=== 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: b1f098bdf757a0d911df37584392ac41aeb8a328 ################# 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 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # 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: 28.03.2024 # # \ 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; \ rm -rf build; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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 ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(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_f90cstub.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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/01 job id: 53252614 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/02 job id: 53252615 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/03 job id: 53252616 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/04 job id: 53252617 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/05 job id: 53252618 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/06 job id: 53252619 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/07 job id: 53252620 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/08 job id: 53252621 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/09 job id: 53252622 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/10 job id: 53252623 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/11 job id: 53252624 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/12 job id: 53252625 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/13 job id: 53252626 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/14 job id: 53252627 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/15 job id: 53252628 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/16 job id: 53252629 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/17 job id: 53252630 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/18 job id: 53252631 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/19 job id: 53252632 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/20 job id: 53252633 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/21 job id: 53252634 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/22 job id: 53252635 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/23 job id: 53252636 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/24 job id: 53252637 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/25 job id: 53252638 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/26 job id: 53252639 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/27 job id: 53252640 --- 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/b1f098bdf757a0d911df37584392ac41aeb8a328_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 172669. 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 0.017 0.031 129.391 129.392 farming_run 1 2.0 128.720 128.722 129.363 129.367 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.538294E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. 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 0.005 0.026 110.677 110.677 qs_energies 1 2.0 0.047 0.093 110.480 110.481 mp2_main 1 3.0 0.000 0.000 108.636 108.685 mp2_gpw_main 1 4.0 0.080 0.097 107.693 107.694 mp2_ri_gpw_compute_in 1 5.0 0.187 0.203 89.180 89.353 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 57.302 57.461 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.174 43.746 48.671 get_2c_integrals 1 6.0 0.008 0.009 31.430 31.687 integrate_v_rspace 273 8.0 0.443 0.460 25.606 30.132 fft_wrap_pw1pw2 5465 10.4 0.066 0.070 28.049 28.623 grid_integrate_task_list 273 9.0 20.835 26.328 20.835 26.328 fft_wrap_pw1pw2_100 2178 11.4 1.297 1.454 25.503 26.050 compute_2c_integrals 1 7.0 0.002 0.002 22.140 22.144 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.985 21.875 mp2_eri_2c_integrate_gpw 1 9.0 2.379 2.430 20.983 21.874 rpa_ri_compute_en 1 5.0 0.020 0.032 18.341 18.638 fft3d_s 5443 12.4 16.927 17.443 16.947 17.464 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.308 10.767 15.308 calculate_wavefunction 272 8.0 5.302 5.433 12.779 13.736 calc_potential_gpw 544 9.5 0.004 0.004 11.061 12.271 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.957 10.985 rpa_num_int 1 6.0 0.001 0.021 10.506 10.515 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.447 10.470 cp_fm_cholesky_decompose 12 8.2 9.440 9.696 9.440 9.696 potential_pw2rs 545 10.0 0.106 0.108 8.478 9.554 calc_mat_Q 8 8.0 0.000 0.000 9.445 9.537 contract_S_to_Q 8 9.0 0.000 0.000 8.869 8.959 cholesky_decomp 1 7.0 0.000 0.000 8.480 8.734 collocate_single_gaussian 272 10.0 0.039 0.042 7.827 8.602 parallel_gemm_fm 14 9.1 0.000 0.000 8.496 8.594 parallel_gemm_fm_cosma 14 10.1 8.496 8.594 8.496 8.594 create_integ_mat 1 6.0 0.015 0.028 7.577 7.586 array2fm 1 7.0 0.000 0.000 6.690 7.058 pw_poisson_solve 545 10.5 0.010 0.012 4.840 6.051 pw_gather_s 2722 12.2 3.882 5.103 3.882 5.103 pw_scatter_s 2720 12.7 4.458 4.897 4.458 4.897 pw_poisson_set 548 11.5 0.017 0.019 3.488 4.354 pw_copy 4911 11.6 2.723 3.343 2.723 3.343 array2fm_buffer_send 1 8.0 2.952 3.174 2.952 3.174 pw_derive 1641 12.5 2.069 2.602 2.069 2.602 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.080 1.233 2.023 2.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.694377, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 200775. 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 0.023 0.035 427.237 427.240 farming_run 1 2.0 425.911 425.916 427.197 427.201 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.279906E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27355934. MP_SendRecv 2876 2171443. 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 0.011 0.033 223.225 223.226 qs_energies 1 2.0 0.000 0.000 223.009 223.016 scf_env_do_scf 1 3.0 0.000 0.000 115.284 115.285 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.399 114.406 rebuild_ks_matrix 4 6.0 0.000 0.000 114.398 114.405 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.398 114.405 hfx_ks_matrix 4 8.0 0.001 0.001 114.054 114.058 integrate_four_center 4 9.0 0.152 0.471 114.053 114.058 mp2_main 1 3.0 0.000 0.000 107.441 107.448 mp2_gpw_main 1 4.0 0.047 0.072 106.598 106.609 integrate_four_center_main 4 10.0 0.115 0.529 101.914 105.921 integrate_four_center_bin 265 11.0 101.798 105.902 101.798 105.902 init_scf_loop 1 4.0 0.000 0.000 97.499 97.499 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 79.780 80.858 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.814 57.890 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.161 44.545 49.380 integrate_v_rspace 95 8.0 0.398 0.569 28.195 32.865 fft_wrap_pw1pw2 1868 10.4 0.028 0.035 30.561 30.883 fft_wrap_pw1pw2_100 730 11.4 0.629 0.719 28.275 28.596 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.665 28.385 grid_integrate_task_list 95 9.0 23.495 28.376 23.495 28.376 ao_to_mo_and_store_B_mult_1 91 7.0 10.593 25.965 10.593 25.965 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.847 1.895 24.931 24.941 get_2c_integrals 1 6.0 0.000 0.000 22.889 22.901 compute_2c_integrals 1 7.0 0.002 0.003 22.053 22.056 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.740 21.929 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.862 21.739 21.927 fft3d_s 1823 12.4 19.258 19.894 19.271 19.907 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.784 17.784 calc_potential_gpw 182 9.5 0.002 0.002 12.332 12.937 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.715 11.412 calculate_wavefunction 91 8.0 1.983 2.026 10.005 10.793 potential_pw2rs 186 10.0 0.033 0.035 9.095 10.326 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.579 8.842 9.445 collocate_single_gaussian 91 10.0 0.017 0.023 8.298 9.063 local_gemm 172 8.0 8.286 8.893 8.286 8.893 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.522 7.867 8.214 mp_sync 37 10.5 4.099 8.202 4.099 8.202 integrate_four_center_load 4 10.0 0.000 0.000 6.759 6.763 hfx_load_balance 1 11.0 0.000 0.000 6.759 6.763 mp2_ri_gpw_compute_en_ener 172 7.0 6.330 6.392 6.330 6.392 mp_sendrecv_dm3 2068 8.0 5.893 6.243 5.893 6.243 pw_poisson_solve 186 10.4 0.005 0.006 5.548 6.035 pw_gather_s 912 12.2 4.878 5.430 4.878 5.430 pw_poisson_set 189 11.4 0.009 0.010 4.107 4.556 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.597415, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1534.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 445.620224E+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 3690 72628. 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 0.019 0.117 56.209 56.210 qs_mol_dyn_low 1 2.0 0.013 0.082 55.457 55.465 qs_forces 11 3.9 0.002 0.003 54.996 54.998 qs_energies 11 4.9 0.021 0.023 53.388 53.402 scf_env_do_scf 11 5.9 0.000 0.001 47.261 47.261 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 45.306 45.306 dbcsr_multiply_generic 2286 12.5 0.094 0.098 35.257 35.717 qs_scf_new_mos 108 7.5 0.000 0.001 35.235 35.528 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.235 35.528 ot_scf_mini 108 9.5 0.002 0.002 33.474 33.629 multiply_cannon 2286 13.5 0.188 0.195 27.248 28.929 multiply_cannon_loop 2286 14.5 1.815 1.908 26.581 28.286 velocity_verlet 10 3.0 0.005 0.031 27.063 27.071 ot_mini 108 10.5 0.001 0.001 20.498 20.759 qs_ot_get_derivative 108 11.5 0.001 0.001 17.361 17.553 mp_waitall_1 245248 16.5 9.049 15.196 9.049 15.196 multiply_cannon_metrocomm3 54864 15.5 0.071 0.076 6.412 13.057 multiply_cannon_multrec 54864 15.5 3.629 5.610 7.784 11.180 qs_ot_get_p 119 10.4 0.001 0.001 8.246 8.512 rebuild_ks_matrix 119 8.3 0.000 0.000 7.964 8.120 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.010 7.964 8.120 mp_sum_l 7287 12.8 6.086 7.930 6.086 7.930 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.009 7.163 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.136 6.574 multiply_cannon_sync_h2d 54864 15.5 5.126 6.124 5.126 6.124 dbcsr_mm_accdrv_process 76910 16.1 1.831 2.867 4.069 5.851 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.549 5.661 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.325 5.363 init_scf_run 11 5.9 0.000 0.001 4.885 4.885 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.885 4.885 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.497 4.497 sum_up_and_integrate 119 10.3 0.001 0.002 4.426 4.432 integrate_v_rspace 119 11.3 0.002 0.003 4.415 4.423 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.283 4.284 cp_fm_redistribute_end 50 14.0 2.186 4.257 2.192 4.261 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.126 4.256 calculate_rho_elec 119 8.7 0.011 0.017 4.125 4.256 cp_fm_diag_elpa_base 50 14.0 2.063 4.154 2.067 4.162 calculate_dm_sparse 119 9.5 0.000 0.000 3.393 3.559 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.946 3.162 apply_single 119 13.6 0.000 0.000 2.945 3.162 acc_transpose_blocks 54864 15.5 0.217 0.234 2.541 3.088 calculate_first_density_matrix 1 7.0 0.000 0.001 2.985 2.991 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.974 2.947 jit_kernel_multiply 13 15.8 2.175 2.910 2.175 2.910 ot_diis_step 108 11.5 0.006 0.006 2.810 2.810 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.511 2.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.450 2.465 density_rs2pw 119 9.7 0.003 0.004 2.110 2.234 mp_sum_d 4139 12.0 1.617 2.182 1.617 2.182 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.099 2.161 grid_integrate_task_list 119 12.3 2.000 2.116 2.000 2.116 init_scf_loop 11 6.9 0.000 0.000 1.936 1.936 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.808 1.847 wfi_extrapolate 11 7.9 0.001 0.001 1.828 1.828 potential_pw2rs 119 12.3 0.004 0.004 1.805 1.816 make_m2s 4572 13.5 0.054 0.056 1.704 1.753 make_images 4572 14.5 0.133 0.139 1.621 1.669 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.570 1.643 mp_alltoall_d11v 2130 13.8 1.402 1.533 1.402 1.533 transfer_rs2pw 487 10.6 0.006 0.006 1.373 1.465 mp_waitany 12084 13.8 1.250 1.431 1.250 1.431 grid_collocate_task_list 119 9.7 1.350 1.430 1.350 1.430 acc_transpose_blocks_sync 164592 16.5 1.193 1.425 1.193 1.425 fft3d_ps 1201 13.6 0.365 0.466 1.344 1.410 acc_transpose_blocks_kernels 54864 16.5 0.248 0.381 1.087 1.395 transfer_pw2rs 487 13.2 0.005 0.006 1.335 1.344 fft_wrap_pw1pw2_140 487 12.2 0.048 0.054 1.204 1.276 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.845 1.260 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.210000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 486.133760E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 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 0.036 0.072 39.304 39.305 qs_mol_dyn_low 1 2.0 0.003 0.003 38.831 38.840 qs_forces 11 3.9 0.002 0.002 38.695 38.696 qs_energies 11 4.9 0.001 0.001 37.037 37.041 scf_env_do_scf 11 5.9 0.000 0.001 31.594 31.595 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.269 29.270 dbcsr_multiply_generic 2286 12.5 0.101 0.105 22.714 23.055 qs_scf_new_mos 108 7.5 0.001 0.001 20.884 21.113 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.883 21.112 ot_scf_mini 108 9.5 0.002 0.003 19.961 20.127 multiply_cannon 2286 13.5 0.209 0.216 17.362 18.893 velocity_verlet 10 3.0 0.005 0.027 18.034 18.036 multiply_cannon_loop 2286 14.5 1.186 1.238 16.224 17.892 ot_mini 108 10.5 0.001 0.001 12.534 12.769 mp_waitall_1 200699 16.5 5.572 10.903 5.572 10.903 qs_ot_get_derivative 108 11.5 0.001 0.001 10.194 10.365 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 4.237 9.832 multiply_cannon_multrec 27432 