=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 577b676415f3d2d1924bda4bef9a238f96821cb1 ################# 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 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_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 SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/01 job id: 51462637 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/02 job id: 51462638 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/03 job id: 51462639 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/04 job id: 51462640 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/05 job id: 51462641 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/06 job id: 51462642 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/07 job id: 51462643 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/08 job id: 51462644 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/09 job id: 51462645 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/10 job id: 51462646 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/11 job id: 51462647 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/12 job id: 51462648 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/13 job id: 51462649 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/14 job id: 51462650 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/15 job id: 51462652 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/16 job id: 51462653 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/17 job id: 51462654 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/18 job id: 51462655 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/19 job id: 51462656 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/20 job id: 51462657 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/21 job id: 51462658 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/22 job id: 51462659 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/23 job id: 51462660 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/24 job id: 51462661 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/25 job id: 51462662 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/26 job id: 51462663 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/27 job id: 51462664 --- 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/577b676415f3d2d1924bda4bef9a238f96821cb1_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 0.019 0.035 133.845 133.846 farming_run 1 2.0 133.313 133.315 133.812 133.816 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.486021E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.025 115.218 115.219 qs_energies 1 2.0 0.001 0.001 115.032 115.034 mp2_main 1 3.0 0.000 0.000 112.968 112.970 mp2_gpw_main 1 4.0 0.040 0.047 111.984 111.986 mp2_ri_gpw_compute_in 1 5.0 0.173 0.174 93.144 93.286 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.551 54.692 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 40.972 45.755 get_2c_integrals 1 6.0 0.008 0.009 37.743 38.421 integrate_v_rspace 273 8.0 0.438 0.453 24.569 29.029 pw_transfer 6555 10.6 0.387 0.405 26.861 27.442 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.451 25.811 grid_integrate_task_list 273 9.0 20.477 25.361 20.477 25.361 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 22.992 23.337 compute_2c_integrals 1 7.0 0.002 0.002 19.530 19.530 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.683 19.260 mp2_eri_2c_integrate_gpw 1 9.0 2.371 2.422 18.681 19.259 cp_fm_cholesky_decompose 12 8.2 18.184 18.911 18.184 18.911 rpa_ri_compute_en 1 5.0 0.019 0.022 18.722 18.838 cholesky_decomp 1 7.0 0.000 0.000 17.060 17.740 fft3d_s 5443 13.4 16.140 16.378 16.162 16.399 ao_to_mo_and_store_B_mult_1 272 7.0 10.759 15.302 10.759 15.302 calculate_wavefunction 272 8.0 5.394 5.478 12.306 13.009 rpa_num_int 1 6.0 0.000 0.005 10.708 10.717 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.647 10.679 calc_mat_Q 8 8.0 0.000 0.000 9.473 9.569 contract_S_to_Q 8 9.0 0.000 0.000 8.896 8.992 calc_potential_gpw 544 9.5 0.004 0.005 8.230 8.830 parallel_gemm_fm 14 9.1 0.000 0.000 8.488 8.600 parallel_gemm_fm_cosma 14 10.1 8.488 8.600 8.488 8.600 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.117 8.489 potential_pw2rs 545 10.0 0.106 0.108 7.535 8.162 create_integ_mat 1 6.0 0.027 0.040 7.780 7.790 collocate_single_gaussian 272 10.0 0.039 0.042 7.384 7.685 array2fm 1 7.0 0.000 0.000 6.771 7.228 pw_scatter_s 2720 13.7 4.320 4.539 4.320 4.539 pw_gather_s 2722 13.2 3.447 3.863 3.447 3.863 array2fm_buffer_send 1 8.0 2.998 3.200 2.998 3.200 pw_poisson_solve 545 10.5 1.118 1.169 2.226 2.492 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.985613, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 0.038 522.524 522.525 farming_run 1 2.0 521.742 521.747 522.482 522.484 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.232130E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27633238. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 319.081 319.082 qs_energies 1 2.0 0.001 0.009 318.855 318.855 mp2_main 1 3.0 0.000 0.000 203.569 203.571 mp2_gpw_main 1 4.0 0.050 0.073 202.673 202.678 mp2_ri_gpw_compute_en 1 5.0 0.037 0.041 128.535 130.556 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.765 1.909 126.896 126.900 scf_env_do_scf 1 3.0 0.000 0.000 114.985 114.985 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.058 114.066 rebuild_ks_matrix 4 6.0 0.000 0.000 114.057 114.065 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 114.057 114.065 hfx_ks_matrix 4 8.0 0.001 0.001 113.697 113.702 integrate_four_center 4 9.0 0.152 0.466 113.696 113.701 mp2_ri_gpw_compute_en_expansio 172 7.0 0.468 0.502 111.329 112.072 local_gemm 172 8.0 110.862 111.598 110.862 111.598 integrate_four_center_main 4 10.0 0.152 0.630 101.926 105.639 integrate_four_center_bin 265 11.0 101.775 105.637 101.775 105.637 init_scf_loop 1 4.0 0.000 0.000 96.773 96.773 mp2_ri_gpw_compute_in 1 5.0 0.065 0.077 73.987 75.002 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.761 54.776 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 41.447 46.522 integrate_v_rspace 95 8.0 0.397 0.568 27.932 32.853 pw_transfer 2240 10.6 0.146 0.178 29.552 29.954 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.548 28.983 grid_integrate_task_list 95 9.0 23.333 28.465 23.333 28.465 ao_to_mo_and_store_B_mult_1 91 7.0 10.647 26.852 10.647 26.852 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.289 26.720 get_2c_integrals 1 6.0 0.000 0.000 20.133 20.161 compute_2c_integrals 1 7.0 0.002 0.003 19.114 19.122 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.716 18.986 mp2_eri_2c_integrate_gpw 1 9.0 1.726 1.861 18.715 18.986 fft3d_s 1823 13.4 18.493 18.869 18.507 18.882 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.211 18.211 calculate_wavefunction 91 8.0 2.011 2.053 9.619 9.849 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.518 7.711 9.210 potential_pw2rs 186 10.0 0.033 0.034 8.437 8.991 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.208 8.519 calc_potential_gpw 182 9.5 0.002 0.002 7.959 8.153 collocate_single_gaussian 91 10.0 0.017 0.021 7.787 7.990 mp_sendrecv_dm3 2068 8.0 5.735 7.275 5.735 7.275 mp_sync 37 10.5 3.751 6.809 3.751 6.809 integrate_four_center_load 4 10.0 0.000 0.000 6.750 6.754 hfx_load_balance 1 11.0 0.000 0.000 6.750 6.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.667232, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1492.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 445.087744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 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 0.013 0.042 54.622 54.624 qs_mol_dyn_low 1 2.0 0.005 0.012 54.329 54.339 qs_forces 11 3.9 0.002 0.002 54.261 54.264 qs_energies 11 4.9 0.002 0.002 52.759 52.782 scf_env_do_scf 11 5.9 0.000 0.001 46.470 46.471 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.366 44.366 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.231 34.672 qs_scf_new_mos 108 7.5 0.000 0.000 34.296 34.590 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.296 34.590 ot_scf_mini 108 9.5 0.002 0.002 32.608 32.775 multiply_cannon 2286 13.5 0.187 0.194 26.656 28.151 multiply_cannon_loop 2286 14.5 1.800 1.921 25.997 27.517 velocity_verlet 10 3.0 0.002 0.005 26.557 26.558 ot_mini 108 10.5 0.001 0.001 19.761 20.015 qs_ot_get_derivative 108 11.5 0.001 0.001 16.705 16.872 mp_waitall_1 245248 16.5 8.694 14.801 8.694 14.801 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 6.193 12.720 multiply_cannon_multrec 54864 15.5 3.642 5.725 7.867 11.281 qs_ot_get_p 119 10.4 0.001 0.001 8.197 8.482 rebuild_ks_matrix 119 8.3 0.000 0.000 7.870 8.016 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.869 8.016 mp_sum_l 7287 12.8 5.627 7.156 5.627 7.156 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.932 7.060 multiply_cannon_sync_h2d 54864 15.5 5.172 6.583 5.172 6.583 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.774 6.212 dbcsr_mm_accdrv_process 76910 16.1 1.818 2.863 4.139 5.872 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.550 5.674 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.272 5.318 init_scf_run 11 5.9 0.000 0.001 5.016 5.016 scf_env_initial_rho_setup 11 6.9 0.003 0.003 5.016 5.016 sum_up_and_integrate 119 10.3 0.001 0.002 4.450 4.456 integrate_v_rspace 119 11.3 0.002 0.003 4.439 4.447 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.446 4.447 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.144 4.264 calculate_rho_elec 119 8.7 0.011 0.016 4.143 4.263 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.254 4.254 cp_fm_redistribute_end 50 14.0 2.169 4.227 2.175 4.230 cp_fm_diag_elpa_base 50 14.0 2.049 4.119 2.053 4.122 calculate_dm_sparse 119 9.5 0.000 0.000 3.156 3.311 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.950 3.163 apply_single 119 13.6 0.000 0.000 2.950 3.163 jit_kernel_multiply 13 15.8 2.259 2.982 2.259 2.982 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.829 2.956 calculate_first_density_matrix 1 7.0 0.000 0.000 2.787 2.793 acc_transpose_blocks 54864 15.5 0.219 0.242 2.229 2.768 ot_diis_step 108 11.5 0.006 0.006 2.750 2.751 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.479 2.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.365 2.367 density_rs2pw 119 9.7 0.004 0.004 2.169 2.250 wfi_extrapolate 11 7.9 0.001 0.001 2.155 2.155 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.073 2.119 grid_integrate_task_list 119 12.3 2.017 2.107 2.017 2.107 init_scf_loop 11 6.9 0.000 0.000 2.087 2.087 mp_sum_d 4137 12.0 1.437 2.036 1.437 2.036 potential_pw2rs 119 12.3 0.004 0.004 1.825 1.834 pw_transfer 1439 11.6 0.052 0.057 1.681 1.749 make_m2s 4572 13.5 0.053 0.055 1.648 1.691 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.605 1.677 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.633 1.676 make_images 4572 14.5 0.132 0.139 1.567 1.610 transfer_rs2pw 487 10.6 0.005 0.006 1.366 1.463 mp_alltoall_d11v 2130 13.8 1.315 1.434 1.315 1.434 acc_transpose_blocks_sync 164592 16.5 1.183 1.427 1.183 1.427 grid_collocate_task_list 119 9.7 1.342 1.403 1.342 1.403 fft3d_ps 1201 14.6 0.370 0.476 1.318 1.382 mp_waitany 12084 13.8 1.234 1.378 1.234 1.378 transfer_pw2rs 487 13.2 0.006 0.007 1.335 1.343 fft_wrap_pw1pw2_140 487 13.2 0.107 0.115 1.241 1.316 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.786 1.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.624000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 479.744000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 750669. 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 0.007 0.026 38.178 38.178 qs_mol_dyn_low 1 2.0 0.004 0.007 38.010 38.017 qs_forces 11 3.9 0.004 0.010 37.913 37.915 qs_energies 11 4.9 0.002 0.002 36.255 36.262 scf_env_do_scf 11 5.9 0.000 0.001 31.300 31.300 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.929 28.929 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.815 22.171 qs_scf_new_mos 108 7.5 0.001 0.001 20.425 20.650 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.424 20.649 ot_scf_mini 108 9.5 0.002 0.003 19.508 19.676 multiply_cannon 2286 13.5 0.208 0.216 16.824 18.339 velocity_verlet 10 3.0 0.002 0.005 18.135 18.136 multiply_cannon_loop 2286 14.5 1.181 1.238 15.689 17.127 ot_mini 108 10.5 0.001 0.001 12.052 12.285 mp_waitall_1 200699 16.5 5.483 10.668 5.483 10.668 qs_ot_get_derivative 108 11.5 0.001 0.001 9.693 9.862 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.038 9.456 multiply_cannon_multrec 27432 15.5 1.826 4.126 6.131 8.815 rebuild_ks_matrix 119 8.3 0.000 0.000 6.689 6.828 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.689 6.828 dbcsr_mm_accdrv_process 47894 16.0 3.563 5.826 4.225 6.285 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.884 6.010 qs_ot_get_p 119 10.4 0.001 0.001 4.553 4.778 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.679 4.529 mp_sum_l 7287 12.8 1.995 4.062 1.995 4.062 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.991 4.044 apply_single 119 13.6 0.000 0.000 2.991 4.044 init_scf_run 11 5.9 0.000 0.001 3.748 3.748 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.747 3.748 sum_up_and_integrate 119 10.3 0.001 0.001 3.635 3.642 integrate_v_rspace 119 11.3 0.002 0.003 3.621 3.628 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.548 3.574 calculate_rho_elec 119 8.7 0.021 0.024 3.547 3.574 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.941 2.960 make_m2s 4572 13.5 0.051 0.053 2.542 2.874 make_images 4572 14.5 0.205 0.244 2.452 2.787 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.549 2.549 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.297 2.383 init_scf_loop 11 6.9 0.000 0.000 2.351 2.351 ot_diis_step 108 11.5 0.011 0.011 2.303 2.304 calculate_first_density_matrix 1 7.0 0.000 0.001 2.256 2.259 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.240 2.241 cp_fm_redistribute_end 50 14.0 1.136 2.217 1.139 2.220 multiply_cannon_sync_h2d 27432 15.5 1.697 2.197 1.697 2.197 cp_fm_diag_elpa_base 50 14.0 1.049 2.130 1.077 2.168 calculate_dm_sparse 119 9.5 0.000 0.001 2.055 2.131 acc_transpose_blocks 27432 15.5 0.113 0.121 1.607 1.945 grid_integrate_task_list 119 12.3 1.842 1.939 1.842 1.939 density_rs2pw 119 9.7 0.004 0.004 1.848 1.