=== This is the CP2K Performance-Test === Updating 613b5c7f4..9dbea54b0 Fast-forward src/start/cp2k_shell.F | 75 +++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 68 insertions(+), 7 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 9dbea54b02bae51086f4cc159fb0e2da2f450cf9 ################# 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/01 job id: 53100668 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/02 job id: 53100669 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/03 job id: 53100670 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/04 job id: 53100671 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/05 job id: 53100672 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/06 job id: 53100673 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/07 job id: 53100674 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/08 job id: 53100675 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/09 job id: 53100676 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/10 job id: 53100678 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/11 job id: 53100680 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/12 job id: 53100683 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/13 job id: 53100686 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/14 job id: 53100689 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/15 job id: 53100693 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/16 job id: 53100696 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/17 job id: 53100697 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/18 job id: 53100698 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/19 job id: 53100699 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/20 job id: 53100700 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/21 job id: 53100702 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/22 job id: 53100703 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/23 job id: 53100706 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/24 job id: 53100707 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/25 job id: 53100708 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/26 job id: 53100709 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/27 job id: 53100710 --- 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 0.025 0.035 128.388 128.388 farming_run 1 2.0 127.824 127.827 128.354 128.357 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.540305E+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 0.025 110.353 110.353 qs_energies 1 2.0 0.001 0.001 110.135 110.137 mp2_main 1 3.0 0.000 0.000 108.338 108.340 mp2_gpw_main 1 4.0 0.029 0.037 107.442 107.444 mp2_ri_gpw_compute_in 1 5.0 0.174 0.178 88.276 88.761 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.161 57.647 mp2_eri_3c_integrate_gpw 272 7.0 0.150 0.164 43.594 48.849 get_2c_integrals 1 6.0 0.008 0.009 30.684 30.940 integrate_v_rspace 273 8.0 0.439 0.458 25.513 30.141 fft_wrap_pw1pw2 5465 10.4 0.067 0.072 28.060 28.707 grid_integrate_task_list 273 9.0 20.759 26.319 20.759 26.319 fft_wrap_pw1pw2_100 2178 11.4 1.315 1.454 25.508 26.161 compute_2c_integrals 1 7.0 0.002 0.002 21.365 21.366 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.847 21.098 mp2_eri_2c_integrate_gpw 1 9.0 2.388 2.424 20.845 21.095 rpa_ri_compute_en 1 5.0 0.020 0.039 19.045 19.316 fft3d_s 5443 12.4 16.862 17.553 16.883 17.574 ao_to_mo_and_store_B_mult_1 272 7.0 10.775 15.307 10.775 15.307 calculate_wavefunction 272 8.0 5.275 5.444 12.752 13.660 calc_potential_gpw 544 9.5 0.004 0.004 10.907 11.739 rpa_num_int 1 6.0 0.002 0.031 10.566 10.566 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.818 10.508 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.459 10.491 potential_pw2rs 545 10.0 0.107 0.108 8.450 10.239 cp_fm_cholesky_decompose 12 8.2 9.471 9.727 9.471 9.727 calc_mat_Q 8 8.0 0.000 0.000 9.448 9.542 contract_S_to_Q 8 9.0 0.000 0.000 8.873 8.960 cholesky_decomp 1 7.0 0.000 0.000 8.508 8.765 parallel_gemm_fm 14 9.1 0.000 0.000 8.494 8.625 parallel_gemm_fm_cosma 14 10.1 8.494 8.625 8.494 8.625 collocate_single_gaussian 272 10.0 0.039 0.042 7.831 8.549 create_integ_mat 1 6.0 0.024 0.028 7.887 7.887 array2fm 1 7.0 0.000 0.000 6.816 7.338 pw_poisson_solve 545 10.5 0.010 0.011 4.678 4.954 pw_scatter_s 2720 12.7 4.489 4.906 4.489 4.906 pw_gather_s 2722 12.2 3.914 4.870 3.914 4.870 pw_poisson_set 548 11.5 0.019 0.019 3.370 3.609 array2fm_buffer_send 1 8.0 2.987 3.182 2.987 3.182 pw_copy 4911 11.6 2.623 2.796 2.623 2.796 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.082 1.235 2.026 2.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.440920, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2810.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 0.024 0.036 426.302 426.303 farming_run 1 2.0 425.537 425.547 426.265 426.267 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.270456E+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 13118677. 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 0.012 0.032 222.468 222.468 qs_energies 1 2.0 0.011 0.029 222.217 222.224 scf_env_do_scf 1 3.0 0.000 0.000 114.247 114.247 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.332 113.341 rebuild_ks_matrix 4 6.0 0.000 0.000 113.331 113.339 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 113.331 113.339 hfx_ks_matrix 4 8.0 0.001 0.001 112.982 112.985 integrate_four_center 4 9.0 0.153 0.466 112.981 112.985 mp2_main 1 3.0 0.000 0.000 107.668 107.685 mp2_gpw_main 1 4.0 0.054 0.085 106.810 106.819 integrate_four_center_main 4 10.0 0.119 0.557 101.984 104.818 integrate_four_center_bin 264 11.0 101.865 104.721 101.865 104.721 init_scf_loop 1 4.0 0.000 0.000 96.510 96.510 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 79.901 81.068 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.797 57.956 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.164 44.694 49.642 integrate_v_rspace 95 8.0 0.397 0.560 28.353 33.148 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.451 31.236 mp2_ri_gpw_compute_en 1 5.0 0.054 0.062 26.749 28.981 fft_wrap_pw1pw2_100 730 11.4 0.633 0.750 28.155 28.939 grid_integrate_task_list 95 9.0 23.654 28.637 23.654 28.637 ao_to_mo_and_store_B_mult_1 91 7.0 10.427 27.370 10.427 27.370 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.902 24.941 24.951 get_2c_integrals 1 6.0 0.000 0.000 23.014 23.045 compute_2c_integrals 1 7.0 0.002 0.003 22.182 22.196 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 21.784 22.028 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.851 21.783 22.028 fft3d_s 1823 12.4 19.195 19.977 19.209 19.990 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.735 17.735 calc_potential_gpw 182 9.5 0.002 0.002 12.481 13.423 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.798 11.385 calculate_wavefunction 91 8.0 1.981 2.025 10.003 10.742 potential_pw2rs 186 10.0 0.033 0.035 9.024 10.594 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.575 8.723 9.426 collocate_single_gaussian 91 10.0 0.017 0.022 8.267 9.222 local_gemm 172 8.0 8.167 8.858 8.167 8.858 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.531 7.983 8.323 mp_sync 37 10.5 2.992 7.736 2.992 7.736 integrate_four_center_load 4 10.0 0.000 0.000 6.745 6.749 hfx_load_balance 1 11.0 0.000 0.000 6.745 6.749 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.424 6.344 6.424 pw_poisson_solve 186 10.4 0.005 0.006 5.657 6.397 mp_sendrecv_dm3 2068 8.0 6.004 6.370 6.004 6.370 pw_gather_s 912 12.2 4.765 5.466 4.765 5.466 pw_poisson_set 189 11.4 0.009 0.011 4.186 4.881 pw_scatter_s 910 12.7 4.116 4.825 4.116 4.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.805516, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1522.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 445.521920E+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 0.011 0.026 56.178 56.180 qs_mol_dyn_low 1 2.0 0.016 0.107 55.779 55.787 qs_forces 11 3.9 0.039 0.060 55.265 55.277 qs_energies 11 4.9 0.006 0.022 53.724 53.780 scf_env_do_scf 11 5.9 0.000 0.001 47.088 47.088 scf_env_do_scf_inner_loop 108 6.5 0.006 0.021 45.143 45.144 qs_scf_new_mos 108 7.5 0.000 0.001 35.015 35.333 qs_scf_loop_do_ot 108 8.5 0.000 0.000 35.015 35.333 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.795 35.258 ot_scf_mini 108 9.5 0.002 0.002 33.326 33.552 multiply_cannon 2286 13.5 0.181 0.199 26.725 28.034 multiply_cannon_loop 2286 14.5 1.822 1.946 26.063 27.406 velocity_verlet 10 3.0 0.002 0.008 26.621 26.626 ot_mini 108 10.5 0.001 0.001 20.173 20.412 qs_ot_get_derivative 108 11.5 0.001 0.001 17.153 17.389 mp_waitall_1 245248 16.5 8.951 15.063 8.951 15.063 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.310 13.028 multiply_cannon_multrec 54864 15.5 3.619 5.627 7.712 11.047 qs_ot_get_p 119 10.4 0.001 0.001 8.336 8.674 rebuild_ks_matrix 119 8.3 0.000 0.000 7.956 8.116 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.955 8.116 mp_sum_l 7287 12.8 6.058 7.658 6.058 7.658 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.010 7.145 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.839 6.322 multiply_cannon_sync_h2d 54864 15.5 5.102 6.139 5.102 6.139 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.700 5.845 dbcsr_mm_accdrv_process 76910 16.1 1.840 2.906 4.007 5.715 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.370 5.437 init_scf_run 11 5.9 0.000 0.001 5.357 5.357 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.357 5.357 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.544 4.544 sum_up_and_integrate 119 10.3 0.001 0.002 4.449 4.456 integrate_v_rspace 119 11.3 0.002 0.003 4.438 4.447 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.326 4.326 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.177 4.315 calculate_rho_elec 119 8.7 0.011 0.016 4.177 4.314 cp_fm_redistribute_end 50 14.0 2.207 4.301 2.213 4.304 cp_fm_diag_elpa_base 50 14.0 2.085 4.201 2.089 4.206 calculate_first_density_matrix 1 7.0 0.002 0.017 3.450 3.455 calculate_dm_sparse 119 9.5 0.000 0.000 3.080 3.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.926 3.105 apply_single 119 13.6 0.000 0.000 2.925 3.105 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.947 2.987 jit_kernel_multiply 13 15.8 2.102 2.779 2.102 2.779 acc_transpose_blocks 54864 15.5 0.224 0.249 2.234 2.769 ot_diis_step 108 11.5 0.006 0.006 2.710 2.710 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.605 2.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.331 2.345 density_rs2pw 119 9.7 0.003 0.004 2.188 2.276 grid_integrate_task_list 119 12.3 2.011 2.106 2.011 2.106 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.048 2.098 mp_sum_d 4139 12.0 1.398 1.999 1.398 1.999 init_scf_loop 11 6.9 0.000 0.001 1.926 1.927 wfi_extrapolate 11 7.9 0.001 0.001 1.838 1.838 potential_pw2rs 119 12.3 0.004 0.004 1.821 1.835 make_m2s 4572 13.5 0.053 0.055 1.700 1.752 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.607 1.682 make_images 4572 14.5 0.133 0.139 1.618 1.669 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.558 1.594 transfer_rs2pw 487 10.6 0.006 0.007 1.460 1.581 mp_alltoall_d11v 2130 13.8 1.354 1.573 1.354 1.573 mp_waitany 12084 13.8 1.320 1.475 1.320 1.475 fft3d_ps 1201 13.6 0.361 0.461 1.382 1.453 acc_transpose_blocks_sync 164592 16.5 1.205 1.449 1.205 1.449 grid_collocate_task_list 119 9.7 1.352 1.425 1.352 1.425 transfer_pw2rs 487 13.2 0.005 0.006 1.359 1.371 fft_wrap_pw1pw2_140 487 12.2 0.049 0.055 1.251 1.326 make_basis_sm 11 9.8 0.000 0.000 1.264 1.269 parallel_gemm_fm 81 9.0 0.000 0.000 1.202 1.251 parallel_gemm_fm_cosma 81 10.0 1.202 1.251 1.202 1.251 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.755 1.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.180000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.727273, yerr=0.862439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 485.625856E+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 1108327. 