15.5 1.826 4.152 6.419 9.359 dbcsr_mm_accdrv_process 47894 16.0 3.566 6.079 4.511 6.995 rebuild_ks_matrix 119 8.3 0.000 0.000 6.670 6.805 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.670 6.805 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.862 5.984 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.838 5.012 qs_ot_get_p 119 10.4 0.001 0.001 4.538 4.763 mp_sum_l 7287 12.8 2.311 4.431 2.311 4.431 init_scf_run 11 5.9 0.000 0.001 4.136 4.136 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.135 4.136 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.301 4.042 apply_single 119 13.6 0.000 0.000 3.301 4.042 sum_up_and_integrate 119 10.3 0.001 0.001 3.561 3.569 integrate_v_rspace 119 11.3 0.002 0.003 3.547 3.555 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.438 3.467 calculate_rho_elec 119 8.7 0.021 0.024 3.437 3.466 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.924 2.942 make_m2s 4572 13.5 0.052 0.054 2.591 2.933 make_images 4572 14.5 0.206 0.244 2.501 2.846 calculate_first_density_matrix 1 7.0 0.000 0.000 2.675 2.676 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.530 2.531 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.294 2.382 init_scf_loop 11 6.9 0.000 0.000 2.306 2.307 ot_diis_step 108 11.5 0.011 0.011 2.292 2.292 calculate_dm_sparse 119 9.5 0.000 0.001 2.206 2.291 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.226 2.227 cp_fm_redistribute_end 50 14.0 1.129 2.204 1.132 2.206 multiply_cannon_sync_h2d 27432 15.5 1.709 2.196 1.709 2.196 cp_fm_diag_elpa_base 50 14.0 1.043 2.117 1.071 2.160 acc_transpose_blocks 27432 15.5 0.111 0.116 1.730 2.107 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.077 2.079 jit_kernel_multiply 9 16.2 0.886 2.032 0.886 2.032 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.895 1.934 grid_integrate_task_list 119 12.3 1.835 1.930 1.835 1.930 density_rs2pw 119 9.7 0.003 0.004 1.741 1.823 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.596 1.625 make_images_data 4572 15.5 0.047 0.054 1.215 1.528 prepare_preconditioner 11 7.9 0.000 0.000 1.432 1.458 make_preconditioner 11 8.9 0.000 0.000 1.432 1.458 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.075 1.451 wfi_extrapolate 11 7.9 0.001 0.001 1.405 1.405 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.389 1.397 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.327 1.376 grid_collocate_task_list 119 9.7 1.290 1.353 1.290 1.353 mp_alltoall_d11v 2130 13.8 1.188 1.347 1.188 1.347 fft3d_ps 1201 13.6 0.511 0.565 1.314 1.343 potential_pw2rs 119 12.3 0.006 0.006 1.314 1.320 fft_wrap_pw1pw2_140 487 12.2 0.048 0.051 1.280 1.311 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.191 1.237 transfer_rs2pw 487 10.6 0.005 0.006 1.020 1.113 qs_energies_init_hamiltonians 11 5.9 0.007 0.022 1.056 1.057 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.780 1.045 mp_sum_d 4139 12.0 0.576 1.008 0.576 1.008 mp_allgather_i34 2286 14.5 0.560 0.969 0.560 0.969 acc_transpose_blocks_sync 82296 16.5 0.814 0.937 0.814 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.877 transfer_pw2rs 487 13.2 0.004 0.005 0.858 0.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.305000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.181818, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 514.306048E+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 3629 73909. 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 0.010 0.118 33.235 33.236 qs_mol_dyn_low 1 2.0 0.003 0.003 32.808 32.817 qs_forces 11 3.9 0.020 0.045 32.640 32.641 qs_energies 11 4.9 0.001 0.001 31.097 31.116 scf_env_do_scf 11 5.9 0.000 0.001 25.056 25.056 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.855 22.855 dbcsr_multiply_generic 2286 12.5 0.095 0.098 18.381 18.454 multiply_cannon 2286 13.5 0.195 0.200 15.124 15.792 qs_scf_new_mos 108 7.5 0.001 0.001 15.426 15.439 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.425 15.438 multiply_cannon_loop 2286 14.5 0.864 0.890 14.361 15.062 velocity_verlet 10 3.0 0.072 0.114 14.757 14.759 ot_scf_mini 108 9.5 0.002 0.002 14.668 14.679 ot_mini 108 10.5 0.001 0.001 9.166 9.178 multiply_cannon_multrec 18288 15.5 1.934 2.894 7.828 8.222 qs_ot_get_derivative 108 11.5 0.001 0.001 7.693 7.704 dbcsr_mm_accdrv_process 38222 16.0 5.740 6.583 5.798 6.648 rebuild_ks_matrix 119 8.3 0.000 0.000 5.870 5.883 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.869 5.883 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.166 5.178 init_scf_run 11 5.9 0.000 0.001 4.776 4.776 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.776 4.776 mp_waitall_1 158411 16.6 2.707 3.836 2.707 3.836 calculate_first_density_matrix 1 7.0 0.000 0.000 3.607 3.608 qs_ot_get_derivative_taylor 59 13.0 0.001 0.002 2.846 3.545 sum_up_and_integrate 119 10.3 0.001 0.001 3.425 3.430 integrate_v_rspace 119 11.3 0.003 0.003 3.413 3.419 qs_ot_get_p 119 10.4 0.001 0.001 3.294 3.315 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.084 3.090 calculate_rho_elec 119 8.7 0.031 0.031 3.083 3.089 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.513 2.514 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.108 2.460 apply_single 119 13.6 0.000 0.000 2.108 2.460 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.563 2.455 calculate_dm_sparse 119 9.5 0.000 0.001 2.412 2.421 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.343 2.348 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.192 2.201 init_scf_loop 11 6.9 0.000 0.000 2.185 2.186 acc_transpose_blocks 18288 15.5 0.077 0.080 2.077 2.180 grid_integrate_task_list 119 12.3 1.812 1.905 1.812 1.905 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.898 1.898 make_m2s 4572 13.5 0.044 0.045 1.739 1.880 density_rs2pw 119 9.7 0.003 0.004 1.655 1.806 make_images 4572 14.5 0.192 0.205 1.653 1.793 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.750 1.753 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.694 1.702 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.661 1.671 cp_fm_diag_elpa_base 50 14.0 1.638 1.651 1.659 1.670 mp_sum_l 7287 12.8 1.240 1.642 1.240 1.642 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.593 1.599 ot_diis_step 108 11.5 0.011 0.011 1.459 1.459 prepare_preconditioner 11 7.9 0.000 0.000 1.435 1.438 make_preconditioner 11 8.9 0.000 0.000 1.435 1.438 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.321 1.376 fft_wrap_pw1pw2_140 487 12.2 0.060 0.062 1.298 1.305 grid_collocate_task_list 119 9.7 1.233 1.302 1.233 1.302 fft3d_ps 1201 13.6 0.534 0.552 1.279 1.289 acc_transpose_blocks_kernels 18288 16.5 0.217 0.226 1.235 1.265 potential_pw2rs 119 12.3 0.007 0.008 1.223 1.226 multiply_cannon_sync_h2d 18288 15.5 1.037 1.161 1.037 1.161 wfi_extrapolate 11 7.9 0.001 0.001 1.118 1.118 transfer_rs2pw 487 10.6 0.005 0.005 0.958 1.096 jit_kernel_transpose 5 15.6 1.017 1.045 1.017 1.045 qs_energies_init_hamiltonians 11 5.9 0.002 0.013 1.017 1.035 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.937 0.957 make_images_data 4572 15.5 0.047 0.051 0.738 0.898 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.645 0.823 acc_transpose_blocks_sync 54864 16.5 0.747 0.819 0.747 0.819 transfer_pw2rs 487 13.2 0.004 0.004 0.781 0.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.772 0.774 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.464 0.770 mp_alltoall_d11v 2130 13.8 0.644 0.762 0.644 0.762 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.676 0.733 mp_waitany 9880 13.7 0.543 0.708 0.543 0.708 transfer_rs2pw_140 130 11.5 0.119 0.138 0.537 0.678 mp_alltoall_z22v 1201 15.6 0.600 0.667 0.600 0.667 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.236000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=489.727273, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 554.729472E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.036 36.541 36.542 qs_mol_dyn_low 1 2.0 0.003 0.003 36.173 36.203 qs_forces 11 3.9 0.002 0.002 36.098 36.099 qs_energies 11 4.9 0.001 0.001 34.394 34.400 scf_env_do_scf 11 5.9 0.000 0.001 29.033 29.033 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.236 26.236 dbcsr_multiply_generic 2286 12.5 0.101 0.104 20.485 20.614 qs_scf_new_mos 108 7.5 0.001 0.001 18.204 18.259 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.204 18.258 velocity_verlet 10 3.0 0.006 0.009 18.015 18.017 ot_scf_mini 108 9.5 0.002 0.003 17.165 17.211 multiply_cannon 2286 13.5 0.221 0.230 16.691 17.163 multiply_cannon_loop 2286 14.5 1.508 1.621 15.709 16.251 ot_mini 108 10.5 0.001 0.001 10.584 10.643 multiply_cannon_multrec 27432 15.5 2.511 3.204 9.241 9.579 qs_ot_get_derivative 108 11.5 0.001 0.001 8.705 8.751 dbcsr_mm_accdrv_process 47916 15.9 6.162 7.674 6.629 8.038 rebuild_ks_matrix 119 8.3 0.000 0.000 6.308 6.352 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.308 6.351 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.591 5.631 init_scf_run 11 5.9 0.000 0.001 4.025 4.025 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.024 4.025 qs_ot_get_p 119 10.4 0.001 0.001 3.707 3.774 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.276 3.623 sum_up_and_integrate 119 10.3 0.001 0.001 3.458 3.467 integrate_v_rspace 119 11.3 0.003 0.003 3.447 3.455 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.317 3.377 calculate_rho_elec 119 8.7 0.040 0.046 3.317 3.377 acc_transpose_blocks 27432 15.5 0.113 0.118 2.644 2.963 init_scf_loop 11 6.9 0.000 0.000 2.773 2.774 calculate_first_density_matrix 1 7.0 0.000 0.000 2.611 2.616 make_m2s 4572 13.5 0.054 0.056 2.351 2.450 mp_waitall_1 137007 16.6 1.906 2.405 1.906 2.405 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.375 2.396 make_images 4572 14.5 0.273 0.335 2.242 2.338 calculate_dm_sparse 119 9.5 0.000 0.000 2.243 2.297 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.059 2.271 apply_single 119 13.6 0.000 0.000 2.059 2.271 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.243 2.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.135 2.136 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.947 1.958 prepare_preconditioner 11 7.9 0.000 0.000 1.942 1.951 make_preconditioner 11 8.9 0.000 0.000 1.942 1.951 grid_integrate_task_list 119 12.3 1.820 1.913 1.820 1.913 density_rs2pw 119 9.7 0.003 0.003 1.764 1.886 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.683 1.879 ot_diis_step 108 11.5 0.012 0.012 1.835 1.835 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.829 1.830 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.758 1.790 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.573 1.574 acc_transpose_blocks_sync 82296 16.5 1.413 1.561 1.413 1.561 cp_fm_redistribute_end 50 14.0 0.789 1.549 0.793 1.550 cp_fm_diag_elpa_base 50 14.0 0.731 1.477 0.755 1.518 mp_sum_l 7287 12.8 1.020 1.468 1.020 1.468 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.430 1.466 fft3d_ps 1201 13.6 0.577 0.641 1.419 1.443 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.832 1.412 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.336 1.351 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.314 1.332 grid_collocate_task_list 119 9.7 1.249 1.327 1.249 1.327 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 1.090 1.270 potential_pw2rs 119 12.3 0.008 0.009 1.258 1.264 jit_kernel_multiply 6 16.0 0.400 1.245 0.400 1.245 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 1.061 1.062 transfer_rs2pw 487 10.6 0.005 0.005 0.934 1.058 mp_alltoall_d11v 2130 13.8 0.808 1.007 0.808 1.007 jit_kernel_transpose 5 15.6 0.820 1.001 0.820 1.001 make_images_data 4572 15.5 0.047 0.051 0.864 0.969 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.737 0.893 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.802 0.877 dbcsr_complete_redistribute 329 12.2 0.108 0.125 0.690 0.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.811 mp_alltoall_z22v 1201 15.6 0.726 0.760 0.726 0.760 transfer_pw2rs 487 13.2 0.003 0.004 0.751 0.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.542000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.454545, yerr=2.147457 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 629.227520E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.009 0.058 28.767 28.768 qs_mol_dyn_low 1 2.0 0.003 0.003 28.414 28.421 qs_forces 11 3.9 0.003 0.003 28.304 28.305 qs_energies 11 4.9 0.001 0.001 26.613 26.615 scf_env_do_scf 11 5.9 0.000 0.001 21.539 21.539 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.251 19.251 velocity_verlet 10 3.0 0.014 0.035 14.315 14.320 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.685 13.748 qs_scf_new_mos 108 7.5 0.001 0.001 11.869 11.897 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.869 11.896 multiply_cannon 2286 13.5 0.225 0.234 11.159 11.666 ot_scf_mini 108 9.5 0.002 0.002 11.166 11.190 multiply_cannon_loop 2286 14.5 0.645 0.662 10.228 10.370 multiply_cannon_multrec 9144 15.5 1.840 2.062 6.498 6.734 ot_mini 108 10.5 0.001 0.001 6.515 6.544 rebuild_ks_matrix 119 8.3 0.000 0.000 5.663 5.685 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.663 5.684 qs_ot_get_derivative 108 11.5 0.001 0.001 5.221 5.246 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.039 5.059 dbcsr_mm_accdrv_process 12550 15.8 3.696 4.620 4.548 4.662 init_scf_run 11 5.9 0.000 0.001 3.621 3.621 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.621 3.621 sum_up_and_integrate 119 10.3 0.001 0.001 3.296 3.301 integrate_v_rspace 119 11.3 0.003 0.003 3.286 3.291 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.174 3.179 calculate_rho_elec 119 8.7 0.060 0.061 3.174 3.179 qs_ot_get_p 119 10.4 0.001 0.001 2.758 2.796 calculate_first_density_matrix 1 7.0 0.000 0.000 2.517 2.517 init_scf_loop 11 6.9 0.000 0.000 2.267 2.268 make_m2s 4572 13.5 0.034 0.035 1.771 1.935 grid_integrate_task_list 119 12.3 1.869 1.919 1.869 1.919 calculate_dm_sparse 119 9.5 0.000 0.000 1.885 1.906 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.866 1.