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.838 1.840 pw_transfer 1439 11.6 0.065 0.069 1.805 1.839 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.714 1.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.646 1.687 make_images_data 4572 15.5 0.047 0.054 1.212 1.562 prepare_preconditioner 11 7.9 0.000 0.000 1.486 1.512 make_preconditioner 11 8.9 0.000 0.000 1.486 1.512 jit_kernel_multiply 8 16.5 0.605 1.480 0.605 1.480 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.393 1.447 wfi_extrapolate 11 7.9 0.001 0.001 1.433 1.433 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.077 1.432 fft_wrap_pw1pw2_140 487 13.2 0.128 0.132 1.374 1.408 potential_pw2rs 119 12.3 0.006 0.006 1.388 1.392 fft3d_ps 1201 14.6 0.520 0.577 1.345 1.375 mp_alltoall_d11v 2130 13.8 1.169 1.363 1.169 1.363 grid_collocate_task_list 119 9.7 1.290 1.339 1.290 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.195 1.242 transfer_rs2pw 487 10.6 0.005 0.005 1.037 1.089 mp_sum_d 4137 12.0 0.544 1.001 0.544 1.001 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.949 0.952 mp_allgather_i34 2286 14.5 0.560 0.936 0.560 0.936 acc_transpose_blocks_sync 82296 16.5 0.810 0.936 0.810 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.875 transfer_pw2rs 487 13.2 0.004 0.005 0.866 0.873 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.657 0.871 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.700 0.767 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.178000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=457.000000, yerr=1.858641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 512.061440E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.034 31.657 31.657 qs_mol_dyn_low 1 2.0 0.004 0.016 31.427 31.436 qs_forces 11 3.9 0.003 0.009 31.362 31.364 qs_energies 11 4.9 0.002 0.006 29.815 29.818 scf_env_do_scf 11 5.9 0.001 0.001 25.177 25.177 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 22.799 22.799 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.750 16.816 qs_scf_new_mos 108 7.5 0.001 0.001 15.211 15.233 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.211 15.232 velocity_verlet 10 3.0 0.002 0.006 14.957 14.959 ot_scf_mini 108 9.5 0.002 0.003 14.457 14.475 multiply_cannon 2286 13.5 0.194 0.200 13.621 14.320 multiply_cannon_loop 2286 14.5 0.859 0.899 12.858 13.590 ot_mini 108 10.5 0.001 0.001 8.950 8.970 qs_ot_get_derivative 108 11.5 0.001 0.001 7.473 7.491 multiply_cannon_multrec 18288 15.5 1.873 2.827 6.951 7.228 rebuild_ks_matrix 119 8.3 0.000 0.000 5.910 5.925 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 5.909 5.925 dbcsr_mm_accdrv_process 38222 16.0 4.927 5.701 4.984 5.765 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.206 5.221 mp_waitall_1 158411 16.6 2.540 3.700 2.540 3.700 sum_up_and_integrate 119 10.3 0.001 0.002 3.495 3.500 integrate_v_rspace 119 11.3 0.003 0.003 3.482 3.489 init_scf_run 11 5.9 0.000 0.001 3.439 3.440 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.439 3.439 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.837 3.405 qs_ot_get_p 119 10.4 0.001 0.001 3.281 3.305 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.219 3.223 calculate_rho_elec 119 8.7 0.031 0.032 3.218 3.223 multiply_cannon_metrocomm3 18288 15.5 0.049 0.050 1.435 2.439 init_scf_loop 11 6.9 0.001 0.003 2.362 2.362 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.967 2.341 apply_single 119 13.6 0.000 0.000 1.967 2.341 calculate_first_density_matrix 1 7.0 0.000 0.003 2.230 2.232 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.194 2.202 grid_integrate_task_list 119 12.3 1.807 1.927 1.807 1.927 density_rs2pw 119 9.7 0.004 0.004 1.785 1.909 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.902 1.902 make_m2s 4572 13.5 0.044 0.045 1.745 1.891 calculate_dm_sparse 119 9.5 0.000 0.001 1.831 1.841 pw_transfer 1439 11.6 0.065 0.068 1.825 1.832 make_images 4572 14.5 0.191 0.203 1.660 1.804 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.732 1.742 acc_transpose_blocks 18288 15.5 0.078 0.081 1.629 1.715 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.684 1.694 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.664 1.677 cp_fm_diag_elpa_base 50 14.0 1.641 1.653 1.662 1.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.664 1.665 prepare_preconditioner 11 7.9 0.000 0.000 1.615 1.618 make_preconditioner 11 8.9 0.000 0.000 1.615 1.618 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.473 1.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.521 1.527 ot_diis_step 108 11.5 0.011 0.011 1.463 1.463 fft_wrap_pw1pw2_140 487 13.2 0.179 0.182 1.412 1.422 mp_sum_l 7287 12.8 0.995 1.374 0.995 1.374 grid_collocate_task_list 119 9.7 1.235 1.334 1.235 1.334 potential_pw2rs 119 12.3 0.007 0.008 1.303 1.307 fft3d_ps 1201 14.6 0.545 0.565 1.287 1.295 multiply_cannon_sync_h2d 18288 15.5 1.043 1.230 1.043 1.230 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.172 1.176 wfi_extrapolate 11 7.9 0.001 0.002 1.156 1.156 transfer_rs2pw 487 10.6 0.005 0.005 0.957 1.092 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.969 0.971 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.936 0.955 make_images_data 4572 15.5 0.047 0.051 0.746 0.914 mp_alltoall_d11v 2130 13.8 0.641 0.850 0.641 0.850 hybrid_alltoall_any 4725 16.4 0.058 0.119 0.646 0.835 acc_transpose_blocks_sync 54864 16.5 0.740 0.815 0.740 0.815 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 0.791 0.806 transfer_pw2rs 487 13.2 0.004 0.004 0.785 0.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.775 0.777 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.672 0.748 cp_fm_cholesky_invert 11 10.9 0.681 0.685 0.681 0.685 mp_alltoall_z22v 1201 16.6 0.596 0.664 0.596 0.664 transfer_rs2pw_140 130 11.5 0.116 0.129 0.526 0.653 mp_waitany 9880 13.7 0.535 0.650 0.535 0.650 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.657000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=487.909091, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 547.467264E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.025 35.915 35.916 qs_mol_dyn_low 1 2.0 0.003 0.003 35.754 35.764 qs_forces 11 3.9 0.002 0.002 35.658 35.658 qs_energies 11 4.9 0.001 0.002 33.963 33.967 scf_env_do_scf 11 5.9 0.000 0.001 28.917 28.918 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.595 25.595 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.815 19.924 velocity_verlet 10 3.0 0.033 0.047 18.260 18.262 qs_scf_new_mos 108 7.5 0.001 0.001 17.598 17.655 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.598 17.654 ot_scf_mini 108 9.5 0.002 0.003 16.573 16.628 multiply_cannon 2286 13.5 0.219 0.227 16.238 16.522 multiply_cannon_loop 2286 14.5 1.525 1.586 15.302 15.570 ot_mini 108 10.5 0.001 0.001 10.209 10.271 multiply_cannon_multrec 27432 15.5 2.455 3.116 9.138 9.435 qs_ot_get_derivative 108 11.5 0.001 0.001 8.355 8.410 dbcsr_mm_accdrv_process 47916 15.9 6.277 7.839 6.581 7.954 rebuild_ks_matrix 119 8.3 0.000 0.000 6.272 6.330 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.272 6.329 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.557 5.610 init_scf_run 11 5.9 0.000 0.001 3.713 3.713 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.713 3.713 qs_ot_get_p 119 10.4 0.001 0.001 3.528 3.605 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.114 3.509 sum_up_and_integrate 119 10.3 0.001 0.001 3.438 3.447 integrate_v_rspace 119 11.3 0.003 0.003 3.427 3.436 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.316 3.357 calculate_rho_elec 119 8.7 0.040 0.046 3.316 3.357 init_scf_loop 11 6.9 0.000 0.000 3.303 3.304 acc_transpose_blocks 27432 15.5 0.118 0.121 2.460 2.535 prepare_preconditioner 11 7.9 0.000 0.000 2.481 2.488 make_preconditioner 11 8.9 0.000 0.000 2.481 2.488 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.094 2.416 calculate_first_density_matrix 1 7.0 0.000 0.000 2.342 2.344 make_m2s 4572 13.5 0.054 0.056 2.200 2.326 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.293 2.323 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.098 2.220 apply_single 119 13.6 0.000 0.000 2.098 2.220 mp_waitall_1 137007 16.6 1.658 2.217 1.658 2.217 make_images 4572 14.5 0.272 0.331 2.091 2.217 calculate_dm_sparse 119 9.5 0.000 0.000 2.153 2.201 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.138 2.150 pw_transfer 1439 11.6 0.065 0.068 1.961 1.996 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.910 1.911 grid_integrate_task_list 119 12.3 1.824 1.909 1.824 1.909 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.869 1.907 density_rs2pw 119 9.7 0.004 0.004 1.762 1.866 ot_diis_step 108 11.5 0.012 0.012 1.817 1.817 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.752 1.770 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.743 1.743 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.561 1.600 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.512 1.524 cp_fm_diag_elpa_base 50 14.0 1.479 1.499 1.510 1.522 acc_transpose_blocks_sync 82296 16.5 1.413 1.497 1.413 1.497 fft3d_ps 1201 14.6 0.591 0.651 1.357 1.380 grid_collocate_task_list 119 9.7 1.247 1.344 1.247 1.344 wfi_extrapolate 11 7.9 0.001 0.001 1.319 1.319 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.742 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.276 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.254 1.262 potential_pw2rs 119 12.3 0.008 0.009 1.252 1.254 mp_sum_l 7287 12.8 0.944 1.226 0.944 1.226 jit_kernel_multiply 6 16.8 0.240 1.169 0.240 1.169 cp_fm_upper_to_full 72 14.2 0.814 1.155 0.814 1.155 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 1.079 1.079 dbcsr_complete_redistribute 329 12.2 0.115 0.140 0.729 0.999 transfer_rs2pw 487 10.6 0.004 0.005 0.852 0.955 make_images_data 4572 15.5 0.048 0.053 0.809 0.935 acc_transpose_blocks_kernels 27432 16.5 0.273 0.280 0.899 0.918 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.876 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.692 0.875 mp_alltoall_d11v 2130 13.8 0.719 0.860 0.719 0.860 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.545 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.808 cp_fm_cholesky_invert 11 10.9 0.718 0.721 0.718 0.721 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.916000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=520.727273, yerr=3.277774 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 609.849344E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 28.471 28.471 qs_mol_dyn_low 1 2.0 0.003 0.003 28.309 28.317 qs_forces 11 3.9 0.002 0.003 28.253 28.254 qs_energies 11 4.9 0.002 0.002 26.555 26.558 scf_env_do_scf 11 5.9 0.000 0.001 21.911 21.911 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.493 19.494 velocity_verlet 10 3.0 0.002 0.003 14.567 14.570 dbcsr_multiply_generic 2286 12.5 0.095 0.098 13.073 13.183 qs_scf_new_mos 108 7.5 0.001 0.001 11.850 11.877 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.849 11.877 ot_scf_mini 108 9.5 0.002 0.002 11.142 11.167 multiply_cannon 2286 13.5 0.225 0.231 10.515 10.997 multiply_cannon_loop 2286 14.5 0.642 0.669 9.610 9.810 ot_mini 108 10.5 0.001 0.001 6.460 6.489 multiply_cannon_multrec 9144 15.5 1.679 1.852 5.941 6.157 rebuild_ks_matrix 119 8.3 0.000 0.000 5.772 5.797 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.771 5.797 qs_ot_get_derivative 108 11.5 0.001 0.001 5.179 5.206 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.136 5.159 dbcsr_mm_accdrv_process 12550 15.8 3.735 4.188 4.155 4.230 sum_up_and_integrate 119 10.3 0.001 0.001 3.413 3.422 integrate_v_rspace 119 11.3 0.003 0.003 3.403 3.412 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.336 3.343 calculate_rho_elec 119 8.7 0.060 0.061 3.335 3.343 init_scf_run 11 5.9 0.000 0.001 3.208 3.208 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.208 3.208 qs_ot_get_p 119 10.4 0.001 0.001 2.809 2.843 init_scf_loop 11 6.9 0.000 0.000 2.399 2.400 pw_transfer 1439 11.6 0.066 0.068 2.081 2.090 calculate_first_density_matrix 1 7.0 0.000 0.000 2.082 2.082 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.987 1.998 grid_integrate_task_list 119 12.3 1.872 1.934 1.872 1.934 make_m2s 4572 13.5 0.033 0.035 1.759 1.914 mp_waitall_1 115863 16.7 1.446 1.893 1.446 1.893 density_rs2pw 119 9.7 0.003 0.003 1.736 1.869 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.830 1.832 make_images 4572 14.5 0.268 0.300 1.670 1.822 calculate_dm_sparse 119 9.5 0.000 0.000 1.733 1.751 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.694 1.712 prepare_preconditioner 11 7.9 0.000 0.000 1.679 1.683 make_preconditioner 11 8.9 0.000 0.000 1.679 1.683 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.655 1.665 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.575 1.603 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.590 1.591 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.528 1.529 acc_transpose_blocks 9144 15.5 0.042 0.044 1.432 1.472 jit_kernel_multiply 8 15.7 0.381 1.409 0.381 1.409 grid_collocate_task_list 119 9.7 1.297 1.395 1.297 1.