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 0.006 0.027 38.670 38.670 qs_mol_dyn_low 1 2.0 0.003 0.003 38.488 38.496 qs_forces 11 3.9 0.004 0.007 38.422 38.423 qs_energies 11 4.9 0.001 0.002 36.764 36.770 scf_env_do_scf 11 5.9 0.000 0.001 31.550 31.551 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.253 29.253 dbcsr_multiply_generic 2286 12.5 0.102 0.104 22.180 22.526 qs_scf_new_mos 108 7.5 0.001 0.001 20.638 20.866 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.638 20.865 ot_scf_mini 108 9.5 0.002 0.003 19.729 19.892 multiply_cannon 2286 13.5 0.208 0.216 17.017 18.463 velocity_verlet 10 3.0 0.001 0.002 18.079 18.080 multiply_cannon_loop 2286 14.5 1.186 1.244 15.867 17.477 ot_mini 108 10.5 0.001 0.001 12.232 12.463 mp_waitall_1 200699 16.5 5.605 10.846 5.605 10.846 qs_ot_get_derivative 108 11.5 0.001 0.001 9.878 10.043 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.071 9.623 multiply_cannon_multrec 27432 15.5 1.815 4.109 6.166 9.067 rebuild_ks_matrix 119 8.3 0.000 0.000 6.734 6.882 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.733 6.882 dbcsr_mm_accdrv_process 47894 16.0 3.495 5.720 4.272 6.406 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.927 6.062 qs_ot_get_p 119 10.4 0.001 0.001 4.652 4.883 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.714 4.655 mp_sum_l 7287 12.8 2.111 4.108 2.111 4.108 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.090 4.079 apply_single 119 13.6 0.000 0.000 3.090 4.079 init_scf_run 11 5.9 0.000 0.001 4.004 4.005 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.004 4.004 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.610 3.655 calculate_rho_elec 119 8.7 0.021 0.024 3.609 3.654 sum_up_and_integrate 119 10.3 0.001 0.001 3.620 3.628 integrate_v_rspace 119 11.3 0.002 0.003 3.606 3.614 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.951 2.968 make_m2s 4572 13.5 0.052 0.054 2.593 2.936 make_images 4572 14.5 0.206 0.244 2.504 2.849 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.555 2.556 calculate_first_density_matrix 1 7.0 0.000 0.000 2.524 2.528 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.324 2.408 ot_diis_step 108 11.5 0.011 0.011 2.304 2.304 init_scf_loop 11 6.9 0.003 0.013 2.271 2.272 multiply_cannon_sync_h2d 27432 15.5 1.682 2.241 1.682 2.241 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.240 2.240 calculate_dm_sparse 119 9.5 0.000 0.001 2.146 2.230 cp_fm_redistribute_end 50 14.0 1.136 2.216 1.139 2.219 cp_fm_diag_elpa_base 50 14.0 1.048 2.129 1.075 2.162 acc_transpose_blocks 27432 15.5 0.116 0.122 1.668 2.010 density_rs2pw 119 9.7 0.003 0.004 1.902 1.977 grid_integrate_task_list 119 12.3 1.834 1.933 1.834 1.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.910 1.912 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.715 1.756 jit_kernel_multiply 9 16.1 0.717 1.750 0.717 1.750 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.672 1.702 make_images_data 4572 15.5 0.048 0.054 1.226 1.545 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.091 1.469 wfi_extrapolate 11 7.9 0.001 0.001 1.422 1.422 fft3d_ps 1201 13.6 0.511 0.564 1.391 1.416 potential_pw2rs 119 12.3 0.006 0.006 1.375 1.382 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.356 1.368 prepare_preconditioner 11 7.9 0.000 0.000 1.335 1.364 make_preconditioner 11 8.9 0.000 0.000 1.335 1.364 fft_wrap_pw1pw2_140 487 12.2 0.048 0.050 1.322 1.353 grid_collocate_task_list 119 9.7 1.289 1.343 1.289 1.343 mp_alltoall_d11v 2130 13.8 1.225 1.317 1.225 1.317 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.239 1.288 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.197 1.244 transfer_rs2pw 487 10.6 0.005 0.006 1.129 1.203 mp_sum_d 4139 12.0 0.579 1.064 0.579 1.064 mp_allgather_i34 2286 14.5 0.571 0.981 0.571 0.981 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.712 0.937 acc_transpose_blocks_sync 82296 16.5 0.814 0.935 0.814 0.935 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.928 0.930 transfer_pw2rs 487 13.2 0.004 0.005 0.903 0.910 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.862 0.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.670000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.909091, yerr=1.975051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 515.760128E+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 0.046 0.082 34.701 34.703 qs_mol_dyn_low 1 2.0 0.038 0.146 33.946 34.383 qs_forces 11 3.9 0.026 0.079 33.805 33.845 qs_energies 11 4.9 0.005 0.019 32.209 32.262 scf_env_do_scf 11 5.9 0.001 0.001 25.759 25.759 scf_env_do_scf_inner_loop 108 6.5 0.012 0.043 23.415 23.415 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.427 17.504 qs_scf_new_mos 108 7.5 0.001 0.001 15.679 15.694 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.679 15.693 velocity_verlet 10 3.0 0.007 0.030 15.112 15.114 ot_scf_mini 108 9.5 0.002 0.003 14.915 14.933 multiply_cannon 2286 13.5 0.192 0.198 14.129 14.855 multiply_cannon_loop 2286 14.5 0.861 0.885 13.339 14.079 ot_mini 108 10.5 0.001 0.001 9.168 9.188 qs_ot_get_derivative 108 11.5 0.001 0.001 7.661 7.680 multiply_cannon_multrec 18288 15.5 1.876 2.859 7.109 7.484 rebuild_ks_matrix 119 8.3 0.000 0.000 6.039 6.083 qs_ks_build_kohn_sham_matrix 119 9.3 0.023 0.070 6.038 6.083 dbcsr_mm_accdrv_process 38222 16.0 5.055 5.900 5.141 5.966 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.326 5.370 init_scf_run 11 5.9 0.000 0.002 4.808 4.808 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.808 4.808 mp_waitall_1 158411 16.6 2.911 4.024 2.911 4.024 calculate_first_density_matrix 1 7.0 0.000 0.002 3.577 3.579 sum_up_and_integrate 119 10.3 0.001 0.002 3.546 3.551 integrate_v_rspace 119 11.3 0.003 0.004 3.532 3.541 qs_ot_get_p 119 10.4 0.001 0.001 3.494 3.522 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.954 3.501 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.276 3.302 calculate_rho_elec 119 8.7 0.031 0.032 3.275 3.301 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.628 2.561 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.971 2.376 apply_single 119 13.6 0.000 0.000 1.970 2.376 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.347 2.353 init_scf_loop 11 6.9 0.004 0.017 2.325 2.325 calculate_dm_sparse 119 9.5 0.000 0.001 2.059 2.067 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.048 2.049 make_m2s 4572 13.5 0.044 0.045 1.850 1.995 grid_integrate_task_list 119 12.3 1.803 1.921 1.803 1.921 make_images 4572 14.5 0.192 0.204 1.764 1.909 density_rs2pw 119 9.7 0.003 0.004 1.809 1.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.897 1.901 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.789 1.800 cp_fm_diag_elpa_base 50 14.0 1.765 1.777 1.787 1.798 acc_transpose_blocks 18288 15.5 0.080 0.082 1.706 1.789 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.735 1.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.725 1.732 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.664 1.675 prepare_preconditioner 11 7.9 0.000 0.000 1.535 1.552 make_preconditioner 11 8.9 0.001 0.005 1.535 1.552 transfer_rs2pw 487 10.6 0.005 0.005 1.311 1.535 mp_sum_l 7287 12.8 1.096 1.510 1.096 1.510 ot_diis_step 108 11.5 0.011 0.012 1.492 1.492 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.416 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.391 1.394 qs_energies_init_hamiltonians 11 5.9 0.003 0.013 1.333 1.386 fft3d_ps 1201 13.6 0.535 0.554 1.349 1.360 potential_pw2rs 119 12.3 0.007 0.008 1.340 1.346 grid_collocate_task_list 119 9.7 1.232 1.330 1.232 1.330 fft_wrap_pw1pw2_140 487 12.2 0.061 0.064 1.310 1.322 parallel_gemm_fm 81 9.0 0.000 0.000 1.284 1.286 parallel_gemm_fm_cosma 81 10.0 1.284 1.286 1.284 1.286 make_basis_sm 11 9.8 0.000 0.000 1.258 1.262 wfi_extrapolate 11 7.9 0.004 0.016 1.173 1.173 multiply_cannon_sync_h2d 18288 15.5 1.032 1.170 1.032 1.170 mp_waitany 9880 13.7 0.804 1.045 0.804 1.045 transfer_rs2pw_140 130 11.5 0.121 0.130 0.794 1.021 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.955 0.973 qs_env_update_s_mstruct 11 6.9 0.000 0.002 0.842 0.965 make_images_data 4572 15.5 0.048 0.052 0.794 0.951 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.867 0.925 transfer_pw2rs 487 13.2 0.004 0.004 0.866 0.870 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.692 0.866 mp_alltoall_d11v 2130 13.8 0.745 0.857 0.745 0.857 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.476 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.786 acc_transpose_blocks_sync 54864 16.5 0.738 0.783 0.738 0.783 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.670 0.745 mp_alltoall_z22v 1201 15.6 0.656 0.728 0.656 0.728 jit_kernel_transpose 5 15.6 0.650 0.707 0.650 0.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.703000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.545455, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 554.041344E+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 0.120 0.165 36.635 36.645 qs_mol_dyn_low 1 2.0 0.016 0.091 36.061 36.078 qs_forces 11 3.9 0.002 0.002 35.783 35.785 qs_energies 11 4.9 0.004 0.020 34.085 34.094 scf_env_do_scf 11 5.9 0.001 0.001 28.724 28.724 scf_env_do_scf_inner_loop 108 6.5 0.010 0.035 25.780 25.780 dbcsr_multiply_generic 2286 12.5 0.101 0.104 20.263 20.400 qs_scf_new_mos 108 7.5 0.001 0.001 17.906 17.967 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.905 17.966 velocity_verlet 10 3.0 0.026 0.159 17.841 17.866 multiply_cannon 2286 13.5 0.216 0.224 16.571 16.957 ot_scf_mini 108 9.5 0.002 0.003 16.877 16.930 multiply_cannon_loop 2286 14.5 1.547 1.646 15.634 16.126 ot_mini 108 10.5 0.001 0.001 10.483 10.553 multiply_cannon_multrec 27432 15.5 2.460 3.140 9.318 9.593 qs_ot_get_derivative 108 11.5 0.001 0.001 8.618 8.674 dbcsr_mm_accdrv_process 47916 15.9 5.990 7.480 6.758 8.077 rebuild_ks_matrix 119 8.3 0.000 0.000 6.259 6.316 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 6.259 6.315 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.544 5.596 init_scf_run 11 5.9 0.001 0.008 3.988 3.989 scf_env_initial_rho_setup 11 6.9 0.006 0.007 3.988 3.988 qs_ot_get_p 119 10.4 0.002 0.002 3.577 3.657 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.186 3.630 sum_up_and_integrate 119 10.3 0.001 0.001 3.405 3.412 integrate_v_rspace 119 11.3 0.002 0.003 3.394 3.401 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.252 3.299 calculate_rho_elec 119 8.7 0.040 0.047 3.252 3.299 init_scf_loop 11 6.9 0.023 0.039 2.924 2.924 calculate_first_density_matrix 1 7.0 0.000 0.003 2.573 2.575 acc_transpose_blocks 27432 15.5 0.119 0.123 2.452 2.549 mp_waitall_1 137007 16.6 1.828 2.445 1.828 2.445 make_m2s 4572 13.5 0.054 0.056 2.239 2.353 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.311 2.339 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.181 2.333 apply_single 119 13.6 0.000 0.000 2.181 2.333 calculate_dm_sparse 119 9.5 0.000 0.000 2.272 2.326 make_images 4572 14.5 0.273 0.335 2.131 2.243 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.158 2.168 prepare_preconditioner 11 7.9 0.000 0.000 2.024 2.055 make_preconditioner 11 8.9 0.009 0.069 2.024 2.055 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.752 1.984 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.954 1.956 grid_integrate_task_list 119 12.3 1.838 1.917 1.838 1.917 jit_kernel_multiply 9 15.6 0.700 1.869 0.700 1.869 ot_diis_step 108 11.5 0.012 0.012 1.826 1.826 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.781 1.797 density_rs2pw 119 9.7 0.003 0.004 1.690 1.795 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.760 1.760 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.681 1.711 acc_transpose_blocks_sync 82296 16.5 1.465 1.569 1.465 1.569 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.526 1.527 cp_fm_redistribute_end 50 14.0 0.765 1.505 0.768 1.506 cp_fm_diag_elpa_base 50 14.0 0.712 1.437 0.736 1.479 mp_sum_l 7287 12.8 1.086 1.456 1.086 1.456 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.389 1.422 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.374 1.387 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.837 1.387 fft3d_ps 1201 13.6 0.575 0.638 1.343 1.369 grid_collocate_task_list 119 9.7 1.254 1.342 1.254 1.342 wfi_extrapolate 11 7.9 0.001 0.001 1.312 1.312 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.288 1.306 potential_pw2rs 119 12.3 0.008 0.009 1.199 1.202 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.086 1.088 transfer_rs2pw 487 10.6 0.005 0.005 0.915 1.025 make_images_data 4572 15.5 0.048 0.051 0.833 0.955 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.710 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.876 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.837 0.866 mp_alltoall_d11v 2130 13.8 0.766 0.862 0.766 0.862 dbcsr_complete_redistribute 329 12.2 0.108 0.126 0.650 0.810 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.645000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.090909, yerr=2.678349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 621.449216E+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 0.060 0.083 28.783 28.783 qs_mol_dyn_low 1 2.0 0.003 0.003 28.406 28.447 qs_forces 11 3.9 0.002 0.003 28.336 28.337 qs_energies 11 4.9 0.001 0.002 26.646 26.648 scf_env_do_scf 11 5.9 0.000 0.001 21.543 21.544 scf_env_do_scf_inner_loop 108 6.5 0.029 0.032 19.178 19.178 velocity_verlet 10 3.0 0.012 0.013 14.313 14.316 dbcsr_multiply_generic 2286 12.5 0.095 0.099 13.497 13.559 qs_scf_new_mos 108 7.5 0.001 0.001 11.755 11.778 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.754 11.777 multiply_cannon 2286 13.5 0.225 0.233 10.976 11.433 ot_scf_mini 108 9.5 0.002 0.002 11.047 11.075 multiply_cannon_loop 2286 14.5 0.644 0.665 10.086 10.223 multiply_cannon_multrec 9144 15.5 1.785 1.972 6.443 6.657 ot_mini 108 10.5 0.001 0.001 6.356 6.386 rebuild_ks_matrix 119 8.3 0.000 0.000 5.658 5.676 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.658 5.676 qs_ot_get_derivative 108 11.5 0.001 0.001 5.075 5.104 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.030 5.048 dbcsr_mm_accdrv_process 12550 15.8 3.779 4.558 4.549 4.656 init_scf_run 11 5.9 0.000 0.001 3.682 3.682 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.681 3.682 sum_up_and_integrate 119 10.3 0.001 0.001 3.331 3.337 integrate_v_rspace 119 11.3 0.003 0.003 3.321 3.326 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.184 3.192 calculate_rho_elec 119 8.7 0.060 0.061 3.183 3.191 qs_ot_get_p 119 10.4 0.001 0.001 2.858 2.898 calculate_first_density_matrix 1 7.0 0.000 0.000 2.563 2.563 init_scf_loop 11 6.9 0.064 0.065 2.345 2.347 calculate_dm_sparse 119 9.5 0.000 0.000 1.965 1.982 grid_integrate_task_list 119 12.3 1.868 1.930 1.868 1.930 make_m2s 4572 13.5 0.034 0.035 1.752 1.875 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.840 1.842 mp_waitall_1 115863 16.7 1.274 1.812 1.274 1.812 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.802 1.804 make_images 4572 14.5 0.269 0.301 1.663 1.783 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.733 1.754 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.727 1.736 density_rs2pw 119 9.7 0.003 0.003 1.569 1.688 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.636 1.644 acc_transpose_blocks 9144 15.5 0.043 0.044 1.574 1.608 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.599 1.599 jit_kernel_multiply 8 15.7 0.730 1.555 0.730 1.555 prepare_preconditioner 11 7.9 0.000 0.000 1.523 1.526 make_preconditioner 11 8.9 0.000 0.000 1.523 1.