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.853 1.855 make_images 4572 14.5 0.270 0.300 1.681 1.844 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.797 1.800 mp_waitall_1 115863 16.7 1.283 1.735 1.283 1.735 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.712 1.722 acc_transpose_blocks 9144 15.5 0.040 0.041 1.664 1.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.687 1.693 density_rs2pw 119 9.7 0.003 0.004 1.561 1.666 jit_kernel_multiply 8 15.6 0.813 1.580 0.813 1.580 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.557 1.558 prepare_preconditioner 11 7.9 0.000 0.000 1.526 1.531 make_preconditioner 11 8.9 0.000 0.000 1.526 1.531 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.422 1.447 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.427 1.437 grid_collocate_task_list 119 9.7 1.296 1.396 1.296 1.396 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.320 1.335 fft3d_ps 1201 13.6 0.637 0.651 1.324 1.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.281 1.287 ot_diis_step 108 11.5 0.012 0.013 1.281 1.281 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.240 1.281 apply_single 119 13.6 0.000 0.000 1.240 1.281 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.269 1.277 cp_fm_diag_elpa_base 50 14.0 1.243 1.258 1.267 1.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.219 1.219 potential_pw2rs 119 12.3 0.010 0.010 1.110 1.112 wfi_extrapolate 11 7.9 0.001 0.001 1.055 1.055 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.761 0.985 make_images_data 4572 15.5 0.042 0.047 0.772 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.931 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.885 0.919 mp_alltoall_d11v 2130 13.8 0.784 0.859 0.784 0.859 transfer_rs2pw 487 10.6 0.004 0.005 0.753 0.846 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.817 0.824 jit_kernel_transpose 5 15.6 0.768 0.801 0.768 0.801 acc_transpose_blocks_sync 27432 16.5 0.723 0.752 0.723 0.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.744 0.746 cp_fm_cholesky_invert 11 10.9 0.739 0.744 0.739 0.744 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.665 0.717 mp_allgather_i34 2286 14.5 0.228 0.659 0.228 0.659 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.325 0.643 transfer_pw2rs 487 13.2 0.003 0.004 0.626 0.627 mp_alltoall_z22v 1201 15.6 0.563 0.596 0.563 0.596 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.768000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.272727, yerr=8.645450 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 789.737472E+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 3629 74142. 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 0.017 0.075 40.754 40.754 qs_mol_dyn_low 1 2.0 0.003 0.003 40.209 40.270 qs_forces 11 3.9 0.002 0.002 40.136 40.137 qs_energies 11 4.9 0.001 0.001 38.181 38.184 scf_env_do_scf 11 5.9 0.001 0.001 31.862 31.862 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.967 25.969 velocity_verlet 10 3.0 0.001 0.002 21.758 21.764 dbcsr_multiply_generic 2286 12.5 0.103 0.105 19.397 19.609 qs_scf_new_mos 108 7.5 0.001 0.001 16.879 16.974 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.878 16.973 multiply_cannon 2286 13.5 0.322 0.356 15.639 16.614 ot_scf_mini 108 9.5 0.002 0.002 15.717 15.815 multiply_cannon_loop 2286 14.5 0.863 0.888 14.352 15.285 ot_mini 108 10.5 0.001 0.001 9.606 9.721 multiply_cannon_multrec 9144 15.5 3.413 4.740 8.917 8.997 qs_ot_get_derivative 108 11.5 0.001 0.001 7.577 7.675 rebuild_ks_matrix 119 8.3 0.000 0.000 6.983 7.123 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.983 7.123 dbcsr_mm_accdrv_process 12550 15.8 4.536 5.922 5.364 6.673 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.301 6.428 init_scf_loop 11 6.9 0.000 0.000 5.868 5.871 prepare_preconditioner 11 7.9 0.000 0.000 4.864 4.878 make_preconditioner 11 8.9 0.000 0.000 4.864 4.878 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.904 4.768 init_scf_run 11 5.9 0.000 0.001 4.299 4.299 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.299 4.299 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.855 3.862 calculate_rho_elec 119 8.7 0.118 0.121 3.854 3.862 sum_up_and_integrate 119 10.3 0.001 0.001 3.659 3.665 integrate_v_rspace 119 11.3 0.004 0.004 3.649 3.656 qs_ot_get_p 119 10.4 0.001 0.001 3.208 3.344 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.831 3.212 mp_waitall_1 94719 16.7 2.165 3.075 2.165 3.075 cp_fm_upper_to_full 72 14.2 2.069 2.937 2.069 2.937 calculate_first_density_matrix 1 7.0 0.000 0.000 2.894 2.895 calculate_dm_sparse 119 9.5 0.000 0.000 2.714 2.731 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.157 2.457 apply_single 119 13.6 0.000 0.000 2.157 2.457 make_m2s 4572 13.5 0.038 0.038 2.246 2.386 acc_transpose_blocks 9144 15.5 0.043 0.043 2.279 2.341 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.310 2.316 make_images 4572 14.5 0.355 0.391 2.126 2.266 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.323 2.167 grid_integrate_task_list 119 12.3 2.077 2.142 2.077 2.142 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.024 2.073 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.030 2.031 dbcsr_complete_redistribute 329 12.2 0.259 0.266 1.479 2.023 ot_diis_step 108 11.5 0.014 0.014 2.000 2.000 fft_wrap_pw1pw2_140 487 12.2 0.179 0.181 1.974 1.982 density_rs2pw 119 9.7 0.003 0.003 1.906 1.925 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.857 1.860 mp_sum_l 7287 12.8 1.007 1.783 1.007 1.783 fft3d_ps 1201 13.6 0.859 0.881 1.762 1.768 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.754 1.754 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.737 1.747 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.185 1.728 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.709 1.718 grid_collocate_task_list 119 9.7 1.519 1.560 1.519 1.560 mp_alltoall_i22 627 13.8 0.982 1.537 0.982 1.537 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.502 1.502 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.955 1.496 wfi_extrapolate 11 7.9 0.001 0.001 1.344 1.344 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.303 1.323 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.220 1.220 cp_fm_diag_elpa_base 50 14.0 1.135 1.148 1.218 1.218 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.007 1.202 mp_alltoall_d11v 2130 13.8 1.135 1.164 1.135 1.164 acc_transpose_blocks_sync 27432 16.5 1.106 1.158 1.106 1.158 potential_pw2rs 119 12.3 0.013 0.014 1.154 1.155 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 1.115 1.142 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.084 1.097 make_images_data 4572 15.5 0.046 0.048 0.930 1.097 cp_fm_cholesky_invert 11 10.9 1.080 1.087 1.080 1.087 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.990 1.025 jit_kernel_transpose 5 15.6 0.995 1.024 0.995 1.024 qs_create_task_list 11 7.9 0.000 0.000 0.933 0.944 generate_qs_task_list 11 8.9 0.369 0.389 0.932 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.892 0.905 jit_kernel_multiply 5 15.3 0.799 0.900 0.799 0.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.754000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=743.454545, yerr=14.797811 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 495.755264E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 84.181 84.183 qs_mol_dyn_low 1 2.0 0.003 0.007 83.780 83.789 qs_forces 11 3.9 0.003 0.003 83.380 83.381 qs_energies 11 4.9 0.001 0.001 80.528 80.545 scf_env_do_scf 11 5.9 0.000 0.001 71.000 71.001 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.771 65.772 dbcsr_multiply_generic 2055 12.4 0.106 0.110 53.143 53.391 qs_scf_new_mos 99 7.5 0.000 0.001 49.167 49.342 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.167 49.342 ot_scf_mini 99 9.5 0.002 0.002 46.767 46.930 multiply_cannon 2055 13.4 0.182 0.190 43.702 44.697 multiply_cannon_loop 2055 14.4 1.811 1.885 42.786 43.825 velocity_verlet 10 3.0 0.003 0.016 42.483 42.487 ot_mini 99 10.5 0.001 0.001 27.908 28.030 qs_ot_get_derivative 99 11.5 0.001 0.001 21.011 21.149 multiply_cannon_multrec 49320 15.4 11.120 11.993 17.435 18.280 rebuild_ks_matrix 110 8.3 0.000 0.000 14.294 14.468 qs_ks_build_kohn_sham_matrix 110 9.3 0.010 0.011 14.293 14.468 mp_waitall_1 220248 16.4 11.618 12.957 11.618 12.957 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.540 12.705 qs_ot_get_p 110 10.4 0.001 0.001 10.118 10.280 multiply_cannon_sync_h2d 49320 15.4 9.341 10.009 9.341 10.009 multiply_cannon_metrocomm3 49320 15.4 0.079 0.085 7.028 8.106 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.433 8.008 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.219 7.776 apply_single 110 13.6 0.000 0.001 7.218 7.776 init_scf_run 11 5.9 0.000 0.001 7.421 7.421 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.421 7.421 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.777 6.799 sum_up_and_integrate 110 10.3 0.002 0.003 6.758 6.773 integrate_v_rspace 110 11.3 0.003 0.003 6.733 6.754 ot_diis_step 99 11.5 0.006 0.006 6.603 6.603 dbcsr_mm_accdrv_process 87628 16.1 3.173 3.392 6.187 6.592 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.387 6.495 mp_sum_l 6594 12.7 5.160 6.245 5.160 6.245 qs_rho_update_rho_low 110 7.6 0.000 0.001 5.989 6.133 calculate_rho_elec 110 8.6 0.020 0.024 5.989 6.132 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.778 5.778 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.314 5.315 cp_fm_redistribute_end 48 14.0 2.697 5.289 2.707 5.293 init_scf_loop 11 6.9 0.000 0.000 5.198 5.199 cp_fm_diag_elpa_base 48 14.0 2.576 5.177 2.584 5.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.078 4.080 calculate_dm_sparse 110 9.5 0.001 0.001 3.890 4.028 wfi_extrapolate 11 7.9 0.001 0.001 3.927 3.927 make_m2s 4110 13.4 0.061 0.066 3.822 3.925 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.841 3.862 make_images 4110 14.4 0.178 0.190 3.727 3.834 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.698 3.736 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.468 3.536 grid_integrate_task_list 110 12.3 3.256 3.415 3.256 3.415 calculate_first_density_matrix 1 7.0 0.000 0.000 3.394 3.399 acc_transpose_blocks 49320 15.4 0.211 0.230 3.033 3.216 density_rs2pw 110 9.6 0.004 0.004 3.023 3.206 prepare_preconditioner 11 7.9 0.000 0.000 3.053 3.078 make_preconditioner 11 8.9 0.000 0.000 3.053 3.078 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.789 2.839 jit_kernel_multiply 13 15.9 2.731 2.797 2.731 2.797 fft_wrap_pw1pw2 1111 11.6 0.013 0.018 2.563 2.626 mp_alltoall_d11v 2046 13.8 2.158 2.601 2.158 2.601 potential_pw2rs 110 12.3 0.006 0.006 2.373 2.386 grid_collocate_task_list 110 9.6 2.157 2.275 2.157 2.275 fft_wrap_pw1pw2_140 451 12.1 0.093 0.101 2.190 2.251 fft3d_ps 1111 13.6 0.731 0.816 2.099 2.161 mp_waitany 14300 13.8 1.823 2.017 1.823 2.017 mp_sum_d 3893 11.9 1.502 1.999 1.502 1.999 transfer_rs2pw 451 10.6 0.006 0.006 1.819 1.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.891 1.917 make_images_data 4110 15.4 0.043 0.046 1.737 1.883 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.773 1.805 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.518 1.791 transfer_pw2rs 451 13.1 0.005 0.007 1.680 1.687 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.183000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=471.363636, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 590.716928E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 1899808. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.036 69.528 69.529 qs_mol_dyn_low 1 2.0 0.021 0.145 69.271 69.298 qs_forces 11 3.9 0.003 0.003 68.729 68.731 qs_energies 11 4.9 0.001 0.001 65.381 65.385 scf_env_do_scf 11 5.9 0.001 0.002 56.608 56.609 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.555 49.556 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.370 38.553 velocity_verlet 10 3.0 0.028 0.063 35.804 35.823 qs_scf_new_mos 99 7.5 0.001 0.001 33.704 33.877 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.704 33.877 multiply_cannon 2055 13.4 0.225 0.247 31.661 32.790 ot_scf_mini 99 9.5 0.003 0.003 32.028 32.192 multiply_cannon_loop 2055 14.4 1.156 1.176 30.381 31.362 ot_mini 99 10.5 0.001 0.001 18.613 18.771 multiply_cannon_multrec 24660 15.4 6.955 8.699 14.190 16.162 rebuild_ks_matrix 110 8.3 0.000 0.001 13.679 13.794 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.679 13.793 qs_ot_get_derivative 99 11.5 0.001 0.001 12.854 13.015 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.033 12.133 mp_waitall_1 176588 16.5 7.755 9.964 7.755 9.964 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.259 8.053 dbcsr_mm_accdrv_process 52282 16.1 5.622 6.669 7.065 7.416 multiply_cannon_sync_h2d 24660 15.4 6.302 7.331 6.302 7.331 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.417 7.051 apply_single 110 13.6 0.000 0.001 6.416 7.051 qs_ot_get_p 110 10.4 0.001 0.001 6.818 7.022 init_scf_loop 11 6.9 0.000 0.000 7.013 7.014 sum_up_and_integrate 110 10.3 0.008 0.051 6.407 6.421 integrate_v_rspace 110 11.3 0.003 0.003 6.374 6.392 init_scf_run 11 5.9 0.000 0.001 6.295 6.296 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.295 6.295 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.822 5.830 calculate_rho_elec 110 8.6 0.042 0.071 5.821 5.830 ot_diis_step 99 11.5 0.010 0.010 5.710 5.710 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.829 5.551 prepare_preconditioner 11 7.9 0.000 0.000 4.890 4.916 make_preconditioner 11 8.9 0.000 0.000 4.890 4.916 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.789 4.809 make_m2s 4110 13.4 0.057 0.059 4.288 4.