8.209 9.628 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.750 8.860 mp_waitall_1 137007 16.6 6.063 7.792 6.063 7.792 qs_ot_get_p 119 10.4 0.005 0.026 7.420 7.522 qs_rho_update_rho_low 119 7.7 0.001 0.001 6.436 6.461 calculate_rho_elec 119 8.7 0.040 0.046 6.435 6.460 init_scf_run 11 5.9 0.000 0.001 6.459 6.460 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.459 6.460 sum_up_and_integrate 119 10.3 0.001 0.001 6.445 6.452 init_scf_loop 11 6.9 0.000 0.000 6.440 6.442 integrate_v_rspace 119 11.3 0.003 0.003 6.434 6.441 make_m2s 4572 13.5 0.055 0.058 4.846 5.136 prepare_preconditioner 11 7.9 0.000 0.000 5.101 5.110 make_preconditioner 11 8.9 0.000 0.000 5.101 5.110 make_images 4572 14.5 0.273 0.334 4.735 5.022 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.553 4.971 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.364 4.830 density_rs2pw 119 9.7 0.004 0.004 4.716 4.807 qs_ot_p2m_diag 50 11.0 0.015 0.023 4.564 4.575 calculate_first_density_matrix 1 7.0 0.000 0.001 4.036 4.038 potential_pw2rs 119 12.3 0.009 0.009 4.009 4.030 cp_dbcsr_syevd 50 12.0 0.004 0.004 3.988 3.989 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.818 3.821 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 3.572 3.601 pw_transfer 1439 11.6 0.065 0.069 3.442 3.488 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 3.350 3.398 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.337 3.384 cp_fm_diag_elpa 50 13.0 0.000 0.000 3.326 3.334 cp_fm_diag_elpa_base 50 14.0 3.259 3.285 3.320 3.328 transfer_rs2pw 487 10.6 0.004 0.005 3.079 3.168 mp_sum_l 7287 12.8 2.247 3.012 2.247 3.012 transfer_pw2rs 487 13.2 0.004 0.004 2.905 2.926 fft3d_ps 1201 14.6 0.564 0.614 2.872 2.917 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.397 2.847 apply_single 119 13.6 0.000 0.000 2.397 2.846 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 1.173 2.810 make_images_data 4572 15.5 0.048 0.052 2.211 2.673 calculate_dm_sparse 119 9.5 0.000 0.000 2.520 2.587 acc_transpose_blocks 27432 15.5 0.119 0.125 2.397 2.560 jit_kernel_multiply 6 16.1 0.432 2.499 0.432 2.499 fft_wrap_pw1pw2_140 487 13.2 0.243 0.257 2.282 2.332 hybrid_alltoall_any 4725 16.4 0.066 0.155 1.847 2.331 wfi_extrapolate 11 7.9 0.001 0.001 2.330 2.331 mp_sendrecv_dv 8211 12.7 2.201 2.239 2.201 2.239 mp_alltoall_z22v 1201 16.6 2.186 2.226 2.186 2.226 make_images_sizes 4572 15.5 0.005 0.005 1.506 2.201 mp_alltoall_i44 4572 16.5 1.502 2.196 1.502 2.196 cp_fm_cholesky_invert 11 10.9 2.170 2.181 2.170 2.181 ot_diis_step 108 11.5 0.012 0.012 2.172 2.172 dbcsr_complete_redistribute 329 12.2 0.187 0.210 1.625 1.973 grid_integrate_task_list 119 12.3 1.810 1.881 1.810 1.881 mp_alltoall_d11v 2130 13.8 1.815 1.876 1.815 1.876 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.801 1.865 acc_transpose_blocks_sync 82296 16.5 1.389 1.536 1.389 1.536 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.446 1.460 cp_fm_upper_to_full 72 14.2 0.968 1.437 0.968 1.437 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.040 1.386 grid_collocate_task_list 119 9.7 1.250 1.325 1.250 1.325 transfer_rs2pw_50 119 11.7 0.138 0.144 1.235 1.304 transfer_pw2rs_50 119 14.3 0.127 0.135 1.186 1.276 arnoldi_extremal 119 11.4 0.002 0.002 1.227 1.263 arnoldi_normal_ev 119 12.4 0.002 0.003 1.225 1.261 cp_fm_cholesky_decompose 22 10.9 1.223 1.227 1.223 1.227 yz_to_x 606 15.1 0.049 0.057 1.193 1.224 mp_allgather_i34 2286 14.5 1.031 1.185 1.031 1.185 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.173 1.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=57.506000, 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/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 619.008000E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.049 31.031 31.039 qs_mol_dyn_low 1 2.0 0.003 0.003 30.814 30.823 qs_forces 11 3.9 0.002 0.003 29.856 29.858 qs_energies 11 4.9 0.002 0.002 28.142 28.146 scf_env_do_scf 11 5.9 0.000 0.001 23.286 23.286 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.644 20.645 velocity_verlet 10 3.0 0.002 0.002 16.243 16.302 dbcsr_multiply_generic 2286 12.5 0.094 0.099 13.595 13.670 qs_scf_new_mos 108 7.5 0.001 0.001 12.636 12.667 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.636 12.667 ot_scf_mini 108 9.5 0.002 0.002 11.890 11.920 multiply_cannon 2286 13.5 0.224 0.231 10.691 11.243 multiply_cannon_loop 2286 14.5 0.642 0.671 9.706 9.931 ot_mini 108 10.5 0.001 0.001 6.692 6.730 multiply_cannon_multrec 9144 15.5 1.665 1.854 5.978 6.186 rebuild_ks_matrix 119 8.3 0.000 0.000 5.953 5.987 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.952 5.987 qs_ot_get_derivative 108 11.5 0.001 0.001 5.351 5.381 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.308 5.337 dbcsr_mm_accdrv_process 12550 15.8 3.600 4.217 4.202 4.281 sum_up_and_integrate 119 10.3 0.001 0.001 3.594 3.598 integrate_v_rspace 119 11.3 0.003 0.003 3.584 3.588 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.557 3.564 calculate_rho_elec 119 8.7 0.060 0.061 3.556 3.564 init_scf_run 11 5.9 0.000 0.001 3.400 3.400 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.400 3.400 qs_ot_get_p 119 10.4 0.002 0.002 3.232 3.272 init_scf_loop 11 6.9 0.000 0.000 2.619 2.619 mp_waitall_1 115863 16.7 1.784 2.298 1.784 2.298 make_m2s 4572 13.5 0.034 0.035 2.029 2.219 calculate_first_density_matrix 1 7.0 0.001 0.001 2.152 2.153 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.129 2.131 make_images 4572 14.5 0.270 0.302 1.939 2.128 pw_transfer 1439 11.6 0.065 0.068 2.093 2.105 density_rs2pw 119 9.7 0.004 0.004 1.903 2.029 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.999 2.013 grid_integrate_task_list 119 12.3 1.878 1.941 1.878 1.941 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.877 1.897 prepare_preconditioner 11 7.9 0.000 0.000 1.866 1.871 make_preconditioner 11 8.9 0.000 0.000 1.866 1.871 cp_dbcsr_syevd 50 12.0 0.004 0.006 1.858 1.859 calculate_dm_sparse 119 9.5 0.000 0.000 1.788 1.808 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.751 1.778 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.637 1.638 fft_wrap_pw1pw2_140 487 13.2 0.322 0.331 1.622 1.637 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.527 1.535 cp_fm_diag_elpa_base 50 14.0 1.498 1.514 1.524 1.532 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.487 1.503 acc_transpose_blocks 9144 15.5 0.041 0.042 1.444 1.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.435 1.443 fft3d_ps 1201 14.6 0.566 0.576 1.414 1.425 grid_collocate_task_list 119 9.7 1.304 1.395 1.304 1.395 potential_pw2rs 119 12.3 0.010 0.011 1.369 1.375 ot_diis_step 108 11.5 0.013 0.013 1.320 1.320 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.262 1.296 apply_single 119 13.6 0.000 0.000 1.261 1.295 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.233 1.234 wfi_extrapolate 11 7.9 0.001 0.001 1.184 1.184 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.903 1.175 make_images_data 4572 15.5 0.042 0.046 0.912 1.151 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.140 1.145 jit_kernel_multiply 6 15.4 0.562 1.072 0.562 1.072 transfer_rs2pw 487 10.6 0.004 0.005 0.910 1.037 mp_alltoall_d11v 2130 13.8 0.907 1.017 0.907 1.017 cp_fm_cholesky_invert 11 10.9 0.973 0.977 0.973 0.977 mp_sum_dm 438 4.9 0.903 0.966 0.903 0.966 md_write_output 11 3.9 0.059 0.930 0.061 0.945 md_output 10 3.0 0.000 0.000 0.061 0.944 update_particle_set 20 4.0 0.000 0.000 0.882 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.919 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.889 0.898 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.364 0.790 mp_allgather_i34 2286 14.5 0.300 0.780 0.300 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.767 0.772 transfer_pw2rs 487 13.2 0.003 0.004 0.763 0.767 acc_transpose_blocks_sync 27432 16.5 0.719 0.767 0.719 0.767 mp_alltoall_z22v 1201 16.6 0.722 0.755 0.722 0.755 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.683 0.732 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.668 0.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.039000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.727273, yerr=5.118432 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 785.862656E+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 63723. MP_Allreduce 10154 429. 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 0.013 0.032 45.337 45.338 qs_mol_dyn_low 1 2.0 0.003 0.003 45.119 45.129 qs_forces 11 3.9 0.002 0.002 45.051 45.052 qs_energies 11 4.9 0.001 0.002 43.023 43.028 scf_env_do_scf 11 5.9 0.001 0.001 36.625 36.625 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.327 28.328 velocity_verlet 10 3.0 0.002 0.002 25.388 25.393 dbcsr_multiply_generic 2286 12.5 0.103 0.106 20.193 20.401 qs_scf_new_mos 108 7.5 0.001 0.001 18.418 18.527 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.417 18.527 ot_scf_mini 108 9.5 0.002 0.002 17.221 17.327 multiply_cannon 2286 13.5 0.301 0.308 15.769 16.693 multiply_cannon_loop 2286 14.5 0.864 0.879 14.438 15.287 ot_mini 108 10.5 0.001 0.001 10.251 10.377 multiply_cannon_multrec 9144 15.5 3.399 4.914 9.122 9.230 init_scf_loop 11 6.9 0.000 0.000 8.265 8.268 qs_ot_get_derivative 108 11.5 0.001 0.001 8.118 8.220 rebuild_ks_matrix 119 8.3 0.000 0.000 7.389 7.539 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.389 7.539 prepare_preconditioner 11 7.9 0.000 0.000 7.281 7.295 make_preconditioner 11 8.9 0.000 0.000 7.281 7.295 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.808 7.169 dbcsr_mm_accdrv_process 12550 15.8 4.567 6.870 5.585 6.903 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.662 6.798 cp_fm_upper_to_full 72 14.2 3.159 4.576 3.159 4.576 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.370 4.437 calculate_rho_elec 119 8.7 0.118 0.121 4.369 4.436 init_scf_run 11 5.9 0.000 0.001 4.286 4.286 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.286 4.286 qs_ot_get_p 119 10.4 0.005 0.006 3.908 4.051 sum_up_and_integrate 119 10.3 0.001 0.001 3.989 3.996 integrate_v_rspace 119 11.3 0.003 0.003 3.979 3.986 mp_waitall_1 94719 16.7 2.823 3.646 2.823 3.646 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.972 3.380 dbcsr_complete_redistribute 329 12.2 0.287 0.294 2.105 2.949 make_m2s 4572 13.5 0.038 0.039 2.722 2.924 pw_transfer 1439 11.6 0.068 0.069 2.807 2.816 make_images 4572 14.5 0.354 0.386 2.600 2.802 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.709 2.719 calculate_first_density_matrix 1 7.0 0.001 0.001 2.654 2.716 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.740 2.570 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.260 2.560 apply_single 119 13.6 0.000 0.000 2.260 2.560 calculate_dm_sparse 119 9.5 0.000 0.000 2.345 2.407 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.363 2.363 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.348 