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.367 1.377 fft3d_ps 1201 14.6 0.647 0.661 1.340 1.351 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.335 1.346 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.306 1.314 cp_fm_diag_elpa_base 50 14.0 1.280 1.297 1.305 1.313 ot_diis_step 108 11.5 0.013 0.013 1.265 1.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.207 1.243 apply_single 119 13.6 0.000 0.000 1.207 1.243 potential_pw2rs 119 12.3 0.010 0.011 1.226 1.229 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.224 1.226 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.121 1.125 wfi_extrapolate 11 7.9 0.001 0.001 1.076 1.076 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.757 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.924 make_images_data 4572 15.5 0.042 0.046 0.754 0.913 transfer_rs2pw 487 10.6 0.004 0.004 0.758 0.891 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.487 0.874 mp_alltoall_d11v 2130 13.8 0.750 0.843 0.750 0.843 cp_fm_cholesky_invert 11 10.9 0.835 0.838 0.835 0.838 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.816 0.822 acc_transpose_blocks_sync 27432 16.5 0.720 0.768 0.720 0.768 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.758 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.672 0.724 acc_transpose_blocks_kernels 9144 16.5 0.119 0.120 0.653 0.662 transfer_pw2rs 487 13.2 0.003 0.004 0.636 0.637 mp_allgather_i34 2286 14.5 0.222 0.632 0.222 0.632 mp_alltoall_z22v 1201 16.6 0.566 0.600 0.566 0.600 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.471000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=579.181818, yerr=3.712832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 741.928960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 42.701 42.701 qs_mol_dyn_low 1 2.0 0.003 0.003 42.473 42.480 qs_forces 11 3.9 0.022 0.023 42.409 42.410 qs_energies 11 4.9 0.002 0.003 40.378 40.381 scf_env_do_scf 11 5.9 0.001 0.001 34.453 34.453 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.702 26.704 velocity_verlet 10 3.0 0.002 0.002 23.978 23.984 dbcsr_multiply_generic 2286 12.5 0.104 0.106 18.877 19.065 qs_scf_new_mos 108 7.5 0.001 0.001 17.055 17.154 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.055 17.153 ot_scf_mini 108 9.5 0.002 0.002 15.893 15.996 multiply_cannon 2286 13.5 0.295 0.307 14.983 15.937 multiply_cannon_loop 2286 14.5 0.863 0.889 13.739 14.671 ot_mini 108 10.5 0.001 0.001 9.661 9.781 multiply_cannon_multrec 9144 15.5 3.401 4.782 8.761 8.899 init_scf_loop 11 6.9 0.000 0.000 7.725 7.729 qs_ot_get_derivative 108 11.5 0.001 0.001 7.625 7.727 rebuild_ks_matrix 119 8.3 0.000 0.000 7.254 7.401 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.254 7.400 prepare_preconditioner 11 7.9 0.000 0.000 6.762 6.776 make_preconditioner 11 8.9 0.000 0.000 6.761 6.776 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.533 6.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.321 6.655 dbcsr_mm_accdrv_process 12550 15.8 4.548 6.495 5.227 6.582 cp_fm_upper_to_full 72 14.2 3.183 4.567 3.183 4.567 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.161 4.166 calculate_rho_elec 119 8.7 0.117 0.120 4.160 4.166 sum_up_and_integrate 119 10.3 0.001 0.001 3.844 3.850 integrate_v_rspace 119 11.3 0.004 0.004 3.834 3.840 init_scf_run 11 5.9 0.000 0.001 3.765 3.766 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.765 3.765 qs_ot_get_p 119 10.4 0.001 0.001 3.312 3.458 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.775 3.196 mp_waitall_1 94719 16.7 2.203 3.185 2.203 3.185 pw_transfer 1439 11.6 0.069 0.070 3.002 3.007 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.904 2.909 dbcsr_complete_redistribute 329 12.2 0.357 0.359 2.036 2.848 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.738 2.555 fft_wrap_pw1pw2_140 487 13.2 0.650 0.652 2.483 2.490 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.205 2.487 apply_single 119 13.6 0.000 0.000 2.205 2.487 make_m2s 4572 13.5 0.037 0.037 2.318 2.465 make_images 4572 14.5 0.352 0.383 2.198 2.345 calculate_first_density_matrix 1 7.0 0.000 0.000 2.312 2.313 mp_alltoall_i22 627 13.8 1.476 2.291 1.476 2.291 calculate_dm_sparse 119 9.5 0.000 0.000 2.226 2.245 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.245 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.319 2.230 density_rs2pw 119 9.7 0.003 0.004 2.204 2.225 grid_integrate_task_list 119 12.3 2.081 2.129 2.081 2.129 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.037 2.090 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.043 2.044 ot_diis_step 108 11.5 0.014 0.014 2.006 2.006 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.940 1.941 acc_transpose_blocks 9144 15.5 0.044 0.045 1.825 1.881 qs_energies_init_hamiltonians 11 5.9 0.013 0.029 1.879 1.879 fft3d_ps 1201 14.6 0.867 0.885 1.825 1.829 mp_sum_l 7287 12.8 1.029 1.819 1.029 1.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.664 1.687 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.588 1.589 grid_collocate_task_list 119 9.7 1.523 1.549 1.523 1.549 cp_fm_cholesky_invert 11 10.9 1.389 1.392 1.389 1.392 wfi_extrapolate 11 7.9 0.001 0.001 1.389 1.389 potential_pw2rs 119 12.3 0.014 0.014 1.344 1.345 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.331 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.311 1.312 cp_fm_diag_elpa_base 50 14.0 1.168 1.220 1.310 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.225 1.233 qs_env_update_s_mstruct 11 6.9 0.065 0.088 1.192 1.225 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.022 1.204 acc_transpose_blocks_sync 27432 16.5 1.106 1.152 1.106 1.152 make_images_data 4572 15.5 0.046 0.050 0.963 1.151 mp_alltoall_d11v 2130 13.8 1.127 1.148 1.127 1.148 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.996 1.051 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.945 qs_create_task_list 11 7.9 0.000 0.000 0.925 0.937 generate_qs_task_list 11 8.9 0.368 0.389 0.925 0.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.701000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=698.000000, yerr=12.202831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 494.026752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.027 80.544 80.545 qs_mol_dyn_low 1 2.0 0.003 0.007 80.301 80.310 qs_forces 11 3.9 0.003 0.003 80.234 80.235 qs_energies 11 4.9 0.002 0.002 77.402 77.432 scf_env_do_scf 11 5.9 0.000 0.001 68.708 68.711 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.305 63.305 dbcsr_multiply_generic 2055 12.4 0.104 0.107 50.337 50.763 qs_scf_new_mos 99 7.5 0.000 0.001 46.489 46.633 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.488 46.633 ot_scf_mini 99 9.5 0.002 0.002 44.160 44.243 multiply_cannon 2055 13.4 0.182 0.186 42.301 43.471 velocity_verlet 10 3.0 0.001 0.005 42.660 42.660 multiply_cannon_loop 2055 14.4 1.767 1.810 41.332 42.586 ot_mini 99 10.5 0.001 0.001 26.322 26.416 qs_ot_get_derivative 99 11.5 0.001 0.001 19.522 19.634 multiply_cannon_multrec 49320 15.4 11.406 12.128 17.366 18.084 rebuild_ks_matrix 110 8.3 0.000 0.001 14.335 14.533 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.335 14.532 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.569 12.734 mp_waitall_1 220248 16.4 10.553 11.615 10.553 11.615 multiply_cannon_sync_h2d 49320 15.4 9.605 10.166 9.605 10.166 qs_ot_get_p 110 10.4 0.001 0.001 9.537 9.668 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.499 8.064 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.464 7.870 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.193 7.611 apply_single 110 13.6 0.000 0.000 7.192 7.611 sum_up_and_integrate 110 10.3 0.002 0.003 6.885 6.898 integrate_v_rspace 110 11.3 0.003 0.003 6.860 6.879 init_scf_run 11 5.9 0.000 0.001 6.671 6.671 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.671 6.671 ot_diis_step 99 11.5 0.006 0.006 6.558 6.558 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.382 6.403 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.225 6.358 calculate_rho_elec 110 8.6 0.020 0.024 6.225 6.358 dbcsr_mm_accdrv_process 87628 16.1 3.028 3.106 5.829 6.081 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.586 5.586 init_scf_loop 11 6.9 0.000 0.000 5.378 5.378 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.179 5.246 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.034 5.057 cp_fm_diag_elpa_base 48 14.0 5.022 5.043 5.032 5.055 mp_sum_l 6594 12.7 3.773 4.421 3.773 4.421 wfi_extrapolate 11 7.9 0.001 0.001 4.006 4.006 make_m2s 4110 13.4 0.060 0.065 3.806 3.939 make_images 4110 14.4 0.177 0.189 3.712 3.848 calculate_dm_sparse 110 9.5 0.001 0.001 3.687 3.795 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.503 3.507 density_rs2pw 110 9.6 0.004 0.005 3.226 3.447 grid_integrate_task_list 110 12.3 3.265 3.404 3.265 3.404 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.302 3.281 prepare_preconditioner 11 7.9 0.000 0.000 3.232 3.251 make_preconditioner 11 8.9 0.000 0.000 3.232 3.251 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.174 3.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.035 3.076 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.021 3.064 pw_transfer 1331 11.6 0.055 0.070 2.978 3.052 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.889 2.967 acc_transpose_blocks 49320 15.4 0.206 0.213 2.524 2.584 fft_wrap_pw1pw2_140 451 13.1 0.309 0.334 2.490 2.572 calculate_first_density_matrix 1 7.0 0.000 0.001 2.553 2.571 jit_kernel_multiply 13 15.9 2.526 2.536 2.526 2.536 potential_pw2rs 110 12.3 0.006 0.006 2.501 2.516 mp_alltoall_d11v 2046 13.8 2.025 2.447 2.025 2.447 grid_collocate_task_list 110 9.6 2.152 2.279 2.152 2.279 fft3d_ps 1111 14.6 0.797 0.884 2.183 2.236 mp_waitany 14300 13.8 1.794 2.052 1.794 2.052 transfer_rs2pw 451 10.6 0.005 0.006 1.822 2.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.895 1.930 mp_sum_d 3891 11.9 1.377 1.909 1.377 1.909 make_images_data 4110 15.4 0.043 0.047 1.752 1.877 hybrid_alltoall_any 4261 16.3 0.083 0.484 1.511 1.828 cp_fm_cholesky_invert 11 10.9 1.762 1.765 1.762 1.765 transfer_pw2rs 451 13.1 0.006 0.007 1.691 1.696 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.638 1.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.545000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=468.636364, yerr=3.198657 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 584.400896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2864676. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.092 70.869 70.871 qs_mol_dyn_low 1 2.0 0.005 0.008 70.243 70.253 qs_forces 11 3.9 0.003 0.003 70.171 70.174 qs_energies 11 4.9 0.002 0.009 66.807 66.811 scf_env_do_scf 11 5.9 0.000 0.001 58.125 58.129 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 50.470 50.471 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.143 38.310 velocity_verlet 10 3.0 0.003 0.006 36.543 36.545 qs_scf_new_mos 99 7.5 0.001 0.001 34.176 34.338 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.175 34.337 ot_scf_mini 99 9.5 0.003 0.003 32.495 32.656 multiply_cannon 2055 13.4 0.224 0.243 31.398 32.512 multiply_cannon_loop 2055 14.4 1.157 1.182 30.123 31.064 ot_mini 99 10.5 0.001 0.001 18.537 18.701 multiply_cannon_multrec 24660 15.4 6.985 8.783 14.035 16.014 rebuild_ks_matrix 110 8.3 0.000 0.000 13.791 13.912 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.791 13.911 qs_ot_get_derivative 99 11.5 0.001 0.001 12.758 12.922 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.139 12.248 mp_waitall_1 176588 16.5 7.661 10.082 7.661 10.082 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.118 8.029 init_scf_loop 11 6.9 0.001 0.005 7.615 7.616 qs_ot_get_p 110 10.4 0.001 0.005 7.275 7.461 multiply_cannon_sync_h2d 24660 15.4 6.480 7.362 6.480 7.362 dbcsr_mm_accdrv_process 52282 16.1 5.545 6.506 6.879 7.189 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.450 7.122 apply_single 110 13.6 0.000 0.001 6.450 7.122 sum_up_and_integrate 110 10.3 0.002 0.003 6.581 6.595 integrate_v_rspace 110 11.3 0.002 0.003 6.554 6.570 init_scf_run 11 5.9 0.000 0.001 6.218 6.218 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.217 6.218 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.151 6.160 calculate_rho_elec 110 8.6 0.039 0.047 6.150 6.160 ot_diis_step 99 11.5 0.010 0.010 5.731 5.731 prepare_preconditioner 11 7.9 0.000 0.000 5.552 5.569 make_preconditioner 11 8.9 0.000 0.001 5.552 5.569 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.785 5.497 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.146 5.308 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.243 5.263 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.768 4.769 make_m2s 4110 13.4 0.057 0.059 4.307 4.721 make_images 4110 14.4 0.406 0.462 4.195 4.604 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.131 4.143 cp_fm_diag_elpa_base 48 14.0 4.081 4.101 4.128 4.140 pw_transfer 1331 11.6 0.066 0.071 3.735 3.884 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.629 3.777 wfi_extrapolate 11 7.9 0.001 0.003 3.583 3.583 density_rs2pw 110 9.6 0.004 0.004 3.376 3.577 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.355 3.432 grid_integrate_task_list 110 12.3 3.157 3.332 3.157 3.332 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.283 3.285 fft_wrap_pw1pw2_140 451 13.1 0.359 0.380 3.022 3.176 calculate_dm_sparse 110 9.5 0.001 0.001 2.985 3.012 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.934 2.975 fft3d_ps 1111 14.6 1.116 1.357 2.722 2.887 hybrid_alltoall_any 4261 16.3 0.105 0.458 2.019 2.752 make_images_data 4110 15.4 0.049 0.053 2.306 2.740 cp_fm_cholesky_invert 11 10.9 2.579 2.586 2.579 2.586 calculate_first_density_matrix 1 7.0 0.000 0.002 2.536 2.539 mp_sum_l 6594 12.7 1.815 2.502 1.815 2.502 potential_pw2rs 110 12.3 0.008 0.009 2.443 2.465 grid_collocate_task_list 110 9.6 2.168 2.323 2.168 2.323 mp_alltoall_d11v 2046 13.8 1.834 2.142 1.834 2.142 acc_transpose_blocks 24660 15.4 0.115 0.118 1.985 2.032 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.985 2.004 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.930 1.932 jit_kernel_multiply 10 16.3 0.971 1.890 0.971 1.890 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.811 1.822 transfer_rs2pw 451 10.6 0.006 0.007 1.616 1.776 mp_allgather_i34 2055 14.4 0.583 1.580 0.583 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.543 1.555 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.753 1.522 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.493 mp_sum_d 3891 11.9 1.002 1.423 1.002 1.423 dbcsr_complete_redistribute 325 12.2 0.230 0.288 1.158 1.421 mp_alltoall_z22v 1111 16.6 1.335 1.421 1.335 1.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.871000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.363636, yerr=7.163959 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.218432E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.057 60.848 60.849 qs_mol_dyn_low 1 2.0 0.005 0.012 60.495 60.504 qs_forces 11 3.9 0.003 0.003 60.420 60.422 qs_energies 11 4.9 0.002 0.005 57.222 57.226 scf_env_do_scf 11 5.9 0.001 0.001 49.303 49.303 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.780 40.780 velocity_verlet 10 3.0 0.002 0.005 32.682 32.684 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.767 29.050 qs_scf_new_mos 99 7.5 0.001 0.001 25.964 26.059 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.964 26.058 ot_scf_mini 99 9.5 0.002 0.003 24.709 24.828 multiply_cannon 2055 13.4 0.213 0.221 22.400 23.589 multiply_cannon_loop 2055 14.4 0.811 0.862 21.238 22.252 ot_mini 99 10.5 0.001 0.001 13.781 13.900 rebuild_ks_matrix 110 8.3 0.000 0.000 12.238 12.400 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.237 12.400 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.764 10.912 multiply_cannon_multrec 16440 15.4 3.696 4.640 9.889 10.767 mp_waitall_1 139946 16.5 6.862 9.987 6.862 9.987 qs_ot_get_derivative 99 11.5 0.001 0.001 9.354 9.474 init_scf_loop 11 6.9 0.001 0.010 8.482 8.483 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.192 7.187 prepare_preconditioner 11 7.9 0.000 0.000 6.728 6.746 make_preconditioner 11 8.9 0.000 0.000 6.728 6.746 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.079 6.434 sum_up_and_integrate 110 10.3 0.001 0.003 6.372 6.387 integrate_v_rspace 110 11.3 0.003 0.003 6.346 6.361 qs_ot_get_p 110 10.4 0.001 0.001 6.085 6.225 dbcsr_mm_accdrv_process 34862 16.1 5.396 5.712 6.037 6.191 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.861 5.871 calculate_rho_elec 110 8.6 0.059 0.060 5.861 5.870 init_scf_run 11 5.9 0.000 0.001 5.487 5.487 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.487 5.487 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.922 5.332 apply_single 110 13.6 0.000 0.000 4.921 5.332 make_m2s 4110 13.4 0.049 0.051 4.246 4.587 make_images 4110 14.4 0.397 0.517 4.129 4.470 ot_diis_step 99 11.5 0.010 0.011 4.396 4.396 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.348 4.351 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.967 3.967 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.182 3.819 multiply_cannon_sync_h2d 16440 15.4 3.264 3.737 3.264 3.737 pw_transfer 1331 11.6 0.065 0.072 3.682 3.701 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.575 3.596 grid_integrate_task_list 110 12.3 3.204 3.431 3.204 3.431 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.377 3.385 cp_fm_diag_elpa_base 48 14.0 3.307 3.339 3.374 3.381 density_rs2pw 110 9.6 0.004 0.004 3.101 3.251 fft_wrap_pw1pw2_140 451 13.1 0.460 0.475 3.065 3.091 wfi_extrapolate 11 7.9 0.001 0.003 3.006 3.006 make_images_data 4110 15.4 0.046 0.051 2.369 2.826 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.789 2.791 hybrid_alltoall_any 4261 16.3 0.108 0.383 2.094 2.748 cp_fm_cholesky_invert 11 10.9 2.648 2.654 2.648 2.654 calculate_dm_sparse 110 9.5 0.001 0.001 2.539 2.568 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.479 2.552 fft3d_ps 1111 14.6 1.121 1.154 2.494 2.511 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.447 2.507 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.879 2.438 calculate_first_density_matrix 1 7.0 0.000 0.002 2.386 2.388 grid_collocate_task_list 110 9.6 2.220 2.362 2.220 2.362 mp_irecv_dv 48980 15.7 0.804 2.309 0.804 2.309 potential_pw2rs 110 12.3 0.011 0.011 2.212 2.228 mp_sum_l 6594 12.7 1.513 2.170 1.513 2.170 mp_alltoall_d11v 2046 13.8 1.764 2.139 1.764 2.139 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.974 1.976 dbcsr_complete_redistribute 325 12.2 0.313 0.341 1.447 1.910 cp_fm_upper_to_full 70 14.2 1.397 1.781 1.397 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.666 1.682 acc_transpose_blocks 16440 15.4 0.078 0.081 1.589 1.679 cp_fm_cholesky_decompose 22 10.9 1.601 1.617 1.601 1.617 mp_allgather_i34 2055 14.4 0.476 1.583 0.476 1.583 transfer_rs2pw 451 10.6 0.005 0.006 1.347 1.520 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.474 1.485 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.473 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.958 1.410 mp_waitany 17072 13.8 1.123 1.323 1.123 1.323 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.288 1.296 transfer_pw2rs 451 13.1 0.005 0.005 1.231 1.244 rs_gather_matrices 110 12.3 0.136 0.149 0.877 1.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.849000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.454545, yerr=8.456989 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 724.987904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 66.775 66.777 qs_mol_dyn_low 1 2.0 0.007 0.018 66.460 66.475 qs_forces 11 3.9 0.004 0.006 66.214 66.218 qs_energies 11 4.9 0.003 0.007 62.806 62.813 scf_env_do_scf 11 5.9 0.001 0.001 54.428 54.430 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.677 42.677 velocity_verlet 10 3.0 0.002 0.005 37.561 37.570 dbcsr_multiply_generic 2055 12.4 0.117 0.123 30.635 30.832 qs_scf_new_mos 99 7.5 0.001 0.001 27.906 28.009 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.906 28.008 ot_scf_mini 99 9.5 0.003 0.003 26.204 26.307 multiply_cannon 2055 13.4 0.239 0.256 23.572 24.234 multiply_cannon_loop 2055 14.4 1.390 1.453 22.022 22.566 ot_mini 99 10.5 0.001 0.001 14.764 14.898 multiply_cannon_multrec 24660 15.4 4.106 6.764 13.062 14.216 rebuild_ks_matrix 110 8.3 0.000 0.000 12.081 12.187 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.024 12.081 12.187 init_scf_loop 11 6.9 0.007 0.029 11.708 11.710 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.675 10.772 qs_ot_get_derivative 99 11.5 0.001 0.001 10.600 10.706 prepare_preconditioner 11 7.9 0.000 0.000 9.889 9.903 make_preconditioner 11 8.9 0.000 0.001 9.889 9.903 dbcsr_mm_accdrv_process 52304 16.0 7.818 9.188 8.798 9.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.133 9.581 qs_ot_get_p 110 10.4 0.002 0.012 6.315 6.469 sum_up_and_integrate 110 10.3 0.002 0.003 6.307 6.319 integrate_v_rspace 110 11.3 0.003 0.005 6.281 6.292 mp_waitall_1 121746 16.5 4.110 6.090 4.110 6.090 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.928 5.941 calculate_rho_elec 110 8.6 0.077 0.081 5.928 5.941 make_m2s 4110 13.4 0.059 0.061 5.404 5.853 make_images 4110 14.4 0.571 0.691 5.262 5.708 init_scf_run 11 5.9 0.000 0.001 5.571 5.572 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.571 5.572 cp_fm_upper_to_full 70 14.2 3.354 4.854 3.354 4.854 qs_ot_p2m_diag 48 11.0 0.057 0.071 4.357 4.375 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.044 4.129 apply_single 110 13.6 0.000 0.000 4.043 4.129 ot_diis_step 99 11.5 0.011 0.011 4.127 4.127 pw_transfer 1331 11.6 0.065 0.072 3.825 3.864 dbcsr_complete_redistribute 325 12.2 0.410 0.455 2.675 3.807 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.791 3.793 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.718 3.761 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.498 3.554 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.361 3.424 grid_integrate_task_list 110 12.3 3.277 3.407 3.277 3.407 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.128 3.242 density_rs2pw 110 9.6 0.004 0.004 3.058 3.221 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.213 3.220 cp_fm_diag_elpa_base 48 14.0 3.061 3.118 3.210 3.218 fft_wrap_pw1pw2_140 451 13.1 0.531 0.544 3.172 3.217 make_images_data 4110 15.4 0.049 0.052 2.633 3.100 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.309 3.099 calculate_dm_sparse 110 9.5 0.001 0.001 2.999 3.029 wfi_extrapolate 11 7.9 0.001 0.004 3.020 3.020 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.301 2.967 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.747 2.844 mp_alltoall_i22 605 13.7 1.657 2.827 1.657 2.827 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.814 2.816 cp_fm_cholesky_invert 11 10.9 2.678 2.687 2.678 2.687 fft3d_ps 1111 14.6 1.153 1.189 2.546 2.563 multiply_cannon_sync_h2d 24660 15.4 2.370 2.557 2.370 2.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.487 acc_transpose_blocks 24660 15.4 0.110 0.113 2.366 2.467 calculate_first_density_matrix 1 7.0 0.001 0.002 2.454 2.458 grid_collocate_task_list 110 9.6 2.264 2.388 2.264 2.388 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.296 2.301 potential_pw2rs 110 12.3 0.012 0.013 2.116 2.124 mp_alltoall_d11v 2046 13.8 1.780 1.988 1.780 1.988 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.809 1.850 cp_fm_cholesky_decompose 22 10.9 1.678 1.719 1.678 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.713 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.648 1.658 mp_sum_l 6594 12.7 0.979 1.579 0.979 1.579 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.543 1.563 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.848 1.525 acc_transpose_blocks_sync 73980 16.4 1.354 1.445 1.354 1.445 mp_irecv_dv 62702 16.1 0.743 1.444 0.743 1.444 transfer_rs2pw 451 10.6 0.005 0.006 1.265 1.434 jit_kernel_multiply 8 16.2 0.641 1.430 0.641 1.430 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.777000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.090909, yerr=8.836438 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 855.879680E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.038 54.755 54.755 qs_mol_dyn_low 1 2.0 0.003 0.003 54.483 54.492 qs_forces 11 3.9 0.004 0.005 54.417 54.418 qs_energies 11 4.9 0.003 0.004 50.780 50.784 scf_env_do_scf 11 5.9 0.000 0.001 42.499 42.499 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.940 34.941 velocity_verlet 10 3.0 0.002 0.002 30.877 30.880 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.039 23.228 qs_scf_new_mos 99 7.5 0.001 0.001 20.392 20.435 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.391 20.435 ot_scf_mini 99 9.5 0.002 0.002 19.141 19.163 multiply_cannon 2055 13.4 0.241 0.251 17.564 19.028 multiply_cannon_loop 2055 14.4 0.606 0.636 16.299 16.696 rebuild_ks_matrix 110 8.3 0.000 0.000 11.595 11.615 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.594 11.615 ot_mini 99 10.5 0.001 0.001 10.646 10.663 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.313 10.332 multiply_cannon_multrec 8220 15.4 3.175 4.323 7.574 8.526 init_scf_loop 11 6.9 0.000 0.000 7.514 7.515 mp_waitall_1 103326 16.6 5.722 7.482 5.722 7.482 qs_ot_get_derivative 99 11.5 0.001 0.001 6.983 7.004 sum_up_and_integrate 110 10.3 0.001 0.002 6.161 6.176 integrate_v_rspace 110 11.3 0.003 0.004 6.134 6.149 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.997 6.013 calculate_rho_elec 110 8.6 0.112 0.113 5.997 6.012 prepare_preconditioner 11 7.9 0.000 0.000 5.879 5.883 make_preconditioner 11 8.9 0.000 0.000 5.879 5.883 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.473 5.547 dbcsr_mm_accdrv_process 17442 15.9 3.131 4.155 4.259 5.185 init_scf_run 11 5.9 0.000 0.001 5.084 5.084 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.083 5.083 qs_ot_get_p 110 10.4 0.001 0.001 4.559 4.585 make_m2s 4110 13.4 0.037 0.039 4.153 4.439 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.992 4.427 make_images 4110 14.4 0.644 0.704 4.023 4.309 pw_transfer 1331 11.6 0.066 0.071 4.033 4.045 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.925 3.940 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.712 3.767 apply_single 110 13.6 0.000 0.000 3.712 3.767 ot_diis_step 99 11.5 0.012 0.012 3.641 3.641 grid_integrate_task_list 110 12.3 3.384 3.505 3.384 3.505 fft_wrap_pw1pw2_140 451 13.1 0.717 0.726 3.421 3.434 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.221 3.225 density_rs2pw 110 9.6 0.004 0.004 2.980 3.186 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.886 2.886 cp_fm_cholesky_invert 11 10.9 2.791 2.795 2.791 2.795 hybrid_alltoall_any 4261 16.3 0.199 0.852 2.267 2.781 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.745 2.745 wfi_extrapolate 11 7.9 0.001 0.001 2.685 2.685 make_images_data 4110 15.4 0.040 0.046 2.228 2.