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.420 1.446 fft_wrap_pw1pw2_140 487 12.2 0.093 0.096 1.429 1.439 grid_collocate_task_list 119 9.7 1.297 1.388 1.297 1.388 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.358 1.362 fft3d_ps 1201 13.6 0.636 0.653 1.339 1.349 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.329 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.296 1.304 cp_fm_diag_elpa_base 50 14.0 1.269 1.285 1.294 1.302 ot_diis_step 108 11.5 0.013 0.013 1.268 1.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.214 1.244 apply_single 119 13.6 0.000 0.000 1.214 1.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.210 1.211 potential_pw2rs 119 12.3 0.010 0.010 1.144 1.147 wfi_extrapolate 11 7.9 0.001 0.001 1.069 1.069 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.753 0.929 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.923 make_images_data 4572 15.5 0.043 0.046 0.742 0.882 transfer_rs2pw 487 10.6 0.004 0.005 0.758 0.878 mp_alltoall_d11v 2130 13.8 0.800 0.850 0.800 0.850 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.820 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.796 0.798 cp_fm_cholesky_invert 11 10.9 0.748 0.753 0.748 0.753 acc_transpose_blocks_sync 27432 16.5 0.720 0.753 0.720 0.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.750 0.753 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.660 0.707 jit_kernel_transpose 5 15.6 0.678 0.679 0.678 0.679 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.328 0.673 transfer_pw2rs 487 13.2 0.003 0.004 0.649 0.651 mp_alltoall_z22v 1201 15.6 0.578 0.615 0.578 0.615 mp_allgather_i34 2286 14.5 0.203 0.603 0.203 0.603 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.783000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.272727, yerr=4.974314 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 772.706304E+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 0.013 0.032 41.070 41.071 qs_mol_dyn_low 1 2.0 0.003 0.003 40.835 40.842 qs_forces 11 3.9 0.002 0.003 40.769 40.770 qs_energies 11 4.9 0.017 0.017 38.783 38.786 scf_env_do_scf 11 5.9 0.001 0.001 32.640 32.640 scf_env_do_scf_inner_loop 108 6.5 0.005 0.008 26.662 26.663 velocity_verlet 10 3.0 0.002 0.003 22.130 22.136 dbcsr_multiply_generic 2286 12.5 0.103 0.104 19.445 19.609 qs_scf_new_mos 108 7.5 0.001 0.001 17.443 17.541 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.442 17.541 ot_scf_mini 108 9.5 0.002 0.002 16.268 16.364 multiply_cannon 2286 13.5 0.301 0.308 15.231 16.113 multiply_cannon_loop 2286 14.5 0.860 0.878 13.962 14.847 ot_mini 108 10.5 0.001 0.001 9.919 10.033 multiply_cannon_multrec 9144 15.5 3.398 4.766 8.876 9.005 qs_ot_get_derivative 108 11.5 0.001 0.001 7.852 7.954 rebuild_ks_matrix 119 8.3 0.000 0.000 6.999 7.140 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.998 7.139 dbcsr_mm_accdrv_process 12550 15.8 4.789 6.674 5.345 6.708 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.297 6.425 init_scf_loop 11 6.9 0.032 0.034 5.951 5.955 prepare_preconditioner 11 7.9 0.000 0.000 4.977 4.989 make_preconditioner 11 8.9 0.000 0.000 4.976 4.989 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.981 4.873 init_scf_run 11 5.9 0.001 0.007 4.057 4.057 scf_env_initial_rho_setup 11 6.9 0.020 0.021 4.056 4.057 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.915 3.925 calculate_rho_elec 119 8.7 0.118 0.122 3.915 3.925 sum_up_and_integrate 119 10.3 0.001 0.001 3.698 3.705 integrate_v_rspace 119 11.3 0.003 0.003 3.687 3.695 qs_ot_get_p 119 10.4 0.001 0.002 3.405 3.546 mp_waitall_1 94719 16.7 2.426 3.502 2.426 3.502 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.940 3.421 cp_fm_upper_to_full 72 14.2 2.102 3.002 2.102 3.002 make_m2s 4572 13.5 0.037 0.038 2.546 2.702 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.286 2.671 apply_single 119 13.6 0.000 0.000 2.286 2.671 calculate_first_density_matrix 1 7.0 0.000 0.000 2.595 2.598 make_images 4572 14.5 0.352 0.388 2.425 2.581 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.382 2.407 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.384 2.390 calculate_dm_sparse 119 9.5 0.000 0.000 2.308 2.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.175 2.177 dbcsr_complete_redistribute 329 12.2 0.261 0.270 1.561 2.130 grid_integrate_task_list 119 12.3 2.071 2.108 2.071 2.108 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.053 2.106 ot_diis_step 108 11.5 0.014 0.014 2.037 2.038 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.020 2.027 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.983 1.985 density_rs2pw 119 9.7 0.003 0.003 1.954 1.976 mp_sum_l 7287 12.8 1.120 1.948 1.120 1.948 acc_transpose_blocks 9144 15.5 0.045 0.046 1.869 1.922 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.826 1.875 fft3d_ps 1201 13.6 0.858 0.878 1.837 1.844 qs_energies_init_hamiltonians 11 5.9 0.003 0.003 1.793 1.794 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.242 1.792 mp_alltoall_i22 627 13.8 1.057 1.668 1.057 1.668 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.627 1.628 grid_collocate_task_list 119 9.7 1.515 1.543 1.515 1.543 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.990 1.536 wfi_extrapolate 11 7.9 0.001 0.001 1.380 1.380 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.137 1.378 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.322 1.345 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.327 1.327 cp_fm_diag_elpa_base 50 14.0 1.240 1.253 1.325 1.325 make_images_data 4572 15.5 0.046 0.049 1.099 1.302 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.301 potential_pw2rs 119 12.3 0.013 0.014 1.209 1.212 mp_alltoall_d11v 2130 13.8 1.168 1.189 1.168 1.189 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.125 1.140 cp_fm_cholesky_invert 11 10.9 1.130 1.137 1.130 1.137 acc_transpose_blocks_sync 27432 16.5 1.095 1.135 1.095 1.135 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.032 qs_create_task_list 11 7.9 0.001 0.001 0.928 0.939 generate_qs_task_list 11 8.9 0.368 0.388 0.928 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.912 0.925 jit_kernel_multiply 5 15.5 0.526 0.923 0.526 0.923 mp_alltoall_z22v 1201 15.6 0.843 0.873 0.843 0.873 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.781 0.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.071000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.636364, yerr=12.715579 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 498.495488E+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 2142509. 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 0.007 0.028 84.838 84.839 qs_mol_dyn_low 1 2.0 0.003 0.003 84.574 84.602 qs_forces 11 3.9 0.003 0.003 84.367 84.368 qs_energies 11 4.9 0.004 0.006 81.511 81.535 scf_env_do_scf 11 5.9 0.000 0.001 71.250 71.251 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.008 66.008 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.809 53.181 qs_scf_new_mos 99 7.5 0.000 0.001 49.265 49.408 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.264 49.407 ot_scf_mini 99 9.5 0.002 0.002 46.799 46.913 multiply_cannon 2055 13.4 0.174 0.181 43.259 44.234 multiply_cannon_loop 2055 14.4 1.809 1.889 42.312 43.200 velocity_verlet 10 3.0 0.001 0.002 42.759 42.761 ot_mini 99 10.5 0.001 0.001 27.763 27.875 qs_ot_get_derivative 99 11.5 0.001 0.001 20.873 20.975 multiply_cannon_multrec 49320 15.4 11.121 11.888 17.326 18.156 rebuild_ks_matrix 110 8.3 0.000 0.001 14.414 14.518 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.414 14.518 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.650 12.741 mp_waitall_1 220248 16.4 11.669 12.583 11.669 12.583 qs_ot_get_p 110 10.4 0.001 0.004 10.334 10.499 multiply_cannon_sync_h2d 49320 15.4 9.313 9.898 9.313 9.898 init_scf_run 11 5.9 0.000 0.002 8.161 8.162 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.161 8.161 multiply_cannon_metrocomm3 49320 15.4 0.081 0.086 7.084 8.138 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.578 8.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.233 7.815 apply_single 110 13.6 0.000 0.001 7.232 7.815 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.885 6.918 sum_up_and_integrate 110 10.3 0.002 0.003 6.841 6.853 integrate_v_rspace 110 11.3 0.003 0.003 6.815 6.834 ot_diis_step 99 11.5 0.005 0.005 6.610 6.610 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.403 6.469 dbcsr_mm_accdrv_process 87628 16.1 3.176 3.430 6.078 6.465 mp_sum_l 6594 12.7 5.237 6.184 5.237 6.184 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.050 6.173 calculate_rho_elec 110 8.6 0.020 0.024 6.049 6.173 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.959 5.960 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.475 5.476 cp_fm_redistribute_end 48 14.0 2.777 5.448 2.787 5.451 cp_fm_diag_elpa_base 48 14.0 2.653 5.337 2.662 5.354 init_scf_loop 11 6.9 0.000 0.001 5.211 5.211 calculate_first_density_matrix 1 7.0 0.000 0.000 4.113 4.120 calculate_dm_sparse 110 9.5 0.000 0.001 3.990 4.091 make_m2s 4110 13.4 0.061 0.065 3.847 3.951 wfi_extrapolate 11 7.9 0.001 0.001 3.945 3.945 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.766 3.906 make_images 4110 14.4 0.179 0.190 3.751 3.858 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.747 3.750 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.393 3.463 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.381 3.419 grid_integrate_task_list 110 12.3 3.241 3.410 3.241 3.410 density_rs2pw 110 9.6 0.004 0.004 3.089 3.294 prepare_preconditioner 11 7.9 0.000 0.000 3.048 3.073 make_preconditioner 11 8.9 0.000 0.000 3.048 3.073 acc_transpose_blocks 49320 15.4 0.219 0.238 2.716 2.952 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.784 2.817 jit_kernel_multiply 13 15.9 2.619 2.732 2.619 2.732 mp_alltoall_d11v 2046 13.8 2.201 2.725 2.201 2.725 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 2.575 2.630 potential_pw2rs 110 12.3 0.006 0.006 2.445 2.461 grid_collocate_task_list 110 9.6 2.159 2.309 2.159 2.309 fft_wrap_pw1pw2_140 451 12.1 0.092 0.097 2.191 2.249 parallel_gemm_fm 81 9.0 0.000 0.000 2.231 2.236 parallel_gemm_fm_cosma 81 10.0 2.231 2.236 2.231 2.236 mp_sum_d 3893 11.9 1.546 2.181 1.546 2.181 fft3d_ps 1111 13.6 0.726 0.808 2.114 2.159 mp_waitany 14300 13.8 1.822 2.099 1.822 2.099 make_basis_sm 11 9.8 0.000 0.000 2.072 2.075 transfer_rs2pw 451 10.6 0.006 0.007 1.846 2.027 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.915 make_images_data 4110 15.4 0.044 0.047 1.757 1.894 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.817 1.846 hybrid_alltoall_any 4261 16.3 0.085 0.488 1.544 1.822 transfer_pw2rs 451 13.1 0.006 0.007 1.745 1.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.839000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.545455, yerr=2.349609 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 589.463552E+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 1281609. 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 0.023 0.066 68.140 68.141 qs_mol_dyn_low 1 2.0 0.004 0.008 67.676 67.686 qs_forces 11 3.9 0.003 0.003 67.567 67.569 qs_energies 11 4.9 0.002 0.002 64.253 64.257 scf_env_do_scf 11 5.9 0.000 0.001 55.816 55.816 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.878 48.878 dbcsr_multiply_generic 2055 12.4 0.115 0.120 37.936 38.107 velocity_verlet 10 3.0 0.007 0.045 35.011 35.013 qs_scf_new_mos 99 7.5 0.001 0.001 33.296 33.415 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.296 33.414 multiply_cannon 2055 13.4 0.224 0.244 31.349 32.307 ot_scf_mini 99 9.5 0.003 0.003 31.627 31.737 multiply_cannon_loop 2055 14.4 1.157 1.180 30.222 31.136 ot_mini 99 10.5 0.001 0.001 18.387 18.504 multiply_cannon_multrec 24660 15.4 6.949 8.261 14.087 15.408 rebuild_ks_matrix 110 8.3 0.000 0.001 13.470 13.566 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.469 13.565 qs_ot_get_derivative 99 11.5 0.001 0.001 12.631 12.738 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.848 11.939 mp_waitall_1 176588 16.5 7.562 10.132 7.562 10.132 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.195 7.689 multiply_cannon_sync_h2d 24660 15.4 6.337 7.306 6.337 7.306 dbcsr_mm_accdrv_process 52282 16.1 5.539 6.343 6.969 7.284 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.423 7.052 apply_single 110 13.6 0.000 0.001 6.423 7.052 init_scf_loop 11 6.9 0.030 0.037 6.899 6.900 qs_ot_get_p 110 10.4 0.024 0.056 6.688 6.821 sum_up_and_integrate 110 10.3 0.001 0.003 6.218 6.229 integrate_v_rspace 110 11.3 0.002 0.003 6.191 6.201 init_scf_run 11 5.9 0.000 0.001 6.098 6.099 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.098 6.099 ot_diis_step 99 11.5 0.010 0.010 5.709 5.709 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.645 5.652 calculate_rho_elec 110 8.6 0.039 0.047 5.644 5.651 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.786 5.527 prepare_preconditioner 11 7.9 0.000 0.000 4.799 4.846 make_preconditioner 11 8.9 0.000 0.000 4.799 4.846 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.741 4.761 make_m2s 4110 13.4 0.057 0.060 4.211 4.