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.543 4.645 make_images 4110 14.4 0.408 0.466 4.177 4.624 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.322 4.322 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.766 3.766 cp_fm_redistribute_end 48 14.0 1.899 3.733 1.906 3.736 cp_fm_diag_elpa_base 48 14.0 1.765 3.578 1.823 3.661 wfi_extrapolate 11 7.9 0.001 0.001 3.462 3.462 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.319 3.399 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 3.235 3.385 density_rs2pw 110 9.6 0.004 0.004 3.042 3.371 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.366 3.367 grid_integrate_task_list 110 12.3 3.158 3.295 3.158 3.295 calculate_dm_sparse 110 9.5 0.001 0.001 3.118 3.149 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.015 3.070 fft_wrap_pw1pw2_140 451 12.1 0.111 0.120 2.719 2.861 make_images_data 4110 15.4 0.048 0.053 2.304 2.773 fft3d_ps 1111 13.6 1.074 1.287 2.612 2.750 hybrid_alltoall_any 4261 16.3 0.106 0.471 1.995 2.732 calculate_first_density_matrix 1 7.0 0.000 0.001 2.714 2.717 cp_fm_cholesky_invert 11 10.9 2.406 2.420 2.406 2.420 mp_sum_l 6594 12.7 1.797 2.413 1.797 2.413 grid_collocate_task_list 110 9.6 2.168 2.318 2.168 2.318 potential_pw2rs 110 12.3 0.008 0.008 2.264 2.281 mp_alltoall_d11v 2046 13.8 1.845 2.224 1.845 2.224 jit_kernel_multiply 10 16.2 1.072 2.111 1.072 2.111 acc_transpose_blocks 24660 15.4 0.113 0.116 2.015 2.068 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.967 1.986 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.963 1.964 transfer_rs2pw 451 10.6 0.007 0.007 1.559 1.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.808 1.823 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.681 1.696 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.817 1.670 mp_allgather_i34 2055 14.4 0.588 1.570 0.588 1.570 mp_irecv_dv 57340 16.2 0.685 1.545 0.685 1.545 mp_waitany 10164 13.8 1.235 1.507 1.235 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.009 1.391 1.494 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.529000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.636364, yerr=6.442370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 665.956352E+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 3480 77281. 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 0.007 0.026 59.317 59.318 qs_mol_dyn_low 1 2.0 0.003 0.003 58.967 59.047 qs_forces 11 3.9 0.005 0.009 58.877 58.878 qs_energies 11 4.9 0.015 0.028 55.700 55.707 scf_env_do_scf 11 5.9 0.001 0.001 47.212 47.212 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.709 39.709 velocity_verlet 10 3.0 0.001 0.001 31.245 31.247 dbcsr_multiply_generic 2055 12.4 0.108 0.112 29.566 29.856 qs_scf_new_mos 99 7.5 0.001 0.001 25.350 25.452 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.350 25.452 multiply_cannon 2055 13.4 0.213 0.221 23.245 24.466 ot_scf_mini 99 9.5 0.002 0.002 24.099 24.209 multiply_cannon_loop 2055 14.4 0.816 0.856 22.104 23.069 ot_mini 99 10.5 0.001 0.001 13.839 13.949 rebuild_ks_matrix 110 8.3 0.000 0.000 12.176 12.320 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.176 12.319 multiply_cannon_multrec 16440 15.4 3.595 4.554 10.286 11.226 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.731 10.857 mp_waitall_1 139946 16.5 7.064 10.167 7.064 10.167 qs_ot_get_derivative 99 11.5 0.001 0.001 9.401 9.509 init_scf_loop 11 6.9 0.000 0.000 7.468 7.468 multiply_cannon_metrocomm3 16440 15.4 0.044 0.046 4.482 7.415 dbcsr_mm_accdrv_process 34862 16.1 5.928 6.345 6.533 6.764 sum_up_and_integrate 110 10.3 0.001 0.002 6.136 6.153 integrate_v_rspace 110 11.3 0.003 0.003 6.110 6.126 init_scf_run 11 5.9 0.000 0.001 6.094 6.094 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.094 6.094 prepare_preconditioner 11 7.9 0.000 0.000 5.642 5.659 make_preconditioner 11 8.9 0.000 0.000 5.642 5.659 qs_ot_get_p 110 10.4 0.001 0.001 5.425 5.561 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.475 5.484 calculate_rho_elec 110 8.6 0.058 0.058 5.475 5.484 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.983 5.437 apply_single 110 13.6 0.000 0.000 4.983 5.436 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.102 5.347 make_m2s 4110 13.4 0.049 0.051 4.188 4.545 make_images 4110 14.4 0.400 0.521 4.072 4.430 ot_diis_step 99 11.5 0.010 0.011 4.408 4.408 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.255 3.949 multiply_cannon_sync_h2d 16440 15.4 3.221 3.799 3.221 3.799 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.784 3.787 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.413 3.413 grid_integrate_task_list 110 12.3 3.179 3.376 3.179 3.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.338 3.339 calculate_first_density_matrix 1 7.0 0.000 0.000 3.097 3.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.007 3.067 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.048 3.057 wfi_extrapolate 11 7.9 0.001 0.001 2.904 2.904 density_rs2pw 110 9.6 0.004 0.004 2.729 2.881 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.863 2.873 cp_fm_diag_elpa_base 48 14.0 2.821 2.837 2.861 2.871 make_images_data 4110 15.4 0.045 0.049 2.358 2.787 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.110 2.773 calculate_dm_sparse 110 9.5 0.001 0.001 2.681 2.721 fft_wrap_pw1pw2_140 451 12.1 0.127 0.131 2.612 2.624 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.454 2.515 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.900 2.485 fft3d_ps 1111 13.6 1.093 1.099 2.361 2.373 cp_fm_cholesky_invert 11 10.9 2.361 2.372 2.361 2.372 mp_irecv_dv 48980 15.7 0.823 2.348 0.823 2.348 grid_collocate_task_list 110 9.6 2.218 2.337 2.218 2.337 mp_sum_l 6594 12.7 1.509 2.139 1.509 2.139 mp_alltoall_d11v 2046 13.8 1.792 2.122 1.792 2.122 potential_pw2rs 110 12.3 0.010 0.011 1.996 2.005 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 1.937 1.940 acc_transpose_blocks 16440 15.4 0.075 0.077 1.778 1.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.661 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.618 1.634 dbcsr_complete_redistribute 325 12.2 0.298 0.325 1.305 1.606 mp_allgather_i34 2055 14.4 0.452 1.523 0.452 1.523 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.475 transfer_rs2pw 451 10.6 0.005 0.006 1.258 1.422 cp_fm_cholesky_decompose 22 10.9 1.330 1.357 1.330 1.357 mp_waitany 17072 13.8 1.127 1.325 1.127 1.325 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.266 1.274 cp_fm_upper_to_full 70 14.2 0.941 1.199 0.941 1.199 rs_gather_matrices 110 12.3 0.138 0.150 0.884 1.194 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.318000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.818182, yerr=8.266231 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 742.699008E+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 3480 77279. 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 0.010 0.028 64.080 64.081 qs_mol_dyn_low 1 2.0 0.003 0.003 63.705 63.715 qs_forces 11 3.9 0.003 0.003 63.598 63.598 qs_energies 11 4.9 0.001 0.001 60.213 60.216 scf_env_do_scf 11 5.9 0.000 0.001 51.204 51.210 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.938 41.938 velocity_verlet 10 3.0 0.021 0.047 34.640 34.642 dbcsr_multiply_generic 2055 12.4 0.115 0.121 32.063 32.276 qs_scf_new_mos 99 7.5 0.001 0.001 27.672 27.786 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.671 27.786 multiply_cannon 2055 13.4 0.238 0.258 24.989 26.184 ot_scf_mini 99 9.5 0.002 0.003 25.995 26.111 multiply_cannon_loop 2055 14.4 1.380 1.481 23.563 24.179 ot_mini 99 10.5 0.001 0.001 15.359 15.503 multiply_cannon_multrec 24660 15.4 4.283 7.460 13.944 15.490 rebuild_ks_matrix 110 8.3 0.000 0.000 11.929 12.032 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.928 12.031 qs_ot_get_derivative 99 11.5 0.001 0.001 11.185 11.303 dbcsr_mm_accdrv_process 52304 16.0 8.330 9.950 9.502 10.691 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.547 10.634 init_scf_loop 11 6.9 0.000 0.000 9.227 9.227 prepare_preconditioner 11 7.9 0.000 0.000 7.407 7.422 make_preconditioner 11 8.9 0.000 0.000 7.407 7.422 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.238 7.046 mp_waitall_1 121746 16.5 4.567 6.541 4.567 6.541 init_scf_run 11 5.9 0.000 0.001 6.286 6.286 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.285 6.286 sum_up_and_integrate 110 10.3 0.001 0.002 6.047 6.060 integrate_v_rspace 110 11.3 0.003 0.004 6.020 6.033 make_m2s 4110 13.4 0.060 0.062 5.404 5.796 qs_ot_get_p 110 10.4 0.001 0.001 5.547 5.701 make_images 4110 14.4 0.577 0.697 5.262 5.649 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.522 5.532 calculate_rho_elec 110 8.6 0.077 0.081 5.521 5.532 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.080 4.245 apply_single 110 13.6 0.000 0.000 4.080 4.245 ot_diis_step 99 11.5 0.011 0.011 4.134 4.134 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.676 3.695 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.528 3.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.531 3.532 grid_integrate_task_list 110 12.3 3.290 3.432 3.290 3.432 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.308 3.372 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.581 3.310 calculate_first_density_matrix 1 7.0 0.000 0.000 3.287 3.290 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.156 3.190 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.138 3.165 calculate_dm_sparse 110 9.5 0.001 0.001 3.085 3.126 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.120 3.120 hybrid_alltoall_any 4261 16.3 0.122 0.466 2.258 3.096 make_images_data 4110 15.4 0.048 0.052 2.710 3.087 acc_transpose_blocks 24660 15.4 0.105 0.109 2.570 2.969 cp_fm_upper_to_full 70 14.2 2.058 2.906 2.058 2.906 wfi_extrapolate 11 7.9 0.001 0.001 2.901 2.902 density_rs2pw 110 9.6 0.003 0.004 2.645 2.778 dbcsr_complete_redistribute 325 12.2 0.376 0.441 2.049 2.756 fft_wrap_pw1pw2_140 451 12.1 0.145 0.148 2.700 2.730 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.595 2.604 cp_fm_diag_elpa_base 48 14.0 2.531 2.556 2.591 2.603 multiply_cannon_sync_h2d 24660 15.4 2.386 2.540 2.386 2.540 fft3d_ps 1111 13.6 1.134 1.180 2.407 2.431 grid_collocate_task_list 110 9.6 2.267 2.398 2.267 2.398 cp_fm_cholesky_invert 11 10.9 2.378 2.388 2.378 2.388 qs_energies_init_hamiltonians 11 5.9 0.003 0.014 2.247 2.248 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.540 2.226 jit_kernel_multiply 8 16.0 0.824 2.146 0.824 2.146 mp_alltoall_d11v 2046 13.8 1.816 2.061 1.816 2.061 potential_pw2rs 110 12.3 0.012 0.012 1.857 1.860 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.159 1.827 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.780 1.813 mp_alltoall_i22 605 13.7 1.031 1.742 1.031 1.742 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.719 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.655 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.638 mp_allgather_i34 2055 14.4 0.508 1.612 0.508 1.612 mp_sum_l 6594 12.7 0.978 1.583 0.978 1.583 acc_transpose_blocks_sync 73980 16.4 1.365 1.552 1.365 1.552 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.847 1.512 mp_irecv_dv 62702 16.1 0.742 1.433 0.742 1.433 cp_fm_cholesky_decompose 22 10.9 1.330 1.376 1.330 1.