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.523 2.524 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.512 fft3d_ps 1111 14.6 1.293 1.308 2.470 2.483 grid_collocate_task_list 110 9.6 2.362 2.480 2.362 2.480 multiply_cannon_sync_h2d 8220 15.4 2.372 2.473 2.372 2.473 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.327 2.335 cp_fm_diag_elpa_base 48 14.0 2.272 2.294 2.325 2.333 calculate_first_density_matrix 1 7.0 0.000 0.000 2.304 2.305 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.119 2.133 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.983 2.022 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.788 2.000 potential_pw2rs 110 12.3 0.015 0.015 1.972 1.975 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.885 1.895 mp_alltoall_d11v 2046 13.8 1.691 1.855 1.691 1.855 cp_fm_cholesky_decompose 22 10.9 1.652 1.662 1.652 1.662 qs_env_update_s_mstruct 11 6.9 0.031 0.034 1.548 1.654 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.607 1.609 dbcsr_complete_redistribute 325 12.2 0.602 0.631 1.448 1.542 mp_allgather_i34 2055 14.4 0.434 1.518 0.434 1.518 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.428 1.442 acc_transpose_blocks 8220 15.4 0.041 0.042 1.373 1.413 jit_kernel_multiply 8 15.4 0.815 1.362 0.815 1.362 transfer_rs2pw 451 10.6 0.005 0.005 1.099 1.334 qs_create_task_list 11 7.9 0.013 0.014 1.221 1.308 mp_waitany 9240 13.8 1.049 1.300 1.049 1.300 generate_qs_task_list 11 8.9 0.374 0.440 1.208 1.294 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.187 1.214 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.755000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=803.272727, yerr=11.855884 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 1.390846E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.061 86.825 86.826 qs_mol_dyn_low 1 2.0 0.003 0.003 86.417 86.426 qs_forces 11 3.9 0.003 0.003 86.353 86.353 qs_energies 11 4.9 0.002 0.002 82.233 82.235 scf_env_do_scf 11 5.9 0.000 0.001 72.293 72.293 velocity_verlet 10 3.0 0.002 0.002 55.599 55.605 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.153 44.154 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.194 29.313 init_scf_loop 11 6.9 0.000 0.000 28.070 28.072 qs_scf_new_mos 99 7.5 0.001 0.001 26.392 26.428 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.391 26.428 prepare_preconditioner 11 7.9 0.000 0.000 26.071 26.079 make_preconditioner 11 8.9 0.000 0.000 26.071 26.079 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.303 25.549 ot_scf_mini 99 9.5 0.002 0.002 24.611 24.639 multiply_cannon 2055 13.4 0.333 0.352 22.299 23.043 multiply_cannon_loop 2055 14.4 0.824 0.839 20.540 20.836 cp_fm_upper_to_full 70 14.2 12.757 18.242 12.757 18.242 ot_mini 99 10.5 0.001 0.001 13.963 13.995 rebuild_ks_matrix 110 8.3 0.000 0.001 13.845 13.885 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.845 13.884 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.563 12.599 dbcsr_complete_redistribute 325 12.2 1.012 1.035 7.314 10.502 multiply_cannon_multrec 8220 15.4 4.078 4.244 9.759 9.886 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.300 9.498 qs_ot_get_derivative 99 11.5 0.001 0.001 9.408 9.438 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.752 8.921 mp_alltoall_i22 605 13.7 5.374 8.597 5.374 8.597 mp_waitall_1 84994 16.7 7.299 8.147 7.299 8.147 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.396 7.433 calculate_rho_elec 110 8.6 0.222 0.223 7.396 7.432 sum_up_and_integrate 110 10.3 0.002 0.002 6.834 6.849 integrate_v_rspace 110 11.3 0.004 0.004 6.806 6.820 pw_transfer 1331 11.6 0.075 0.075 5.754 5.759 dbcsr_mm_accdrv_process 11614 15.7 3.940 4.151 5.533 5.755 make_m2s 4110 13.4 0.043 0.043 5.219 5.710 init_scf_run 11 5.9 0.000 0.001 5.652 5.652 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.651 5.651 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.636 5.642 make_images 4110 14.4 0.881 0.937 5.030 5.521 qs_ot_get_p 110 10.4 0.001 0.001 5.392 5.421 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.707 5.175 apply_single 110 13.6 0.000 0.000 4.707 5.175 cp_fm_cholesky_invert 11 10.9 5.082 5.086 5.082 5.086 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.754 5.084 fft_wrap_pw1pw2_140 451 13.1 1.368 1.373 4.960 4.969 ot_diis_step 99 11.5 0.015 0.015 4.539 4.539 density_rs2pw 110 9.6 0.004 0.004 3.886 3.891 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.866 3.875 grid_integrate_task_list 110 12.3 3.671 3.732 3.671 3.732 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.687 3.687 hybrid_alltoall_any 4261 16.3 0.264 0.562 2.764 3.491 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.976 3.439 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.354 3.355 fft3d_ps 1111 14.6 1.882 1.888 3.341 3.350 make_images_data 4110 15.4 0.045 0.048 2.731 3.326 wfi_extrapolate 11 7.9 0.001 0.001 3.311 3.311 calculate_dm_sparse 110 9.5 0.001 0.001 3.145 3.174 multiply_cannon_sync_h2d 8220 15.4 3.121 3.157 3.121 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.839 2.840 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.779 2.780 cp_fm_diag_elpa_base 48 14.0 2.256 2.446 2.778 2.778 grid_collocate_task_list 110 9.6 2.662 2.674 2.662 2.674 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.609 2.625 potential_pw2rs 110 12.3 0.021 0.021 2.393 2.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.353 2.367 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.204 2.260 calculate_first_density_matrix 1 7.0 0.000 0.000 2.224 2.225 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.086 2.179 mp_alltoall_d11v 2046 13.8 1.964 2.011 1.964 2.011 cp_fm_cholesky_decompose 22 10.9 1.911 1.944 1.911 1.944 qs_create_task_list 11 7.9 0.001 0.001 1.893 1.941 generate_qs_task_list 11 8.9 0.735 0.791 1.892 1.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.880 1.884 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.739 1.775 acc_transpose_blocks 8220 15.4 0.041 0.042 1.722 1.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.826000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1242.363636, yerr=59.605175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 1.103378E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 12049536 0.0% 0.0% 100.0% average stack size 0.0 0.0 564.9 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 623.132672E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10439712 MPI messages size (bytes): total size 4.526312E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.566750E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3611304 96569262080 32768 < size <= 131072 1307392 74803314688 131072 < size <= 4194304 5190114 3200150010616 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57182. MP_Allreduce 11186 792. MP_Sync 88 MP_Alltoall 2242 2530004. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.046 207.426 207.428 qs_mol_dyn_low 1 2.0 0.007 0.020 206.839 206.853 qs_forces 11 3.9 0.005 0.006 206.686 206.689 qs_energies 11 4.9 0.002 0.004 201.128 201.145 scf_env_do_scf 11 5.9 0.001 0.001 184.771 184.775 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 164.566 164.568 dbcsr_multiply_generic 2529 12.6 0.180 0.185 126.604 127.874 qs_scf_new_mos 118 7.6 0.001 0.001 125.208 125.408 qs_scf_loop_do_ot 118 8.6 0.001 0.001 125.207 125.407 velocity_verlet 10 3.0 0.014 0.045 124.238 124.241 ot_scf_mini 118 9.6 0.003 0.003 118.444 118.589 multiply_cannon 2529 13.6 0.235 0.244 102.388 104.369 multiply_cannon_loop 2529 14.6 2.416 2.454 100.290 102.328 ot_mini 118 10.6 0.001 0.001 67.141 67.311 multiply_cannon_multrec 60696 15.6 32.116 34.016 41.940 44.078 qs_ot_get_derivative 118 11.6 0.001 0.001 41.854 42.024 rebuild_ks_matrix 129 8.3 0.001 0.001 33.538 33.764 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.019 33.538 33.764 mp_waitall_1 269446 16.5 29.046 32.243 29.046 32.243 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.143 30.333 qs_ot_get_p 129 10.4 0.001 0.001 28.910 29.120 multiply_cannon_sync_h2d 60696 15.6 26.487 28.198 26.487 28.198 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.597 26.250 apply_single 129 13.6 0.001 0.001 24.596 26.250 ot_diis_step 118 11.6 0.008 0.008 24.980 24.981 qs_ot_p2m_diag 84 11.4 0.080 0.092 22.343 22.417 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 20.220 20.337 init_scf_loop 11 6.9 0.000 0.000 20.133 20.134 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.482 19.482 multiply_cannon_metrocomm3 60696 15.6 0.116 0.121 16.322 18.028 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.545 16.573 cp_fm_diag_elpa_base 84 14.4 16.478 16.510 16.541 16.570 prepare_preconditioner 11 7.9 0.000 0.000 15.521 15.559 make_preconditioner 11 8.9 0.000 0.000 15.521 15.559 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.765 14.964 make_m2s 5058 13.6 0.105 0.114 13.823 14.209 make_images 5058 14.6 0.398 0.417 13.638 14.026 sum_up_and_integrate 129 10.3 0.002 0.004 13.915 13.930 integrate_v_rspace 129 11.3 0.003 0.004 13.857 13.875 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.041 13.164 calculate_rho_elec 129 8.7 0.045 0.064 13.041 13.164 init_scf_run 11 5.9 0.000 0.001 12.199 12.200 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.199 12.200 mp_sum_l 8016 12.9 9.138 10.266 9.138 10.266 dbcsr_mm_accdrv_process 125516 16.2 4.781 4.908 9.388 9.905 wfi_extrapolate 11 7.9 0.001 0.004 8.979 8.979 cp_fm_cholesky_invert 11 10.9 8.777 8.785 8.777 8.785 calculate_dm_sparse 129 9.5 0.001 0.001 8.566 8.667 multiply_cannon_metrocomm1 60696 15.6 0.093 0.096 6.372 8.521 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.181 8.258 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.137 8.248 make_images_data 5058 15.6 0.067 0.074 6.737 7.722 pw_transfer 1559 11.6 0.074 0.094 7.536 7.718 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.332 7.516 grid_integrate_task_list 129 12.3 7.118 7.485 7.118 7.485 hybrid_alltoall_any 5245 16.5 0.294 2.265 5.923 7.371 density_rs2pw 129 9.7 0.006 0.006 6.630 7.197 fft_wrap_pw1pw2_140 527 13.2 0.863 0.906 6.482 6.661 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.569 6.580 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.732 5.818 fft3d_ps 1301 14.7 2.210 2.869 5.322 5.659 mp_alltoall_d11v 2429 14.1 4.280 5.572 4.280 5.572 grid_collocate_task_list 129 9.7 4.870 5.263 4.870 5.263 cp_fm_cholesky_decompose 22 10.9 4.569 4.583 4.569 4.583 mp_sum_d 4511 12.1 3.666 4.405 3.666 4.405 potential_pw2rs 129 12.3 0.009 0.010 4.365 4.384 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.428000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=590.727273, yerr=7.398905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 837.771264E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 4636633. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 189.971 189.972 qs_mol_dyn_low 1 2.0 0.004 0.007 189.614 189.630 qs_forces 11 3.9 0.004 0.004 189.399 189.401 qs_energies 11 4.9 0.028 0.029 182.654 182.669 scf_env_do_scf 11 5.9 0.001 0.001 166.310 166.320 scf_env_do_scf_inner_loop 118 6.6 0.006 0.014 133.778 133.780 velocity_verlet 10 3.0 0.002 0.006 120.063 120.067 dbcsr_multiply_generic 2527 12.6 0.189 0.194 97.817 99.031 qs_scf_new_mos 118 7.6 0.001 0.001 94.805 95.331 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.804 95.330 ot_scf_mini 118 9.6 0.004 0.004 89.960 90.616 multiply_cannon 2527 13.6 0.508 0.582 77.657 82.422 multiply_cannon_loop 2527 14.6 1.573 1.641 74.234 76.952 ot_mini 118 10.6 0.001 0.001 50.575 51.152 mp_waitall_1 216438 16.6 24.048 38.836 24.048 38.836 multiply_cannon_multrec 30324 15.6 21.011 24.722 31.655 35.665 rebuild_ks_matrix 129 8.3 0.001 0.001 32.481 33.074 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.481 33.073 init_scf_loop 11 6.9 0.000 0.000 32.442 32.444 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.232 29.770 qs_ot_get_derivative 118 11.6 0.001 0.002 28.449 29.088 multiply_cannon_metrocomm3 30324 15.6 0.100 0.106 15.763 28.769 prepare_preconditioner 11 7.9 0.000 0.000 28.144 28.205 make_preconditioner 11 8.9 0.000 0.000 28.144 28.205 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.844 27.391 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.187 23.396 apply_single 129 13.6 0.001 0.001 22.186 23.395 qs_ot_get_p 129 10.4 0.001 0.001 21.296 22.028 ot_diis_step 118 11.6 0.014 0.015 21.943 21.945 multiply_cannon_sync_h2d 30324 15.6 18.061 20.758 18.061 20.758 cp_fm_cholesky_invert 11 10.9 16.405 16.417 16.405 16.417 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.329 16.364 make_m2s 5054 13.6 0.087 0.092 14.176 15.876 make_images 5054 14.6 1.169 1.377 13.959 15.660 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.125 15.126 sum_up_and_integrate 129 10.3 0.002 0.004 14.118 14.147 integrate_v_rspace 129 11.3 0.003 0.004 14.057 14.089 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.580 13.614 calculate_rho_elec 129 8.7 0.088 0.105 13.579 13.613 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.917 11.957 cp_fm_diag_elpa_base 83 14.4 11.662 11.748 11.910 11.946 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.212 11.677 init_scf_run 11 5.9 0.000 0.001 11.467 11.468 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.466 11.468 multiply_cannon_metrocomm4 27797 15.6 0.100 0.115 3.949 10.938 dbcsr_mm_accdrv_process 62734 16.2 5.502 6.210 10.091 10.653 mp_irecv_dv 70031 16.3 3.743 10.537 3.743 10.537 make_images_data 5054 15.6 0.067 0.078 8.262 10.314 hybrid_alltoall_any 5240 16.5 0.350 1.522 7.123 10.111 pw_transfer 1559 11.6 0.085 0.106 8.702 8.762 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.476 8.540 wfi_extrapolate 11 7.9 0.001 0.001 8.384 8.384 grid_integrate_task_list 129 12.3 7.230 7.632 7.230 7.632 fft_wrap_pw1pw2_140 527 13.2 0.940 0.968 7.482 7.556 density_rs2pw 129 9.7 0.006 0.006 7.134 7.539 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.412 7.260 cp_fm_cholesky_decompose 22 10.9 6.844 6.937 6.844 6.937 calculate_dm_sparse 129 9.5 0.001 0.001 6.500 6.662 mp_sum_l 8010 12.9 4.173 6.250 4.173 6.250 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.115 6.124 fft3d_ps 1301 14.7 2.854 3.038 6.014 6.075 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.430 5.508 grid_collocate_task_list 129 9.7 5.047 5.493 5.047 5.493 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.206 5.371 mp_alltoall_d11v 2423 14.1 4.402 5.177 4.402 5.177 mp_allgather_i34 2527 14.6 1.939 4.916 1.939 4.916 potential_pw2rs 129 12.3 0.015 0.018 4.519 4.533 mp_sum_d 4501 12.1 2.644 3.981 2.644 3.981 dbcsr_complete_redistribute 395 12.7 0.765 0.850 3.139 3.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.972000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=794.000000, yerr=2.374103 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+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 6755942624 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 932.204544E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. 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 0.023 0.047 174.338 174.339 qs_mol_dyn_low 1 2.0 0.017 0.043 173.854 173.876 qs_forces 11 3.9 0.007 0.026 173.624 173.627 qs_energies 11 4.9 0.003 0.011 167.014 167.026 scf_env_do_scf 11 5.9 0.004 0.025 151.448 151.450 scf_env_do_scf_inner_loop 117 6.6 0.062 0.080 116.996 117.005 velocity_verlet 10 3.0 0.039 0.108 111.791 111.800 dbcsr_multiply_generic 2507 12.6 0.184 0.188 81.854 83.023 qs_scf_new_mos 117 7.6 0.001 0.001 79.868 80.187 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.867 80.186 ot_scf_mini 117 9.6 0.004 0.007 75.634 76.000 multiply_cannon 2507 13.6 0.507 0.531 62.281 67.236 multiply_cannon_loop 2507 14.6 1.125 1.182 59.272 61.899 ot_mini 117 10.6 0.001 0.001 42.603 42.958 mp_waitall_1 170520 16.6 24.533 34.822 24.533 34.822 init_scf_loop 11 6.9 0.001 0.004 34.349 34.387 rebuild_ks_matrix 128 8.3 0.001 0.001 30.158 30.624 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 30.157 30.623 prepare_preconditioner 11 7.9 0.000 0.000 30.354 30.392 make_preconditioner 11 8.9 0.000 0.001 30.354 30.392 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.035 29.416 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.165 27.590 multiply_cannon_multrec 20056 15.6 13.113 15.892 22.445 25.276 multiply_cannon_metrocomm3 20056 15.6 0.061 0.066 15.282 24.969 qs_ot_get_derivative 117 11.6 0.001 0.002 23.106 23.481 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.553 20.609 apply_single 128 13.6 0.001 0.001 19.553 20.609 ot_diis_step 117 11.6 0.017 0.018 19.385 19.385 qs_ot_get_p 128 10.4 0.001 0.001 18.472 18.928 make_m2s 5014 13.6 0.078 0.085 14.489 15.803 make_images 5014 14.6 1.140 1.238 14.253 15.561 multiply_cannon_sync_h2d 20056 15.6 13.637 15.299 13.637 15.299 qs_ot_p2m_diag 83 11.4 0.268 0.294 14.123 14.131 cp_fm_cholesky_invert 11 10.9 14.043 14.052 14.043 14.052 sum_up_and_integrate 128 10.3 0.002 0.004 13.914 13.938 integrate_v_rspace 128 11.3 0.003 0.004 13.854 13.880 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.596 13.632 calculate_rho_elec 128 8.7 0.129 0.143 13.595 13.632 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.068 13.072 init_scf_run 11 5.9 0.000 0.001 10.431 10.432 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.431 10.431 make_images_data 5014 15.6 0.064 0.074 8.652 10.407 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.889 9.916 cp_fm_diag_elpa_base 83 14.4 9.482 9.635 9.886 9.912 hybrid_alltoall_any 5200 16.5 0.446 2.024 7.422 9.911 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.120 9.398 multiply_cannon_metrocomm4 17549 15.6 0.065 0.073 3.486 9.219 mp_irecv_dv 50230 16.2 3.356 8.957 3.356 8.957 dbcsr_mm_accdrv_process 41502 16.2 5.622 5.968 8.779 8.938 pw_transfer 1547 11.6 0.084 0.100 8.774 8.873 fft_wrap_pw1pw2 1291 12.7 0.009 0.010 8.552 8.658 fft_wrap_pw1pw2_140 523 13.2 1.064 1.103 7.613 7.725 grid_integrate_task_list 128 12.3 7.296 7.647 7.296 7.647 wfi_extrapolate 11 7.9 0.001 0.003 7.430 7.430 density_rs2pw 128 9.7 0.006 0.006 6.883 7.234 cp_fm_upper_to_full 105 14.8 5.727 7.228 5.727 7.228 cp_fm_cholesky_decompose 22 10.9 7.203 7.221 7.203 7.221 dbcsr_complete_redistribute 395 12.7 1.166 1.196 4.493 6.268 fft3d_ps 1291 14.7 2.820 3.057 5.862 5.947 calculate_dm_sparse 128 9.5 0.001 0.001 5.831 5.926 grid_collocate_task_list 128 9.7 5.157 5.630 5.157 5.630 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.434 5.440 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.686 5.311 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.348 5.112 mp_sum_l 7950 12.9 3.264 4.977 3.264 4.977 mp_alltoall_d11v 2415 14.1 4.371 4.824 4.371 4.824 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.002 4.540 4.657 mp_allgather_i34 2507 14.6 1.514 4.406 1.514 4.406 potential_pw2rs 128 12.3 0.020 0.021 4.324 4.334 transfer_fm_to_dbcsr 11 9.9 0.020 0.026 2.299 4.026 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.909 3.945 mp_alltoall_i22 716 14.1 1.945 3.862 1.945 3.862 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.822 3.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.490 3.535 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.339000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=881.272727, yerr=7.249964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.139044E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57894. MP_Allreduce 11137 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.032 188.029 188.030 qs_mol_dyn_low 1 2.0 0.003 0.003 187.614 187.630 qs_forces 11 3.9 0.004 0.004 187.322 187.329 qs_energies 11 4.9 0.002 0.002 180.220 180.232 scf_env_do_scf 11 5.9 0.001 0.001 163.633 163.645 velocity_verlet 10 3.0 0.012 0.020 125.169 125.179 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 117.574 117.575 dbcsr_multiply_generic 2527 12.6 0.193 0.199 80.801 81.494 qs_scf_new_mos 118 7.6 0.001 0.001 80.815 81.125 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.814 81.124 ot_scf_mini 118 9.6 0.003 0.003 76.224 76.527 multiply_cannon 2527 13.6 0.555 0.587 56.025 58.827 multiply_cannon_loop 2527 14.6 1.835 1.905 52.352 54.076 init_scf_loop 11 6.9 0.000 0.000 45.937 45.938 ot_mini 118 10.6 0.001 0.001 43.038 43.339 prepare_preconditioner 11 7.9 0.000 0.000 41.843 41.864 make_preconditioner 11 8.9 0.000 0.000 41.843 41.864 make_full_inverse_cholesky 11 9.9 0.011 0.020 35.534 40.535 multiply_cannon_multrec 30324 15.6 13.841 19.032 26.823 31.668 rebuild_ks_matrix 129 8.3 0.001 0.001 29.639 29.936 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 29.639 29.936 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.765 27.026 mp_waitall_1 149060 16.7 17.165 26.451 17.165 26.451 qs_ot_get_derivative 118 11.6 0.001 0.002 23.402 23.705 make_m2s 5054 13.6 0.093 0.097 20.411 21.162 make_images 5054 14.6 1.940 2.269 20.098 20.847 qs_ot_get_p 129 10.4 0.001 0.001 19.231 19.577 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.058 19.570 apply_single 129 13.6 0.001 0.001 19.058 19.570 ot_diis_step 118 11.6 0.018 0.019 19.507 19.509 cp_fm_upper_to_full 105 14.8 11.453 16.860 11.453 16.860 cp_fm_cholesky_invert 11 10.9 15.928 15.937 15.928 15.937 multiply_cannon_metrocomm3 30324 15.6 0.050 0.052 6.268 15.116 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.009 15.057 sum_up_and_integrate 129 10.3 0.002 0.003 14.138 14.167 integrate_v_rspace 129 11.3 0.003 0.004 14.078 14.108 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.910 13.956 calculate_rho_elec 129 8.7 0.171 0.187 13.909 13.956 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.611 13.613 dbcsr_mm_accdrv_process 62756 16.2 8.555 9.251 12.549 13.069 dbcsr_complete_redistribute 395 12.7 1.481 1.596 8.993 12.698 make_images_data 5054 15.6 0.066 0.073 10.998 12.618 hybrid_alltoall_any 5240 16.5 0.535 2.217 9.674 11.934 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.645 11.374 multiply_cannon_sync_h2d 30324 15.6 10.520 11.367 10.520 11.367 init_scf_run 11 5.9 0.000 0.001 10.679 10.680 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.679 10.680 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.411 10.425 cp_fm_diag_elpa_base 83 14.4 9.437 9.751 10.404 10.417 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.749 9.967 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.286 9.945 pw_transfer 1559 11.6 0.086 0.102 9.314 9.373 mp_alltoall_i22 718 14.1 5.583 9.288 5.583 9.288 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.086 9.151 fft_wrap_pw1pw2_140 527 13.2 1.262 1.303 8.090 8.167 grid_integrate_task_list 129 12.3 7.566 7.901 7.566 7.901 wfi_extrapolate 11 7.9 0.001 0.001 7.602 7.602 cp_fm_cholesky_decompose 22 10.9 7.448 7.540 7.448 7.540 density_rs2pw 129 9.7 0.005 0.006 6.888 7.200 multiply_cannon_metrocomm4 25270 15.6 0.079 0.088 2.781 7.052 mp_irecv_dv 76685 16.2 2.626 6.774 2.626 6.774 calculate_dm_sparse 129 9.5 0.001 0.001 6.268 6.370 fft3d_ps 1301 14.7 3.019 3.099 6.106 6.180 mp_alltoall_d11v 2423 14.1 4.946 5.650 4.946 5.650 grid_collocate_task_list 129 9.7 5.329 5.635 5.329 5.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.389 5.436 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.581 4.672 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.331 4.440 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.437 4.439 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.263 4.339 potential_pw2rs 129 12.3 0.023 0.024 4.307 4.