2.351 mp_alltoall_i22 627 13.8 1.458 2.329 1.458 2.329 density_rs2pw 119 9.7 0.004 0.004 2.286 2.317 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.496 2.312 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.468 2.287 fft_wrap_pw1pw2_140 487 13.2 0.576 0.580 2.251 2.265 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.187 2.240 grid_integrate_task_list 119 12.3 2.107 2.172 2.107 2.172 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.071 2.118 ot_diis_step 108 11.5 0.014 0.014 2.095 2.095 mp_sum_l 7287 12.8 1.155 1.994 1.155 1.994 acc_transpose_blocks 9144 15.5 0.044 0.045 1.952 1.970 cp_dbcsr_syevd 50 12.0 0.008 0.008 1.965 1.965 fft3d_ps 1201 14.6 0.598 0.613 1.804 1.808 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.800 1.800 cp_fm_cholesky_invert 11 10.9 1.778 1.783 1.778 1.783 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.618 1.618 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.314 1.617 cp_fm_diag_elpa_base 50 14.0 1.464 1.520 1.616 1.616 grid_collocate_task_list 119 9.7 1.541 1.587 1.541 1.587 make_images_data 4572 15.5 0.046 0.049 1.283 1.563 wfi_extrapolate 11 7.9 0.001 0.001 1.505 1.505 mp_alltoall_d11v 2130 13.8 1.379 1.477 1.379 1.477 potential_pw2rs 119 12.3 0.014 0.014 1.420 1.422 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.392 1.415 jit_kernel_multiply 6 15.5 0.988 1.391 0.988 1.391 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.384 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.130 1.149 acc_transpose_blocks_sync 27432 16.5 1.099 1.118 1.099 1.118 mp_alltoall_z22v 1201 16.6 1.069 1.090 1.069 1.090 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.036 qs_create_task_list 11 7.9 0.001 0.001 0.952 0.963 generate_qs_task_list 11 8.9 0.378 0.400 0.952 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.941 0.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.338000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=738.545455, yerr=13.350953 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 502.063104E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1242805. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 82.815 82.817 qs_mol_dyn_low 1 2.0 0.003 0.003 82.540 82.551 qs_forces 11 3.9 0.003 0.003 82.463 82.464 qs_energies 11 4.9 0.002 0.002 79.531 79.551 scf_env_do_scf 11 5.9 0.000 0.001 70.424 70.426 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 64.425 64.426 dbcsr_multiply_generic 2055 12.4 0.106 0.109 49.676 50.016 qs_scf_new_mos 99 7.5 0.000 0.001 46.613 46.712 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.612 46.712 ot_scf_mini 99 9.5 0.002 0.002 44.320 44.403 velocity_verlet 10 3.0 0.002 0.003 42.957 42.958 multiply_cannon 2055 13.4 0.187 0.192 41.867 42.762 multiply_cannon_loop 2055 14.4 1.766 1.805 40.757 41.702 ot_mini 99 10.5 0.001 0.001 25.603 25.693 qs_ot_get_derivative 99 11.5 0.001 0.001 18.792 18.894 multiply_cannon_multrec 49320 15.4 11.611 12.404 17.535 18.489 rebuild_ks_matrix 110 8.3 0.000 0.001 14.966 15.142 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 14.966 15.141 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.138 13.289 mp_waitall_1 220248 16.4 9.993 10.805 9.993 10.805 qs_ot_get_p 110 10.4 0.002 0.003 10.421 10.556 multiply_cannon_sync_h2d 49320 15.4 9.785 10.445 9.785 10.445 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.238 7.783 apply_single 110 13.6 0.000 0.001 7.238 7.783 sum_up_and_integrate 110 10.3 0.002 0.003 7.368 7.383 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.821 7.365 integrate_v_rspace 110 11.3 0.003 0.003 7.343 7.362 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.215 7.232 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 5.775 7.072 init_scf_run 11 5.9 0.000 0.001 6.941 6.941 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.941 6.941 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.779 6.929 calculate_rho_elec 110 8.6 0.021 0.025 6.778 6.929 ot_diis_step 99 11.5 0.006 0.006 6.606 6.607 cp_dbcsr_syevd 48 12.0 0.003 0.004 6.416 6.417 dbcsr_mm_accdrv_process 87628 16.1 2.956 3.087 5.792 6.094 init_scf_loop 11 6.9 0.000 0.000 5.963 5.964 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.714 5.742 cp_fm_diag_elpa_base 48 14.0 5.696 5.725 5.712 5.739 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.383 5.447 make_m2s 4110 13.4 0.062 0.066 4.254 4.365 make_images 4110 14.4 0.178 0.189 4.156 4.270 wfi_extrapolate 11 7.9 0.001 0.001 4.224 4.224 density_rs2pw 110 9.6 0.004 0.005 3.816 3.983 mp_sum_l 6594 12.7 3.098 3.919 3.098 3.919 prepare_preconditioner 11 7.9 0.000 0.000 3.754 3.778 make_preconditioner 11 8.9 0.000 0.000 3.754 3.778 calculate_dm_sparse 110 9.5 0.001 0.001 3.666 3.744 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.609 3.613 pw_transfer 1331 11.6 0.056 0.065 3.486 3.593 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.541 3.586 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.397 3.506 grid_integrate_task_list 110 12.3 3.262 3.396 3.262 3.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.200 3.251 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.073 3.152 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.980 3.024 fft_wrap_pw1pw2_140 451 13.1 0.387 0.429 2.820 2.928 potential_pw2rs 110 12.3 0.005 0.006 2.898 2.924 fft3d_ps 1111 14.6 0.799 0.894 2.695 2.781 mp_alltoall_d11v 2046 13.8 2.209 2.678 2.209 2.678 calculate_first_density_matrix 1 7.0 0.000 0.001 2.605 2.611 jit_kernel_multiply 13 15.9 2.558 2.574 2.558 2.574 acc_transpose_blocks 49320 15.4 0.208 0.216 2.484 2.555 transfer_rs2pw 451 10.6 0.005 0.006 2.155 2.308 grid_collocate_task_list 110 9.6 2.162 2.270 2.162 2.270 cp_fm_cholesky_invert 11 10.9 2.111 2.115 2.111 2.115 mp_waitany 14300 13.8 1.866 2.092 1.866 2.092 make_images_data 4110 15.4 0.044 0.047 1.895 2.046 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.961 1.984 transfer_pw2rs 451 13.1 0.006 0.007 1.951 1.976 hybrid_alltoall_any 4261 16.3 0.085 0.488 1.645 1.905 mp_sum_d 3889 11.9 1.373 1.884 1.373 1.884 mp_alltoall_z22v 1111 16.6 1.565 1.733 1.565 1.733 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.666 1.688 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.817000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.874798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 595.083264E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2798173. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 90.561 90.562 qs_mol_dyn_low 1 2.0 0.003 0.003 90.205 90.281 qs_forces 11 3.9 0.003 0.003 90.018 90.018 qs_energies 11 4.9 0.002 0.007 86.315 86.321 scf_env_do_scf 11 5.9 0.000 0.001 76.115 76.118 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.721 63.723 velocity_verlet 10 3.0 0.001 0.002 47.341 47.345 dbcsr_multiply_generic 2055 12.4 0.120 0.124 43.293 43.652 qs_scf_new_mos 99 7.5 0.001 0.001 43.245 43.405 qs_scf_loop_do_ot 99 8.5 0.001 0.001 43.244 43.405 ot_scf_mini 99 9.5 0.003 0.003 41.360 41.517 multiply_cannon 2055 13.4 0.225 0.245 33.116 34.250 multiply_cannon_loop 2055 14.4 1.174 1.200 31.145 32.114 ot_mini 99 10.5 0.001 0.001 22.384 22.529 qs_ot_get_derivative 99 11.5 0.001 0.001 16.350 16.519 rebuild_ks_matrix 110 8.3 0.000 0.000 16.209 16.360 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 16.209 16.360 multiply_cannon_multrec 24660 15.4 6.824 8.401 14.146 15.986 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.345 14.483 mp_waitall_1 176588 16.5 10.988 13.987 10.988 13.987 init_scf_loop 11 6.9 0.000 0.000 12.328 12.331 qs_ot_get_p 110 10.4 0.003 0.006 11.611 11.763 prepare_preconditioner 11 7.9 0.000 0.000 9.970 9.994 make_preconditioner 11 8.9 0.000 0.000 9.970 9.994 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.471 9.673 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.883 9.141 sum_up_and_integrate 110 10.3 0.002 0.003 8.761 8.770 integrate_v_rspace 110 11.3 0.002 0.003 8.734 8.744 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.694 8.718 calculate_rho_elec 110 8.6 0.039 0.047 8.693 8.718 qs_ot_p2m_diag 48 11.0 0.029 0.044 8.497 8.525 cp_dbcsr_syevd 48 12.0 0.003 0.004 7.799 7.800 init_scf_run 11 5.9 0.000 0.001 7.548 7.549 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.548 7.549 dbcsr_mm_accdrv_process 52282 16.1 5.543 6.299 7.158 7.495 multiply_cannon_sync_h2d 24660 15.4 6.205 7.242 6.205 7.242 make_m2s 4110 13.4 0.058 0.060 6.746 7.207 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.619 7.187 apply_single 110 13.6 0.000 0.001 6.619 7.187 make_images 4110 14.4 0.406 0.461 6.633 7.090 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.644 6.659 cp_fm_diag_elpa_base 48 14.0 6.558 6.603 6.640 6.655 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.815 6.535 ot_diis_step 99 11.5 0.010 0.011 5.919 5.920 density_rs2pw 110 9.6 0.004 0.005 5.777 5.911 pw_transfer 1331 11.6 0.067 0.075 5.364 5.590 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.257 5.487 cp_fm_cholesky_invert 11 10.9 4.992 5.010 4.992 5.010 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.766 4.859 wfi_extrapolate 11 7.9 0.001 0.001 4.623 4.623 fft3d_ps 1111 14.6 1.115 1.344 4.363 4.604 potential_pw2rs 110 12.3 0.008 0.009 4.443 4.509 make_images_data 4110 15.4 0.049 0.053 3.464 4.020 hybrid_alltoall_any 4261 16.3 0.107 0.457 2.935 3.701 mp_sum_l 6594 12.7 2.799 3.686 2.799 3.686 fft_wrap_pw1pw2_140 451 13.1 0.450 0.464 3.426 3.651 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.632 3.636 calculate_dm_sparse 110 9.5 0.001 0.001 3.343 3.374 grid_integrate_task_list 110 12.3 3.102 3.343 3.102 3.343 transfer_rs2pw 451 10.6 0.006 0.007 3.162 3.289 cp_fm_cholesky_decompose 22 10.9 3.201 3.208 3.201 3.208 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.091 3.127 mp_alltoall_d11v 2046 13.8 2.709 3.076 2.709 3.076 mp_alltoall_z22v 1111 16.6 2.808 2.972 2.808 2.972 calculate_first_density_matrix 1 7.0 0.000 0.001 2.768 2.769 transfer_pw2rs 451 13.1 0.005 0.006 2.679 2.760 make_images_sizes 4110 15.4 0.005 0.005 1.825 2.604 mp_alltoall_i44 4110 16.4 1.821 2.599 1.821 2.599 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.496 2.518 mp_allgather_i34 2055 14.4 1.278 2.512 1.278 2.512 grid_collocate_task_list 110 9.6 2.174 2.317 2.174 2.317 dbcsr_complete_redistribute 325 12.2 0.231 0.291 1.747 2.108 acc_transpose_blocks 24660 15.4 0.116 0.119 2.058 2.103 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.029 2.042 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.943 1.944 jit_kernel_multiply 9 16.3 1.253 1.940 1.253 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=90.562000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.363636, yerr=7.450110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 665.665536E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 