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.081000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.727273, yerr=9.696212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 863.748096E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.030 0.044 53.980 53.981 qs_mol_dyn_low 1 2.0 0.003 0.003 53.716 53.753 qs_forces 11 3.9 0.003 0.003 53.576 53.577 qs_energies 11 4.9 0.001 0.001 49.959 49.964 scf_env_do_scf 11 5.9 0.000 0.001 41.502 41.503 scf_env_do_scf_inner_loop 99 6.5 0.053 0.073 34.415 34.416 velocity_verlet 10 3.0 0.024 0.033 29.863 29.869 dbcsr_multiply_generic 2055 12.4 0.106 0.107 23.671 23.767 qs_scf_new_mos 99 7.5 0.001 0.001 20.318 20.389 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.317 20.388 multiply_cannon 2055 13.4 0.237 0.248 18.137 19.527 ot_scf_mini 99 9.5 0.002 0.002 19.068 19.109 multiply_cannon_loop 2055 14.4 0.602 0.620 16.870 17.173 rebuild_ks_matrix 110 8.3 0.000 0.000 11.450 11.481 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.450 11.481 ot_mini 99 10.5 0.001 0.001 10.661 10.693 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.189 10.220 multiply_cannon_multrec 8220 15.4 3.210 4.406 7.851 8.825 mp_waitall_1 103326 16.6 5.690 7.142 5.690 7.142 init_scf_loop 11 6.9 0.000 0.000 7.042 7.043 qs_ot_get_derivative 99 11.5 0.001 0.001 6.979 7.017 sum_up_and_integrate 110 10.3 0.001 0.002 5.985 6.000 integrate_v_rspace 110 11.3 0.003 0.003 5.958 5.973 dbcsr_mm_accdrv_process 17442 15.9 3.294 4.492 4.503 5.607 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.588 5.604 calculate_rho_elec 110 8.6 0.115 0.116 5.588 5.603 prepare_preconditioner 11 7.9 0.000 0.000 5.342 5.349 make_preconditioner 11 8.9 0.000 0.000 5.342 5.349 init_scf_run 11 5.9 0.000 0.001 5.322 5.322 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.322 5.322 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.936 5.009 qs_ot_get_p 110 10.4 0.001 0.001 4.524 4.549 make_m2s 4110 13.4 0.038 0.039 4.156 4.450 make_images 4110 14.4 0.649 0.710 4.026 4.320 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.963 4.287 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.762 3.836 apply_single 110 13.6 0.000 0.000 3.761 3.836 ot_diis_step 99 11.5 0.012 0.012 3.661 3.661 grid_integrate_task_list 110 12.3 3.388 3.516 3.388 3.516 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.280 3.289 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.145 3.148 calculate_dm_sparse 110 9.5 0.001 0.001 2.836 2.871 fft_wrap_pw1pw2_140 451 12.1 0.191 0.194 2.840 2.858 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.811 2.812 density_rs2pw 110 9.6 0.004 0.004 2.630 2.763 calculate_first_density_matrix 1 7.0 0.000 0.000 2.738 2.739 qs_energies_init_hamiltonians 11 5.9 0.010 0.037 2.725 2.726 hybrid_alltoall_any 4261 16.3 0.202 0.838 2.195 2.720 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.647 2.648 make_images_data 4110 15.4 0.041 0.048 2.234 2.621 grid_collocate_task_list 110 9.6 2.369 2.555 2.369 2.555 cp_fm_cholesky_invert 11 10.9 2.506 2.513 2.506 2.513 wfi_extrapolate 11 7.9 0.001 0.001 2.492 2.492 multiply_cannon_sync_h2d 8220 15.4 2.389 2.484 2.389 2.484 fft3d_ps 1111 13.6 1.273 1.298 2.440 2.453 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.306 2.314 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.293 2.313 cp_fm_diag_elpa_base 48 14.0 2.252 2.273 2.304 2.312 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.088 2.154 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.999 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.874 1.891 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.791 1.800 potential_pw2rs 110 12.3 0.015 0.015 1.794 1.798 mp_alltoall_d11v 2046 13.8 1.570 1.729 1.570 1.729 acc_transpose_blocks 8220 15.4 0.038 0.039 1.691 1.713 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.529 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.590 1.594 mp_allgather_i34 2055 14.4 0.439 1.542 0.439 1.542 dbcsr_complete_redistribute 325 12.2 0.524 0.556 1.415 1.500 cp_fm_cholesky_decompose 22 10.9 1.419 1.432 1.419 1.432 qs_create_task_list 11 7.9 0.000 0.000 1.204 1.306 generate_qs_task_list 11 8.9 0.376 0.443 1.204 1.306 transfer_rs2pw 451 10.6 0.005 0.005 1.081 1.239 jit_kernel_multiply 6 15.4 0.887 1.205 0.887 1.205 mp_waitany 9240 13.8 1.040 1.192 1.040 1.192 copy_dbcsr_to_fm 151 11.3 0.002 0.003 1.113 1.127 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.466 1.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.981000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=814.909091, yerr=12.609193 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 1.346867E+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 3469 77990. 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 0.082 0.131 78.976 78.976 qs_mol_dyn_low 1 2.0 0.003 0.003 78.565 78.582 qs_forces 11 3.9 0.003 0.003 78.493 78.494 qs_energies 11 4.9 0.001 0.001 74.410 74.412 scf_env_do_scf 11 5.9 0.000 0.001 63.856 63.857 velocity_verlet 10 3.0 0.001 0.001 48.011 48.017 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.442 43.443 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.116 30.226 qs_scf_new_mos 99 7.5 0.001 0.001 26.590 26.638 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.589 26.637 ot_scf_mini 99 9.5 0.002 0.002 24.789 24.831 multiply_cannon 2055 13.4 0.333 0.363 23.125 23.788 multiply_cannon_loop 2055 14.4 0.825 0.832 21.365 21.735 init_scf_loop 11 6.9 0.000 0.000 20.343 20.344 prepare_preconditioner 11 7.9 0.000 0.000 18.355 18.364 make_preconditioner 11 8.9 0.000 0.000 18.355 18.364 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.522 17.856 ot_mini 99 10.5 0.001 0.001 14.250 14.291 rebuild_ks_matrix 110 8.3 0.000 0.001 13.396 13.446 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.396 13.446 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.160 12.205 cp_fm_upper_to_full 70 14.2 8.288 11.729 8.288 11.729 multiply_cannon_multrec 8220 15.4 4.071 4.242 9.994 10.173 qs_ot_get_derivative 99 11.5 0.001 0.001 9.637 9.679 mp_waitall_1 84994 16.7 7.690 8.390 7.690 8.390 dbcsr_complete_redistribute 325 12.2 1.018 1.047 5.486 7.469 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.816 6.864 calculate_rho_elec 110 8.6 0.227 0.227 6.816 6.863 sum_up_and_integrate 110 10.3 0.002 0.002 6.547 6.562 integrate_v_rspace 110 11.3 0.004 0.004 6.519 6.533 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.457 6.435 init_scf_run 11 5.9 0.000 0.001 6.156 6.156 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.155 6.156 dbcsr_mm_accdrv_process 11614 15.7 3.865 4.218 5.774 6.064 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.817 5.771 make_m2s 4110 13.4 0.043 0.044 5.239 5.711 make_images 4110 14.4 0.885 0.938 5.050 5.520 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.136 5.490 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.762 5.483 apply_single 110 13.6 0.000 0.000 4.762 5.483 mp_alltoall_i22 605 13.7 3.440 5.439 3.440 5.439 qs_ot_get_p 110 10.4 0.001 0.001 5.248 5.298 ot_diis_step 99 11.5 0.015 0.015 4.594 4.594 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.457 4.470 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 3.902 3.917 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 3.819 3.819 cp_fm_cholesky_invert 11 10.9 3.756 3.765 3.756 3.765 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.282 3.746 grid_integrate_task_list 110 12.3 3.671 3.738 3.671 3.738 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.665 3.672 calculate_dm_sparse 110 9.5 0.001 0.001 3.390 3.423 hybrid_alltoall_any 4261 16.3 0.265 0.585 2.778 3.390 fft3d_ps 1111 13.6 1.837 1.860 3.316 3.328 density_rs2pw 110 9.6 0.004 0.004 3.289 3.307 make_images_data 4110 15.4 0.044 0.047 2.699 3.296 wfi_extrapolate 11 7.9 0.001 0.001 3.197 3.197 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.163 3.164 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.148 3.149 multiply_cannon_sync_h2d 8220 15.4 3.126 3.145 3.126 3.145 calculate_first_density_matrix 1 7.0 0.000 0.000 2.840 2.849 grid_collocate_task_list 110 9.6 2.662 2.676 2.662 2.676 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.636 2.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.571 2.587 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.550 2.551 cp_fm_diag_elpa_base 48 14.0 2.253 2.324 2.549 2.549 qs_env_update_s_mstruct 11 6.9 0.049 0.075 2.308 2.366 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.192 potential_pw2rs 110 12.3 0.021 0.021 2.102 2.106 mp_alltoall_d11v 2046 13.8 2.027 2.081 2.027 2.081 acc_transpose_blocks 8220 15.4 0.039 0.040 1.891 2.053 jit_kernel_multiply 10 15.3 1.711 1.980 1.711 1.980 qs_create_task_list 11 7.9 0.001 0.003 1.881 1.925 generate_qs_task_list 11 8.9 0.733 0.785 1.880 1.924 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.891 1.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.839 1.845 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.742 1.766 cp_fm_cholesky_decompose 22 10.9 1.733 1.755 1.733 1.755 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.633 1.641 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=78.976000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1225.909091, yerr=47.907970 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 633.806848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67369. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2001976. 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 0.008 0.027 204.625 204.626 qs_mol_dyn_low 1 2.0 0.031 0.222 204.027 204.042 qs_forces 11 3.9 0.004 0.005 203.598 203.600 qs_energies 11 4.9 0.001 0.002 198.006 198.023 scf_env_do_scf 11 5.9 0.001 0.001 181.239 181.256 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.572 162.574 dbcsr_multiply_generic 2507 12.6 0.194 0.215 126.462 127.245 qs_scf_new_mos 117 7.6 0.001 0.001 123.813 124.062 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.813 124.062 velocity_verlet 10 3.0 0.013 0.096 121.270 121.277 ot_scf_mini 117 9.6 0.003 0.003 117.049 117.249 multiply_cannon 2507 13.6 0.238 0.248 102.227 104.523 multiply_cannon_loop 2507 14.6 2.383 2.449 100.037 102.226 ot_mini 117 10.6 0.001 0.001 66.424 66.619 multiply_cannon_multrec 60168 15.6 31.662 33.880 41.850 43.627 qs_ot_get_derivative 117 11.6 0.001 0.001 41.700 41.894 rebuild_ks_matrix 128 8.3 0.001 0.001 33.391 33.568 qs_ks_build_kohn_sham_matrix 128 9.3 0.014 0.017 33.390 33.567 mp_waitall_1 267128 16.5 29.041 32.083 29.041 32.083 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.015 30.179 qs_ot_get_p 128 10.4 0.001 0.001 28.316 28.632 multiply_cannon_sync_h2d 60168 15.6 26.385 28.206 26.385 28.206 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.225 25.134 apply_single 128 13.6 0.001 0.001 24.224 25.134 ot_diis_step 117 11.6 0.008 0.008 24.476 24.478 qs_ot_p2m_diag 83 11.4 0.079 0.093 21.644 21.766 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.820 19.967 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.837 18.838 multiply_cannon_metrocomm3 60168 15.6 0.118 0.122 16.178 18.618 init_scf_loop 11 6.9 0.000 0.000 18.588 18.589 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.770 15.808 cp_fm_diag_elpa_base 83 14.4 15.722 15.766 15.764 15.800 make_m2s 5014 13.6 0.105 0.112 13.707 14.087 prepare_preconditioner 11 7.9 0.000 0.000 13.949 13.986 make_preconditioner 11 8.9 0.000 0.000 13.949 13.986 make_images 5014 14.6 0.400 0.421 13.524 13.916 sum_up_and_integrate 128 10.3 0.002 0.004 13.601 13.615 integrate_v_rspace 128 11.3 0.004 0.004 13.542 13.559 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.219 13.354 init_scf_run 11 5.9 0.000 0.001 12.693 12.693 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.693 12.693 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.536 12.652 calculate_rho_elec 128 8.7 0.045 0.064 12.536 12.651 mp_sum_l 7950 12.9 9.276 10.361 9.276 10.361 dbcsr_mm_accdrv_process 124484 16.2 4.708 4.847 9.759 10.325 wfi_extrapolate 11 7.9 0.001 0.001 8.855 8.855 calculate_dm_sparse 128 9.5 0.001 0.001 8.618 8.769 multiply_cannon_metrocomm1 60168 15.6 0.090 0.094 6.482 8.487 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.308 8.416 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.251 8.376 cp_fm_cholesky_invert 11 10.9 7.901 7.919 7.901 7.919 make_images_data 5014 15.6 0.066 0.072 6.732 7.715 grid_integrate_task_list 128 12.3 7.060 7.365 7.060 7.365 hybrid_alltoall_any 5200 16.5 0.295 2.264 5.900 7.201 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.183 7.194 density_rs2pw 128 9.7 0.005 0.006 6.144 6.739 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.389 6.702 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.269 6.355 mp_alltoall_d11v 2415 14.1 4.380 5.835 4.380 5.835 fft_wrap_pw1pw2_140 523 12.2 0.233 0.244 5.545 5.714 fft3d_ps 1291 13.7 2.094 2.620 5.213 5.480 grid_collocate_task_list 128 9.7 4.845 5.365 4.845 5.365 mp_sum_d 4469 12.1 3.504 4.305 3.504 4.305 potential_pw2rs 128 12.3 0.009 0.010 4.148 4.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.626000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=7.917968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/16/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 2.183246E+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 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 856.924160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4009 67655. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5076776. 