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.435 4.522 make_images 4110 14.4 0.406 0.468 4.101 4.414 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.271 4.271 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.733 3.733 cp_fm_redistribute_end 48 14.0 1.881 3.704 1.887 3.706 cp_fm_diag_elpa_base 48 14.0 1.758 3.569 1.814 3.646 wfi_extrapolate 11 7.9 0.001 0.001 3.407 3.407 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.277 3.340 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.290 3.292 grid_integrate_task_list 110 12.3 3.168 3.285 3.168 3.285 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.066 3.204 calculate_dm_sparse 110 9.5 0.001 0.001 3.034 3.061 density_rs2pw 110 9.6 0.004 0.004 2.890 3.047 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.961 3.001 fft_wrap_pw1pw2_140 451 12.1 0.109 0.116 2.613 2.749 make_images_data 4110 15.4 0.049 0.053 2.297 2.636 calculate_first_density_matrix 1 7.0 0.000 0.000 2.594 2.597 fft3d_ps 1111 13.6 1.068 1.276 2.445 2.572 hybrid_alltoall_any 4261 16.3 0.106 0.463 1.975 2.539 cp_fm_cholesky_invert 11 10.9 2.352 2.367 2.352 2.367 mp_sum_l 6594 12.7 1.750 2.338 1.750 2.338 grid_collocate_task_list 110 9.6 2.175 2.322 2.175 2.322 jit_kernel_multiply 11 16.2 1.059 2.178 1.059 2.178 potential_pw2rs 110 12.3 0.008 0.009 2.106 2.116 acc_transpose_blocks 24660 15.4 0.118 0.123 2.062 2.096 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.962 1.982 mp_alltoall_d11v 2046 13.8 1.771 1.976 1.771 1.976 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.854 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.782 1.794 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.795 1.742 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.600 1.613 transfer_rs2pw 451 10.6 0.007 0.007 1.402 1.551 mp_irecv_dv 57340 16.2 0.661 1.515 0.661 1.515 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.488 mp_allgather_i34 2055 14.4 0.434 1.463 0.434 1.463 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.141000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.090909, yerr=6.126760 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 666.603520E+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 0.038 0.099 60.461 60.463 qs_mol_dyn_low 1 2.0 0.033 0.132 59.690 60.004 qs_forces 11 3.9 0.010 0.015 58.889 58.933 qs_energies 11 4.9 0.003 0.006 55.630 55.683 scf_env_do_scf 11 5.9 0.001 0.002 46.697 46.698 scf_env_do_scf_inner_loop 99 6.5 0.007 0.023 39.232 39.233 velocity_verlet 10 3.0 0.002 0.005 31.786 31.809 dbcsr_multiply_generic 2055 12.4 0.110 0.114 28.986 29.277 qs_scf_new_mos 99 7.5 0.001 0.001 24.862 24.951 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.861 24.950 multiply_cannon 2055 13.4 0.212 0.220 22.696 24.021 ot_scf_mini 99 9.5 0.004 0.016 23.609 23.718 multiply_cannon_loop 2055 14.4 0.818 0.843 21.539 22.540 ot_mini 99 10.5 0.001 0.001 13.566 13.670 rebuild_ks_matrix 110 8.3 0.000 0.000 12.159 12.305 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 12.159 12.304 multiply_cannon_multrec 16440 15.4 3.702 4.944 10.098 11.265 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.663 10.795 mp_waitall_1 139946 16.5 6.805 9.550 6.805 9.550 qs_ot_get_derivative 99 11.5 0.001 0.001 9.146 9.253 init_scf_loop 11 6.9 0.002 0.007 7.427 7.429 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.232 6.914 dbcsr_mm_accdrv_process 34862 16.1 5.582 5.956 6.241 6.444 init_scf_run 11 5.9 0.000 0.002 6.296 6.296 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.295 6.296 sum_up_and_integrate 110 10.3 0.002 0.002 6.155 6.192 integrate_v_rspace 110 11.3 0.003 0.003 6.129 6.166 prepare_preconditioner 11 7.9 0.000 0.000 5.655 5.670 make_preconditioner 11 8.9 0.001 0.002 5.655 5.670 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.482 5.511 calculate_rho_elec 110 8.6 0.058 0.058 5.482 5.511 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.956 5.390 apply_single 110 13.6 0.000 0.000 4.956 5.390 qs_ot_get_p 110 10.4 0.002 0.007 5.246 5.358 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.098 5.321 make_m2s 4110 13.4 0.049 0.051 4.162 4.537 make_images 4110 14.4 0.396 0.515 4.048 4.422 ot_diis_step 99 11.5 0.011 0.011 4.389 4.389 multiply_cannon_sync_h2d 16440 15.4 3.252 3.980 3.252 3.980 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.100 3.774 qs_ot_p2m_diag 48 11.0 0.042 0.046 3.634 3.639 grid_integrate_task_list 110 12.3 3.192 3.359 3.192 3.359 calculate_first_density_matrix 1 7.0 0.003 0.024 3.325 3.328 cp_dbcsr_syevd 48 12.0 0.003 0.008 3.267 3.267 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.042 3.105 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.017 3.022 density_rs2pw 110 9.6 0.004 0.004 2.725 2.925 wfi_extrapolate 11 7.9 0.002 0.009 2.875 2.875 make_images_data 4110 15.4 0.046 0.050 2.318 2.807 calculate_dm_sparse 110 9.5 0.001 0.001 2.721 2.761 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.066 2.737 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.714 2.726 cp_fm_diag_elpa_base 48 14.0 2.671 2.690 2.711 2.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.641 2.701 fft_wrap_pw1pw2_140 451 12.1 0.132 0.175 2.617 2.683 multiply_cannon_metrocomm4 14385 15.4 0.048 0.054 0.879 2.524 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.409 2.462 mp_irecv_dv 48980 15.7 0.801 2.386 0.801 2.386 grid_collocate_task_list 110 9.6 2.223 2.365 2.223 2.365 cp_fm_cholesky_invert 11 10.9 2.348 2.360 2.348 2.360 fft3d_ps 1111 13.6 1.086 1.104 2.347 2.359 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.216 2.268 mp_sum_l 6594 12.7 1.516 2.059 1.516 2.059 mp_alltoall_d11v 2046 13.8 1.813 2.026 1.813 2.026 potential_pw2rs 110 12.3 0.011 0.014 1.998 2.004 dbcsr_complete_redistribute 325 12.2 0.311 0.349 1.413 1.725 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.695 1.709 acc_transpose_blocks 16440 15.4 0.078 0.081 1.607 1.676 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.657 1.673 transfer_rs2pw 451 10.6 0.005 0.006 1.432 1.614 mp_waitany 17072 13.8 1.307 1.542 1.307 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.508 mp_allgather_i34 2055 14.4 0.466 1.481 0.466 1.481 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.298 1.433 cp_fm_cholesky_decompose 22 10.9 1.309 1.330 1.309 1.330 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.274 make_basis_sm 11 9.8 0.000 0.001 1.260 1.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.463000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.454545, yerr=9.237903 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 745.660416E+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 0.025 0.028 63.935 63.936 qs_mol_dyn_low 1 2.0 0.003 0.003 62.872 62.881 qs_forces 11 3.9 0.003 0.003 62.787 62.787 qs_energies 11 4.9 0.001 0.001 59.409 59.412 scf_env_do_scf 11 5.9 0.000 0.001 50.962 50.968 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.578 41.580 velocity_verlet 10 3.0 0.016 0.029 34.876 34.878 dbcsr_multiply_generic 2055 12.4 0.118 0.124 30.865 31.099 qs_scf_new_mos 99 7.5 0.001 0.001 27.265 27.386 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.265 27.385 ot_scf_mini 99 9.5 0.002 0.003 25.579 25.704 multiply_cannon 2055 13.4 0.240 0.261 23.899 24.892 multiply_cannon_loop 2055 14.4 1.408 1.492 22.348 23.005 ot_mini 99 10.5 0.001 0.001 14.804 14.966 multiply_cannon_multrec 24660 15.4 4.092 6.810 13.176 14.417 rebuild_ks_matrix 110 8.3 0.000 0.000 11.927 12.035 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.926 12.034 qs_ot_get_derivative 99 11.5 0.001 0.001 10.608 10.741 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.543 10.633 dbcsr_mm_accdrv_process 52304 16.0 7.700 8.863 8.927 9.869 init_scf_loop 11 6.9 0.027 0.044 9.344 9.349 prepare_preconditioner 11 7.9 0.000 0.000 7.520 7.557 make_preconditioner 11 8.9 0.000 0.000 7.520 7.557 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.357 7.142 sum_up_and_integrate 110 10.3 0.001 0.002 6.087 6.101 integrate_v_rspace 110 11.3 0.003 0.003 6.061 6.074 mp_waitall_1 121746 16.5 4.071 6.051 4.071 6.051 qs_ot_get_p 110 10.4 0.001 0.001 5.663 5.833 make_m2s 4110 13.4 0.059 0.061 5.287 5.773 init_scf_run 11 5.9 0.000 0.001 5.745 5.745 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.745 5.745 make_images 4110 14.4 0.577 0.699 5.149 5.631 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.561 5.571 calculate_rho_elec 110 8.6 0.077 0.081 5.561 5.571 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.079 4.175 apply_single 110 13.6 0.000 0.000 4.079 4.175 ot_diis_step 99 11.5 0.011 0.011 4.156 4.156 qs_ot_p2m_diag 48 11.0 0.054 0.062 3.780 3.796 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.594 3.666 grid_integrate_task_list 110 12.3 3.286 3.450 3.286 3.450 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.334 3.411 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.223 3.224 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.157 3.191 calculate_dm_sparse 110 9.5 0.001 0.001 3.116 3.155 hybrid_alltoall_any 4261 16.3 0.122 0.462 2.272 3.104 make_images_data 4110 15.4 0.049 0.053 2.579 3.063 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.412 3.025 wfi_extrapolate 11 7.9 0.001 0.001 2.921 2.921 cp_fm_upper_to_full 70 14.2 2.069 2.916 2.069 2.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.889 2.892 dbcsr_complete_redistribute 325 12.2 0.381 0.456 2.148 2.862 density_rs2pw 110 9.6 0.004 0.004 2.682 2.809 fft_wrap_pw1pw2_140 451 12.1 0.145 0.149 2.707 2.753 calculate_first_density_matrix 1 7.0 0.000 0.000 2.727 2.730 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.680 2.691 cp_fm_diag_elpa_base 48 14.0 2.615 2.639 2.677 2.687 acc_transpose_blocks 24660 15.4 0.112 0.115 2.466 2.600 multiply_cannon_sync_h2d 24660 15.4 2.393 2.570 2.393 2.570 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.491 2.527 fft3d_ps 1111 13.6 1.136 1.188 2.425 2.453 grid_collocate_task_list 110 9.6 2.266 2.404 2.266 2.404 cp_fm_cholesky_invert 11 10.9 2.393 2.403 2.393 2.403 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.599 2.286 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.222 2.223 mp_alltoall_d11v 2046 13.8 1.841 2.117 1.841 2.117 potential_pw2rs 110 12.3 0.012 0.012 1.894 1.898 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.154 1.854 mp_alltoall_i22 605 13.7 1.081 1.820 1.081 1.820 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.787 1.819 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.711 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.677 1.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.640 jit_kernel_multiply 8 15.9 0.877 1.609 0.877 1.609 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.841 1.554 acc_transpose_blocks_sync 73980 16.4 1.404 1.520 1.404 1.520 mp_sum_l 6594 12.7 0.980 1.514 0.980 1.514 mp_irecv_dv 62702 16.1 0.735 1.470 0.735 1.470 cp_fm_cholesky_decompose 22 10.9 1.343 1.399 1.343 1.399 mp_allgather_i34 2055 14.4 0.624 1.324 0.624 1.324 transfer_rs2pw 451 10.6 0.005 0.006 1.152 1.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=63.936000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=704.636364, yerr=11.941317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 859.987968E+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 0.023 0.039 54.520 54.521 qs_mol_dyn_low 1 2.0 0.003 0.003 54.182 54.192 qs_forces 11 3.9 0.003 0.003 54.069 54.070 qs_energies 11 4.9 0.001 0.001 50.417 50.421 scf_env_do_scf 11 5.9 0.000 0.001 42.201 42.201 scf_env_do_scf_inner_loop 99 6.5 0.006 0.027 35.016 35.016 velocity_verlet 10 3.0 0.002 0.002 30.174 30.180 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.642 23.769 qs_scf_new_mos 99 7.5 0.001 0.001 20.847 20.899 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.846 20.899 ot_scf_mini 99 9.5 0.002 0.002 19.588 19.622 multiply_cannon 2055 13.4 0.240 0.253 17.985 19.377 multiply_cannon_loop 2055 14.4 0.607 0.625 16.711 17.070 rebuild_ks_matrix 110 8.3 0.000 0.000 11.527 11.550 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.526 11.550 ot_mini 99 10.5 0.001 0.001 10.906 10.931 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.236 10.258 multiply_cannon_multrec 8220 15.4 3.184 4.441 7.731 8.610 mp_waitall_1 103326 16.6 5.946 7.508 5.946 7.508 qs_ot_get_derivative 99 11.5 0.001 0.001 7.224 7.259 init_scf_loop 11 6.9 0.035 0.044 7.138 7.139 sum_up_and_integrate 110 10.3 0.001 0.002 6.050 6.061 integrate_v_rspace 110 11.3 0.004 0.023 6.023 6.034 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.672 5.688 calculate_rho_elec 110 8.6 0.115 0.116 5.672 5.688 prepare_preconditioner 11 7.9 0.000 0.000 5.435 5.454 make_preconditioner 11 8.9 0.000 0.000 5.435 5.454 dbcsr_mm_accdrv_process 17442 15.9 3.203 4.114 4.407 5.307 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.037 5.121 init_scf_run 11 5.9 0.000 0.002 5.110 5.110 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.110 5.110 qs_ot_get_p 110 10.4 0.001 0.001 4.752 4.775 make_m2s 4110 13.4 0.038 0.039 4.245 4.538 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.041 4.432 make_images 4110 14.4 0.649 0.707 4.117 4.410 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.773 3.831 apply_single 110 13.6 0.000 0.000 3.773 3.830 ot_diis_step 99 11.5 0.012 0.012 3.659 3.659 grid_integrate_task_list 110 12.3 3.377 3.499 3.377 3.499 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.340 3.344 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.319 3.326 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.001 3.002 fft_wrap_pw1pw2_140 451 12.1 0.191 0.195 2.862 2.872 hybrid_alltoall_any 4261 16.3 0.200 0.858 2.274 2.799 density_rs2pw 110 9.6 0.004 0.004 2.670 2.749 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.686 2.686 make_images_data 4110 15.4 0.042 0.048 2.270 2.