63.468 63.469 qs_mol_dyn_low 1 2.0 0.003 0.003 63.066 63.077 qs_forces 11 3.9 0.004 0.012 62.992 62.993 qs_energies 11 4.9 0.002 0.002 59.732 59.736 scf_env_do_scf 11 5.9 0.000 0.001 51.127 51.128 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.928 41.929 velocity_verlet 10 3.0 0.002 0.002 34.142 34.144 dbcsr_multiply_generic 2055 12.4 0.109 0.113 30.278 30.628 qs_scf_new_mos 99 7.5 0.001 0.001 26.830 26.945 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.829 26.944 ot_scf_mini 99 9.5 0.002 0.003 25.536 25.678 multiply_cannon 2055 13.4 0.212 0.223 23.416 24.887 multiply_cannon_loop 2055 14.4 0.817 0.835 22.244 23.677 ot_mini 99 10.5 0.001 0.001 14.547 14.691 rebuild_ks_matrix 110 8.3 0.000 0.000 12.420 12.594 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.420 12.594 multiply_cannon_multrec 16440 15.4 3.604 4.647 10.427 11.424 mp_waitall_1 139946 16.5 7.578 11.303 7.578 11.303 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.917 11.075 qs_ot_get_derivative 99 11.5 0.001 0.001 10.049 10.197 init_scf_loop 11 6.9 0.000 0.000 9.164 9.165 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.645 8.228 prepare_preconditioner 11 7.9 0.000 0.000 7.375 7.393 make_preconditioner 11 8.9 0.000 0.000 7.375 7.393 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.682 7.035 dbcsr_mm_accdrv_process 34862 16.1 5.943 6.353 6.671 6.836 sum_up_and_integrate 110 10.3 0.002 0.003 6.533 6.548 integrate_v_rspace 110 11.3 0.003 0.004 6.507 6.522 qs_ot_get_p 110 10.4 0.002 0.003 6.041 6.203 init_scf_run 11 5.9 0.000 0.002 6.178 6.179 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.178 6.178 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.985 5.993 calculate_rho_elec 110 8.6 0.059 0.059 5.984 5.993 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.973 5.432 apply_single 110 13.6 0.000 0.000 4.973 5.432 make_m2s 4110 13.4 0.051 0.052 4.456 4.796 make_images 4110 14.4 0.399 0.522 4.340 4.680 ot_diis_step 99 11.5 0.011 0.011 4.454 4.454 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.252 4.263 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.358 4.021 multiply_cannon_sync_h2d 16440 15.4 3.225 3.875 3.225 3.875 cp_dbcsr_syevd 48 12.0 0.005 0.005 3.852 3.853 pw_transfer 1331 11.6 0.066 0.073 3.723 3.735 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.616 3.629 grid_integrate_task_list 110 12.3 3.185 3.434 3.185 3.434 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.402 3.403 density_rs2pw 110 9.6 0.004 0.005 3.197 3.385 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.254 3.263 cp_fm_diag_elpa_base 48 14.0 3.184 3.214 3.252 3.262 wfi_extrapolate 11 7.9 0.001 0.001 3.122 3.122 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.039 3.097 fft_wrap_pw1pw2_140 451 13.1 0.578 0.587 3.062 3.077 cp_fm_cholesky_invert 11 10.9 3.006 3.012 3.006 3.012 calculate_first_density_matrix 1 7.0 0.001 0.001 2.966 2.967 make_images_data 4110 15.4 0.046 0.050 2.533 2.927 hybrid_alltoall_any 4261 16.3 0.109 0.384 2.176 2.866 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.754 2.832 calculate_dm_sparse 110 9.5 0.001 0.001 2.576 2.618 fft3d_ps 1111 14.6 1.093 1.102 2.559 2.572 mp_sum_l 6594 12.7 1.781 2.555 1.781 2.555 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.897 2.521 mp_irecv_dv 48980 15.7 0.821 2.381 0.821 2.381 potential_pw2rs 110 12.3 0.011 0.012 2.365 2.376 grid_collocate_task_list 110 9.6 2.219 2.342 2.219 2.342 mp_alltoall_d11v 2046 13.8 1.866 2.144 1.866 2.144 dbcsr_complete_redistribute 325 12.2 0.326 0.360 1.541 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.973 1.974 cp_fm_cholesky_decompose 22 10.9 1.820 1.842 1.820 1.842 cp_fm_upper_to_full 70 14.2 1.412 1.790 1.412 1.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.718 acc_transpose_blocks 16440 15.4 0.078 0.080 1.589 1.638 transfer_rs2pw 451 10.6 0.005 0.006 1.439 1.603 mp_allgather_i34 2055 14.4 0.486 1.587 0.486 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.474 1.492 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.480 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.010 1.468 mp_waitany 17072 13.8 1.196 1.417 1.196 1.417 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.372 1.380 transfer_pw2rs 451 13.1 0.005 0.005 1.342 1.353 mp_alltoall_z22v 1111 16.6 1.251 1.271 1.251 1.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.469000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.000000, yerr=7.722458 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 730.468352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 68.931 68.932 qs_mol_dyn_low 1 2.0 0.003 0.003 68.687 68.699 qs_forces 11 3.9 0.003 0.004 68.609 68.611 qs_energies 11 4.9 0.001 0.002 65.167 65.173 scf_env_do_scf 11 5.9 0.000 0.001 56.583 56.585 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 44.202 44.203 velocity_verlet 10 3.0 0.002 0.002 38.821 38.823 dbcsr_multiply_generic 2055 12.4 0.115 0.122 31.324 31.605 qs_scf_new_mos 99 7.5 0.001 0.001 29.083 29.200 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.083 29.200 ot_scf_mini 99 9.5 0.002 0.003 27.357 27.466 multiply_cannon 2055 13.4 0.236 0.250 23.740 24.991 multiply_cannon_loop 2055 14.4 1.392 1.487 22.287 23.072 ot_mini 99 10.5 0.001 0.001 15.356 15.498 multiply_cannon_multrec 24660 15.4 4.046 6.643 13.077 14.269 rebuild_ks_matrix 110 8.3 0.000 0.000 12.263 12.381 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.262 12.381 init_scf_loop 11 6.9 0.000 0.000 12.332 12.333 qs_ot_get_derivative 99 11.5 0.001 0.001 11.095 11.214 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.827 10.933 prepare_preconditioner 11 7.9 0.000 0.000 10.573 10.589 make_preconditioner 11 8.9 0.000 0.000 10.573 10.589 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.783 10.222 dbcsr_mm_accdrv_process 52304 16.0 7.701 9.337 8.874 9.823 qs_ot_get_p 110 10.4 0.002 0.003 6.815 6.965 mp_waitall_1 121746 16.5 4.614 6.797 4.614 6.797 sum_up_and_integrate 110 10.3 0.001 0.002 6.519 6.532 integrate_v_rspace 110 11.3 0.003 0.003 6.493 6.506 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.130 6.139 calculate_rho_elec 110 8.6 0.078 0.081 6.130 6.138 make_m2s 4110 13.4 0.061 0.062 5.769 6.079 make_images 4110 14.4 0.580 0.698 5.626 5.935 init_scf_run 11 5.9 0.000 0.002 5.775 5.775 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.775 5.775 cp_fm_upper_to_full 70 14.2 3.328 4.818 3.328 4.818 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.792 4.807 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.103 4.226 apply_single 110 13.6 0.000 0.000 4.102 4.226 ot_diis_step 99 11.5 0.011 0.012 4.214 4.215 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.204 4.205 pw_transfer 1331 11.6 0.066 0.076 3.867 3.902 dbcsr_complete_redistribute 325 12.2 0.417 0.470 2.771 3.887 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.761 3.799 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.676 3.730 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.543 3.609 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.589 3.599 cp_fm_diag_elpa_base 48 14.0 3.432 3.492 3.586 3.596 grid_integrate_task_list 110 12.3 3.283 3.476 3.283 3.476 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.459 3.415 density_rs2pw 110 9.6 0.004 0.004 3.212 3.365 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.176 3.282 make_images_data 4110 15.4 0.049 0.053 2.869 3.225 fft_wrap_pw1pw2_140 451 13.1 0.610 0.630 3.167 3.212 cp_fm_cholesky_invert 11 10.9 3.167 3.179 3.167 3.179 wfi_extrapolate 11 7.9 0.001 0.001 3.160 3.160 calculate_dm_sparse 110 9.5 0.001 0.001 3.079 3.114 hybrid_alltoall_any 4261 16.3 0.123 0.464 2.483 3.103 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.928 2.929 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.781 2.870 mp_alltoall_i22 605 13.7 1.649 2.825 1.649 2.825 fft3d_ps 1111 14.6 1.086 1.117 2.643 2.656 multiply_cannon_sync_h2d 24660 15.4 2.356 2.601 2.356 2.601 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.536 2.572 acc_transpose_blocks 24660 15.4 0.114 0.118 2.416 2.549 calculate_first_density_matrix 1 7.0 0.000 0.001 2.513 2.516 grid_collocate_task_list 110 9.6 2.265 2.400 2.265 2.400 potential_pw2rs 110 12.3 0.012 0.013 2.300 2.306 mp_alltoall_d11v 2046 13.8 1.962 2.283 1.962 2.283 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.271 2.273 cp_fm_cholesky_decompose 22 10.9 1.882 1.919 1.882 1.919 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.871 1.899 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.679 1.691 mp_sum_l 6594 12.7 1.132 1.691 1.132 1.691 multiply_cannon_metrocomm4 20550 15.4 0.063 0.067 0.884 1.665 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.606 1.622 mp_allgather_i34 2055 14.4 0.546 1.578 0.546 1.578 transfer_rs2pw 451 10.6 0.005 0.006 1.381 1.559 mp_irecv_dv 62702 16.1 0.779 1.491 0.779 1.491 acc_transpose_blocks_sync 73980 16.4 1.371 1.488 1.371 1.488 mp_alltoall_z22v 1111 16.6 1.340 1.400 1.340 1.400 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.932000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.090909, yerr=7.597303 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 865.619968E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.036 81.744 81.746 qs_mol_dyn_low 1 2.0 0.003 0.003 81.439 81.449 qs_forces 11 3.9 0.003 0.003 81.352 81.353 qs_energies 11 4.9 0.001 0.002 77.293 77.299 scf_env_do_scf 11 5.9 0.000 0.001 67.253 67.254 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 52.502 52.504 velocity_verlet 10 3.0 0.001 0.002 45.086 45.090 qs_scf_new_mos 99 7.5 0.001 0.001 33.411 33.515 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.410 33.515 ot_scf_mini 99 9.5 0.002 0.002 31.819 31.865 dbcsr_multiply_generic 2055 12.4 0.111 0.114 29.767 30.181 multiply_cannon 2055 13.4 0.241 0.255 20.625 23.225 multiply_cannon_loop 2055 14.4 0.608 0.631 18.940 20.039 ot_mini 99 10.5 0.001 0.001 15.548 15.603 rebuild_ks_matrix 110 8.3 0.000 0.000 14.518 14.694 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.518 14.694 init_scf_loop 11 6.9 0.000 0.000 14.680 14.681 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.959 13.117 prepare_preconditioner 11 7.9 0.000 0.000 12.629 12.638 make_preconditioner 11 8.9 0.000 0.000 12.629 12.638 mp_waitall_1 103326 16.6 10.689 12.532 10.689 12.532 make_full_inverse_cholesky 11 9.9 0.000 0.000 12.025 12.185 qs_ot_get_derivative 99 11.5 0.001 0.001 11.483 11.541 qs_ot_get_p 110 10.4 0.002 0.003 11.449 11.511 qs_ot_p2m_diag 48 11.0 0.081 0.085 8.638 8.645 multiply_cannon_multrec 8220 15.4 3.182 4.336 7.629 8.545 sum_up_and_integrate 110 10.3 0.002 0.002 8.495 8.506 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.470 8.489 calculate_rho_elec 110 8.6 0.114 0.114 8.469 8.489 integrate_v_rspace 110 11.3 0.003 0.003 8.469 8.479 