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 0.022 0.108 183.768 183.769 qs_mol_dyn_low 1 2.0 0.005 0.017 183.163 183.176 qs_forces 11 3.9 0.004 0.005 182.938 182.939 qs_energies 11 4.9 0.001 0.002 176.096 176.107 scf_env_do_scf 11 5.9 0.001 0.001 158.906 158.922 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.767 132.770 velocity_verlet 10 3.0 0.008 0.036 112.570 112.585 dbcsr_multiply_generic 2507 12.6 0.186 0.192 98.557 99.855 qs_scf_new_mos 117 7.6 0.001 0.001 94.615 95.193 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.615 95.192 ot_scf_mini 117 9.6 0.004 0.004 89.783 90.426 multiply_cannon 2507 13.6 0.504 0.565 78.349 82.983 multiply_cannon_loop 2507 14.6 1.562 1.626 75.044 77.978 ot_mini 117 10.6 0.001 0.001 50.588 51.142 mp_waitall_1 214728 16.6 24.417 38.615 24.417 38.615 multiply_cannon_multrec 30084 15.6 20.895 26.411 31.811 37.611 rebuild_ks_matrix 128 8.3 0.001 0.001 32.313 33.125 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 32.312 33.125 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 15.915 29.877 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.084 29.820 qs_ot_get_derivative 117 11.6 0.001 0.002 28.597 29.241 init_scf_loop 11 6.9 0.000 0.000 26.045 26.047 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.093 23.248 apply_single 128 13.6 0.001 0.001 22.093 23.248 qs_ot_get_p 128 10.4 0.001 0.001 21.223 21.951 ot_diis_step 117 11.6 0.014 0.015 21.800 21.802 prepare_preconditioner 11 7.9 0.000 0.000 21.680 21.746 make_preconditioner 11 8.9 0.000 0.000 21.680 21.746 multiply_cannon_sync_h2d 30084 15.6 17.919 21.000 17.919 21.000 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.561 20.960 qs_ot_p2m_diag 83 11.4 0.188 0.217 16.277 16.310 make_m2s 5014 13.6 0.087 0.093 14.161 15.789 make_images 5014 14.6 1.170 1.375 13.946 15.576 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.077 15.078 sum_up_and_integrate 128 10.3 0.002 0.004 13.649 13.683 integrate_v_rspace 128 11.3 0.003 0.004 13.589 13.627 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.929 12.962 calculate_rho_elec 128 8.7 0.088 0.106 12.929 12.961 init_scf_run 11 5.9 0.000 0.001 12.438 12.439 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.438 12.439 cp_fm_cholesky_invert 11 10.9 11.901 11.935 11.901 11.935 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.318 11.817 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.577 11.613 cp_fm_diag_elpa_base 83 14.4 11.401 11.445 11.567 11.599 dbcsr_mm_accdrv_process 62242 16.2 5.589 6.561 10.360 11.043 multiply_cannon_metrocomm4 27577 15.6 0.107 0.120 3.888 10.738 mp_irecv_dv 69486 16.3 3.676 10.331 3.676 10.331 make_images_data 5014 15.6 0.065 0.076 8.388 10.247 hybrid_alltoall_any 5200 16.5 0.351 1.533 7.352 10.178 wfi_extrapolate 11 7.9 0.001 0.001 8.211 8.211 grid_integrate_task_list 128 12.3 7.182 7.481 7.182 7.481 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.299 7.372 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.441 7.226 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.951 6.960 density_rs2pw 128 9.7 0.006 0.006 6.478 6.887 calculate_dm_sparse 128 9.5 0.001 0.001 6.747 6.881 fft_wrap_pw1pw2_140 523 12.2 0.248 0.258 6.421 6.475 mp_sum_l 7950 12.9 4.255 6.341 4.255 6.341 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.957 6.173 fft3d_ps 1291 13.7 2.766 2.957 5.751 5.802 cp_fm_cholesky_decompose 22 10.9 5.579 5.656 5.579 5.656 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.374 5.463 mp_alltoall_d11v 2415 14.1 4.868 5.398 4.868 5.398 grid_collocate_task_list 128 9.7 4.998 5.310 4.998 5.310 mp_allgather_i34 2507 14.6 1.828 4.978 1.828 4.978 potential_pw2rs 128 12.3 0.015 0.017 4.161 4.177 mp_sum_d 4474 12.1 2.737 4.102 2.737 4.102 calculate_first_density_matrix 1 7.0 0.000 0.000 4.040 4.046 dbcsr_complete_redistribute 395 12.7 0.738 0.892 3.249 3.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.643 3.704 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=183.769000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=816.000000, yerr=4.572646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.031262E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 67631. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.092 168.951 168.953 qs_mol_dyn_low 1 2.0 0.003 0.003 168.450 168.465 qs_forces 11 3.9 0.004 0.005 168.335 168.336 qs_energies 11 4.9 0.001 0.002 161.827 161.834 scf_env_do_scf 11 5.9 0.002 0.013 145.399 145.401 scf_env_do_scf_inner_loop 116 6.6 0.008 0.013 116.617 116.618 velocity_verlet 10 3.0 0.021 0.036 105.271 105.274 dbcsr_multiply_generic 2485 12.5 0.182 0.187 83.261 84.295 qs_scf_new_mos 116 7.6 0.001 0.001 80.393 80.704 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.392 80.703 ot_scf_mini 116 9.6 0.003 0.004 76.234 76.581 multiply_cannon 2485 13.5 0.505 0.526 63.574 68.357 multiply_cannon_loop 2485 14.5 1.120 1.177 60.651 63.296 ot_mini 116 10.6 0.001 0.001 43.026 43.412 mp_waitall_1 169034 16.6 25.326 35.012 25.326 35.012 rebuild_ks_matrix 127 8.3 0.001 0.001 30.131 30.632 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.016 30.130 30.632 init_scf_loop 11 6.9 0.000 0.000 28.672 28.674 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.159 27.616 multiply_cannon_multrec 19880 15.5 12.830 15.413 22.812 25.342 multiply_cannon_metrocomm3 19880 15.5 0.059 0.063 15.804 24.958 prepare_preconditioner 11 7.9 0.000 0.000 24.584 24.626 make_preconditioner 11 8.9 0.000 0.000 24.584 24.626 qs_ot_get_derivative 116 11.6 0.001 0.002 23.293 23.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.666 23.636 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.828 21.022 apply_single 127 13.6 0.001 0.001 19.828 21.022 ot_diis_step 116 11.6 0.018 0.018 19.624 19.625 qs_ot_get_p 127 10.4 0.001 0.001 18.636 19.078 make_m2s 4970 13.5 0.078 0.084 14.753 15.929 make_images 4970 14.5 1.137 1.224 14.520 15.690 multiply_cannon_sync_h2d 19880 15.5 13.439 14.914 13.439 14.914 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.422 14.431 sum_up_and_integrate 127 10.3 0.002 0.003 13.515 13.542 integrate_v_rspace 127 11.3 0.003 0.004 13.456 13.483 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.370 13.371 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.778 12.808 calculate_rho_elec 127 8.7 0.131 0.145 12.777 12.808 cp_fm_cholesky_invert 11 10.9 11.801 11.822 11.801 11.822 init_scf_run 11 5.9 0.000 0.001 11.405 11.406 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.405 11.405 make_images_data 4970 15.5 0.061 0.070 8.949 10.448 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.074 10.102 cp_fm_diag_elpa_base 82 14.4 9.832 9.887 10.071 10.099 hybrid_alltoall_any 5155 16.4 0.447 2.030 7.693 9.638 dbcsr_mm_accdrv_process 41158 16.2 6.109 6.656 9.445 9.620 multiply_cannon_metrocomm4 17395 15.5 0.066 0.075 3.535 9.470 mp_irecv_dv 49801 16.2 3.404 9.208 3.404 9.208 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.888 9.143 grid_integrate_task_list 127 12.3 7.243 7.755 7.243 7.755 fft_wrap_pw1pw2 1281 11.7 0.018 0.024 7.430 7.532 wfi_extrapolate 11 7.9 0.001 0.001 7.382 7.382 density_rs2pw 127 9.7 0.005 0.006 6.306 6.679 fft_wrap_pw1pw2_140 519 12.2 0.280 0.287 6.539 6.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.299 6.305 calculate_dm_sparse 127 9.5 0.001 0.001 5.999 6.095 cp_fm_cholesky_decompose 22 10.9 5.930 5.982 5.930 5.982 fft3d_ps 1281 13.7 2.742 2.962 5.752 5.825 dbcsr_complete_redistribute 393 12.7 1.032 1.099 4.345 5.512 grid_collocate_task_list 127 9.7 5.114 5.492 5.114 5.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.373 5.484 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.808 5.461 mp_alltoall_d11v 2401 14.1 4.504 4.921 4.504 4.921 cp_fm_upper_to_full 104 14.8 3.889 4.874 3.889 4.874 mp_sum_l 7884 12.9 3.196 4.854 3.196 4.854 mp_allgather_i34 2485 14.5 1.438 4.792 1.438 4.792 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.143 4.284 potential_pw2rs 127 12.3 0.019 0.021 4.047 4.055 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.926 3.956 calculate_first_density_matrix 1 7.0 0.004 0.032 3.836 3.838 qs_energies_init_hamiltonians 11 5.9 0.000 0.005 3.749 3.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.479 3.524 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=168.953000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=980.727273, yerr=4.329411 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.096679E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 67627. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496376. 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 0.075 0.121 178.046 178.047 qs_mol_dyn_low 1 2.0 0.015 0.098 177.276 177.373 qs_forces 11 3.9 0.004 0.004 177.033 177.039 qs_energies 11 4.9 0.003 0.015 169.658 169.669 scf_env_do_scf 11 5.9 0.001 0.001 151.463 151.480 scf_env_do_scf_inner_loop 116 6.6 0.017 0.119 114.386 114.387 velocity_verlet 10 3.0 0.008 0.028 113.945 113.953 dbcsr_multiply_generic 2485 12.5 0.188 0.194 81.482 82.277 qs_scf_new_mos 116 7.6 0.001 0.001 78.805 79.164 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.804 79.163 ot_scf_mini 116 9.6 0.003 0.004 74.288 74.655 multiply_cannon 2485 13.5 0.548 0.586 56.700 60.031 multiply_cannon_loop 2485 14.5 1.780 1.918 53.245 55.199 ot_mini 116 10.6 0.001 0.001 42.758 43.097 init_scf_loop 11 6.9 0.000 0.003 36.952 36.953 prepare_preconditioner 11 7.9 0.000 0.000 32.787 32.819 make_preconditioner 11 8.9 0.000 0.001 32.787 32.819 multiply_cannon_multrec 29820 15.5 13.410 19.351 26.716 32.223 make_full_inverse_cholesky 11 9.9 0.017 0.023 28.026 31.566 rebuild_ks_matrix 127 8.3 0.001 0.001 29.516 29.857 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.515 29.856 mp_waitall_1 146592 16.7 17.288 26.796 17.288 26.796 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.392 26.697 qs_ot_get_derivative 116 11.6 0.001 0.002 23.060 23.409 make_m2s 4970 13.5 0.093 0.098 20.444 21.350 make_images 4970 14.5 1.989 2.219 20.137 21.040 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.198 19.766 apply_single 127 13.6 0.001 0.001 19.198 19.766 ot_diis_step 116 11.6 0.017 0.018 19.586 19.588 qs_ot_get_p 127 10.4 0.001 0.001 17.713 18.114 multiply_cannon_metrocomm3 29820 15.5 0.049 0.051 6.490 15.175 sum_up_and_integrate 127 10.3 0.011 0.075 13.873 13.939 integrate_v_rspace 127 11.3 0.003 0.004 13.803 13.880 qs_ot_p2m_diag 82 11.4 0.339 0.386 13.662 13.715 dbcsr_mm_accdrv_process 61748 16.2 8.429 9.376 12.879 13.465 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.240 13.316 calculate_rho_elec 127 8.7 0.173 0.188 13.240 13.315 cp_fm_upper_to_full 104 14.8 8.720 12.398 8.720 12.398 make_images_data 4970 15.5 0.064 0.070 11.003 12.334 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.281 12.282 cp_fm_cholesky_invert 11 10.9 11.964 11.986 11.964 11.986 init_scf_run 11 5.9 0.000 0.001 11.809 11.810 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.809 11.810 hybrid_alltoall_any 5155 16.4 0.526 2.189 9.879 11.693 multiply_cannon_sync_h2d 29820 15.5 10.452 11.229 10.452 11.229 dbcsr_complete_redistribute 393 12.7 1.427 1.562 7.569 10.370 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.533 9.787 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.135 9.151 cp_fm_diag_elpa_base 82 14.4 8.663 8.802 9.129 9.145 copy_fm_to_dbcsr 208 11.6 0.004 0.032 6.130 8.891 fft_wrap_pw1pw2 1281 11.7 0.018 0.024 7.991 8.091 grid_integrate_task_list 127 12.3 7.493 7.743 7.493 7.743 multiply_cannon_metrocomm4 24850 15.5 0.078 0.090 2.842 7.539 wfi_extrapolate 11 7.9 0.001 0.001 7.472 7.472 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 4.739 7.456 mp_irecv_dv 75445 16.2 2.687 7.258 2.687 7.258 fft_wrap_pw1pw2_140 519 12.2 0.330 0.351 7.120 7.252 mp_alltoall_i22 712 14.1 4.076 6.850 4.076 6.850 calculate_dm_sparse 127 9.5 0.001 0.001 6.563 6.638 density_rs2pw 127 9.7 0.005 0.005 6.234 6.587 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.304 6.332 fft3d_ps 1281 13.7 2.922 2.974 6.183 6.257 cp_fm_cholesky_decompose 22 10.9 5.996 6.041 5.996 6.041 mp_alltoall_d11v 2401 14.1 4.956 5.574 4.956 5.574 grid_collocate_task_list 127 9.7 5.267 5.542 5.267 5.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.211 5.315 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.502 4.622 qs_energies_init_hamiltonians 11 5.9 0.002 0.021 4.508 4.509 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.197 4.271 potential_pw2rs 127 12.3 0.022 0.022 4.208 4.260 calculate_first_density_matrix 1 7.0 0.001 0.008 3.996 4.005 acc_transpose_blocks 29820 15.5 0.138 0.142 3.769 3.906 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.765 3.796 dbcsr_special_finalize 12425 15.5 0.065 0.069 3.521 3.685 dbcsr_merge_single_wm 12425 16.5 0.322 0.344 3.431 3.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=178.047000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1039.363636, yerr=12.586231 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 1.433674E+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 751306. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4031 67282. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848449. 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 0.069 0.111 165.321 165.322 qs_mol_dyn_low 1 2.0 0.003 0.003 164.688 164.727 qs_forces 11 3.9 0.011 0.023 164.287 164.290 qs_energies 11 4.9 0.001 0.001 156.890 156.904 scf_env_do_scf 11 5.9 0.001 0.001 138.709 138.721 scf_env_do_scf_inner_loop 118 6.6 0.008 0.016 108.055 108.056 velocity_verlet 10 3.0 0.309 0.310 106.737 106.741 dbcsr_multiply_generic 2527 12.6 0.185 0.192 73.497 73.773 qs_scf_new_mos 118 7.6 0.001 0.001 71.584 71.655 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.583 71.654 ot_scf_mini 118 9.6 0.003 0.004 67.149 67.188 multiply_cannon 2527 13.6 0.567 0.607 54.854 58.353 multiply_cannon_loop 2527 14.6 0.813 0.835 51.958 52.560 ot_mini 118 10.6 0.001 0.001 37.693 37.737 mp_waitall_1 126780 16.7 24.750 30.774 24.750 30.774 init_scf_loop 11 6.9 0.000 0.000 30.499 30.501 rebuild_ks_matrix 129 8.3 0.001 0.001 29.236 29.310 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.235 29.309 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.707 26.781 prepare_preconditioner 11 7.9 0.000 0.000 26.588 26.605 make_preconditioner 11 8.9 0.000 0.000 26.588 26.605 make_full_inverse_cholesky 11 9.9 0.022 0.027 24.691 24.985 multiply_cannon_multrec 10108 15.6 10.483 14.875 18.463 21.565 qs_ot_get_derivative 118 11.6 0.002 0.002 20.570 20.608 multiply_cannon_metrocomm3 10108 15.6 0.025 0.026 12.577 19.826 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.110 17.219 apply_single 129 13.6 0.001 0.001 17.109 17.219 ot_diis_step 118 11.6 0.020 0.020 17.050 17.051 qs_ot_get_p 129 10.4 0.001 0.001 16.728 16.777 make_m2s 5054 13.6 0.065 0.068 14.759 15.563 make_images 5054 14.6 2.187 2.604 14.449 15.246 cp_fm_cholesky_invert 11 10.9 14.574 14.585 14.574 14.585 sum_up_and_integrate 129 10.3 0.002 0.002 13.835 13.887 integrate_v_rspace 129 11.3 0.004 0.004 13.774 13.825 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.503 13.528 calculate_rho_elec 129 8.7 0.259 0.270 13.503 13.527 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.935 12.951 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.771 