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.620 2.623 calculate_dm_sparse 110 9.5 0.001 0.001 2.566 2.601 cp_fm_cholesky_invert 11 10.9 2.516 2.524 2.516 2.524 wfi_extrapolate 11 7.9 0.001 0.001 2.513 2.513 calculate_first_density_matrix 1 7.0 0.000 0.000 2.503 2.504 fft3d_ps 1111 13.6 1.276 1.286 2.479 2.489 grid_collocate_task_list 110 9.6 2.369 2.489 2.369 2.489 multiply_cannon_sync_h2d 8220 15.4 2.378 2.486 2.378 2.486 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.442 2.450 cp_fm_diag_elpa_base 48 14.0 2.386 2.412 2.440 2.448 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.254 2.269 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.151 2.214 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.000 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.900 1.920 mp_alltoall_d11v 2046 13.8 1.724 1.874 1.724 1.874 potential_pw2rs 110 12.3 0.015 0.015 1.828 1.831 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.503 1.623 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.621 1.623 dbcsr_complete_redistribute 325 12.2 0.517 0.547 1.458 1.543 mp_allgather_i34 2055 14.4 0.443 1.534 0.443 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.516 1.526 acc_transpose_blocks 8220 15.4 0.040 0.041 1.453 1.497 cp_fm_cholesky_decompose 22 10.9 1.451 1.476 1.451 1.476 qs_create_task_list 11 7.9 0.012 0.019 1.229 1.332 generate_qs_task_list 11 8.9 0.376 0.443 1.217 1.313 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.818 1.246 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.220 1.241 jit_kernel_multiply 6 15.4 0.884 1.236 0.884 1.236 transfer_rs2pw 451 10.6 0.005 0.005 1.105 1.199 mp_waitany 9240 13.8 1.045 1.138 1.045 1.138 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.521000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.818182, yerr=11.368363 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 1.363386E+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 0.033 0.043 78.421 78.421 qs_mol_dyn_low 1 2.0 0.003 0.003 78.021 78.030 qs_forces 11 3.9 0.003 0.003 77.713 77.714 qs_energies 11 4.9 0.002 0.002 73.642 73.643 scf_env_do_scf 11 5.9 0.000 0.001 63.702 63.703 velocity_verlet 10 3.0 0.003 0.009 48.181 48.217 scf_env_do_scf_inner_loop 99 6.5 0.016 0.020 43.276 43.278 dbcsr_multiply_generic 2055 12.4 0.122 0.125 29.469 29.558 qs_scf_new_mos 99 7.5 0.001 0.001 26.289 26.322 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.288 26.322 ot_scf_mini 99 9.5 0.002 0.002 24.501 24.528 multiply_cannon 2055 13.4 0.339 0.374 22.433 23.228 multiply_cannon_loop 2055 14.4 0.826 0.838 20.640 20.954 init_scf_loop 11 6.9 0.009 0.009 20.356 20.361 prepare_preconditioner 11 7.9 0.000 0.000 18.379 18.386 make_preconditioner 11 8.9 0.000 0.000 18.379 18.386 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.537 17.895 ot_mini 99 10.5 0.001 0.001 13.902 13.932 rebuild_ks_matrix 110 8.3 0.000 0.001 13.481 13.515 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.481 13.515 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.243 12.274 cp_fm_upper_to_full 70 14.2 8.350 11.819 8.350 11.819 multiply_cannon_multrec 8220 15.4 4.054 4.244 9.767 9.911 qs_ot_get_derivative 99 11.5 0.001 0.001 9.339 9.369 mp_waitall_1 84994 16.7 7.376 8.139 7.376 8.139 dbcsr_complete_redistribute 325 12.2 0.939 0.959 5.367 7.477 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.826 6.861 calculate_rho_elec 110 8.6 0.227 0.227 6.825 6.861 sum_up_and_integrate 110 10.3 0.002 0.002 6.582 6.602 integrate_v_rspace 110 11.3 0.004 0.004 6.554 6.574 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.370 6.471 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.827 5.911 make_m2s 4110 13.4 0.043 0.044 5.329 5.823 dbcsr_mm_accdrv_process 11614 15.7 3.878 4.123 5.565 5.813 init_scf_run 11 5.9 0.000 0.001 5.715 5.715 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.715 5.715 make_images 4110 14.4 0.885 0.937 5.142 5.636 mp_alltoall_i22 605 13.7 3.456 5.579 3.456 5.579 qs_ot_get_p 110 10.4 0.001 0.001 5.318 5.353 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.708 5.171 apply_single 110 13.6 0.000 0.000 4.707 5.171 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.771 5.072 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.551 4.567 ot_diis_step 99 11.5 0.015 0.016 4.546 4.547 fft_wrap_pw1pw2_140 451 12.1 0.363 0.364 3.991 4.017 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.742 3.748 cp_fm_cholesky_invert 11 10.9 3.738 3.747 3.738 3.747 grid_integrate_task_list 110 12.3 3.674 3.741 3.674 3.741 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.646 3.647 hybrid_alltoall_any 4261 16.3 0.266 0.569 2.768 3.515 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.010 3.468 fft3d_ps 1111 13.6 1.839 1.895 3.412 3.427 make_images_data 4110 15.4 0.045 0.048 2.770 3.387 density_rs2pw 110 9.6 0.004 0.004 3.293 3.318 calculate_dm_sparse 110 9.5 0.001 0.001 3.267 3.296 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.227 3.228 wfi_extrapolate 11 7.9 0.001 0.001 3.184 3.184 multiply_cannon_sync_h2d 8220 15.4 3.124 3.140 3.124 3.140 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.910 2.913 grid_collocate_task_list 110 9.6 2.664 2.677 2.664 2.677 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.641 2.641 cp_fm_diag_elpa_base 48 14.0 2.338 2.414 2.639 2.639 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.623 2.637 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.420 2.431 calculate_first_density_matrix 1 7.0 0.000 0.000 2.419 2.422 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.169 2.213 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.179 potential_pw2rs 110 12.3 0.021 0.021 2.128 2.133 mp_alltoall_d11v 2046 13.8 2.012 2.073 2.012 2.073 qs_create_task_list 11 7.9 0.001 0.001 1.879 1.924 generate_qs_task_list 11 8.9 0.732 0.786 1.878 1.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.842 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.780 1.800 acc_transpose_blocks 8220 15.4 0.041 0.042 1.778 1.797 cp_fm_cholesky_decompose 22 10.9 1.736 1.771 1.736 1.771 jit_kernel_multiply 10 15.1 1.488 1.764 1.488 1.764 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.701 1.721 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.619 1.625 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=78.421000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1237.272727, yerr=48.781686 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.259136E+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 1583009. 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 0.022 0.071 206.089 206.091 qs_mol_dyn_low 1 2.0 0.014 0.060 205.442 205.456 qs_forces 11 3.9 0.004 0.005 204.244 204.245 qs_energies 11 4.9 0.002 0.002 198.477 198.498 scf_env_do_scf 11 5.9 0.001 0.001 180.934 180.951 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.293 162.295 dbcsr_multiply_generic 2507 12.6 0.173 0.177 126.155 127.024 qs_scf_new_mos 117 7.6 0.001 0.001 123.391 123.735 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.391 123.734 velocity_verlet 10 3.0 0.034 0.059 121.936 121.953 ot_scf_mini 117 9.6 0.003 0.003 116.677 116.995 multiply_cannon 2507 13.6 0.237 0.246 102.185 104.518 multiply_cannon_loop 2507 14.6 2.381 2.442 100.091 102.557 ot_mini 117 10.6 0.001 0.001 66.314 66.676 multiply_cannon_multrec 60168 15.6 32.030 34.372 41.922 43.893 qs_ot_get_derivative 117 11.6 0.001 0.001 41.118 41.468 rebuild_ks_matrix 128 8.3 0.001 0.001 33.629 33.862 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.629 33.861 mp_waitall_1 267128 16.5 28.916 32.345 28.916 32.345 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.221 30.441 multiply_cannon_sync_h2d 60168 15.6 26.545 28.860 26.545 28.860 qs_ot_get_p 128 10.4 0.001 0.003 28.107 28.427 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.666 25.998 apply_single 128 13.6 0.001 0.001 24.666 25.998 ot_diis_step 117 11.6 0.008 0.008 24.951 24.952 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.563 21.655 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.576 19.840 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.722 18.723 init_scf_loop 11 6.9 0.000 0.001 18.561 18.562 multiply_cannon_metrocomm3 60168 15.6 0.121 0.124 16.253 18.184 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.683 15.723 cp_fm_diag_elpa_base 83 14.4 15.634 15.672 15.676 15.714 make_m2s 5014 13.6 0.105 0.114 13.879 14.276 make_images 5014 14.6 0.404 0.423 13.699 14.105 prepare_preconditioner 11 7.9 0.000 0.000 13.955 13.992 make_preconditioner 11 8.9 0.000 0.000 13.955 13.992 sum_up_and_integrate 128 10.3 0.002 0.004 13.627 13.644 integrate_v_rspace 128 11.3 0.003 0.004 13.568 13.587 init_scf_run 11 5.9 0.000 0.001 13.475 13.476 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.475 13.476 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.230 13.384 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.521 12.645 calculate_rho_elec 128 8.7 0.045 0.064 12.520 12.645 mp_sum_l 7950 12.9 8.884 10.211 8.884 10.211 dbcsr_mm_accdrv_process 124484 16.2 4.696 4.851 9.459 9.959 wfi_extrapolate 11 7.9 0.001 0.001 8.874 8.874 calculate_dm_sparse 128 9.5 0.001 0.001 8.663 8.734 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.066 8.231 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.023 8.118 cp_fm_cholesky_invert 11 10.9 7.916 7.934 7.916 7.934 make_images_data 5014 15.6 0.068 0.075 6.768 7.735 multiply_cannon_metrocomm1 60168 15.6 0.094 0.099 6.284 7.717 grid_integrate_task_list 128 12.3 7.012 7.419 7.012 7.419 hybrid_alltoall_any 5200 16.5 0.297 2.267 5.948 7.210 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.056 7.068 density_rs2pw 128 9.7 0.005 0.006 6.131 6.914 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.390 6.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.007 6.098 fft_wrap_pw1pw2_140 523 12.2 0.234 0.244 5.563 5.729 mp_alltoall_d11v 2415 14.1 4.452 5.615 4.452 5.615 fft3d_ps 1291 13.7 2.093 2.603 5.209 5.430 grid_collocate_task_list 128 9.7 4.845 5.211 4.845 5.211 mp_sum_d 4469 12.1 3.699 4.662 3.699 4.662 calculate_first_density_matrix 1 7.0 0.001 0.010 4.414 4.425 potential_pw2rs 128 12.3 0.009 0.010 4.140 4.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.091000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.818182, yerr=7.107207 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 846.557184E+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 4027 67357. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5459774. 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 0.039 0.063 182.786 182.788 qs_mol_dyn_low 1 2.0 0.008 0.043 182.339 182.381 qs_forces 11 3.9 0.010 0.056 181.883 181.884 qs_energies 11 4.9 0.002 0.012 175.110 175.123 scf_env_do_scf 11 5.9 0.001 0.001 157.961 157.977 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.972 131.975 velocity_verlet 10 3.0 0.010 0.073 112.462 112.474 dbcsr_multiply_generic 2507 12.6 0.186 0.192 97.636 98.931 qs_scf_new_mos 117 7.6 0.001 0.001 93.814 94.328 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.813 94.327 ot_scf_mini 117 9.6 0.004 0.004 89.015 89.574 multiply_cannon 2507 13.6 0.507 0.568 77.508 82.612 multiply_cannon_loop 2507 14.6 1.564 1.624 74.097 76.712 ot_mini 117 10.6 0.001 0.001 50.325 50.829 mp_waitall_1 214728 16.6 24.078 38.472 24.078 38.472 multiply_cannon_multrec 30084 15.6 20.862 26.342 31.634 37.560 rebuild_ks_matrix 128 8.3 0.001 0.001 32.344 33.081 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.343 33.080 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.123 29.790 qs_ot_get_derivative 117 11.6 0.001 0.002 28.306 28.871 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.748 28.435 init_scf_loop 11 6.9 0.000 0.001 25.895 25.896 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.083 23.302 apply_single 128 13.6 0.001 0.001 22.082 23.302 ot_diis_step 117 11.6 0.014 0.015 21.837 21.839 prepare_preconditioner 11 7.9 0.000 0.000 21.526 21.609 make_preconditioner 11 8.9 0.000 0.000 21.526 21.609 qs_ot_get_p 128 10.4 0.001 0.001 20.648 21.