cp_dbcsr_syevd 48 12.0 0.004 0.004 8.041 8.043 make_m2s 4110 13.4 0.039 0.040 6.692 7.092 cp_fm_cholesky_invert 11 10.9 6.994 7.008 6.994 7.008 make_images 4110 14.4 0.650 0.710 6.559 6.961 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.879 6.886 cp_fm_diag_elpa_base 48 14.0 6.783 6.831 6.875 6.884 init_scf_run 11 5.9 0.000 0.002 6.477 6.478 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.477 6.478 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.120 5.654 pw_transfer 1331 11.6 0.066 0.071 5.628 5.651 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 5.520 5.547 dbcsr_mm_accdrv_process 17442 15.9 3.124 4.197 4.304 5.241 density_rs2pw 110 9.6 0.004 0.004 5.068 5.214 make_images_data 4110 15.4 0.041 0.047 3.670 4.450 fft_wrap_pw1pw2_140 451 13.1 0.774 0.787 4.393 4.423 fft3d_ps 1111 14.6 1.143 1.158 4.189 4.209 hybrid_alltoall_any 4261 16.3 0.200 0.850 3.390 4.118 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.927 4.110 apply_single 110 13.6 0.000 0.000 3.926 4.110 ot_diis_step 99 11.5 0.012 0.012 3.982 3.982 wfi_extrapolate 11 7.9 0.001 0.001 3.942 3.942 cp_fm_cholesky_decompose 22 10.9 3.782 3.793 3.782 3.793 potential_pw2rs 110 12.3 0.015 0.015 3.697 3.713 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 3.180 3.644 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.588 3.614 grid_integrate_task_list 110 12.3 3.396 3.529 3.396 3.529 mp_alltoall_d11v 2046 13.8 3.270 3.486 3.270 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.370 3.420 calculate_dm_sparse 110 9.5 0.001 0.001 2.871 2.940 mp_alltoall_z22v 1111 16.6 2.803 2.834 2.803 2.834 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.816 2.818 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.774 2.775 make_images_sizes 4110 15.4 0.005 0.005 1.419 2.653 mp_alltoall_i44 4110 16.4 1.414 2.649 1.414 2.649 transfer_rs2pw 451 10.6 0.005 0.005 2.439 2.605 grid_collocate_task_list 110 9.6 2.404 2.539 2.404 2.539 dbcsr_complete_redistribute 325 12.2 0.545 0.562 2.294 2.525 calculate_first_density_matrix 1 7.0 0.001 0.003 2.396 2.397 multiply_cannon_sync_h2d 8220 15.4 2.346 2.397 2.346 2.397 mp_allgather_i34 2055 14.4 0.852 2.394 0.852 2.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.221 2.290 transfer_pw2rs 451 13.1 0.004 0.005 2.062 2.083 mp_sum_l 6594 12.7 1.717 2.049 1.717 2.049 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.782 1.989 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.900 1.917 mp_waitany 9240 13.8 1.709 1.846 1.709 1.846 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.682 1.720 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.605 1.719 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.633 1.674 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=81.746000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=818.090909, yerr=11.595097 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 1.384223E+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 67098. MP_Allreduce 9752 812. 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 0.041 0.076 91.265 91.282 qs_mol_dyn_low 1 2.0 0.003 0.003 90.869 90.878 qs_forces 11 3.9 0.003 0.004 90.787 90.788 qs_energies 11 4.9 0.002 0.002 86.611 86.614 scf_env_do_scf 11 5.9 0.001 0.001 75.967 75.967 velocity_verlet 10 3.0 0.002 0.002 57.790 57.797 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.533 46.534 dbcsr_multiply_generic 2055 12.4 0.121 0.126 31.143 31.222 init_scf_loop 11 6.9 0.000 0.000 29.355 29.358 qs_scf_new_mos 99 7.5 0.001 0.001 28.660 28.696 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.660 28.695 prepare_preconditioner 11 7.9 0.000 0.000 27.316 27.324 make_preconditioner 11 8.9 0.000 0.000 27.316 27.324 ot_scf_mini 99 9.5 0.002 0.002 26.795 26.829 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.501 26.761 multiply_cannon 2055 13.4 0.330 0.350 23.366 24.116 multiply_cannon_loop 2055 14.4 0.824 0.837 21.503 21.784 cp_fm_upper_to_full 70 14.2 12.638 18.152 12.638 18.152 ot_mini 99 10.5 0.001 0.001 14.932 14.967 rebuild_ks_matrix 110 8.3 0.000 0.001 13.975 14.024 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 13.975 14.023 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.683 12.727 dbcsr_complete_redistribute 325 12.2 1.027 1.060 7.519 10.764 multiply_cannon_multrec 8220 15.4 4.145 4.315 10.268 10.455 qs_ot_get_derivative 99 11.5 0.001 0.001 10.172 10.211 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.405 9.659 mp_waitall_1 84994 16.7 8.316 9.124 8.316 9.124 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.798 9.019 mp_alltoall_i22 605 13.7 5.402 8.679 5.402 8.679 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.370 7.404 calculate_rho_elec 110 8.6 0.224 0.225 7.369 7.404 sum_up_and_integrate 110 10.3 0.002 0.002 6.993 7.008 integrate_v_rspace 110 11.3 0.003 0.004 6.964 6.980 qs_ot_get_p 110 10.4 0.003 0.003 6.376 6.402 make_m2s 4110 13.4 0.044 0.045 5.857 6.399 init_scf_run 11 5.9 0.000 0.001 6.312 6.313 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.312 6.312 make_images 4110 14.4 0.884 0.925 5.666 6.206 dbcsr_mm_accdrv_process 11614 15.7 3.931 4.954 5.974 6.158 cp_fm_cholesky_invert 11 10.9 6.021 6.026 6.021 6.026 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.165 5.447 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.888 5.323 apply_single 110 13.6 0.000 0.000 4.888 5.323 pw_transfer 1331 11.6 0.075 0.076 5.165 5.172 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.048 5.054 ot_diis_step 99 11.5 0.015 0.015 4.731 4.732 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.590 4.598 fft_wrap_pw1pw2_140 451 13.1 1.279 1.281 4.349 4.358 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.047 4.048 hybrid_alltoall_any 4261 16.3 0.264 0.565 3.164 3.963 make_images_data 4110 15.4 0.045 0.049 3.224 3.945 density_rs2pw 110 9.6 0.004 0.004 3.764 3.787 grid_integrate_task_list 110 12.3 3.698 3.775 3.698 3.775 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.699 3.701 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.193 3.656 wfi_extrapolate 11 7.9 0.001 0.001 3.458 3.459 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.402 3.403 cp_fm_diag_elpa_base 48 14.0 2.850 3.054 3.400 3.401 calculate_dm_sparse 110 9.5 0.001 0.001 3.347 3.373 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.295 3.296 multiply_cannon_sync_h2d 8220 15.4 3.128 3.157 3.128 3.157 fft3d_ps 1111 14.6 1.295 1.303 3.102 3.110 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.799 2.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.743 2.762 calculate_first_density_matrix 1 7.0 0.001 0.001 2.734 2.735 grid_collocate_task_list 110 9.6 2.685 2.722 2.685 2.722 potential_pw2rs 110 12.3 0.021 0.021 2.423 2.429 mp_alltoall_d11v 2046 13.8 2.293 2.359 2.293 2.359 cp_fm_cholesky_decompose 22 10.9 2.298 2.315 2.298 2.315 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.213 2.277 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.115 2.227 jit_kernel_multiply 10 15.3 1.846 2.180 1.846 2.180 qs_create_task_list 11 7.9 0.001 0.001 1.903 1.949 generate_qs_task_list 11 8.9 0.740 0.794 1.903 1.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.901 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.864 1.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.282000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1233.090909, yerr=58.746690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.202944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529097. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.051 208.745 208.748 qs_mol_dyn_low 1 2.0 0.005 0.021 208.122 208.183 qs_forces 11 3.9 0.005 0.007 207.981 207.985 qs_energies 11 4.9 0.004 0.024 202.318 202.337 scf_env_do_scf 11 5.9 0.001 0.001 184.730 184.736 scf_env_do_scf_inner_loop 117 6.6 0.006 0.025 163.003 163.007 velocity_verlet 10 3.0 0.010 0.026 125.677 125.680 dbcsr_multiply_generic 2507 12.6 0.181 0.185 123.851 124.452 qs_scf_new_mos 117 7.6 0.001 0.001 122.702 123.072 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.701 123.071 ot_scf_mini 117 9.6 0.003 0.003 116.064 116.370 multiply_cannon 2507 13.6 0.242 0.253 101.558 103.592 multiply_cannon_loop 2507 14.6 2.325 2.460 99.266 101.254 ot_mini 117 10.6 0.001 0.001 64.258 64.609 multiply_cannon_multrec 60168 15.6 32.286 34.800 42.523 44.511 qs_ot_get_derivative 117 11.6 0.001 0.001 39.411 39.751 rebuild_ks_matrix 128 8.3 0.001 0.001 34.250 34.617 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 34.249 34.617 mp_waitall_1 267128 16.5 27.512 31.647 27.512 31.647 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.767 31.090 qs_ot_get_p 128 10.4 0.002 0.003 29.896 30.267 multiply_cannon_sync_h2d 60168 15.6 27.259 28.906 27.259 28.906 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.344 25.155 apply_single 128 13.6 0.001 0.001 24.344 25.155 ot_diis_step 117 11.6 0.008 0.008 24.624 24.626 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.179 23.260 init_scf_loop 11 6.9 0.000 0.000 21.650 21.651 cp_dbcsr_syevd 83 12.4 0.005 0.006 20.746 20.749 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 17.613 17.846 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.646 17.680 cp_fm_diag_elpa_base 83 14.4 17.558 17.617 17.642 17.676 multiply_cannon_metrocomm3 60168 15.6 0.120 0.128 14.780 17.067 prepare_preconditioner 11 7.9 0.000 0.000 16.960 16.988 make_preconditioner 11 8.9 0.000 0.000 16.960 16.988 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.141 16.324 make_m2s 5014 13.6 0.105 0.112 14.082 14.503 sum_up_and_integrate 128 10.3 0.002 0.005 14.485 14.499 integrate_v_rspace 128 11.3 0.004 0.004 14.426 14.443 make_images 5014 14.6 0.398 0.417 13.898 14.324 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.537 13.639 calculate_rho_elec 128 8.7 0.045 0.064 13.536 13.638 init_scf_run 11 5.9 0.000 0.001 13.278 13.278 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.278 13.278 dbcsr_mm_accdrv_process 124484 16.2 4.464 4.866 9.774 10.349 cp_fm_cholesky_invert 11 10.9 9.745 9.753 9.745 9.753 wfi_extrapolate 11 7.9 0.001 0.001 9.374 9.375 calculate_dm_sparse 128 9.5 0.001 0.001 8.582 8.691 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.389 8.581 multiply_cannon_metrocomm1 60168 15.6 0.094 0.099 5.794 8.337 mp_sum_l 7950 12.9 6.942 8.250 6.942 8.250 pw_transfer 1547 11.6 0.075 0.092 7.928 8.115 density_rs2pw 128 9.7 0.006 0.007 7.163 8.010 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.724 7.912 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.793 7.892 make_images_data 5014 15.6 0.067 0.074 6.912 7.833 grid_integrate_task_list 128 12.3 7.062 7.440 7.062 7.440 hybrid_alltoall_any 5200 16.5 0.296 2.275 6.048 7.333 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 7.253 7.264 fft_wrap_pw1pw2_140 523 13.2 1.122 1.166 6.620 6.801 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.340 6.539 fft3d_ps 1291 14.7 2.195 2.845 5.735 6.075 mp_alltoall_d11v 2415 14.1 4.479 5.774 4.479 5.774 grid_collocate_task_list 128 9.7 4.845 5.241 4.845 5.241 potential_pw2rs 128 12.3 0.009 0.010 4.941 4.972 cp_fm_cholesky_decompose 22 10.9 4.924 4.937 4.924 4.937 mp_sum_d 4470 12.1 3.524 4.183 3.524 4.