11.771 init_scf_run 11 5.9 0.000 0.001 11.372 11.372 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.372 11.372 multiply_cannon_sync_h2d 10108 15.6 10.972 11.224 10.972 11.224 hybrid_alltoall_any 5240 16.5 0.853 3.822 8.336 9.764 make_images_data 5054 15.6 0.056 0.066 8.488 9.754 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.872 8.885 cp_fm_diag_elpa_base 83 14.4 8.635 8.709 8.868 8.881 dbcsr_mm_accdrv_process 20926 16.1 3.300 4.284 7.626 8.438 fft_wrap_pw1pw2 1301 11.7 0.017 0.019 8.095 8.119 grid_integrate_task_list 129 12.3 7.818 8.069 7.818 8.069 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.977 8.005 multiply_cannon_metrocomm1 10108 15.6 0.030 0.030 4.477 7.360 wfi_extrapolate 11 7.9 0.001 0.001 7.203 7.203 fft_wrap_pw1pw2_140 527 12.2 0.436 0.442 7.146 7.173 cp_fm_cholesky_decompose 22 10.9 6.486 6.534 6.486 6.534 calculate_dm_sparse 129 9.5 0.001 0.001 6.433 6.513 density_rs2pw 129 9.7 0.005 0.005 6.153 6.310 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.102 6.141 fft3d_ps 1301 13.7 3.128 3.246 6.047 6.087 grid_collocate_task_list 129 9.7 5.619 5.821 5.619 5.821 dbcsr_complete_redistribute 395 12.7 1.998 2.057 5.086 5.406 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 5.279 5.287 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.953 4.991 mp_alltoall_d11v 2423 14.1 4.605 4.869 4.605 4.869 mp_allgather_i34 2527 14.6 1.082 4.535 1.082 4.535 potential_pw2rs 129 12.3 0.026 0.027 3.932 3.941 calculate_first_density_matrix 1 7.0 0.000 0.000 3.933 3.935 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.909 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.701 3.769 multiply_cannon_metrocomm4 7581 15.6 0.025 0.027 1.731 3.746 mp_irecv_dv 29086 15.9 1.692 3.668 1.692 3.668 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.576 3.596 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.457 3.559 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.238 3.544 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.441 3.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=165.322000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1349.636364, yerr=24.955498 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.962784E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 68558. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.071 244.253 244.253 qs_mol_dyn_low 1 2.0 0.003 0.003 243.160 243.199 qs_forces 11 3.9 0.005 0.005 242.746 242.749 qs_energies 11 4.9 0.001 0.001 233.951 233.959 scf_env_do_scf 11 5.9 0.001 0.001 210.789 210.799 velocity_verlet 10 3.0 0.047 0.073 168.598 168.638 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 128.159 128.161 qs_scf_new_mos 116 7.6 0.001 0.001 86.070 86.103 qs_scf_loop_do_ot 116 8.6 0.001 0.001 86.069 86.103 dbcsr_multiply_generic 2485 12.5 0.214 0.224 82.225 82.621 init_scf_loop 11 6.9 0.000 0.000 82.314 82.318 ot_scf_mini 116 9.6 0.003 0.004 81.167 81.176 prepare_preconditioner 11 7.9 0.000 0.000 77.613 77.622 make_preconditioner 11 8.9 0.000 0.000 77.613 77.622 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.866 74.800 multiply_cannon 2485 13.5 0.672 0.703 59.403 60.874 multiply_cannon_loop 2485 14.5 1.038 1.062 55.448 56.649 cp_fm_upper_to_full 104 14.8 34.695 48.739 34.695 48.739 ot_mini 116 10.6 0.001 0.001 43.402 43.406 rebuild_ks_matrix 127 8.3 0.001 0.001 33.259 33.262 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.258 33.261 mp_waitall_1 102768 16.8 27.649 32.194 27.649 32.194 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.901 30.911 dbcsr_complete_redistribute 393 12.7 3.833 3.889 21.675 30.314 copy_fm_to_dbcsr 208 11.6 0.001 0.001 18.146 26.693 qs_ot_get_derivative 116 11.6 0.002 0.002 24.256 24.266 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 15.705 24.135 qs_ot_get_p 127 10.4 0.001 0.001 22.778 22.812 mp_alltoall_i22 712 14.1 13.432 22.128 13.432 22.128 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.684 20.066 make_m2s 4970 13.5 0.074 0.075 18.343 19.814 make_images 4970 14.5 3.094 3.287 17.867 19.336 ot_diis_step 116 11.6 0.022 0.023 19.122 19.122 qs_ot_p2m_diag 82 11.4 0.869 0.874 18.809 18.838 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.517 18.624 apply_single 127 13.6 0.001 0.001 18.517 18.624 multiply_cannon_multrec 9940 15.5 10.215 12.056 18.282 18.426 cp_fm_cholesky_invert 11 10.9 18.103 18.115 18.103 18.115 cp_dbcsr_syevd 82 12.4 0.006 0.006 17.045 17.046 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.608 15.622 calculate_rho_elec 127 8.7 0.479 0.480 15.608 15.621 sum_up_and_integrate 127 10.3 0.002 0.002 15.474 15.565 integrate_v_rspace 127 11.3 0.005 0.005 15.411 15.502 multiply_cannon_sync_h2d 9940 15.5 14.192 14.259 14.192 14.259 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.948 13.949 cp_fm_diag_elpa_base 82 14.4 11.128 12.139 13.943 13.944 hybrid_alltoall_any 5155 16.4 1.308 3.068 10.587 12.867 init_scf_run 11 5.9 0.000 0.001 12.759 12.759 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.759 12.759 make_images_data 4970 15.5 0.062 0.067 10.302 12.425 fft_wrap_pw1pw2 1281 11.7 0.020 0.020 9.844 9.870 dbcsr_mm_accdrv_process 20590 16.0 4.341 6.168 7.818 9.644 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.530 9.537 fft_wrap_pw1pw2_140 519 12.2 0.817 0.820 8.732 8.789 wfi_extrapolate 11 7.9 0.001 0.001 8.776 8.776 grid_integrate_task_list 127 12.3 8.494 8.681 8.494 8.681 qs_energies_init_hamiltonians 11 5.9 0.016 0.024 8.259 8.260 calculate_dm_sparse 127 9.5 0.001 0.001 7.408 7.448 cp_fm_cholesky_decompose 22 10.9 7.340 7.376 7.340 7.376 fft3d_ps 1281 13.7 3.968 4.047 7.149 7.179 mp_alltoall_d11v 2401 14.1 7.015 7.137 7.015 7.137 density_rs2pw 127 9.7 0.005 0.005 6.729 6.774 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.631 6.667 grid_collocate_task_list 127 9.7 6.355 6.430 6.355 6.430 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.164 6.268 qs_env_update_s_mstruct 11 6.9 0.034 0.061 5.066 5.357 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.842 5.285 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=244.253000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2662.454545, yerr=145.474374 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 1.270256E+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 1033 290947. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.076 0.148 85.580 85.581 qs_energies 1 2.0 0.000 0.000 84.731 84.738 ls_scf 1 3.0 0.000 0.000 83.668 83.676 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.662 72.855 multiply_cannon 111 7.7 0.018 0.021 55.914 56.975 multiply_cannon_loop 111 8.7 0.228 0.244 52.512 53.764 ls_scf_main 1 4.0 0.000 0.000 52.373 52.373 density_matrix_trs4 2 5.0 0.002 0.003 46.859 46.935 ls_scf_init_scf 1 4.0 0.000 0.000 28.245 28.246 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.094 27.155 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.014 25.037 mp_waitall_1 11031 10.9 22.490 24.840 22.490 24.840 multiply_cannon_multrec 2664 9.7 8.132 8.928 15.624 17.413 multiply_cannon_sync_h2d 2664 9.7 13.485 15.129 13.485 15.129 make_m2s 222 7.7 0.009 0.011 13.120 13.617 make_images 222 8.7 0.100 0.110 13.098 13.598 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.699 13.560 make_images_data 222 9.7 0.004 0.006 7.657 8.310 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.477 8.167 hybrid_alltoall_any 227 10.6 0.216 1.835 6.606 8.162 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.735 7.111 8.114 dbcsr_mm_accdrv_process_sort 4760 11.4 6.313 7.225 6.313 7.225 calculate_norms 4752 9.8 5.515 6.141 5.515 6.141 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.150 mp_sum_l 887 5.1 3.181 4.540 3.181 4.540 make_images_sizes 222 9.7 0.000 0.000 0.761 3.565 mp_alltoall_i44 222 10.7 0.761 3.565 0.761 3.565 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.057 3.454 mp_irecv_dv 6231 10.9 2.041 3.427 2.041 3.427 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.312 3.425 arnoldi_extremal 4 6.8 0.000 0.000 3.219 3.242 arnoldi_normal_ev 4 7.8 0.001 0.003 3.219 3.242 build_subspace 16 8.4 0.009 0.012 3.130 3.134 ls_scf_post 1 4.0 0.000 0.000 3.051 3.058 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.341 2.770 dbcsr_merge_single_wm 555 10.7 0.455 0.589 2.333 2.761 make_images_pack 222 9.7 2.212 2.633 2.214 2.634 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.319 2.565 dbcsr_sort_data 658 11.4 2.135 2.500 2.135 2.500 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.461 2.069 2.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.278 2.356 buffer_matrices_ensure_size 222 8.7 1.750 2.070 1.750 2.070 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.812 1.813 rebuild_ks_matrix 3 7.3 0.000 0.000 1.802 1.803 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.802 1.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.581000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 2.175910E+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 1033 301899. 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 0.071 0.116 92.589 92.599 qs_energies 1 2.0 0.000 0.005 90.939 90.956 ls_scf 1 3.0 0.009 0.069 89.575 89.591 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.289 75.657 multiply_cannon 111 7.7 0.028 0.043 53.146 57.910 ls_scf_main 1 4.0 0.001 0.021 55.241 55.246 multiply_cannon_loop 111 8.7 0.135 0.146 50.270 54.031 density_matrix_trs4 2 5.0 0.003 0.016 49.606 49.802 mp_waitall_1 9105 10.9 20.989 30.707 20.989 30.707 ls_scf_init_scf 1 4.0 0.000 0.002 30.480 30.518 ls_scf_init_matrix_S 1 5.0 0.010 0.085 29.084 29.164 multiply_cannon_multrec 1332 9.7 13.339 16.545 22.718 27.007 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.574 26.594 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.664 20.984 make_m2s 222 7.7 0.006 0.007 15.217 15.829 make_images 222 8.7 1.374 1.696 15.186 15.799 dbcsr_mm_accdrv_process 4041 10.4 0.335 0.530 8.974 10.561 dbcsr_mm_accdrv_process_sort 4041 11.4 8.490 10.033 8.490 10.033 make_images_data 222 9.7 0.004 0.005 8.808 9.778 hybrid_alltoall_any 227 10.6 0.543 2.550 8.140 9.512 mp_sum_l 887 5.1 5.454 9.057 5.454 9.057 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.175 7.708 mp_irecv_dv 3311 11.0 3.154 7.649 3.154 7.649 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.115 7.188 calculate_norms 2376 9.8 6.067 6.669 6.067 6.669 multiply_cannon_sync_h2d 1332 9.7 4.920 6.500 4.920 6.500 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.960 5.184 arnoldi_extremal 4 6.8 0.000 0.000 4.698 4.724 arnoldi_normal_ev 4 7.8 0.001 0.005 4.698 4.724 build_subspace 16 8.4 0.014 0.020 4.386 4.389 ls_scf_post 1 4.0 0.002 0.019 3.845 3.857 ls_scf_store_result 1 5.0 0.000 0.000 3.512 3.650 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.139 3.368 dbcsr_matrix_vector_mult_local 304 10.0 2.752 3.227 2.754 3.228 mp_allgather_i34 111 8.7 0.852 3.142 0.852 3.142 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.576 2.654 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.164 2.586 dbcsr_data_new 4174 10.1 2.140 2.421 2.140 2.421 make_images_pack 222 9.7 1.835 2.141 1.838 2.143 dbcsr_sort_data 436 11.2 1.833 2.033 1.833 2.033 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.977 2.014 rebuild_ks_matrix 3 7.3 0.000 0.000 1.965 2.001 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.026 1.965 2.001 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.599000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1791.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 2.898592E+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 1033 300682. 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 0.088 0.482 94.771 94.772 qs_energies 1 2.0 0.000 0.001 93.630 93.634 ls_scf 1 3.0 0.001 0.008 92.025 92.028 dbcsr_multiply_generic 111 6.7 0.016 0.018 76.146 76.524 ls_scf_main 1 4.0 0.034 0.294 57.108 57.124 multiply_cannon 111 7.7 0.066 0.119 52.342 56.012 multiply_cannon_loop 111 8.7 0.117 0.130 49.542 53.807 density_matrix_trs4 2 5.0 0.018 0.174 51.058 51.228 mp_waitall_1 7281 11.0 23.539 33.422 23.539 33.422 ls_scf_init_scf 1 4.0 0.002 0.013 31.301 31.340 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.593 29.658 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.029 27.040 multiply_cannon_multrec 888 9.7 12.720 15.367 21.432 24.842 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.944 22.948 make_m2s 222 7.7 0.006 0.007 16.476 17.204 make_images 222 8.7 1.585 1.860 16.438 17.164 make_images_data 222 9.7 0.004 0.004 9.500 10.673 hybrid_alltoall_any 227 10.6 0.644 2.930 9.057 10.393 mp_sum_l 887 5.1 5.668 9.681 5.668 9.681 dbcsr_mm_accdrv_process 3754 10.4 0.353 0.546 8.221 9.509 dbcsr_mm_accdrv_process_sort 3754 11.4 7.763 8.963 7.763 8.963 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.133 7.519 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.483 7.250 mp_irecv_dv 2335 11.1 2.467 7.191 2.467 7.191 multiply_cannon_sync_h2d 888 9.7 6.029 7.122 6.029 7.122 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.740 7.010 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.960 5.246 arnoldi_extremal 4 6.8 0.000 0.000 5.066 5.082 arnoldi_normal_ev 4 7.8 0.001 0.006 5.065 5.082 build_subspace 16 8.4 0.014 0.020 4.763 4.768 calculate_norms 1584 9.8 4.388 4.736 4.388 4.736 mp_allgather_i34 111 8.7 0.880 3.813 0.880 3.813 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.432 3.779 dbcsr_matrix_vector_mult_local 304 10.0 3.035 3.623 3.037 3.625 ls_scf_post 1 4.0 0.000 0.000 3.615 3.621 ls_scf_store_result 1 5.0 0.000 0.000 3.390 3.438 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.878 2.991 dbcsr_data_new 4116 9.9 2.115 2.479 2.115 2.479 make_images_sizes 222 9.7 0.000 0.000 1.151 2.342 mp_alltoall_i44 222 10.7 1.151 2.341 1.151 2.341 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.167 2.206 rebuild_ks_matrix 3 7.3 0.000 0.000 2.149 2.188 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.053 2.149 2.188 dbcsr_sort_data 325 11.1 1.847 2.098 1.847 2.098 make_images_pack 222 9.7 1.644 1.893 1.647 1.