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.420 20.801 multiply_cannon_sync_h2d 30084 15.6 17.956 20.029 17.956 20.029 make_m2s 5014 13.6 0.088 0.093 14.177 15.913 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.755 15.787 make_images 5014 14.6 1.169 1.382 13.965 15.698 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.560 14.561 sum_up_and_integrate 128 10.3 0.002 0.004 13.619 13.646 integrate_v_rspace 128 11.3 0.004 0.009 13.559 13.587 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.873 12.945 calculate_rho_elec 128 8.7 0.088 0.106 12.873 12.944 init_scf_run 11 5.9 0.000 0.001 12.100 12.101 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.099 12.101 cp_fm_cholesky_invert 11 10.9 11.875 11.909 11.875 11.909 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.275 11.667 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.154 11.187 cp_fm_diag_elpa_base 83 14.4 10.979 11.018 11.143 11.172 multiply_cannon_metrocomm4 27577 15.6 0.107 0.122 3.903 10.949 dbcsr_mm_accdrv_process 62242 16.2 5.486 6.484 10.216 10.789 mp_irecv_dv 69486 16.3 3.687 10.533 3.687 10.533 make_images_data 5014 15.6 0.067 0.077 8.378 10.388 hybrid_alltoall_any 5200 16.5 0.352 1.552 7.305 10.134 wfi_extrapolate 11 7.9 0.001 0.003 8.210 8.210 grid_integrate_task_list 128 12.3 7.172 7.468 7.172 7.468 fft_wrap_pw1pw2 1291 11.7 0.019 0.023 7.286 7.382 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.297 7.114 density_rs2pw 128 9.7 0.006 0.006 6.432 6.931 calculate_dm_sparse 128 9.5 0.001 0.001 6.646 6.813 fft_wrap_pw1pw2_140 523 12.2 0.249 0.257 6.419 6.504 mp_sum_l 7950 12.9 4.159 6.486 4.159 6.486 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.288 6.298 fft3d_ps 1291 13.7 2.764 2.935 5.739 5.808 mp_alltoall_d11v 2415 14.1 4.744 5.640 4.744 5.640 cp_fm_cholesky_decompose 22 10.9 5.535 5.599 5.535 5.599 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.291 5.504 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.371 5.449 grid_collocate_task_list 128 9.7 5.004 5.331 5.004 5.331 mp_allgather_i34 2507 14.6 1.919 5.014 1.919 5.014 potential_pw2rs 128 12.3 0.015 0.017 4.120 4.134 mp_sum_d 4483 12.1 2.706 4.107 2.706 4.107 dbcsr_complete_redistribute 395 12.7 0.717 0.880 3.192 3.867 calculate_first_density_matrix 1 7.0 0.000 0.001 3.701 3.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.633 3.699 qs_energies_init_hamiltonians 11 5.9 0.006 0.027 3.664 3.665 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=182.788000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=807.181818, yerr=0.833196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.037435E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67815. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.088 167.809 167.810 qs_mol_dyn_low 1 2.0 0.037 0.150 167.208 167.221 qs_forces 11 3.9 0.043 0.047 166.930 166.965 qs_energies 11 4.9 0.007 0.048 160.403 160.441 scf_env_do_scf 11 5.9 0.001 0.003 144.163 144.164 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 115.643 115.643 velocity_verlet 10 3.0 0.019 0.065 105.004 105.025 dbcsr_multiply_generic 2507 12.6 0.189 0.241 81.507 82.542 qs_scf_new_mos 117 7.6 0.001 0.001 79.394 79.858 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.394 79.857 ot_scf_mini 117 9.6 0.004 0.005 75.236 75.719 multiply_cannon 2507 13.6 0.508 0.529 61.926 66.475 multiply_cannon_loop 2507 14.6 1.131 1.193 59.282 61.805 ot_mini 117 10.6 0.001 0.001 42.317 42.805 mp_waitall_1 170520 16.6 24.527 33.931 24.527 33.931 rebuild_ks_matrix 128 8.3 0.001 0.001 30.040 30.655 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.047 30.039 30.655 init_scf_loop 11 6.9 0.010 0.033 28.419 28.420 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.121 27.681 multiply_cannon_multrec 20056 15.6 13.010 16.327 22.514 25.866 prepare_preconditioner 11 7.9 0.000 0.000 24.440 24.498 make_preconditioner 11 8.9 0.000 0.001 24.440 24.498 multiply_cannon_metrocomm3 20056 15.6 0.061 0.064 15.332 24.214 make_full_inverse_cholesky 11 9.9 0.004 0.030 22.472 23.443 qs_ot_get_derivative 117 11.6 0.001 0.002 22.729 23.213 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.751 20.718 apply_single 128 13.6 0.001 0.001 19.750 20.718 ot_diis_step 117 11.6 0.018 0.018 19.481 19.481 qs_ot_get_p 128 10.4 0.001 0.002 18.356 18.904 make_m2s 5014 13.6 0.079 0.085 14.583 15.439 make_images 5014 14.6 1.136 1.222 14.351 15.207 multiply_cannon_sync_h2d 20056 15.6 13.619 15.173 13.619 15.173 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.168 14.177 sum_up_and_integrate 128 10.3 0.002 0.003 13.563 13.590 integrate_v_rspace 128 11.3 0.003 0.004 13.504 13.529 cp_dbcsr_syevd 83 12.4 0.006 0.007 13.123 13.123 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.787 12.826 calculate_rho_elec 128 8.7 0.132 0.146 12.787 12.825 cp_fm_cholesky_invert 11 10.9 11.658 11.678 11.658 11.678 init_scf_run 11 5.9 0.000 0.001 10.969 10.969 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.968 10.969 make_images_data 5014 15.6 0.063 0.073 8.731 9.984 cp_fm_diag_elpa 83 13.4 0.001 0.001 9.761 9.779 cp_fm_diag_elpa_base 83 14.4 9.515 9.585 9.756 9.774 multiply_cannon_metrocomm4 17549 15.6 0.067 0.078 3.524 9.625 hybrid_alltoall_any 5200 16.5 0.452 2.074 7.615 9.370 mp_irecv_dv 50230 16.2 3.390 9.360 3.390 9.360 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.953 9.317 dbcsr_mm_accdrv_process 41502 16.2 5.808 6.116 8.959 9.126 grid_integrate_task_list 128 12.3 7.313 7.820 7.313 7.820 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.400 7.501 wfi_extrapolate 11 7.9 0.002 0.012 7.273 7.273 density_rs2pw 128 9.7 0.005 0.006 6.272 6.677 fft_wrap_pw1pw2_140 523 12.2 0.285 0.292 6.540 6.656 calculate_dm_sparse 128 9.5 0.001 0.001 5.873 5.962 cp_fm_cholesky_decompose 22 10.9 5.858 5.909 5.858 5.909 fft3d_ps 1291 13.7 2.758 2.979 5.706 5.776 grid_collocate_task_list 128 9.7 5.163 5.626 5.163 5.626 dbcsr_complete_redistribute 395 12.7 1.032 1.095 4.390 5.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.527 5.535 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.571 5.232 cp_fm_upper_to_full 105 14.8 3.913 4.940 3.913 4.940 mp_alltoall_d11v 2415 14.1 4.512 4.921 4.512 4.921 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.591 4.758 mp_sum_l 7950 12.9 3.187 4.498 3.187 4.498 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.224 4.361 mp_allgather_i34 2507 14.6 1.136 4.173 1.136 4.173 potential_pw2rs 128 12.3 0.020 0.022 4.036 4.047 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.959 3.992 qs_energies_init_hamiltonians 11 5.9 0.001 0.008 3.885 3.915 calculate_first_density_matrix 1 7.0 0.001 0.005 3.507 3.509 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.428 3.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=167.810000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=986.090909, yerr=5.333506 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 4.320340E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 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 1.099092E+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 660837893360 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 12496381. 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 0.028 0.042 176.611 176.611 qs_mol_dyn_low 1 2.0 0.003 0.005 176.128 176.142 qs_forces 11 3.9 0.004 0.004 175.983 175.989 qs_energies 11 4.9 0.008 0.010 168.890 168.899 scf_env_do_scf 11 5.9 0.001 0.001 151.311 151.328 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 114.396 114.397 velocity_verlet 10 3.0 0.002 0.002 113.994 113.998 dbcsr_multiply_generic 2485 12.5 0.187 0.192 80.306 80.896 qs_scf_new_mos 116 7.6 0.001 0.001 79.083 79.374 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.082 79.373 ot_scf_mini 116 9.6 0.003 0.003 74.568 74.852 multiply_cannon 2485 13.5 0.566 0.604 55.577 58.529 multiply_cannon_loop 2485 14.5 1.821 1.915 51.958 53.318 ot_mini 116 10.6 0.001 0.001 42.583 42.856 init_scf_loop 11 6.9 0.000 0.001 36.788 36.789 prepare_preconditioner 11 7.9 0.000 0.000 32.726 32.751 make_preconditioner 11 8.9 0.000 0.000 32.726 32.751 make_full_inverse_cholesky 11 9.9 0.017 0.021 27.956 31.525 multiply_cannon_multrec 29820 15.5 13.607 18.780 26.673 31.417 rebuild_ks_matrix 127 8.3 0.001 0.001 29.031 29.243 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.030 29.242 mp_waitall_1 146592 16.7 16.913 26.378 16.913 26.378 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.185 26.371 qs_ot_get_derivative 116 11.6 0.001 0.002 22.889 23.179 make_m2s 4970 13.5 0.092 0.095 20.453 21.403 make_images 4970 14.5 1.987 2.214 20.158 21.108 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.241 19.755 apply_single 127 13.6 0.001 0.001 19.241 19.755 ot_diis_step 116 11.6 0.017 0.019 19.571 19.572 qs_ot_get_p 127 10.4 0.002 0.002 18.163 18.470 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.195 14.718 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.099 14.154 sum_up_and_integrate 127 10.3 0.002 0.003 13.513 13.540 integrate_v_rspace 127 11.3 0.003 0.004 13.454 13.482 dbcsr_mm_accdrv_process 61748 16.2 8.486 9.375 12.636 13.185 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.956 13.000 calculate_rho_elec 127 8.7 0.173 0.188 12.955 13.000 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.710 12.711 cp_fm_upper_to_full 104 14.8 8.722 12.413 8.722 12.413 make_images_data 4970 15.5 0.065 0.071 10.928 12.409 cp_fm_cholesky_invert 11 10.9 11.982 12.004 11.982 12.004 hybrid_alltoall_any 5155 16.4 0.528 2.210 9.838 11.794 init_scf_run 11 5.9 0.002 0.037 11.579 11.580 scf_env_initial_rho_setup 11 6.9 0.030 0.037 11.577 11.580 multiply_cannon_sync_h2d 29820 15.5 10.369 10.924 10.369 10.924 dbcsr_complete_redistribute 393 12.7 1.409 1.494 7.544 10.388 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.534 9.733 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.516 9.529 cp_fm_diag_elpa_base 82 14.4 9.040 9.168 9.509 9.520 copy_fm_to_dbcsr 208 11.6 0.001 0.002 6.152 8.977 grid_integrate_task_list 127 12.3 7.443 7.898 7.443 7.898 fft_wrap_pw1pw2 1281 11.7 0.018 0.023 7.622 7.677 wfi_extrapolate 11 7.9 0.001 0.001 7.508 7.508 transfer_fm_to_dbcsr 11 9.9 0.003 0.011 4.748 7.497 multiply_cannon_metrocomm4 24850 15.5 0.079 0.090 2.762 7.121 mp_alltoall_i22 712 14.1 4.115 6.911 4.115 6.911 mp_irecv_dv 75445 16.2 2.605 6.839 2.605 6.839 fft_wrap_pw1pw2_140 519 12.2 0.319 0.328 6.749 6.827 density_rs2pw 127 9.7 0.005 0.005 6.120 6.528 calculate_dm_sparse 127 9.5 0.001 0.001 6.311 6.390 cp_fm_cholesky_decompose 22 10.9 5.967 6.016 5.967 6.016 fft3d_ps 1281 13.7 2.903 2.955 5.836 5.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.641 5.710 grid_collocate_task_list 127 9.7 5.242 5.625 5.242 5.625 mp_alltoall_d11v 2401 14.1 4.871 5.603 4.871 5.603 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.503 4.566 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.417 4.499 qs_energies_init_hamiltonians 11 5.9 0.004 0.004 4.455 4.455 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.225 4.296 potential_pw2rs 127 12.3 0.022 0.022 3.890 3.900 calculate_first_density_matrix 1 7.0 0.000 0.000 3.842 3.873 dbcsr_special_finalize 12425 15.5 0.066 0.068 3.509 3.647 mp_allgather_i34 2485 14.5 1.696 3.603 1.696 3.603 dbcsr_merge_single_wm 12425 16.5 0.323 0.339 3.418 3.556 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.230 3.536 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=176.611000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1042.272727, yerr=12.225836 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.435787E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 755071. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67810. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.078 164.303 164.303 qs_mol_dyn_low 1 2.0 0.003 0.003 163.711 163.739 qs_forces 11 3.9 0.004 0.004 163.605 163.612 qs_energies 11 4.9 0.003 0.004 156.232 156.239 scf_env_do_scf 11 5.9 0.001 0.001 138.376 138.385 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 107.706 107.708 velocity_verlet 10 3.0 0.017 0.018 106.006 106.010 dbcsr_multiply_generic 2507 12.6 0.184 0.191 72.109 72.474 qs_scf_new_mos 117 7.6 0.001 0.001 71.576 71.661 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.575 71.661 ot_scf_mini 117 9.6 0.003 0.004 67.171 67.263 multiply_cannon 2507 13.6 0.558 0.574 53.729 56.598 multiply_cannon_loop 2507 14.6 0.811 0.845 50.656 51.471 ot_mini 117 10.6 0.001 0.001 37.537 37.633 mp_waitall_1 125778 16.7 24.593 30.789 24.593 30.789 init_scf_loop 11 6.9 0.001 0.001 30.512 30.514 rebuild_ks_matrix 128 8.3 0.001 0.001 28.940 29.044 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.939 29.044 prepare_preconditioner 11 7.9 0.000 0.000 26.689 26.701 make_preconditioner 11 8.9 0.000 0.000 26.689 26.701 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.393 26.489 make_full_inverse_cholesky 11 9.9 0.022 0.028 24.812 25.066 multiply_cannon_multrec 10028 15.6 10.346 15.911 18.100 22.179 qs_ot_get_derivative 117 11.6 0.002 0.002 20.356 20.437 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 12.904 20.111 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.259 17.429 apply_single 128 13.6 0.001 0.001 17.259 17.428 ot_diis_step 117 11.6 0.020 0.020 17.106 17.106 qs_ot_get_p 128 10.4 0.002 0.002 16.958 17.079 make_m2s 5014 13.6 0.063 0.068 14.512 15.418 make_images 5014 14.6 2.172 2.594 14.207 15.109 cp_fm_cholesky_invert 11 10.9 14.709 14.720 14.709 14.720 sum_up_and_integrate 128 10.3 0.002 0.002 13.773 13.815 integrate_v_rspace 128 11.3 0.004 0.004 13.712 13.756 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.422 13.485 calculate_rho_elec 128 8.7 0.257 0.268 13.421 13.484 qs_ot_p2m_diag 83 11.4 0.496 0.503 13.276 13.293 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.033 12.071 multiply_cannon_sync_h2d 10028 15.6 10.739 11.102 10.739 11.102 init_scf_run 11 5.9 0.000 0.004 11.010 11.010 scf_env_initial_rho_setup 11 6.9 0.003 0.004 11.010 11.010 make_images_data 5014 15.6 0.057 0.067 8.248 9.645 hybrid_alltoall_any 5200 16.5 0.848 3.799 8.127 9.302 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.122 9.133 cp_fm_diag_elpa_base 83 14.4 8.887 8.954 9.117 9.129 dbcsr_mm_accdrv_process 20762 16.1 3.273 4.367 7.402 8.151 fft_wrap_pw1pw2 1291 11.7 0.017 0.019 8.047 8.084 grid_integrate_task_list 128 12.3 7.786 7.998 7.786 7.998 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.939 7.993 fft_wrap_pw1pw2_140 523 12.2 0.435 0.441 7.097 7.142 wfi_extrapolate 11 7.9 0.001 0.001 7.108 7.108 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.213 7.038 density_rs2pw 128 9.7 0.005 0.005 6.142 6.613 cp_fm_cholesky_decompose 22 10.9 6.474 6.518 6.474 6.518 calculate_dm_sparse 128 9.5 0.001 0.001 6.178 6.265 fft3d_ps 1291 13.7 3.132 3.259 6.005 6.023 grid_collocate_task_list 128 9.7 5.575 5.957 5.575 5.957 multiply_cannon_metrocomm4 7521 15.6 0.025 0.029 1.865 5.635 mp_irecv_dv 28860 15.9 1.823 5.542 1.823 5.542 dbcsr_complete_redistribute 395 12.7 2.008 2.043 5.100 5.389 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.290 5.330 qs_energies_init_hamiltonians 11 5.9 0.010 0.014 5.311 5.311 mp_alltoall_d11v 2415 14.1 4.571 4.974 4.571 4.974 mp_allgather_i34 2507 14.6 1.252 4.479 1.252 4.479 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.154 4.211 potential_pw2rs 128 12.3 0.025 0.026 3.942 3.950 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.581 3.873 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.584 3.677 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.628 3.675 calculate_first_density_matrix 1 7.0 0.000 0.000 3.645 3.651 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.472 3.557 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.243 3.544 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.451 3.458 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=164.303000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1343.636364, yerr=28.620631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 2.938794E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 67902. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.162 248.958 248.959 qs_mol_dyn_low 1 2.0 0.003 0.003 248.203 248.260 qs_forces 11 3.9 0.005 0.005 247.035 247.037 qs_energies 11 4.9 0.002 0.002 238.269 238.275 scf_env_do_scf 11 5.9 0.001 0.001 215.042 215.055 velocity_verlet 10 3.0 0.006 0.029 172.939 173.078 scf_env_do_scf_inner_loop 118 6.6 0.012 0.016 131.513 131.516 qs_scf_new_mos 118 7.6 0.001 0.001 88.568 88.630 qs_scf_loop_do_ot 118 8.6 0.001 0.001 88.568 88.629 ot_scf_mini 118 9.6 0.003 0.004 83.534 83.574 dbcsr_multiply_generic 2534 12.6 0.218 0.227 82.988 83.443 init_scf_loop 11 6.9 0.001 0.001 83.219 83.224 prepare_preconditioner 11 7.9 0.000 0.000 78.504 78.515 make_preconditioner 11 8.9 0.000 0.000 78.504 78.515 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.826 75.749 multiply_cannon 2534 13.6 0.685 0.712 59.500 60.938 multiply_cannon_loop 2534 14.6 1.057 1.087 55.450 56.778 cp_fm_upper_to_full 106 14.8 34.374 49.453 34.374 49.453 ot_mini 118 10.6 0.001 0.001 44.592 44.650 rebuild_ks_matrix 129 8.3 0.001 0.001 33.837 33.873 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.836 33.873 mp_waitall_1 104780 16.8 28.096 32.347 28.096 32.347 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.506 31.542 dbcsr_complete_redistribute 397 12.7 3.836 3.945 22.794 31.362 copy_fm_to_dbcsr 210 11.7 0.001 0.002 19.163 27.608 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 16.635 24.962 qs_ot_get_derivative 118 11.6 0.002 0.002 24.616 24.653 qs_ot_get_p 129 10.4 0.002 0.002 23.951 24.038 mp_alltoall_i22 720 14.1 14.376 22.991 14.376 22.991 make_m2s 5068 13.6 0.075 0.078 18.820 20.278 multiply_cannon_metrocomm3 10136 15.6 0.025 0.026 18.827 20.182 ot_diis_step 118 11.6 0.022 0.022 19.936 19.936 qs_ot_p2m_diag 84 11.4 0.890 0.894 19.834 19.866 make_images 5068 14.6 3.134 3.349 18.338 19.798 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.208 19.330 apply_single 129 13.6 0.001 0.001 19.208 19.329 multiply_cannon_multrec 10136 15.6 10.471 12.373 18.549 18.619 cp_fm_cholesky_invert 11 10.9 18.369 18.381 18.369 18.381 cp_dbcsr_syevd 84 12.4 0.006 0.006 18.022 18.023 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.854 15.916 calculate_rho_elec 129 8.7 0.486 0.486 15.853 15.915 sum_up_and_integrate 129 10.3 0.002 0.002 15.748 15.835 integrate_v_rspace 129 11.3 0.004 0.005 15.684 15.771 cp_fm_diag_elpa 84 13.4 0.000 0.000 14.787 14.789 cp_fm_diag_elpa_base 84 14.4 11.918 12.944 14.783 14.783 multiply_cannon_sync_h2d 10136 15.6 14.418 14.450 14.418 14.450 hybrid_alltoall_any 5255 16.5 1.334 3.128 10.799 12.902 init_scf_run 11 5.9 0.000 0.001 12.802 12.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.801 12.802 make_images_data 5068 15.6 0.064 0.068 10.587 12.738 fft_wrap_pw1pw2 1301 11.7 0.020 0.020 10.031 10.046 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.827 9.853 dbcsr_mm_accdrv_process 20954 16.1 4.374 6.327 7.822 9.764 fft_wrap_pw1pw2_140 527 12.2 0.830 0.833 8.877 8.903 wfi_extrapolate 11 7.9 0.001 0.001 8.847 8.847 grid_integrate_task_list 129 12.3 8.614 8.811 8.614 8.811 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.181 8.182 cp_fm_cholesky_decompose 22 10.9 7.393 7.450 7.393 7.450 fft3d_ps 1301 13.7 4.011 4.033 7.298 7.318 mp_alltoall_d11v 2429 14.1 7.124 7.240 7.124 7.240 calculate_dm_sparse 129 9.5 0.001 0.001 7.097 7.166 density_rs2pw 129 9.7 0.005 0.005 6.890 6.946 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.524 6.580 grid_collocate_task_list 129 9.7 6.457 6.539 6.457 6.539 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.294 6.438 qs_env_update_s_mstruct 11 6.9 0.001 0.001 5.042 5.280 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.591 4.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=248.959000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2634.000000, yerr=151.647318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 1.262871E+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 0.017 0.036 85.792 85.794 qs_energies 1 2.0 0.000 0.000 85.209 85.221 ls_scf 1 3.0 0.002 0.017 84.195 84.206 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.975 73.165 multiply_cannon 111 7.7 0.018 0.020 56.038 57.044 multiply_cannon_loop 111 8.7 0.231 0.249 52.606 53.788 ls_scf_main 1 4.0 0.000 0.002 52.650 52.652 density_matrix_trs4 2 5.0 0.002 0.003 47.074 47.149 ls_scf_init_scf 1 4.0 0.000 0.000 28.473 28.474 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.261 27.314 mp_waitall_1 11031 10.9 22.499 26.021 22.499 26.021 matrix_sqrt_Newton_Schulz 2 6.5 0.006 0.026 25.149 25.169 multiply_cannon_multrec 2664 9.7 8.172 9.005 15.630 17.430 multiply_cannon_sync_h2d 2664 9.7 13.623 15.161 13.623 15.161 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 9.704 13.800 make_m2s 222 7.7 0.009 0.013 13.257 13.721 make_images 222 8.7 0.101 0.111 13.235 13.700 make_images_data 222 9.7 0.005 0.006 7.771 8.352 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.374 8.247 hybrid_alltoall_any 227 10.6 0.219 1.849 6.677 8.202 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.707 7.072 8.083 dbcsr_mm_accdrv_process_sort 4760 11.4 6.280 7.202 6.280 7.202 calculate_norms 4752 9.8 5.532 6.175 5.532 6.175 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.053 5.272 mp_sum_l 887 5.1 3.135 4.878 3.135 4.878 make_images_sizes 222 9.7 0.000 0.000 0.765 3.771 mp_alltoall_i44 222 10.7 0.765 3.771 0.765 3.771 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.336 3.670 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.474 mp_irecv_dv 6231 10.9 2.039 3.440 2.039 3.440 arnoldi_extremal 4 6.8 0.000 0.000 3.264 3.301 arnoldi_normal_ev 4 7.8 0.005 0.024 3.264 3.301 build_subspace 16 8.4 0.009 0.012 3.138 3.140 ls_scf_post 1 4.0 0.000 0.001 3.070 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.913 dbcsr_special_finalize 555 9.7 0.006 0.006 2.353 2.786 dbcsr_merge_single_wm 555 10.7 0.456 0.573 2.344 2.777 make_images_pack 222 9.7 2.218 2.626 2.220 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.341 2.560 dbcsr_sort_data 658 11.4 2.148 2.529 2.148 2.529 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.452 2.069 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.310 2.394 buffer_matrices_ensure_size 222 8.7 1.749 2.042 1.749 2.042 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.843 1.845 rebuild_ks_matrix 3 7.3 0.000 0.000 1.833 1.835 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.833 1.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.794000, 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/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 2.195812E+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 0.025 0.076 89.779 89.780 qs_energies 1 2.0 0.000 0.000 89.290 89.294 ls_scf 1 3.0 0.000 0.000 87.949 87.952 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.079 74.384 multiply_cannon 111 7.7 0.029 0.042 52.693 56.807 ls_scf_main 1 4.0 0.000 0.000 54.155 54.160 multiply_cannon_loop 111 8.7 0.135 0.147 49.950 53.096 density_matrix_trs4 2 5.0 0.002 0.003 48.417 48.653 ls_scf_init_scf 1 4.0 0.000 0.000 30.092 30.094 mp_waitall_1 9105 10.9 20.803 29.998 20.803 29.998 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.913 29.000 multiply_cannon_multrec 1332 9.7 13.281 17.225 22.607 27.730 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.439 26.448 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.521 20.764 make_m2s 222 7.7 0.007 0.008 15.074 15.622 make_images 222 8.7 1.377 1.700 15.044 15.593 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.538 8.923 10.514 dbcsr_mm_accdrv_process_sort 4041 11.4 8.440 9.976 8.440 9.976 make_images_data 222 9.7 0.004 0.005 8.736 9.605 hybrid_alltoall_any 227 10.6 0.542 2.585 8.123 9.252 mp_sum_l 887 5.1 4.853 7.901 4.853 7.901 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.191 7.822 mp_irecv_dv 3311 11.0 3.170 7.769 3.170 7.769 calculate_norms 2376 9.8 6.073 6.805 6.073 6.805 multiply_cannon_sync_h2d 1332 9.7 4.857 6.202 4.857 6.202 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.626 6.088 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.072 5.332 arnoldi_extremal 4 6.8 0.000 0.000 4.676 4.694 arnoldi_normal_ev 4 7.8 0.057 0.070 4.676 4.694 build_subspace 16 8.4 0.014 0.021 4.366 4.369 ls_scf_post 1 4.0 0.000 0.000 3.702 3.705 ls_scf_store_result 1 5.0 0.000 0.000 3.409 3.525 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.138 3.380 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.224 2.746 3.226 mp_allgather_i34 111 8.7 0.717 2.977 0.717 2.977 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.193 2.848 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.622 2.742 dbcsr_data_new 4174 10.1 2.127 2.425 2.127 2.425 make_images_pack 222 9.7 1.862 2.206 1.864 2.208 dbcsr_sort_data 436 11.2 1.841 2.076 1.841 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.881 1.883 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.780000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1812.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 2.864386E+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 0.111 0.157 94.119 94.121 qs_energies 1 2.0 0.001 0.004 93.144 93.160 ls_scf 1 3.0 0.006 0.048 91.551 91.563 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.050 76.313 ls_scf_main 1 4.0 0.000 0.008 57.136 57.148 multiply_cannon 111 7.7 0.055 0.102 52.613 56.541 multiply_cannon_loop 111 8.7 0.119 0.142 49.773 53.840 density_matrix_trs4 2 5.0 0.002 0.006 51.176 51.390 mp_waitall_1 7281 11.0 23.897 33.698 23.897 33.698 ls_scf_init_scf 1 4.0 0.001 0.008 30.707 30.732 ls_scf_init_matrix_S 1 5.0 0.001 0.004 29.371 29.464 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.844 26.857 multiply_cannon_multrec 888 9.7 12.674 15.086 21.349 24.694 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.294 23.497 make_m2s 222 7.7 0.006 0.007 16.455 17.266 make_images 222 8.7 1.589 1.854 16.418 17.225 make_images_data 222 9.7 0.004 0.005 9.560 10.706 hybrid_alltoall_any 227 10.6 0.642 2.956 9.094 10.343 dbcsr_mm_accdrv_process 3754 10.4 0.328 0.495 8.180 9.412 dbcsr_mm_accdrv_process_sort 3754 11.4 7.724 8.917 7.724 8.917 mp_sum_l 887 5.1 5.332 8.887 5.332 8.887 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.446 7.527 mp_irecv_dv 2335 11.1 2.431 7.479 2.431 7.479 multiply_cannon_sync_h2d 888 9.7 6.103 7.438 6.103 7.438 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.154 7.166 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.691 6.989 arnoldi_extremal 4 6.8 0.000 0.000 5.057 5.074 arnoldi_normal_ev 4 7.8 0.009 0.021 5.057 5.074 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.849 5.050 build_subspace 16 8.4 0.014 0.020 4.751 4.757 calculate_norms 1584 9.8 4.386 4.740 4.386 4.740 mp_allgather_i34 111 8.7 0.920 3.924 0.920 3.924 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.435 3.774 ls_scf_post 1 4.0 0.000 0.004 3.703 3.710 dbcsr_matrix_vector_mult_local 304 10.0 3.028 3.613 3.030 3.615 ls_scf_store_result 1 5.0 0.000 0.000 3.420 3.513 ls_scf_dm_to_ks 2 5.0 0.000 0.006 2.848 2.956 dbcsr_data_new 4116 9.9 2.119 2.473 2.119 2.473 make_images_sizes 222 9.7 0.000 0.000 1.027 2.283 mp_alltoall_i44 222 10.7 1.027 2.283 1.027 2.283 dbcsr_sort_data 325 11.1 1.859 2.173 1.859 2.173 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.987 2.048 rebuild_ks_matrix 3 7.3 0.000 0.000 1.969 2.029 qs_ks_build_kohn_sham_matrix 3 8.3 0.012 0.082 1.969 2.029 make_images_pack 222 9.7 1.651 1.894 1.654 1.