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.030000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.272727, yerr=16.006197 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.521046E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 169.464 169.465 qs_mol_dyn_low 1 2.0 0.003 0.003 169.086 169.098 qs_forces 11 3.9 0.004 0.004 168.998 168.999 qs_energies 11 4.9 0.002 0.002 161.549 161.556 scf_env_do_scf 11 5.9 0.001 0.001 144.368 144.381 velocity_verlet 10 3.0 0.002 0.002 111.393 111.396 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 108.684 108.686 dbcsr_multiply_generic 2507 12.6 0.184 0.191 71.620 71.926 qs_scf_new_mos 117 7.6 0.001 0.001 71.513 71.617 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.512 71.617 ot_scf_mini 117 9.6 0.003 0.004 67.117 67.217 multiply_cannon 2507 13.6 0.563 0.588 53.051 56.346 multiply_cannon_loop 2507 14.6 0.808 0.839 50.148 50.831 ot_mini 117 10.6 0.001 0.001 37.319 37.414 init_scf_loop 11 6.9 0.000 0.000 35.532 35.533 prepare_preconditioner 11 7.9 0.000 0.000 31.594 31.610 make_preconditioner 11 8.9 0.000 0.000 31.594 31.610 mp_waitall_1 125778 16.7 24.539 30.680 24.539 30.680 make_full_inverse_cholesky 11 9.9 0.016 0.027 29.532 29.813 rebuild_ks_matrix 128 8.3 0.001 0.001 29.369 29.484 qs_ks_build_kohn_sham_matrix 128 9.3 0.036 0.038 29.369 29.484 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.771 26.877 multiply_cannon_multrec 10028 15.6 10.375 14.571 17.935 21.044 qs_ot_get_derivative 117 11.6 0.002 0.002 20.410 20.509 multiply_cannon_metrocomm3 10028 15.6 0.024 0.026 12.423 19.477 cp_fm_cholesky_invert 11 10.9 18.043 18.048 18.043 18.048 qs_ot_get_p 128 10.4 0.001 0.001 17.151 17.283 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.926 17.039 apply_single 128 13.6 0.001 0.001 16.926 17.039 ot_diis_step 117 11.6 0.020 0.020 16.839 16.839 make_m2s 5014 13.6 0.065 0.069 14.702 15.502 make_images 5014 14.6 2.150 2.599 14.393 15.190 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.290 14.327 calculate_rho_elec 128 8.7 0.253 0.264 14.290 14.327 sum_up_and_integrate 128 10.3 0.002 0.003 14.237 14.290 integrate_v_rspace 128 11.3 0.004 0.004 14.177 14.229 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.395 13.412 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.231 12.232 multiply_cannon_sync_h2d 10028 15.6 10.854 11.186 10.854 11.186 init_scf_run 11 5.9 0.000 0.001 10.336 10.337 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.336 10.336 pw_transfer 1547 11.6 0.085 0.096 9.954 9.992 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.731 9.775 hybrid_alltoall_any 5200 16.5 0.840 3.775 8.328 9.741 make_images_data 5014 15.6 0.054 0.064 8.458 9.732 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.366 9.380 cp_fm_diag_elpa_base 83 14.4 9.129 9.197 9.363 9.375 fft_wrap_pw1pw2_140 523 13.2 1.675 1.712 8.609 8.653 grid_integrate_task_list 128 12.3 7.770 8.098 7.770 8.098 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.012 8.080 cp_fm_cholesky_decompose 22 10.9 7.901 8.030 7.901 8.030 dbcsr_mm_accdrv_process 20762 16.1 3.358 4.461 7.210 7.890 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.425 7.370 wfi_extrapolate 11 7.9 0.001 0.001 7.308 7.309 density_rs2pw 128 9.7 0.005 0.005 6.995 7.213 fft3d_ps 1291 14.7 3.143 3.241 6.181 6.226 calculate_dm_sparse 128 9.5 0.001 0.001 6.034 6.113 grid_collocate_task_list 128 9.7 5.566 5.739 5.566 5.739 dbcsr_complete_redistribute 395 12.7 2.092 2.150 5.149 5.563 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.261 5.261 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.139 5.150 mp_alltoall_d11v 2415 14.1 4.566 5.000 4.566 5.000 mp_allgather_i34 2507 14.6 1.096 4.523 1.096 4.523 potential_pw2rs 128 12.3 0.026 0.026 4.384 4.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.052 4.108 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.590 3.899 multiply_cannon_metrocomm4 7521 15.6 0.026 0.028 1.736 3.758 mp_irecv_dv 28860 15.9 1.695 3.678 1.695 3.678 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.592 3.652 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.324 3.634 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.555 3.581 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.434 3.517 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.504 3.513 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.465000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1444.454545, yerr=9.335793 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.139621E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59902. MP_Allreduce 10950 1507. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.042 293.176 293.177 qs_mol_dyn_low 1 2.0 0.003 0.003 292.580 292.599 qs_forces 11 3.9 0.004 0.005 292.481 292.485 qs_energies 11 4.9 0.002 0.002 283.595 283.599 scf_env_do_scf 11 5.9 0.001 0.001 260.910 260.916 velocity_verlet 10 3.0 0.002 0.002 210.721 210.730 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 136.088 136.091 init_scf_loop 11 6.9 0.000 0.000 124.545 124.549 prepare_preconditioner 11 7.9 0.000 0.000 119.656 119.680 make_preconditioner 11 8.9 0.000 0.000 119.656 119.679 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.602 116.748 qs_scf_new_mos 116 7.6 0.001 0.001 91.452 91.543 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.451 91.542 ot_scf_mini 116 9.6 0.004 0.004 86.626 86.662 dbcsr_multiply_generic 2485 12.5 0.214 0.225 82.060 82.549 cp_fm_upper_to_full 104 14.8 52.683 75.676 52.683 75.676 multiply_cannon 2485 13.5 0.678 0.703 58.599 59.986 multiply_cannon_loop 2485 14.5 1.024 1.045 54.526 55.994 ot_mini 116 10.6 0.001 0.001 44.742 44.793 dbcsr_complete_redistribute 393 12.7 3.963 3.996 30.061 42.793 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.526 39.254 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.011 36.625 rebuild_ks_matrix 127 8.3 0.001 0.001 34.649 34.696 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 34.648 34.696 mp_alltoall_i22 712 14.1 21.761 34.477 21.761 34.477 cp_fm_cholesky_invert 11 10.9 33.705 33.713 33.705 33.713 mp_waitall_1 102768 16.8 28.355 32.654 28.355 32.654 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.109 32.156 qs_ot_get_p 127 10.4 0.001 0.001 27.028 27.123 qs_ot_get_derivative 116 11.6 0.002 0.002 24.782 24.817 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.805 22.835 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.011 21.012 make_m2s 4970 13.5 0.073 0.077 18.816 20.363 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 18.830 20.254 ot_diis_step 116 11.6 0.021 0.022 19.927 19.927 make_images 4970 14.5 3.024 3.227 18.337 19.887 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.284 19.414 apply_single 127 13.6 0.001 0.001 19.284 19.414 multiply_cannon_multrec 9940 15.5 10.186 11.953 17.921 17.985 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.790 17.791 cp_fm_diag_elpa_base 82 14.4 13.427 15.054 17.785 17.786 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.309 17.338 calculate_rho_elec 127 8.7 0.474 0.475 17.308 17.338 sum_up_and_integrate 127 10.3 0.002 0.002 16.393 16.477 integrate_v_rspace 127 11.3 0.004 0.005 16.330 16.415 multiply_cannon_sync_h2d 9940 15.5 14.213 14.242 14.213 14.242 pw_transfer 1535 11.6 0.095 0.096 13.198 13.205 hybrid_alltoall_any 5155 16.4 1.304 3.039 10.940 13.136 make_images_data 4970 15.5 0.062 0.068 10.764 13.046 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.961 12.966 init_scf_run 11 5.9 0.000 0.001 12.354 12.355 scf_env_initial_rho_setup 11 6.9 0.003 0.004 12.354 12.355 fft_wrap_pw1pw2_140 519 13.2 3.198 3.236 11.514 11.523 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.792 9.824 dbcsr_mm_accdrv_process 20590 16.0 4.158 6.129 7.486 9.375 wfi_extrapolate 11 7.9 0.001 0.001 9.057 9.057 cp_fm_cholesky_decompose 22 10.9 9.015 9.034 9.015 9.034 grid_integrate_task_list 127 12.3 8.493 8.674 8.493 8.674 density_rs2pw 127 9.7 0.005 0.005 8.385 8.437 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.176 8.177 fft3d_ps 1281 14.7 3.947 3.961 7.462 7.495 mp_alltoall_d11v 2401 14.1 6.922 7.037 6.922 7.037 calculate_dm_sparse 127 9.5 0.001 0.001 6.720 6.753 grid_collocate_task_list 127 9.7 6.362 6.434 6.362 6.434 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.313 6.357 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.203 6.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.177000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2841.909091, yerr=149.281585 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 1.251717E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.019 0.059 85.161 85.162 qs_energies 1 2.0 0.000 0.000 84.621 84.628 ls_scf 1 3.0 0.000 0.000 83.715 83.722 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.568 72.739 multiply_cannon 111 7.7 0.017 0.021 55.883 56.940 multiply_cannon_loop 111 8.7 0.229 0.245 52.443 53.836 ls_scf_main 1 4.0 0.000 0.000 52.504 52.505 density_matrix_trs4 2 5.0 0.002 0.003 46.941 47.018 ls_scf_init_scf 1 4.0 0.000 0.000 28.177 28.178 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.052 27.106 mp_waitall_1 11031 10.9 22.449 25.014 22.449 25.014 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.971 24.989 multiply_cannon_multrec 2664 9.7 8.161 8.955 15.553 17.360 multiply_cannon_sync_h2d 2664 9.7 13.598 14.891 13.598 14.891 make_m2s 222 7.7 0.008 0.011 13.117 13.614 make_images 222 8.7 0.098 0.107 13.095 13.595 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.574 12.397 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.440 8.664 make_images_data 222 9.7 0.004 0.005 7.764 8.340 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.709 7.009 7.996 hybrid_alltoall_any 227 10.6 0.216 1.841 6.599 7.796 dbcsr_mm_accdrv_process_sort 4760 11.4 6.215 7.113 6.215 7.113 calculate_norms 4752 9.8 5.534 6.178 5.534 6.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.149 mp_sum_l 887 5.1 3.038 4.892 3.038 4.892 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.074 4.200 mp_irecv_dv 6231 10.9 2.058 4.186 2.058 4.186 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.256 3.853 make_images_sizes 222 9.7 0.000 0.000 0.731 3.649 mp_alltoall_i44 222 10.7 0.730 3.649 0.730 3.649 arnoldi_extremal 4 6.8 0.000 0.000 3.300 3.327 arnoldi_normal_ev 4 7.8 0.001 0.003 3.300 3.327 build_subspace 16 8.4 0.009 0.012 3.208 3.210 ls_scf_post 1 4.0 0.000 0.000 3.034 3.042 ls_scf_store_result 1 5.0 0.000 0.000 2.845 2.891 dbcsr_special_finalize 555 9.7 0.005 0.006 2.269 2.725 dbcsr_merge_single_wm 555 10.7 0.456 0.591 2.261 2.717 make_images_pack 222 9.7 2.207 2.671 2.209 2.672 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.344 2.593 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.477 2.072 2.479 dbcsr_sort_data 658 11.4 2.062 2.453 2.062 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.326 2.426 buffer_matrices_ensure_size 222 8.7 1.745 2.047 1.745 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.770 1.771 rebuild_ks_matrix 3 7.3 0.000 0.000 1.760 1.762 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.760 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.162000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1123.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 2.164724E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.042 90.412 90.414 qs_energies 1 2.0 0.000 0.000 89.919 89.923 ls_scf 1 3.0 0.000 0.000 88.582 88.586 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.302 74.616 multiply_cannon 111 7.7 0.027 0.039 52.594 57.109 ls_scf_main 1 4.0 0.000 0.000 54.425 54.425 multiply_cannon_loop 111 8.7 0.136 0.146 49.854 53.006 density_matrix_trs4 2 5.0 0.002 0.003 48.724 48.904 ls_scf_init_scf 1 4.0 0.000 0.000 30.488 30.489 mp_waitall_1 9105 10.9 20.749 29.924 20.749 29.924 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.321 29.392 multiply_cannon_multrec 1332 9.7 13.436 17.425 22.735 27.852 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.703 26.715 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.289 20.586 make_m2s 222 7.7 0.006 0.007 15.340 15.874 make_images 222 8.7 1.366 1.696 15.310 15.844 dbcsr_mm_accdrv_process 4041 10.4 0.345 0.550 8.892 10.496 make_images_data 222 9.7 0.004 0.005 8.926 9.955 dbcsr_mm_accdrv_process_sort 4041 11.4 8.408 9.946 8.408 9.946 hybrid_alltoall_any 227 10.6 0.539 2.532 8.339 9.532 mp_sum_l 887 5.1 4.911 8.087 4.911 8.087 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.209 7.638 mp_irecv_dv 3311 11.0 3.189 7.589 3.189 7.589 calculate_norms 2376 9.8 6.068 6.803 6.068 6.803 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.753 6.463 multiply_cannon_sync_h2d 1332 9.7 4.890 6.294 4.890 6.294 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.162 5.389 arnoldi_extremal 4 6.8 0.000 0.000 4.675 4.692 arnoldi_normal_ev 4 7.8 0.001 0.005 4.675 4.691 build_subspace 16 8.4 0.014 0.021 4.423 4.425 ls_scf_post 1 4.0 0.000 0.000 3.669 3.674 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.495 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.155 3.407 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.219 2.741 3.221 mp_allgather_i34 111 8.7 0.728 3.157 0.728 3.157 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.654 2.773 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.171 2.759 dbcsr_data_new 4174 10.1 2.119 2.420 2.119 2.420 make_images_pack 222 9.7 1.928 2.331 1.931 2.333 dbcsr_sort_data 436 11.2 1.843 2.043 1.843 2.043 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.862 1.864 rebuild_ks_matrix 3 7.3 0.000 0.000 1.849 1.851 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.849 1.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.414000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1771.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 2.815693E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.052 92.693 92.695 qs_energies 1 2.0 0.000 0.001 92.054 92.082 ls_scf 1 3.0 0.007 0.053 90.615 90.647 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.248 75.555 multiply_cannon 111 7.7 0.038 0.129 51.960 56.744 ls_scf_main 1 4.0 0.001 0.012 56.152 56.157 multiply_cannon_loop 111 8.7 0.116 0.129 49.111 53.033 density_matrix_trs4 2 5.0 0.002 0.004 50.152 50.353 mp_waitall_1 7281 11.0 23.582 33.277 23.582 33.277 ls_scf_init_scf 1 4.0 0.000 0.002 30.708 30.713 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.466 29.