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.748000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.818182, yerr=5.288378 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 838.893568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6313342. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.137 195.722 195.728 qs_mol_dyn_low 1 2.0 0.003 0.005 195.161 195.211 qs_forces 11 3.9 0.004 0.005 195.061 195.062 qs_energies 11 4.9 0.008 0.055 188.256 188.278 scf_env_do_scf 11 5.9 0.004 0.028 171.165 171.176 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 136.962 136.965 velocity_verlet 10 3.0 0.001 0.002 121.758 121.759 dbcsr_multiply_generic 2507 12.6 0.191 0.197 98.127 99.467 qs_scf_new_mos 117 7.6 0.001 0.001 97.770 98.441 qs_scf_loop_do_ot 117 8.6 0.001 0.002 97.770 98.440 ot_scf_mini 117 9.6 0.004 0.005 92.919 93.635 multiply_cannon 2507 13.6 0.503 0.557 77.220 81.852 multiply_cannon_loop 2507 14.6 1.581 1.651 74.107 77.474 ot_mini 117 10.6 0.001 0.001 51.080 51.745 mp_waitall_1 214728 16.6 25.114 38.710 25.114 38.710 multiply_cannon_multrec 30084 15.6 20.731 24.821 31.329 35.834 init_scf_loop 11 6.9 0.003 0.020 34.087 34.088 rebuild_ks_matrix 128 8.3 0.001 0.001 32.673 33.442 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.672 33.441 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.404 30.123 prepare_preconditioner 11 7.9 0.000 0.001 29.632 29.716 make_preconditioner 11 8.9 0.000 0.002 29.632 29.716 qs_ot_get_derivative 117 11.6 0.001 0.002 28.994 29.702 multiply_cannon_metrocomm3 30084 15.6 0.097 0.103 16.259 29.172 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.305 28.841 qs_ot_get_p 128 10.4 0.002 0.003 23.851 24.628 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.162 23.355 apply_single 128 13.6 0.001 0.001 22.161 23.355 ot_diis_step 117 11.6 0.014 0.015 21.901 21.903 multiply_cannon_sync_h2d 30084 15.6 17.769 20.309 17.769 20.309 qs_ot_p2m_diag 83 11.4 0.189 0.217 18.681 18.731 cp_dbcsr_syevd 83 12.4 0.006 0.010 17.396 17.397 cp_fm_cholesky_invert 11 10.9 17.372 17.385 17.372 17.385 make_m2s 5014 13.6 0.091 0.095 14.583 15.871 make_images 5014 14.6 1.174 1.384 14.370 15.659 sum_up_and_integrate 128 10.3 0.002 0.004 14.426 14.463 integrate_v_rspace 128 11.3 0.003 0.004 14.365 14.406 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.030 14.063 cp_fm_diag_elpa_base 83 14.4 13.758 13.850 14.022 14.050 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.793 13.825 calculate_rho_elec 128 8.7 0.087 0.104 13.792 13.824 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.624 12.122 init_scf_run 11 5.9 0.000 0.002 12.009 12.011 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.009 12.010 multiply_cannon_metrocomm4 27577 15.6 0.106 0.122 3.901 11.137 mp_irecv_dv 69486 16.3 3.695 10.726 3.695 10.726 dbcsr_mm_accdrv_process 62242 16.2 5.388 6.009 10.058 10.639 make_images_data 5014 15.6 0.068 0.076 8.610 10.120 hybrid_alltoall_any 5200 16.5 0.353 1.524 7.301 9.570 pw_transfer 1547 11.6 0.085 0.097 8.930 9.003 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.707 8.784 wfi_extrapolate 11 7.9 0.001 0.001 8.526 8.526 density_rs2pw 128 9.7 0.006 0.008 7.475 7.942 fft_wrap_pw1pw2_140 523 13.2 1.215 1.242 7.593 7.689 grid_integrate_task_list 128 12.3 7.165 7.555 7.165 7.555 cp_fm_cholesky_decompose 22 10.9 7.302 7.404 7.302 7.404 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.405 7.145 calculate_dm_sparse 128 9.5 0.001 0.001 6.538 6.684 mp_sum_l 7950 12.9 4.562 6.587 4.562 6.587 fft3d_ps 1291 14.7 2.845 3.017 6.265 6.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.165 6.176 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.480 5.572 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.207 5.416 grid_collocate_task_list 128 9.7 5.005 5.329 5.005 5.329 mp_alltoall_d11v 2415 14.1 4.336 4.980 4.336 4.980 potential_pw2rs 128 12.3 0.016 0.020 4.892 4.917 mp_allgather_i34 2507 14.6 1.640 4.818 1.640 4.818 dbcsr_complete_redistribute 395 12.7 0.773 0.875 3.347 4.198 mp_sum_d 4465 12.1 2.740 4.134 2.740 4.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.728000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.181818, yerr=2.587318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 964.546560E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.136 0.194 184.802 184.818 qs_mol_dyn_low 1 2.0 0.003 0.003 183.793 183.834 qs_forces 11 3.9 0.004 0.004 183.492 183.496 qs_energies 11 4.9 0.002 0.003 176.833 176.843 scf_env_do_scf 11 5.9 0.001 0.001 160.715 160.716 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 123.482 123.483 velocity_verlet 10 3.0 0.001 0.002 117.770 117.778 dbcsr_multiply_generic 2507 12.6 0.181 0.185 84.497 85.554 qs_scf_new_mos 117 7.6 0.001 0.001 85.103 85.460 qs_scf_loop_do_ot 117 8.6 0.001 0.001 85.103 85.459 ot_scf_mini 117 9.6 0.003 0.004 80.762 81.107 multiply_cannon 2507 13.6 0.504 0.521 63.417 69.090 multiply_cannon_loop 2507 14.6 1.137 1.208 60.406 62.945 ot_mini 117 10.6 0.001 0.001 44.093 44.441 mp_waitall_1 170520 16.6 26.794 37.337 26.794 37.337 init_scf_loop 11 6.9 0.000 0.000 37.134 37.135 prepare_preconditioner 11 7.9 0.000 0.000 33.020 33.071 make_preconditioner 11 8.9 0.000 0.000 33.019 33.071 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.555 32.031 rebuild_ks_matrix 128 8.3 0.001 0.001 31.007 31.416 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.007 31.416 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.975 28.350 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 16.520 26.401 multiply_cannon_multrec 20056 15.6 12.804 15.921 22.287 25.437 qs_ot_get_derivative 117 11.6 0.002 0.002 24.263 24.599 qs_ot_get_p 128 10.4 0.001 0.003 21.474 21.857 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.935 20.972 apply_single 128 13.6 0.001 0.001 19.935 20.972 ot_diis_step 117 11.6 0.018 0.030 19.713 19.714 qs_ot_p2m_diag 83 11.4 0.266 0.272 16.876 16.901 make_m2s 5014 13.6 0.080 0.086 15.541 16.641 make_images 5014 14.6 1.156 1.268 15.308 16.407 cp_dbcsr_syevd 83 12.4 0.006 0.007 15.704 15.706 cp_fm_cholesky_invert 11 10.9 15.670 15.680 15.670 15.680 multiply_cannon_sync_h2d 20056 15.6 13.477 14.832 13.477 14.832 sum_up_and_integrate 128 10.3 0.002 0.002 14.510 14.536 integrate_v_rspace 128 11.3 0.003 0.004 14.450 14.477 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.201 14.228 calculate_rho_elec 128 8.7 0.130 0.145 14.201 14.227 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.307 12.330 cp_fm_diag_elpa_base 83 14.4 11.869 12.045 12.303 12.325 init_scf_run 11 5.9 0.000 0.002 10.990 10.990 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.990 10.990 make_images_data 5014 15.6 0.063 0.074 9.391 10.980 hybrid_alltoall_any 5200 16.5 0.451 2.050 8.006 10.344 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.806 10.073 multiply_cannon_metrocomm4 17549 15.6 0.067 0.077 3.562 9.672 mp_irecv_dv 50230 16.2 3.433 9.411 3.433 9.411 pw_transfer 1547 11.6 0.085 0.102 9.266 9.370 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.041 9.153 dbcsr_mm_accdrv_process 41502 16.2 5.823 6.107 8.961 9.092 cp_fm_cholesky_decompose 22 10.9 7.976 8.006 7.976 8.006 fft_wrap_pw1pw2_140 523 13.2 1.304 1.342 7.856 7.986 density_rs2pw 128 9.7 0.006 0.006 7.497 7.830 grid_integrate_task_list 128 12.3 7.348 7.800 7.348 7.800 wfi_extrapolate 11 7.9 0.001 0.001 7.694 7.694 cp_fm_upper_to_full 105 14.8 5.851 7.459 5.851 7.459 dbcsr_complete_redistribute 395 12.7 1.189 1.230 4.929 6.719 fft3d_ps 1291 14.7 2.759 2.982 6.416 6.485 calculate_dm_sparse 128 9.5 0.001 0.001 5.984 6.079 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.779 5.785 grid_collocate_task_list 128 9.7 5.160 5.593 5.160 5.593 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.838 5.536 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.659 5.443 mp_sum_l 7950 12.9 3.722 5.159 3.722 5.159 mp_alltoall_d11v 2415 14.1 4.664 5.123 4.664 5.123 mp_allgather_i34 2507 14.6 1.520 4.996 1.520 4.996 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.762 4.905 potential_pw2rs 128 12.3 0.020 0.022 4.803 4.822 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.230 4.262 transfer_fm_to_dbcsr 11 9.9 0.020 0.025 2.446 4.210 mp_alltoall_i22 716 14.1 2.076 3.998 2.076 3.998 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.767 3.768 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=184.818000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=907.181818, yerr=21.100703 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 4.320341E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 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 1.153151E+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 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 58941. MP_Allreduce 10955 1087. MP_Sync 86 MP_Alltoall 1700 12496378. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 194.102 194.103 qs_mol_dyn_low 1 2.0 0.003 0.003 193.647 193.661 qs_forces 11 3.9 0.004 0.005 193.520 193.527 qs_energies 11 4.9 0.002 0.002 186.375 186.389 scf_env_do_scf 11 5.9 0.001 0.001 168.012 168.024 velocity_verlet 10 3.0 0.002 0.002 126.844 126.848 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 119.904 119.906 qs_scf_new_mos 116 7.6 0.001 0.001 83.415 83.750 qs_scf_loop_do_ot 116 8.6 0.001 0.001 83.415 83.749 dbcsr_multiply_generic 2485 12.5 0.187 0.194 82.119 82.997 ot_scf_mini 116 9.6 0.003 0.003 78.844 79.156 multiply_cannon 2485 13.5 0.540 0.570 57.174 59.916 multiply_cannon_loop 2485 14.5 1.777 1.904 53.280 55.369 init_scf_loop 11 6.9 0.000 0.000 47.963 47.964 prepare_preconditioner 11 7.9 0.000 0.000 43.860 43.885 make_preconditioner 11 8.9 0.000 0.000 43.860 43.885 ot_mini 116 10.6 0.001 0.001 43.304 43.611 make_full_inverse_cholesky 11 9.9 0.011 0.023 37.438 42.392 multiply_cannon_multrec 29820 15.5 13.417 18.165 27.598 32.125 rebuild_ks_matrix 127 8.3 0.001 0.001 29.333 29.670 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.333 29.669 mp_waitall_1 146592 16.7 17.630 27.024 17.630 27.024 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.418 26.729 qs_ot_get_derivative 116 11.6 0.002 0.002 23.773 24.094 qs_ot_get_p 127 10.4 0.002 0.003 21.622 21.943 make_m2s 4970 13.5 0.096 0.099 20.303 21.328 make_images 4970 14.5 1.934 2.278 19.994 21.022 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.893 19.603 apply_single 127 13.6 0.001 0.001 18.892 19.603 ot_diis_step 116 11.6 0.017 0.018 19.411 19.412 cp_fm_cholesky_invert 11 10.9 17.355 17.364 17.355 17.364 qs_ot_p2m_diag 82 11.4 0.339 0.385 17.273 17.337 cp_fm_upper_to_full 104 14.8 11.367 16.725 11.367 16.725 cp_dbcsr_syevd 82 12.4 0.006 0.006 15.809 15.811 multiply_cannon_metrocomm3 29820 15.5 0.049 0.053 6.684 15.511 dbcsr_mm_accdrv_process 61748 16.2 8.315 10.549 13.753 14.255 sum_up_and_integrate 127 10.3 0.002 0.003 14.163 14.187 integrate_v_rspace 127 11.3 0.003 0.004 14.103 14.130 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.015 14.055 calculate_rho_elec 127 8.7 0.171 0.188 14.015 14.055 dbcsr_complete_redistribute 393 12.7 1.497 1.615 9.320 13.081 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.443 12.459 cp_fm_diag_elpa_base 82 14.4 11.450 11.767 12.436 12.452 make_images_data 4970 15.5 0.067 0.074 10.873 12.390 init_scf_run 11 5.9 0.000 0.001 12.345 12.347 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.345 12.347 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.854 11.612 hybrid_alltoall_any 5155 16.4 0.526 2.212 9.869 11.606 multiply_cannon_sync_h2d 29820 15.5 10.386 11.270 10.386 11.270 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.036 10.258 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.400 10.051 mp_alltoall_i22 712 14.1 5.564 9.381 5.564 9.381 pw_transfer 1535 11.6 0.085 0.100 9.170 9.253 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.947 9.037 cp_fm_cholesky_decompose 22 10.9 8.012 8.094 8.012 8.094 fft_wrap_pw1pw2_140 519 13.2 1.432 1.463 7.899 8.000 wfi_extrapolate 11 7.9 0.001 0.001 7.896 7.896 grid_integrate_task_list 127 12.3 7.472 7.752 7.472 7.752 multiply_cannon_metrocomm4 24850 15.5 0.084 0.094 2.799 7.588 mp_irecv_dv 75445 16.2 2.645 7.302 2.645 7.302 density_rs2pw 127 9.7 0.006 0.006 7.053 7.282 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.897 6.972 calculate_dm_sparse 127 9.5 0.001 0.001 6.263 6.355 fft3d_ps 1281 14.7 2.837 2.919 6.149 6.201 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.787 5.859 mp_alltoall_d11v 2401 14.1 5.234 5.604 5.234 5.604 grid_collocate_task_list 127 9.7 5.246 5.576 5.246 5.576 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.576 4.673 potential_pw2rs 127 12.3 0.023 0.023 4.473 4.484 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.459 4.462 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.370 4.429 calculate_first_density_matrix 1 7.0 0.010 0.015 4.236 4.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=194.103000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.363636, yerr=16.783502 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.536938E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.093 0.123 177.695 177.701 qs_mol_dyn_low 1 2.0 0.003 0.003 176.552 176.567 qs_forces 11 3.9 0.004 0.005 176.444 176.449 qs_energies 11 4.9 0.002 0.002 168.954 168.960 scf_env_do_scf 11 5.9 0.001 0.001 151.372 151.384 velocity_verlet 10 3.0 0.016 0.016 114.652 114.656 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 114.413 114.415 qs_scf_new_mos 118 7.6 0.001 0.001 76.364 76.435 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.364 76.434 dbcsr_multiply_generic 2527 12.6 0.184 0.196 73.661 73.958 ot_scf_mini 118 9.6 0.003 0.004 71.833 71.877 multiply_cannon 2527 13.6 0.565 0.587 54.061 57.910 multiply_cannon_loop 2527 14.6 0.811 0.849 50.991 51.680 ot_mini 118 10.6 0.001 0.001 38.449 38.492 init_scf_loop 11 6.9 0.000 0.000 36.801 36.803 prepare_preconditioner 11 7.9 0.000 0.000 32.869 32.903 make_preconditioner 11 8.9 0.000 0.000 32.869 32.903 mp_waitall_1 126780 16.7 25.552 31.982 25.552 31.982 make_full_inverse_cholesky 11 9.9 0.011 0.027 30.727 31.001 rebuild_ks_matrix 129 8.3 0.001 0.001 29.789 29.899 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.788 29.898 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.193 27.287 qs_ot_get_derivative 118 11.6 0.002 0.002 21.360 21.409 multiply_cannon_multrec 10108 15.6 10.456 14.718 18.293 21.313 qs_ot_get_p 129 10.4 0.003 0.003 20.434 20.523 multiply_cannon_metrocomm3 10108 15.6 0.026 0.028 12.394 19.511 cp_fm_cholesky_invert 11 10.9 18.789 18.795 18.789 18.795 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.063 17.220 apply_single 129 13.6 0.001 0.001 17.063 17.220 ot_diis_step 118 11.6 0.020 0.021 17.015 17.017 make_m2s 5054 13.6 0.068 0.071 15.545 16.407 qs_ot_p2m_diag 83 11.4 0.496 0.503 16.335 16.355 make_images 5054 14.6 2.193 2.627 15.233 16.097 cp_dbcsr_syevd 83 12.4 0.006 0.007 15.087 15.088 sum_up_and_integrate 129 10.3 0.002 0.002 14.637 14.691 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.665 14.689 calculate_rho_elec 129 8.7 0.258 0.269 14.664 14.689 integrate_v_rspace 129 11.3 0.004 0.004 14.576 14.633 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.003 12.015 cp_fm_diag_elpa_base 83 14.4 11.729 11.839 11.996 12.007 multiply_cannon_sync_h2d 10108 15.6 10.957 11.233 10.957 11.233 init_scf_run 11 5.9 0.000 0.002 10.704 10.704 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.704 10.704 hybrid_alltoall_any 5240 16.5 0.852 3.817 8.752 10.404 make_images_data 5054 15.6 0.056 0.066 9.011 10.372 pw_transfer 1559 11.6 0.086 0.096 9.793 9.831 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.568 9.611 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.442 8.473 fft_wrap_pw1pw2_140 527 13.2 1.812 1.848 8.367 8.424 cp_fm_cholesky_decompose 22 10.9 8.341 8.415 8.341 8.415 dbcsr_mm_accdrv_process 20926 16.1 3.379 4.389 7.487 8.215 grid_integrate_task_list 129 12.3 7.793 8.126 7.793 8.126 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.745 7.800 density_rs2pw 129 9.7 0.006 0.006 7.219 7.636 wfi_extrapolate 11 7.9 0.001 0.001 7.504 7.505 calculate_dm_sparse 129 9.5 0.001 0.001 6.287 6.348 fft3d_ps 1301 14.7 2.784 2.856 6.277 6.297 grid_collocate_task_list 129 9.7 5.613 5.917 5.613 5.917 dbcsr_complete_redistribute 395 12.7 2.100 2.151 5.502 5.866 mp_alltoall_d11v 2423 14.1 5.104 5.697 5.104 5.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.304 5.343 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.279 5.280 mp_allgather_i34 2527 14.6 1.260 4.889 1.260 4.889 potential_pw2rs 129 12.3 0.026 0.027 4.601 4.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.154 4.210 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.847 3.925 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.573 3.862 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.498 3.827 multiply_cannon_metrocomm4 7581 15.6 0.026 0.029 1.773 3.708 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.626 3.700 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.659 3.693 mp_irecv_dv 29086 15.9 1.733 3.627 1.733 3.627 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.701000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1459.454545, yerr=20.698545 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1979979923456 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.250208E+12 0.0% 0.0% 100.0% flops max/rank 11.788211E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806645728 0.0% 0.0% 100.0% number of processed stacks 1983232 0.0% 0.0% 100.0% average stack size 0.0 0.0 3432.1 marketing flops 145.668111E+12 ------------------------------------------------------------------------------- # multiplications 2537 max memory usage/rank 3.203310E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101480 MPI messages size (bytes): total size 1.145238E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.285357E+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 45968 35701915648 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4042 58572. MP_Allreduce 11182 1500. MP_Sync 87 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.236 0.274 299.396 299.402 qs_mol_dyn_low 1 2.0 0.003 0.003 298.572 298.600 qs_forces 11 3.9 0.004 0.004 298.460 298.467 qs_energies 11 4.9 0.002 0.002 289.490 289.506 scf_env_do_scf 11 5.9 0.001 0.001 266.417 266.426 velocity_verlet 10 3.0 0.002 0.002 215.092 215.101 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 139.831 139.833 init_scf_loop 11 6.9 0.000 0.000 126.309 126.312 prepare_preconditioner 11 7.9 0.000 0.000 121.385 121.406 make_preconditioner 11 8.9 0.000 0.000 121.385 121.405 make_full_inverse_cholesky 11 9.9 0.038 0.039 97.204 118.521 qs_scf_new_mos 118 7.6 0.001 0.001 95.064 95.143 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.063 95.142 ot_scf_mini 118 9.6 0.004 0.004 90.053 90.090 dbcsr_multiply_generic 2537 12.6 0.216 0.229 84.859 85.416 cp_fm_upper_to_full 105 14.8 52.973 76.197 52.973 76.197 multiply_cannon 2537 13.6 0.681 0.707 60.829 61.832 multiply_cannon_loop 2537 14.6 1.067 1.094 56.693 58.065 ot_mini 118 10.6 0.001 0.001 46.411 46.457 dbcsr_complete_redistribute 395 12.7 3.995 4.036 30.334 43.379 copy_fm_to_dbcsr 209 11.7 0.001 0.001 26.740 39.778 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.138 37.097 mp_alltoall_i22 718 14.1 21.937 34.895 21.937 34.895 rebuild_ks_matrix 129 8.3 0.001 0.001 34.824 34.853 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.823 34.852 cp_fm_cholesky_invert 11 10.9 34.659 34.666 34.659 34.666 mp_waitall_1 104900 16.8 29.626 34.493 29.626 34.493 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.358 32.394 qs_ot_get_p 129 10.4 0.003 0.004 28.132 28.174 qs_ot_get_derivative 118 11.6 0.002 0.002 26.299 26.338 qs_ot_p2m_diag 83 11.4 0.879 0.884 23.707 23.743 cp_dbcsr_syevd 83 12.4 0.007 0.007 21.838 21.840 multiply_cannon_metrocomm3 10148 15.6 0.025 0.026 20.262 21.756 make_m2s 5074 13.6 0.078 0.080 18.923 20.454 ot_diis_step 118 11.6 0.022 0.024 20.073 20.074 make_images 5074 14.6 3.096 3.316 18.437 19.968 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.567 19.718 apply_single 129 13.6 0.001 0.001 19.567 19.717 multiply_cannon_multrec 10148 15.6 10.376 12.240 18.305 18.510 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.452 18.453 cp_fm_diag_elpa_base 83 14.4 13.933 15.611 18.447 18.447 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.144 17.164 calculate_rho_elec 129 8.7 0.484 0.485 17.144 17.163 sum_up_and_integrate 129 10.3 0.002 0.002 16.551 16.638 integrate_v_rspace 129 11.3 0.004 0.004 16.488 16.575 multiply_cannon_sync_h2d 10148 15.6 14.414 14.461 14.414 14.461 make_images_data 5074 15.6 0.066 0.071 10.595 12.922 hybrid_alltoall_any 5260 16.5 1.330 3.089 10.637 12.860 init_scf_run 11 5.9 0.000 0.001 12.639 12.641 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.639 12.640 pw_transfer 1559 11.6 0.096 0.098 11.905 11.923 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.666 11.682 fft_wrap_pw1pw2_140 527 13.2 3.068 3.107 10.369 10.385 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.156 10.195 dbcsr_mm_accdrv_process 20966 16.1 4.228 6.457 7.678 9.715 cp_fm_cholesky_decompose 22 10.9 9.539 9.552 9.539 9.552 wfi_extrapolate 11 7.9 0.001 0.001 9.323 9.323 grid_integrate_task_list 129 12.3 8.620 8.803 8.620 8.