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.121000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2280.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 3.390271E+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 0.029 0.047 97.016 97.017 qs_energies 1 2.0 0.000 0.000 96.335 96.342 ls_scf 1 3.0 0.000 0.000 94.678 94.683 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.568 78.781 ls_scf_main 1 4.0 0.000 0.000 58.913 58.917 multiply_cannon 111 7.7 0.052 0.117 51.718 55.611 density_matrix_trs4 2 5.0 0.002 0.003 52.765 52.902 multiply_cannon_loop 111 8.7 0.155 0.168 46.769 50.119 ls_scf_init_scf 1 4.0 0.000 0.000 32.506 32.508 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.318 31.382 mp_waitall_1 6369 11.0 22.686 29.817 22.686 29.817 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.005 28.845 28.856 multiply_cannon_multrec 1332 9.7 14.161 17.265 22.128 25.225 make_m2s 222 7.7 0.008 0.009 21.128 22.353 make_images 222 8.7 3.144 3.624 21.076 22.299 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.183 16.427 make_images_data 222 9.7 0.004 0.005 11.761 13.302 hybrid_alltoall_any 227 10.6 0.797 3.750 11.152 12.983 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.483 7.602 9.141 dbcsr_mm_accdrv_process_sort 3641 11.4 7.162 8.645 7.162 8.645 mp_sum_l 887 5.1 4.147 7.268 4.147 7.268 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.083 6.106 multiply_cannon_sync_h2d 1332 9.7 5.498 6.051 5.498 6.051 mp_irecv_dv 3229 10.9 2.058 6.035 2.058 6.035 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.101 5.658 arnoldi_extremal 4 6.8 0.000 0.000 5.177 5.187 arnoldi_normal_ev 4 7.8 0.002 0.004 5.176 5.187 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.576 5.124 build_subspace 16 8.4 0.014 0.020 4.850 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.599 4.797 calculate_norms 2376 9.8 4.169 4.514 4.169 4.514 mp_allgather_i34 111 8.7 2.041 4.244 2.041 4.244 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.578 3.879 dbcsr_matrix_vector_mult_local 304 10.0 3.195 3.698 3.197 3.700 dbcsr_sort_data 658 11.4 3.072 3.439 3.072 3.439 ls_scf_post 1 4.0 0.000 0.000 3.259 3.266 dbcsr_special_finalize 555 9.7 0.006 0.007 2.837 3.222 dbcsr_merge_single_wm 555 10.7 0.539 0.668 2.828 3.214 ls_scf_store_result 1 5.0 0.000 0.000 3.011 3.065 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.973 3.029 dbcsr_data_release 10477 10.7 1.587 2.474 1.587 2.474 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.017000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2783.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 4.838719E+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 0.037 0.051 92.131 92.131 qs_energies 1 2.0 0.000 0.000 91.279 91.284 ls_scf 1 3.0 0.000 0.000 89.341 89.346 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.838 71.066 ls_scf_main 1 4.0 0.000 0.000 56.503 56.504 multiply_cannon 111 7.7 0.080 0.143 52.406 55.826 multiply_cannon_loop 111 8.7 0.088 0.094 49.839 51.290 density_matrix_trs4 2 5.0 0.002 0.003 49.479 49.573 ls_scf_init_scf 1 4.0 0.000 0.000 29.399 29.404 mp_waitall_1 5436 11.0 24.089 29.102 24.089 29.102 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.179 28.206 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.026 26.095 26.110 multiply_cannon_multrec 444 9.7 13.816 16.361 20.945 23.413 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.727 15.089 make_m2s 222 7.7 0.005 0.006 13.580 14.492 make_images 222 8.7 2.042 2.475 13.512 14.422 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.898 13.584 hybrid_alltoall_any 227 10.6 0.803 3.838 8.082 9.694 make_images_data 222 9.7 0.003 0.004 8.284 9.576 multiply_cannon_sync_h2d 444 9.7 6.739 8.378 6.739 8.378 dbcsr_mm_accdrv_process 3003 10.4 0.358 0.401 6.825 7.933 dbcsr_mm_accdrv_process_sort 3003 11.4 6.453 7.534 6.453 7.534 arnoldi_extremal 4 6.8 0.000 0.000 5.917 5.936 arnoldi_normal_ev 4 7.8 0.043 0.069 5.917 5.935 build_subspace 16 8.4 0.015 0.019 5.469 5.478 mp_sum_l 887 5.1 2.840 5.041 2.840 5.041 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.491 4.686 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.180 4.402 dbcsr_matrix_vector_mult_local 304 10.0 3.711 4.188 3.713 4.190 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.525 3.949 mp_irecv_dv 1241 11.2 1.511 3.920 1.511 3.920 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.961 3.880 calculate_norms 792 9.8 3.626 3.777 3.626 3.777 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.517 3.601 mp_allgather_i34 111 8.7 1.122 3.595 1.122 3.595 ls_scf_post 1 4.0 0.000 0.000 3.439 3.444 make_images_sizes 222 9.7 0.000 0.000 0.876 3.320 mp_alltoall_i44 222 10.7 0.875 3.320 0.875 3.320 ls_scf_store_result 1 5.0 0.000 0.000 3.230 3.265 dbcsr_finalize 304 7.8 0.062 0.076 2.223 2.346 dbcsr_data_new 4608 9.7 1.826 2.306 1.826 2.306 dbcsr_merge_all 275 8.9 0.481 0.537 2.075 2.180 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.041 2.042 rebuild_ks_matrix 3 7.3 0.000 0.000 2.009 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.008 2.009 qs_energies_init_hamiltonians 1 3.0 0.004 0.017 1.923 1.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.131000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 9.056199E+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 0.085 0.109 105.726 105.726 qs_energies 1 2.0 0.000 0.000 104.273 104.285 ls_scf 1 3.0 0.000 0.000 101.288 101.300 dbcsr_multiply_generic 111 6.7 0.023 0.027 75.074 75.225 ls_scf_main 1 4.0 0.000 0.000 63.743 63.744 density_matrix_trs4 2 5.0 0.002 0.003 54.959 55.036 multiply_cannon 111 7.7 0.141 0.216 48.598 50.816 multiply_cannon_loop 111 8.7 0.098 0.100 45.603 46.032 ls_scf_init_scf 1 4.0 0.000 0.000 33.709 33.710 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.150 32.167 matrix_sqrt_Newton_Schulz 2 6.5 0.037 0.074 29.442 29.447 mp_waitall_1 4527 11.1 22.053 25.986 22.053 25.986 make_m2s 222 7.7 0.006 0.006 22.855 23.935 make_images 222 8.7 3.584 3.885 22.746 23.824 multiply_cannon_multrec 444 9.7 17.888 18.584 22.570 23.297 hybrid_alltoall_any 227 10.6 1.659 3.625 12.854 15.895 make_images_data 222 9.7 0.004 0.004 13.161 15.345 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.308 10.761 multiply_cannon_sync_h2d 444 9.7 8.779 8.845 8.779 8.845 arnoldi_extremal 4 6.8 0.000 0.000 7.409 7.452 arnoldi_normal_ev 4 7.8 0.036 0.069 7.409 7.452 build_subspace 16 8.4 0.026 0.037 6.812 6.819 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.498 5.641 dbcsr_matrix_vector_mult_local 304 10.0 5.078 5.381 5.080 5.384 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.891 5.154 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.893 4.985 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.359 4.496 4.627 dbcsr_mm_accdrv_process_sort 1814 11.4 4.160 4.294 4.160 4.294 ls_scf_post 1 4.0 0.000 0.000 3.836 3.847 make_images_sizes 222 9.7 0.000 0.000 1.517 3.837 mp_alltoall_i44 222 10.7 1.516 3.836 1.516 3.836 ls_scf_store_result 1 5.0 0.000 0.000 3.529 3.567 mp_allgather_i34 111 8.7 1.092 3.548 1.092 3.548 calculate_norms 792 9.8 3.248 3.280 3.248 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.095 3.180 dbcsr_merge_all 275 8.9 0.891 0.920 2.879 2.957 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.954 2.954 dbcsr_complete_redistribute 5 7.6 1.428 1.476 2.748 2.856 dbcsr_sort_data 325 11.1 2.451 2.512 2.451 2.512 matrix_ls_to_qs 2 6.0 0.000 0.000 2.394 2.488 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.400 2.401 dbcsr_data_new 6591 9.6 1.845 2.354 1.845 2.354 rebuild_ks_matrix 3 7.3 0.000 0.000 2.333 2.334 qs_ks_build_kohn_sham_matrix 3 8.3 0.036 0.053 2.333 2.334 dbcsr_new_transposed 4 7.5 0.244 0.253 2.307 2.323 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.214 dbcsr_add_anytype 103 7.2 0.860 0.895 2.132 2.214 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.183 2.208 dbcsr_data_release 12724 10.6 1.986 2.192 1.986 2.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.726000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7158.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dbea54b02bae51086f4cc159fb0e2da2f450cf9_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 585.703424E+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 4452845. 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 0.031 0.112 229.924 229.932 qs_mol_dyn_low 1 2.0 0.011 0.062 228.766 228.788 qs_forces 5 3.8 0.005 0.010 228.535 228.537 qs_energies 5 4.8 0.017 0.067 225.510 225.561 scf_env_do_scf 5 5.8 0.001 0.010 206.696 206.697 scf_env_do_scf_inner_loop 105 6.6 0.005 0.085 181.944 181.971 qs_scf_new_mos 105 7.6 0.000 0.001 142.715 142.919 qs_scf_loop_do_ot 105 8.6 0.001 0.001 142.714 142.918 dbcsr_multiply_generic 1445 12.2 0.125 0.134 138.102 138.541 ot_scf_mini 105 9.6 0.004 0.012 132.795 132.948 multiply_cannon 1445 13.2 0.274 0.287 118.826 121.191 multiply_cannon_loop 1445 14.2 2.844 2.997 117.142 118.608 velocity_verlet 4 3.0 0.010 0.077 103.628 103.631 ot_mini 105 10.6 0.001 0.001 60.626 60.760 mp_waitall_1 488190 16.1 35.576 45.600 35.576 45.600 multiply_cannon_multrec 69360 15.2 29.705 34.979 39.888 45.495 qs_ot_get_p 112 10.4 0.001 0.001 42.234 42.548 qs_ot_get_derivative 55 11.6 0.001 0.008 38.774 38.921 multiply_cannon_metrocomm3 69360 15.2 0.202 0.213 25.949 35.664 multiply_cannon_sync_h2d 69360 15.2 29.108 33.653 29.108 33.653 qs_ot_p2m_diag 40 11.0 0.020 0.032 31.277 31.356 rebuild_ks_matrix 110 8.4 0.000 0.000 28.651 28.808 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.066 28.650 28.808 cp_dbcsr_syevd 40 12.0 0.002 0.003 28.020 28.022 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.293 26.452 init_scf_loop 7 6.6 0.001 0.018 24.715 24.718 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.033 23.305 apply_single 62 13.6 0.000 0.000 23.033 23.305 cp_fm_syevd 40 13.0 0.000 0.004 22.475 22.619 ot_new_cg_direction 55 11.6 0.001 0.001 21.069 21.069 prepare_preconditioner 7 7.6 0.000 0.000 19.860 19.900 make_preconditioner 7 8.6 0.002 0.037 19.860 19.900 cp_fm_redistribute_end 40 14.0 8.932 17.801 8.947 17.806 cp_fm_syevd_base 40 14.0 8.850 17.721 8.850 17.721 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.373 16.760 calculate_rho_elec 110 8.6 0.029 0.046 16.373 16.759 init_scf_run 5 5.8 0.000 0.002 16.216 16.217 scf_env_initial_rho_setup 5 6.8 0.002 0.003 16.216 16.217 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.673 14.872 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.800 13.880 mp_sum_l 4764 12.2 12.368 13.128 12.368 13.128 make_full_inverse_cholesky 7 9.6 0.000 0.001 12.494 12.565 acc_transpose_blocks 69360 15.2 0.355 0.372 11.283 12.245 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 11.948 11.969 calculate_dm_sparse 110 9.5 0.000 0.000 11.748 11.927 multiply_cannon_metrocomm1 69360 15.2 0.097 0.103 4.964 11.867 calculate_first_density_matrix 1 7.0 0.001 0.011 11.851 11.865 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.007 11.354 11.428 density_rs2pw 110 9.6 0.004 0.005 10.710 11.164 dbcsr_mm_accdrv_process 154766 15.8 6.239 6.408 10.050 11.034 fft_wrap_pw1pw2 1425 12.5 0.019 0.025 10.734 10.925 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.662 10.751 qs_vxc_create 110 10.4 0.002 0.008 9.927 9.949 check_diag 80 13.5 8.652 8.960 9.545 9.734 fft_wrap_pw1pw2_240 915 14.0 0.199 0.232 9.428 9.621 acc_transpose_blocks_kernels 69360 16.2 0.853 0.930 8.516 9.387 fft3d_pb 915 15.0 2.256 2.441 8.383 8.589 jit_kernel_transpose 5 15.0 7.663 8.486 7.663 8.486 sum_up_and_integrate 60 10.3 0.001 0.012 8.430 8.455 integrate_v_rspace 60 11.3 0.002 0.003 8.412 8.439 cp_fm_cholesky_invert 7 10.6 8.314 8.332 8.314 8.332 transfer_rs2pw 445 10.6 0.008 0.010 7.759 8.234 xc_rho_set_and_dset_create 110 12.4 0.073 0.109 6.991 7.228 make_full_single_inverse 7 9.6 0.001 0.029 7.086 7.127 make_m2s 2890 13.2 0.079 0.091 6.509 7.125 make_images 2890 14.2 0.242 0.263 6.403 7.018 xc_vxc_pw_create 60 11.3 0.036 0.043 6.647 6.673 xc_pw_derive 510 13.4 0.005 0.006 5.797 5.858 mp_alltoall_z22v 2340 16.7 5.556 5.786 5.556 5.786 mp_waitany 7680 13.5 4.656 5.296 4.656 5.296 potential_pw2rs 60 12.3 0.002 0.003 4.663 4.708 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.932000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.200000, yerr=1.939072 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9dbea54b02bae51086f4cc159fb0e2da2f450cf9 Summary: empty Status: OK