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.772000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2311.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 3.426316E+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 1033 301899. 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 0.147 0.216 97.911 97.937 qs_energies 1 2.0 0.000 0.001 97.146 97.152 ls_scf 1 3.0 0.000 0.000 95.109 95.114 dbcsr_multiply_generic 111 6.7 0.017 0.020 78.626 78.854 ls_scf_main 1 4.0 0.000 0.002 58.568 58.569 multiply_cannon 111 7.7 0.048 0.107 51.866 55.648 density_matrix_trs4 2 5.0 0.003 0.009 52.474 52.608 multiply_cannon_loop 111 8.7 0.153 0.170 46.825 49.427 ls_scf_init_scf 1 4.0 0.000 0.001 33.310 33.312 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.859 31.912 mp_waitall_1 6369 11.0 22.989 30.650 22.989 30.650 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.020 29.263 29.280 multiply_cannon_multrec 1332 9.7 13.978 17.023 22.024 24.907 make_m2s 222 7.7 0.007 0.008 21.297 22.687 make_images 222 8.7 3.155 3.621 21.247 22.639 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.402 16.789 make_images_data 222 9.7 0.004 0.004 11.887 13.594 hybrid_alltoall_any 227 10.6 0.797 3.785 11.254 12.954 dbcsr_mm_accdrv_process 3641 10.4 0.298 0.567 7.684 9.234 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.731 7.223 8.731 mp_sum_l 887 5.1 3.910 7.030 3.910 7.030 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.103 6.100 mp_irecv_dv 3229 10.9 2.078 6.030 2.078 6.030 multiply_cannon_sync_h2d 1332 9.7 5.450 5.910 5.450 5.910 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.999 5.427 arnoldi_extremal 4 6.8 0.000 0.000 5.249 5.268 arnoldi_normal_ev 4 7.8 0.004 0.030 5.249 5.268 build_subspace 16 8.4 0.014 0.021 4.867 4.878 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.658 4.827 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.532 4.711 calculate_norms 2376 9.8 4.170 4.530 4.170 4.530 mp_allgather_i34 111 8.7 2.152 4.471 2.152 4.471 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.590 3.887 dbcsr_matrix_vector_mult_local 304 10.0 3.207 3.710 3.210 3.712 dbcsr_sort_data 658 11.4 3.060 3.350 3.060 3.350 dbcsr_special_finalize 555 9.7 0.006 0.007 2.819 3.238 ls_scf_post 1 4.0 0.000 0.000 3.231 3.237 dbcsr_merge_single_wm 555 10.7 0.538 0.665 2.810 3.230 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.987 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.978 3.037 dbcsr_data_release 10477 10.7 1.564 2.417 1.564 2.417 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.188 2.189 rebuild_ks_matrix 3 7.3 0.000 0.000 2.165 2.166 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.165 2.166 qs_energies_init_hamiltonians 1 3.0 0.003 0.011 2.027 2.027 make_images_pack 222 9.7 1.657 2.013 1.659 2.016 dbcsr_finalize 304 7.8 0.049 0.061 1.800 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.937000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2769.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 4.840374E+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 1033 300769. 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 0.036 0.050 92.117 92.117 qs_energies 1 2.0 0.000 0.000 91.328 91.333 ls_scf 1 3.0 0.000 0.000 89.380 89.385 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.937 71.064 multiply_cannon 111 7.7 0.076 0.143 52.594 56.151 ls_scf_main 1 4.0 0.000 0.000 56.057 56.057 multiply_cannon_loop 111 8.7 0.088 0.094 49.975 51.531 density_matrix_trs4 2 5.0 0.002 0.003 49.234 49.320 ls_scf_init_scf 1 4.0 0.000 0.000 29.863 29.867 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.624 28.655 mp_waitall_1 5436 11.0 24.011 28.612 24.011 28.612 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.484 26.499 multiply_cannon_multrec 444 9.7 13.767 16.325 20.982 22.959 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.473 14.995 make_m2s 222 7.7 0.005 0.005 13.633 14.577 make_images 222 8.7 2.043 2.482 13.566 14.508 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.008 14.261 hybrid_alltoall_any 227 10.6 0.801 3.844 8.112 9.742 make_images_data 222 9.7 0.003 0.004 8.344 9.691 multiply_cannon_sync_h2d 444 9.7 6.708 8.390 6.708 8.390 dbcsr_mm_accdrv_process 3003 10.4 0.389 0.493 6.909 8.151 dbcsr_mm_accdrv_process_sort 3003 11.4 6.504 7.658 6.504 7.658 arnoldi_extremal 4 6.8 0.000 0.000 5.879 5.895 arnoldi_normal_ev 4 7.8 0.023 0.033 5.879 5.895 build_subspace 16 8.4 0.015 0.020 5.454 5.466 mp_sum_l 887 5.1 2.708 4.678 2.708 4.678 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.415 4.564 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.194 4.406 dbcsr_matrix_vector_mult_local 304 10.0 3.736 4.209 3.738 4.211 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.509 3.787 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.902 3.759 mp_irecv_dv 1241 11.2 1.495 3.759 1.495 3.759 calculate_norms 792 9.8 3.605 3.745 3.605 3.745 mp_allgather_i34 111 8.7 1.151 3.661 1.151 3.661 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.427 3.478 ls_scf_post 1 4.0 0.000 0.000 3.461 3.466 make_images_sizes 222 9.7 0.000 0.000 0.896 3.352 mp_alltoall_i44 222 10.7 0.895 3.351 0.895 3.351 ls_scf_store_result 1 5.0 0.000 0.000 3.197 3.232 dbcsr_data_new 4608 9.7 1.817 2.310 1.817 2.310 dbcsr_finalize 304 7.8 0.062 0.076 2.196 2.265 dbcsr_merge_all 275 8.9 0.479 0.527 2.053 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.056 2.057 rebuild_ks_matrix 3 7.3 0.000 0.000 2.023 2.024 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.023 2.024 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.933 1.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.117000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3777.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 9.062429E+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 1033 319197. 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 0.118 0.134 107.440 107.440 qs_energies 1 2.0 0.000 0.000 105.877 105.880 ls_scf 1 3.0 0.000 0.000 102.912 102.916 dbcsr_multiply_generic 111 6.7 0.024 0.028 76.519 76.678 ls_scf_main 1 4.0 0.000 0.000 63.510 63.511 density_matrix_trs4 2 5.0 0.002 0.003 54.745 54.811 multiply_cannon 111 7.7 0.192 0.281 50.340 52.508 multiply_cannon_loop 111 8.7 0.098 0.099 47.269 47.781 ls_scf_init_scf 1 4.0 0.000 0.000 35.502 35.503 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.001 34.017 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.171 31.180 mp_waitall_1 4527 11.1 21.987 25.732 21.987 25.732 make_m2s 222 7.7 0.005 0.005 22.574 23.667 make_images 222 8.7 3.577 3.878 22.467 23.558 multiply_cannon_multrec 444 9.7 17.869 18.553 22.629 23.320 hybrid_alltoall_any 227 10.6 1.652 3.605 12.739 15.646 make_images_data 222 9.7 0.003 0.004 12.932 15.079 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.435 12.445 multiply_cannon_sync_h2d 444 9.7 8.788 8.848 8.788 8.848 arnoldi_extremal 4 6.8 0.000 0.000 7.473 7.484 arnoldi_normal_ev 4 7.8 0.024 0.035 7.473 7.484 build_subspace 16 8.4 0.026 0.036 6.899 6.915 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.534 5.713 dbcsr_matrix_vector_mult_local 304 10.0 5.121 5.459 5.123 5.462 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.874 5.127 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.900 4.990 dbcsr_mm_accdrv_process 1814 10.4 0.245 0.416 4.570 4.710 dbcsr_mm_accdrv_process_sort 1814 11.4 4.190 4.314 4.190 4.314 ls_scf_post 1 4.0 0.000 0.000 3.900 3.902 make_images_sizes 222 9.7 0.000 0.000 1.500 3.662 mp_alltoall_i44 222 10.7 1.500 3.661 1.500 3.661 mp_allgather_i34 111 8.7 1.105 3.594 1.105 3.594 ls_scf_store_result 1 5.0 0.000 0.000 3.528 3.564 calculate_norms 792 9.8 3.236 3.279 3.236 3.279 dbcsr_finalize 304 7.8 0.082 0.090 3.082 3.160 dbcsr_merge_all 275 8.9 0.888 0.915 2.867 2.938 qs_energies_init_hamiltonians 1 3.0 0.000 0.000 2.934 2.934 dbcsr_complete_redistribute 5 7.6 1.423 1.471 2.739 2.848 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.585 2.592 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.503 dbcsr_sort_data 325 11.1 2.439 2.499 2.439 2.499 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.365 2.367 dbcsr_new_transposed 4 7.5 0.268 0.305 2.342 2.356 dbcsr_data_new 6591 9.6 1.826 2.354 1.826 2.354 rebuild_ks_matrix 3 7.3 0.000 0.000 2.300 2.301 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.299 2.301 dbcsr_frobenius_norm 74 6.6 2.054 2.130 2.183 2.215 dbcsr_add_d 103 6.2 0.000 0.000 2.127 2.203 dbcsr_add_anytype 103 7.2 0.859 0.890 2.127 2.203 dbcsr_data_release 12724 10.6 1.985 2.199 1.985 2.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.440000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7185.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1f098bdf757a0d911df37584392ac41aeb8a328_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 584.962048E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993460612. 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 0.014 0.035 218.634 218.724 qs_mol_dyn_low 1 2.0 0.007 0.090 217.715 217.805 qs_forces 5 3.8 0.004 0.004 217.442 217.531 qs_energies 5 4.8 0.001 0.002 214.569 214.653 scf_env_do_scf 5 5.8 0.001 0.037 201.227 201.309 scf_env_do_scf_inner_loop 105 6.6 0.007 0.053 177.045 177.117 qs_scf_new_mos 105 7.6 0.000 0.001 139.028 139.174 qs_scf_loop_do_ot 105 8.6 0.001 0.001 139.027 139.174 dbcsr_multiply_generic 1445 12.2 0.135 0.160 132.375 132.744 ot_scf_mini 105 9.6 0.003 0.004 129.160 129.281 multiply_cannon 1445 13.2 0.274 0.286 113.828 115.645 multiply_cannon_loop 1445 14.2 2.853 3.000 112.252 113.455 velocity_verlet 4 3.0 0.012 0.114 102.379 102.426 ot_mini 105 10.6 0.001 0.001 59.573 59.678 multiply_cannon_multrec 69360 15.2 29.837 34.421 40.371 44.913 qs_ot_get_p 112 10.4 0.001 0.001 39.924 40.187 mp_waitall_1 488190 16.1 33.852 40.108 33.852 40.108 qs_ot_get_derivative 55 11.6 0.001 0.001 37.914 38.031 multiply_cannon_sync_h2d 69360 15.2 29.261 33.660 29.261 33.660 multiply_cannon_metrocomm3 69360 15.2 0.199 0.208 25.230 31.970 qs_ot_p2m_diag 40 11.0 0.020 0.030 29.144 29.234 rebuild_ks_matrix 110 8.4 0.000 0.000 27.872 28.041 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.012 27.872 28.041 cp_dbcsr_syevd 40 12.0 0.002 0.002 25.840 25.851 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.586 25.744 init_scf_loop 7 6.6 0.000 0.000 24.148 24.159 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.935 23.179 apply_single 62 13.6 0.000 0.000 22.934 23.179 ot_new_cg_direction 55 11.6 0.001 0.001 20.946 20.954 cp_fm_syevd 40 13.0 0.000 0.000 20.804 20.948 prepare_preconditioner 7 7.6 0.000 0.000 19.426 19.455 make_preconditioner 7 8.6 0.000 0.000 19.426 19.455 cp_fm_redistribute_end 40 14.0 8.111 16.164 8.125 16.169 cp_fm_syevd_base 40 14.0 8.036 16.089 8.036 16.089 qs_rho_update_rho_low 110 7.6 0.000 0.001 15.716 16.059 calculate_rho_elec 110 8.6 0.029 0.051 15.715 16.058 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.522 14.672 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.583 13.678 mp_sum_l 4764 12.2 11.951 12.664 11.951 12.664 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.369 12.443 calculate_dm_sparse 110 9.5 0.000 0.001 11.980 12.112 dbcsr_mm_accdrv_process 154766 15.8 6.227 6.399 10.401 11.384 init_scf_run 5 5.8 0.000 0.000 11.107 11.112 scf_env_initial_rho_setup 5 6.8 0.001 0.001 11.107 11.112 density_rs2pw 110 9.6 0.004 0.005 10.083 10.532 fft_wrap_pw1pw2 1425 12.5 0.018 0.023 10.133 10.304 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.113 10.207 qs_vxc_create 110 10.4 0.002 0.002 9.609 9.626 check_diag 80 13.5 8.650 8.924 9.295 9.437 fft_wrap_pw1pw2_240 915 14.0 0.199 0.219 8.865 9.040 cp_fm_cholesky_invert 7 10.6 8.294 8.312 8.294 8.312 sum_up_and_integrate 60 10.3 0.001 0.002 8.077 8.087 integrate_v_rspace 60 11.3 0.002 0.002 8.060 8.071 fft3d_pb 915 15.0 2.257 2.429 7.812 7.965 transfer_rs2pw 445 10.6 0.008 0.009 7.187 7.659 acc_transpose_blocks 69360 15.2 0.341 0.357 6.971 7.384 xc_rho_set_and_dset_create 110 12.4 0.081 0.115 6.838 7.082 calculate_first_density_matrix 1 7.0 0.000 0.012 6.927 6.942 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.903 6.922 make_full_single_inverse 7 9.6 0.001 0.001 6.789 6.818 make_m2s 2890 13.2 0.078 0.088 6.088 6.663 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.560 6.607 make_images 2890 14.2 0.239 0.259 5.980 6.555 multiply_cannon_metrocomm1 69360 15.2 0.093 0.100 4.480 6.551 xc_vxc_pw_create 60 11.3 0.036 0.037 6.412 6.432 xc_pw_derive 510 13.4 0.005 0.006 5.491 5.550 mp_alltoall_z22v 2340 16.7 4.951 5.192 4.951 5.192 mp_waitany 7680 13.5 4.335 4.938 4.335 4.938 acc_transpose_blocks_kernels 69360 16.2 0.849 0.895 4.244 4.564 multiply_cannon_metrocomm4 67915 15.2 0.181 0.197 2.007 4.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=218.724000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b1f098bdf757a0d911df37584392ac41aeb8a328 Summary: empty Status: OK