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 8.184 8.186 density_rs2pw 129 9.7 0.005 0.006 7.901 7.946 mp_alltoall_d11v 2423 14.1 7.503 7.633 7.503 7.633 calculate_dm_sparse 129 9.5 0.001 0.001 6.900 6.977 fft3d_ps 1301 14.7 2.872 2.882 6.881 6.921 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.531 6.594 grid_collocate_task_list 129 9.7 6.469 6.536 6.469 6.536 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.291 6.389 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.402000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2896.727273, yerr=154.618646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 1.259897E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 33219687. 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 0.027 0.068 85.509 85.516 qs_energies 1 2.0 0.000 0.000 84.966 84.990 ls_scf 1 3.0 0.000 0.000 84.072 84.096 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.823 72.975 multiply_cannon 111 7.7 0.018 0.020 55.967 57.251 multiply_cannon_loop 111 8.7 0.227 0.240 52.549 53.800 ls_scf_main 1 4.0 0.000 0.000 52.578 52.579 density_matrix_trs4 2 5.0 0.002 0.003 46.989 47.040 ls_scf_init_scf 1 4.0 0.000 0.000 28.438 28.439 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.288 27.341 mp_waitall_1 11031 10.9 22.625 25.592 22.625 25.592 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.176 25.195 multiply_cannon_multrec 2664 9.7 8.161 8.937 15.552 17.303 multiply_cannon_sync_h2d 2664 9.7 13.572 15.232 13.572 15.232 make_m2s 222 7.7 0.009 0.012 13.264 13.801 make_images 222 8.7 0.098 0.110 13.242 13.781 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.774 12.157 make_images_data 222 9.7 0.004 0.005 7.746 8.323 hybrid_alltoall_any 227 10.6 0.216 1.844 6.662 8.127 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.428 8.029 dbcsr_mm_accdrv_process 4760 10.4 0.594 0.717 7.012 7.998 dbcsr_mm_accdrv_process_sort 4760 11.4 6.217 7.095 6.217 7.095 calculate_norms 4752 9.8 5.482 6.104 5.482 6.104 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.053 5.155 mp_sum_l 887 5.1 3.071 4.174 3.071 4.174 make_images_sizes 222 9.7 0.000 0.000 0.843 3.814 mp_alltoall_i44 222 10.7 0.843 3.814 0.843 3.814 arnoldi_extremal 4 6.8 0.000 0.000 3.341 3.367 arnoldi_normal_ev 4 7.8 0.001 0.004 3.341 3.366 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.062 3.351 mp_irecv_dv 6231 10.9 2.046 3.323 2.046 3.323 build_subspace 16 8.4 0.009 0.012 3.248 3.250 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.272 3.188 ls_scf_post 1 4.0 0.000 0.000 3.056 3.080 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.893 dbcsr_special_finalize 555 9.7 0.005 0.006 2.324 2.890 dbcsr_merge_single_wm 555 10.7 0.455 0.598 2.315 2.881 dbcsr_sort_data 658 11.4 2.118 2.629 2.118 2.629 make_images_pack 222 9.7 2.205 2.626 2.207 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.358 2.616 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.479 2.069 2.481 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.331 2.396 buffer_matrices_ensure_size 222 8.7 1.753 2.148 1.753 2.148 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.799 rebuild_ks_matrix 3 7.3 0.000 0.000 1.789 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.788 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.516000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 2.162483E+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 266673. MP_Allreduce 3138 10075. 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 0.045 0.060 91.273 91.275 qs_energies 1 2.0 0.000 0.001 90.644 90.655 ls_scf 1 3.0 0.000 0.001 89.158 89.172 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.106 75.370 multiply_cannon 111 7.7 0.027 0.039 52.849 57.734 ls_scf_main 1 4.0 0.000 0.006 55.191 55.192 multiply_cannon_loop 111 8.7 0.136 0.148 49.987 53.518 density_matrix_trs4 2 5.0 0.002 0.005 49.453 49.664 mp_waitall_1 9105 10.9 20.968 30.414 20.968 30.414 ls_scf_init_scf 1 4.0 0.000 0.002 30.256 30.258 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.076 29.155 multiply_cannon_multrec 1332 9.7 13.370 17.108 22.666 27.479 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.005 26.689 26.701 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.461 20.542 make_m2s 222 7.7 0.007 0.008 15.371 15.979 make_images 222 8.7 1.366 1.697 15.340 15.950 dbcsr_mm_accdrv_process 4041 10.4 0.352 0.531 8.892 10.459 dbcsr_mm_accdrv_process_sort 4041 11.4 8.409 9.928 8.409 9.928 make_images_data 222 9.7 0.004 0.005 8.957 9.903 hybrid_alltoall_any 227 10.6 0.543 2.558 8.324 9.739 mp_sum_l 887 5.1 5.405 8.821 5.405 8.821 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.176 7.755 mp_irecv_dv 3311 11.0 3.155 7.698 3.155 7.698 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.110 7.066 calculate_norms 2376 9.8 6.058 6.673 6.058 6.673 multiply_cannon_sync_h2d 1332 9.7 4.934 6.478 4.934 6.478 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.950 5.234 arnoldi_extremal 4 6.8 0.000 0.000 4.755 4.772 arnoldi_normal_ev 4 7.8 0.001 0.005 4.755 4.772 build_subspace 16 8.4 0.014 0.021 4.495 4.500 ls_scf_post 1 4.0 0.002 0.013 3.711 3.722 ls_scf_store_result 1 5.0 0.000 0.000 3.424 3.522 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.201 3.439 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.231 2.743 3.233 mp_allgather_i34 111 8.7 0.850 3.097 0.850 3.097 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.637 2.722 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.191 2.660 dbcsr_data_new 4174 10.1 2.113 2.409 2.113 2.409 make_images_pack 222 9.7 1.820 2.117 1.822 2.119 dbcsr_sort_data 436 11.2 1.847 2.107 1.847 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.876 1.878 rebuild_ks_matrix 3 7.3 0.000 0.000 1.864 1.866 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.864 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.275000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1832.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 2.926088E+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 265448. MP_Allreduce 3138 10896. 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 0.025 0.043 96.461 96.463 qs_energies 1 2.0 0.000 0.000 95.897 95.901 ls_scf 1 3.0 0.000 0.000 94.468 94.473 dbcsr_multiply_generic 111 6.7 0.016 0.016 78.830 79.224 multiply_cannon 111 7.7 0.028 0.040 55.680 59.371 multiply_cannon_loop 111 8.7 0.117 0.130 52.928 57.174 ls_scf_main 1 4.0 0.000 0.000 56.671 56.674 density_matrix_trs4 2 5.0 0.002 0.003 50.658 50.853 ls_scf_init_scf 1 4.0 0.000 0.000 34.113 34.117 mp_waitall_1 7281 11.0 24.082 33.564 24.082 33.564 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.892 32.970 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.560 30.573 multiply_cannon_multrec 888 9.7 12.741 15.450 21.307 24.601 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.467 22.915 make_m2s 222 7.7 0.006 0.007 16.377 17.116 make_images 222 8.7 1.585 1.850 16.339 17.075 make_images_data 222 9.7 0.004 0.004 9.615 10.658 hybrid_alltoall_any 227 10.6 0.641 2.955 9.231 10.375 dbcsr_mm_accdrv_process 3754 10.4 0.299 0.489 8.112 9.324 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.836 7.676 8.836 mp_sum_l 887 5.1 5.079 8.247 5.079 8.247 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.457 7.255 multiply_cannon_sync_h2d 888 9.7 6.047 7.243 6.047 7.243 mp_irecv_dv 2335 11.1 2.442 7.214 2.442 7.214 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.642 6.924 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.914 6.430 arnoldi_extremal 4 6.8 0.000 0.000 5.350 5.367 arnoldi_normal_ev 4 7.8 0.001 0.005 5.349 5.367 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.790 5.056 build_subspace 16 8.4 0.014 0.021 5.036 5.044 calculate_norms 1584 9.8 4.358 4.757 4.358 4.757 compute_matrix_preconditioner 1 6.0 0.000 0.001 4.356 4.362 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.552 3.848 mp_allgather_i34 111 8.7 0.865 3.739 0.865 3.739 acc_transpose_blocks 888 9.7 0.005 0.005 3.394 3.732 ls_scf_post 1 4.0 0.000 0.000 3.684 3.688 acc_transpose_blocks_kernels 888 10.7 0.013 0.014 3.331 3.666 jit_kernel_transpose 1 13.0 3.318 3.652 3.318 3.652 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.619 3.029 3.621 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.509 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.909 3.037 dbcsr_data_new 4116 9.9 2.117 2.463 2.117 2.463 make_images_sizes 222 9.7 0.000 0.000 0.950 2.245 mp_alltoall_i44 222 10.7 0.949 2.245 0.949 2.245 dbcsr_sort_data 325 11.1 1.857 2.113 1.857 2.113 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.463000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2243.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 3.450888E+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 266673. MP_Allreduce 3138 13030. 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 0.056 0.070 98.279 98.281 qs_energies 1 2.0 0.000 0.001 97.552 97.558 ls_scf 1 3.0 0.000 0.001 95.784 95.793 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.200 79.468 ls_scf_main 1 4.0 0.000 0.002 59.449 59.450 multiply_cannon 111 7.7 0.042 0.081 51.933 56.104 density_matrix_trs4 2 5.0 0.002 0.004 53.179 53.325 multiply_cannon_loop 111 8.7 0.153 0.168 46.867 49.869 ls_scf_init_scf 1 4.0 0.000 0.003 33.039 33.042 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.806 31.872 mp_waitall_1 6369 11.0 23.294 29.454 23.294 29.454 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.348 29.361 multiply_cannon_multrec 1332 9.7 14.190 17.294 22.121 24.931 make_m2s 222 7.7 0.007 0.007 21.502 22.963 make_images 222 8.7 3.137 3.602 21.452 22.915 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.370 17.230 make_images_data 222 9.7 0.004 0.004 12.131 13.816 hybrid_alltoall_any 227 10.6 0.798 3.761 11.409 13.343 dbcsr_mm_accdrv_process 3641 10.4 0.309 0.497 7.565 9.104 dbcsr_mm_accdrv_process_sort 3641 11.4 7.116 8.611 7.116 8.611 mp_sum_l 887 5.1 4.171 7.522 4.171 7.522 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.068 5.971 multiply_cannon_sync_h2d 1332 9.7 5.378 5.968 5.378 5.968 mp_irecv_dv 3229 10.9 2.043 5.906 2.043 5.906 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.189 5.872 arnoldi_extremal 4 6.8 0.000 0.000 5.478 5.491 arnoldi_normal_ev 4 7.8 0.001 0.004 5.478 5.491 build_subspace 16 8.4 0.014 0.021 5.135 5.145 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.622 5.127 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.625 4.780 mp_allgather_i34 111 8.7 2.181 4.514 2.181 4.514 calculate_norms 2376 9.8 4.175 4.489 4.175 4.489 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.674 4.005 dbcsr_matrix_vector_mult_local 304 10.0 3.202 3.701 3.204 3.703 dbcsr_sort_data 658 11.4 3.072 3.428 3.072 3.428 ls_scf_post 1 4.0 0.000 0.000 3.297 3.302 dbcsr_special_finalize 555 9.7 0.006 0.007 2.805 3.215 dbcsr_merge_single_wm 555 10.7 0.535 0.659 2.796 3.207 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.067 3.131 ls_scf_store_result 1 5.0 0.000 0.000 3.021 3.080 dbcsr_data_release 10477 10.7 1.596 2.445 1.596 2.445 dbcsr_finalize 304 7.8 0.049 0.061 1.812 1.977 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.281000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2688.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 4.759667E+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 265536. MP_Allreduce 3129 15263. 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 0.038 0.058 92.785 92.787 qs_energies 1 2.0 0.000 0.000 91.967 91.973 ls_scf 1 3.0 0.000 0.000 90.009 90.014 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.205 71.361 ls_scf_main 1 4.0 0.000 0.000 56.801 56.802 multiply_cannon 111 7.7 0.080 0.159 52.989 56.507 multiply_cannon_loop 111 8.7 0.088 0.094 50.397 51.533 density_matrix_trs4 2 5.0 0.002 0.003 49.734 49.813 ls_scf_init_scf 1 4.0 0.000 0.000 29.783 29.785 mp_waitall_1 5436 11.0 24.843 28.717 24.843 28.717 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.506 28.540 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 26.368 26.377 multiply_cannon_multrec 444 9.7 13.701 16.354 20.796 22.185 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.230 15.872 make_m2s 222 7.7 0.005 0.005 13.525 14.442 make_images 222 8.7 2.037 2.478 13.458 14.374 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.154 13.428 make_images_data 222 9.7 0.003 0.004 8.275 9.697 hybrid_alltoall_any 227 10.6 0.801 3.807 8.251 9.472 multiply_cannon_sync_h2d 444 9.7 6.703 7.987 6.703 7.987 dbcsr_mm_accdrv_process 3003 10.4 0.343 0.402 6.783 7.879 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.479 6.417 7.479 arnoldi_extremal 4 6.8 0.000 0.000 6.095 6.105 arnoldi_normal_ev 4 7.8 0.002 0.005 6.095 6.104 build_subspace 16 8.4 0.015 0.020 5.688 5.705 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.541 4.839 mp_sum_l 887 5.1 2.696 4.589 2.696 4.589 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.338 4.479 dbcsr_matrix_vector_mult_local 304 10.0 3.725 4.197 3.727 4.199 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.524 3.887 mp_irecv_dv 1241 11.2 1.511 3.846 1.511 3.846 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.886 3.711 calculate_norms 792 9.8 3.611 3.687 3.611 3.687 mp_allgather_i34 111 8.7 1.155 3.620 1.155 3.620 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.521 3.590 ls_scf_post 1 4.0 0.000 0.000 3.425 3.431 make_images_sizes 222 9.7 0.000 0.000 0.862 3.331 mp_alltoall_i44 222 10.7 0.862 3.330 0.862 3.330 ls_scf_store_result 1 5.0 0.000 0.000 3.221 3.252 dbcsr_finalize 304 7.8 0.062 0.077 2.196 2.273 dbcsr_data_new 4608 9.7 1.819 2.262 1.819 2.262 dbcsr_merge_all 275 8.9 0.480 0.524 2.055 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.040 2.041 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.008 2.009 qs_energies_init_hamiltonians 1 3.0 0.002 0.014 1.942 1.942 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.787000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3718.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_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 8.918012E+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 284089. MP_Allreduce 3123 21388. 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 0.110 0.115 107.084 107.086 qs_energies 1 2.0 0.000 0.000 105.624 105.638 ls_scf 1 3.0 0.000 0.000 102.650 102.664 dbcsr_multiply_generic 111 6.7 0.023 0.027 75.848 75.999 ls_scf_main 1 4.0 0.000 0.000 64.622 64.626 density_matrix_trs4 2 5.0 0.002 0.003 55.639 55.706 multiply_cannon 111 7.7 0.135 0.228 49.054 51.003 multiply_cannon_loop 111 8.7 0.098 0.099 46.198 46.867 ls_scf_init_scf 1 4.0 0.000 0.000 34.226 34.227 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.698 32.725 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.953 29.961 mp_waitall_1 4527 11.1 22.880 26.303 22.880 26.303 make_m2s 222 7.7 0.005 0.005 23.032 24.002 make_images 222 8.7 3.570 3.872 22.925 23.893 multiply_cannon_multrec 444 9.7 17.935 18.625 22.585 23.301 hybrid_alltoall_any 227 10.6 1.657 3.637 13.135 16.103 make_images_data 222 9.7 0.003 0.004 13.404 15.459 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.866 11.396 multiply_cannon_sync_h2d 444 9.7 8.792 8.852 8.792 8.852 arnoldi_extremal 4 6.8 0.000 0.000 7.753 7.764 arnoldi_normal_ev 4 7.8 0.003 0.009 7.753 7.764 build_subspace 16 8.4 0.026 0.035 7.186 7.199 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.699 5.849 dbcsr_matrix_vector_mult_local 304 10.0 5.124 5.443 5.127 5.445 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.086 5.180 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.899 5.159 dbcsr_mm_accdrv_process 1814 10.4 0.316 0.376 4.468 4.612 dbcsr_mm_accdrv_process_sort 1814 11.4 4.129 4.265 4.129 4.265 ls_scf_post 1 4.0 0.000 0.000 3.801 3.816 make_images_sizes 222 9.7 0.000 0.000 1.499 3.695 mp_alltoall_i44 222 10.7 1.499 3.694 1.499 3.694 ls_scf_store_result 1 5.0 0.000 0.000 3.515 3.550 calculate_norms 792 9.8 3.229 3.270 3.229 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.073 3.161 mp_allgather_i34 111 8.7 0.981 3.053 0.981 3.053 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.944 2.944 dbcsr_merge_all 275 8.9 0.884 0.918 2.857 2.941 dbcsr_complete_redistribute 5 7.6 1.434 1.481 2.754 2.869 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.520 dbcsr_sort_data 325 11.1 2.439 2.494 2.439 2.494 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.406 2.408 dbcsr_new_transposed 4 7.5 0.241 0.251 2.391 2.405 dbcsr_data_new 6591 9.6 1.855 2.351 1.855 2.351 rebuild_ks_matrix 3 7.3 0.000 0.000 2.340 2.342 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.005 2.340 2.342 dbcsr_frobenius_norm 74 6.6 2.052 2.129 2.200 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.202 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.201 dbcsr_data_release 12724 10.6 1.985 2.195 1.985 2.195 mp_alltoall_d11v 48 9.2 2.117 2.177 2.117 2.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.086000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6953.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c94756ddc75c22fb829516d4d1ee9958f089b1c9_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.461248E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4802585. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.043 248.305 248.307 qs_mol_dyn_low 1 2.0 0.004 0.029 247.251 247.342 qs_forces 5 3.8 0.006 0.063 247.120 247.126 qs_energies 5 4.8 0.004 0.053 243.870 243.911 scf_env_do_scf 5 5.8 0.002 0.037 227.628 227.633 scf_env_do_scf_inner_loop 105 6.6 0.003 0.038 195.746 195.760 qs_scf_new_mos 105 7.6 0.000 0.001 151.600 151.811 qs_scf_loop_do_ot 105 8.6 0.001 0.001 151.600 151.811 ot_scf_mini 105 9.6 0.003 0.006 141.558 141.752 dbcsr_multiply_generic 1445 12.2 0.127 0.135 137.265 137.980 multiply_cannon 1445 13.2 0.277 0.287 116.806 119.928 multiply_cannon_loop 1445 14.2 2.863 3.017 114.668 116.584 velocity_verlet 4 3.0 0.006 0.017 113.753 113.755 ot_mini 105 10.6 0.001 0.003 61.891 62.064 qs_ot_get_p 112 10.4 0.001 0.003 49.262 49.553 mp_waitall_1 488190 16.1 37.276 46.216 37.276 46.216 multiply_cannon_multrec 69360 15.2 29.389 34.594 39.471 44.317 qs_ot_get_derivative 55 11.6 0.002 0.058 40.051 40.258 qs_ot_p2m_diag 40 11.0 0.020 0.034 37.616 37.702 multiply_cannon_metrocomm3 69360 15.2 0.202 0.210 26.945 34.993 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.292 34.295 multiply_cannon_sync_h2d 69360 15.2 28.607 33.143 28.607 33.143 init_scf_loop 7 6.6 0.001 0.019 31.818 31.821 rebuild_ks_matrix 110 8.4 0.000 0.000 31.338 31.540 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.017 31.337 31.540 cp_fm_syevd 40 13.0 0.000 0.001 29.041 29.194 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.781 28.962 prepare_preconditioner 7 7.6 0.000 0.000 26.540 26.574 make_preconditioner 7 8.6 0.000 0.003 26.540 26.574 cp_fm_redistribute_end 40 14.0 12.113 24.156 12.121 24.159 cp_fm_syevd_base 40 14.0 12.028 24.079 12.028 24.079 apply_preconditioner_dbcsr 62 12.6 0.001 0.009 23.132 23.412 apply_single 62 13.6 0.000 0.000 23.131 23.412 ot_new_cg_direction 55 11.6 0.001 0.003 21.108 21.110 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.951 19.252 calculate_rho_elec 110 8.6 0.030 0.033 18.951 19.251 make_full_inverse_cholesky 7 9.6 0.000 0.000 18.573 18.633 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.090 15.390 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.439 14.574 pw_transfer 1645 12.4 0.079 0.096 13.777 14.012 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 13.639 13.878 density_rs2pw 110 9.6 0.006 0.007 12.987 13.392 init_scf_run 5 5.8 0.000 0.001 13.312 13.312 scf_env_initial_rho_setup 5 6.8 0.002 0.004 13.311 13.312 mp_sum_l 4764 12.2 12.384 13.274 12.384 13.274 cp_fm_cholesky_invert 7 10.6 13.098 13.108 13.098 13.108 fft_wrap_pw1pw2_240 915 15.0 1.128 1.220 11.964 12.179 calculate_dm_sparse 110 9.5 0.001 0.002 11.626 11.825 qs_vxc_create 110 10.4 0.002 0.005 11.583 11.638 dbcsr_mm_accdrv_process 154766 15.8 6.297 6.499 9.953 11.331 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.648 10.719 fft3d_pb 915 16.0 2.374 2.661 10.199 10.474 check_diag 80 13.5 8.595 8.889 9.639 9.792 transfer_rs2pw 445 10.6 0.007 0.008 9.090 9.558 sum_up_and_integrate 60 10.3 0.001 0.003 9.513 9.525 integrate_v_rspace 60 11.3 0.001 0.015 9.496 9.508 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.925 9.408 acc_transpose_blocks 69360 15.2 0.355 0.370 8.709 9.297 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 8.760 8.777 xc_rho_set_and_dset_create 110 12.4 0.077 0.100 8.072 8.325 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.239 8.300 calculate_first_density_matrix 1 7.0 0.001 0.006 8.257 8.270 make_m2s 2890 13.2 0.079 0.088 7.584 8.173 make_images 2890 14.2 0.242 0.262 7.477 8.066 xc_vxc_pw_create 60 11.3 0.039 0.052 7.724 7.777 mp_alltoall_z22v 2340 17.7 7.358 7.669 7.358 7.669 make_full_single_inverse 7 9.6 0.001 0.002 7.571 7.606 xc_pw_derive 510 13.4 0.005 0.006 7.250 7.326 acc_transpose_blocks_kernels 69360 16.2 0.851 0.894 5.942 6.434 jit_kernel_transpose 5 15.0 5.091 5.549 5.091 5.549 potential_pw2rs 60 12.3 0.003 0.003 5.524 5.549 mp_waitany 7680 13.5 4.644 5.461 4.644 5.461 multiply_cannon_metrocomm4 67915 15.2 0.186 0.206 2.052 5.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=248.307000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c94756ddc75c22fb829516d4d1ee9958f089b1c9 Summary: empty Status: OK