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.723 26.735 multiply_cannon_multrec 888 9.7 12.640 14.977 21.229 24.348 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.014 22.791 make_m2s 222 7.7 0.006 0.007 16.529 17.399 make_images 222 8.7 1.584 1.849 16.492 17.360 make_images_data 222 9.7 0.004 0.004 9.690 10.875 hybrid_alltoall_any 227 10.6 0.641 2.952 9.225 10.594 dbcsr_mm_accdrv_process 3754 10.4 0.317 0.490 8.103 9.304 mp_sum_l 887 5.1 5.061 8.901 5.061 8.901 dbcsr_mm_accdrv_process_sort 3754 11.4 7.668 8.814 7.668 8.814 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.488 7.514 mp_irecv_dv 2335 11.1 2.473 7.459 2.473 7.459 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.518 7.003 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.850 6.983 multiply_cannon_sync_h2d 888 9.7 5.974 6.900 5.974 6.900 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.258 5.537 arnoldi_extremal 4 6.8 0.000 0.000 5.109 5.124 arnoldi_normal_ev 4 7.8 0.001 0.005 5.109 5.124 build_subspace 16 8.4 0.014 0.020 4.809 4.815 calculate_norms 1584 9.8 4.407 4.768 4.407 4.768 mp_allgather_i34 111 8.7 0.972 4.045 0.972 4.045 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.447 3.768 ls_scf_post 1 4.0 0.003 0.020 3.748 3.753 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.595 3.021 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.478 3.576 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.879 3.004 dbcsr_data_new 4116 9.9 2.099 2.441 2.099 2.441 make_images_pack 222 9.7 1.701 2.247 1.704 2.250 dbcsr_sort_data 325 11.1 1.905 2.133 1.905 2.133 make_images_sizes 222 9.7 0.000 0.000 0.907 1.996 mp_alltoall_i44 222 10.7 0.906 1.996 0.906 1.996 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.916 rebuild_ks_matrix 3 7.3 0.000 0.000 1.894 1.897 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.021 1.894 1.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.695000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2235.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 3.365028E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.084 97.177 97.178 qs_energies 1 2.0 0.000 0.001 96.495 96.499 ls_scf 1 3.0 0.000 0.001 94.818 94.824 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.404 78.608 ls_scf_main 1 4.0 0.000 0.003 58.913 58.915 multiply_cannon 111 7.7 0.047 0.101 51.605 56.277 density_matrix_trs4 2 5.0 0.002 0.003 52.739 52.864 multiply_cannon_loop 111 8.7 0.152 0.165 46.495 49.165 ls_scf_init_scf 1 4.0 0.000 0.002 32.652 32.655 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.441 31.509 mp_waitall_1 6369 11.0 22.554 29.113 22.554 29.113 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.901 28.913 multiply_cannon_multrec 1332 9.7 14.196 17.601 22.119 25.161 make_m2s 222 7.7 0.007 0.008 21.148 22.635 make_images 222 8.7 3.138 3.594 21.098 22.587 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.000 17.274 make_images_data 222 9.7 0.004 0.004 11.794 13.417 hybrid_alltoall_any 227 10.6 0.796 3.765 11.078 12.903 dbcsr_mm_accdrv_process 3641 10.4 0.304 0.481 7.556 9.080 dbcsr_mm_accdrv_process_sort 3641 11.4 7.114 8.582 7.114 8.582 mp_sum_l 887 5.1 4.064 7.532 4.064 7.532 multiply_cannon_sync_h2d 1332 9.7 5.405 6.130 5.405 6.130 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.069 6.038 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.121 5.978 mp_irecv_dv 3229 10.9 2.045 5.963 2.045 5.963 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.588 5.409 arnoldi_extremal 4 6.8 0.000 0.000 5.324 5.340 arnoldi_normal_ev 4 7.8 0.001 0.005 5.324 5.340 build_subspace 16 8.4 0.014 0.021 4.990 4.997 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.646 4.841 mp_allgather_i34 111 8.7 2.237 4.689 2.237 4.689 calculate_norms 2376 9.8 4.218 4.577 4.218 4.577 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.618 3.945 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.684 3.192 3.686 dbcsr_sort_data 658 11.4 3.053 3.365 3.053 3.365 ls_scf_post 1 4.0 0.000 0.000 3.253 3.257 dbcsr_special_finalize 555 9.7 0.006 0.007 2.796 3.190 dbcsr_merge_single_wm 555 10.7 0.538 0.661 2.787 3.182 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.041 3.112 ls_scf_store_result 1 5.0 0.000 0.000 2.984 3.039 dbcsr_data_release 10477 10.7 1.596 2.418 1.596 2.418 dbcsr_finalize 304 7.8 0.049 0.061 1.807 1.968 make_images_pack 222 9.7 1.623 1.944 1.625 1.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.178000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 4.751249E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.054 91.794 91.795 qs_energies 1 2.0 0.000 0.000 91.068 91.074 ls_scf 1 3.0 0.000 0.000 89.032 89.036 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.640 70.833 ls_scf_main 1 4.0 0.000 0.000 56.532 56.533 multiply_cannon 111 7.7 0.067 0.134 52.782 56.137 multiply_cannon_loop 111 8.7 0.089 0.108 50.211 51.714 density_matrix_trs4 2 5.0 0.002 0.003 49.637 49.701 ls_scf_init_scf 1 4.0 0.000 0.000 29.104 29.106 mp_waitall_1 5436 11.0 24.595 28.598 24.595 28.598 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.844 27.874 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.753 25.765 multiply_cannon_multrec 444 9.7 13.572 16.424 20.652 21.971 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.844 15.859 make_m2s 222 7.7 0.005 0.005 13.363 14.299 make_images 222 8.7 2.037 2.479 13.295 14.231 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.443 13.537 make_images_data 222 9.7 0.003 0.004 8.125 9.597 hybrid_alltoall_any 227 10.6 0.801 3.801 8.060 9.590 multiply_cannon_sync_h2d 444 9.7 6.773 7.955 6.773 7.955 dbcsr_mm_accdrv_process 3003 10.4 0.342 0.425 6.777 7.943 dbcsr_mm_accdrv_process_sort 3003 11.4 6.412 7.518 6.412 7.518 arnoldi_extremal 4 6.8 0.000 0.000 5.845 5.859 arnoldi_normal_ev 4 7.8 0.010 0.017 5.845 5.859 build_subspace 16 8.4 0.015 0.019 5.436 5.443 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.363 4.579 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.156 4.357 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.166 3.701 4.168 mp_sum_l 887 5.1 2.487 4.037 2.487 4.037 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.554 3.900 mp_irecv_dv 1241 11.2 1.542 3.867 1.542 3.867 mp_allgather_i34 111 8.7 1.142 3.694 1.142 3.694 calculate_norms 792 9.8 3.608 3.692 3.608 3.692 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.505 3.557 ls_scf_post 1 4.0 0.000 0.000 3.396 3.401 ls_scf_store_result 1 5.0 0.000 0.000 3.181 3.214 make_images_sizes 222 9.7 0.000 0.000 0.851 3.095 mp_alltoall_i44 222 10.7 0.851 3.095 0.851 3.095 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.704 2.965 dbcsr_finalize 304 7.8 0.062 0.077 2.202 2.302 dbcsr_data_new 4608 9.7 1.817 2.263 1.817 2.263 dbcsr_merge_all 275 8.9 0.477 0.523 2.061 2.139 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.037 2.037 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.021 2.021 rebuild_ks_matrix 3 7.3 0.000 0.000 2.004 2.005 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.004 2.005 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.795000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3756.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.939831E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.093 106.392 106.392 qs_energies 1 2.0 0.000 0.000 105.070 105.080 ls_scf 1 3.0 0.000 0.000 102.105 102.114 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.784 75.934 ls_scf_main 1 4.0 0.000 0.000 63.638 63.639 density_matrix_trs4 2 5.0 0.002 0.003 54.649 54.708 multiply_cannon 111 7.7 0.173 0.258 49.601 52.006 multiply_cannon_loop 111 8.7 0.097 0.098 46.497 47.098 ls_scf_init_scf 1 4.0 0.000 0.000 34.687 34.688 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.127 33.151 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.423 30.437 mp_waitall_1 4527 11.1 21.955 25.717 21.955 25.717 make_m2s 222 7.7 0.005 0.005 22.588 23.784 make_images 222 8.7 3.574 3.870 22.480 23.676 multiply_cannon_multrec 444 9.7 17.832 18.455 22.461 23.098 hybrid_alltoall_any 227 10.6 1.655 3.628 12.766 15.624 make_images_data 222 9.7 0.003 0.004 12.971 15.221 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.396 11.460 multiply_cannon_sync_h2d 444 9.7 8.791 8.852 8.791 8.852 arnoldi_extremal 4 6.8 0.000 0.000 7.354 7.367 arnoldi_normal_ev 4 7.8 0.003 0.009 7.354 7.367 build_subspace 16 8.4 0.026 0.036 6.805 6.819 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.455 5.608 dbcsr_matrix_vector_mult_local 304 10.0 5.043 5.354 5.045 5.356 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.158 5.256 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.825 5.084 dbcsr_mm_accdrv_process 1814 10.4 0.287 0.349 4.454 4.588 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.255 4.120 4.255 mp_allgather_i34 111 8.7 1.220 3.912 1.220 3.912 ls_scf_post 1 4.0 0.000 0.000 3.780 3.790 make_images_sizes 222 9.7 0.000 0.000 1.490 3.600 mp_alltoall_i44 222 10.7 1.490 3.600 1.490 3.600 ls_scf_store_result 1 5.0 0.000 0.000 3.506 3.542 calculate_norms 792 9.8 3.227 3.276 3.227 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.067 3.118 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.934 2.934 dbcsr_merge_all 275 8.9 0.887 0.922 2.853 2.895 dbcsr_complete_redistribute 5 7.6 1.420 1.463 2.716 2.850 matrix_ls_to_qs 2 6.0 0.000 0.000 2.372 2.513 dbcsr_sort_data 325 11.1 2.436 2.486 2.436 2.486 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.461 2.463 rebuild_ks_matrix 3 7.3 0.000 0.000 2.395 2.396 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.395 2.396 dbcsr_data_new 6591 9.6 1.834 2.368 1.834 2.368 dbcsr_new_transposed 4 7.5 0.243 0.253 2.285 2.303 dbcsr_frobenius_norm 74 6.6 2.058 2.140 2.200 2.232 dbcsr_add_d 103 6.2 0.000 0.000 2.129 2.211 dbcsr_add_anytype 103 7.2 0.858 0.891 2.129 2.211 dbcsr_data_release 12724 10.6 1.975 2.170 1.975 2.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.392000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7028.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/577b676415f3d2d1924bda4bef9a238f96821cb1_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 580.833280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1006787032. 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 0.032 0.103 235.027 235.028 qs_mol_dyn_low 1 2.0 0.009 0.088 233.474 233.564 qs_forces 5 3.8 0.005 0.041 233.172 233.177 qs_energies 5 4.8 0.004 0.020 230.103 230.125 scf_env_do_scf 5 5.8 0.001 0.015 217.015 217.018 scf_env_do_scf_inner_loop 105 6.6 0.002 0.007 190.067 190.069 qs_scf_new_mos 105 7.6 0.000 0.000 149.402 149.573 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.402 149.572 ot_scf_mini 105 9.6 0.003 0.004 139.491 139.638 dbcsr_multiply_generic 1445 12.2 0.128 0.137 131.983 132.346 multiply_cannon 1445 13.2 0.272 0.288 112.723 114.829 multiply_cannon_loop 1445 14.2 2.835 2.969 110.953 112.228 velocity_verlet 4 3.0 0.016 0.122 107.242 107.248 ot_mini 105 10.6 0.001 0.002 60.358 60.472 qs_ot_get_p 112 10.4 0.001 0.002 49.158 49.394 multiply_cannon_multrec 69360 15.2 29.759 34.172 39.602 44.501 mp_waitall_1 488190 16.1 35.065 42.146 35.065 42.146 qs_ot_get_derivative 55 11.6 0.001 0.001 38.593 38.729 qs_ot_p2m_diag 40 11.0 0.020 0.031 37.978 38.078 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.610 34.611 multiply_cannon_metrocomm3 69360 15.2 0.198 0.210 25.677 32.906 multiply_cannon_sync_h2d 69360 15.2 29.049 32.786 29.049 32.786 cp_fm_syevd 40 13.0 0.000 0.000 29.346 29.479 rebuild_ks_matrix 110 8.4 0.000 0.001 29.225 29.395 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.031 29.224 29.395 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.777 26.928 init_scf_loop 7 6.6 0.000 0.005 26.916 26.917 cp_fm_redistribute_end 40 14.0 12.241 24.421 12.247 24.424 cp_fm_syevd_base 40 14.0 12.168 24.352 12.168 24.352 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.056 23.320 apply_single 62 13.6 0.000 0.000 23.056 23.320 prepare_preconditioner 7 7.6 0.000 0.000 22.034 22.075 make_preconditioner 7 8.6 0.000 0.002 22.034 22.075 ot_new_cg_direction 55 11.6 0.001 0.002 21.077 21.077 qs_rho_update_rho_low 110 7.6 0.000 0.002 17.318 17.702 calculate_rho_elec 110 8.6 0.029 0.032 17.317 17.702 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.735 14.888 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.751 14.813 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.927 14.043 mp_sum_l 4764 12.2 12.133 12.879 12.133 12.879 density_rs2pw 110 9.6 0.005 0.006 11.592 12.169 pw_transfer 1645 12.4 0.079 0.104 11.816 12.089 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.677 11.957 calculate_dm_sparse 110 9.5 0.000 0.000 11.475 11.681 dbcsr_mm_accdrv_process 154766 15.8 6.250 6.452 9.712 10.530 fft_wrap_pw1pw2_240 915 15.0 0.822 0.925 10.235 10.491 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.403 10.463 qs_vxc_create 110 10.4 0.002 0.004 10.412 10.459 init_scf_run 5 5.8 0.000 0.001 10.339 10.340 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.339 10.340 cp_fm_cholesky_invert 7 10.6 10.260 10.269 10.260 10.269 check_diag 80 13.5 8.686 8.944 9.731 9.878 sum_up_and_integrate 60 10.3 0.001 0.003 8.794 8.810 transfer_rs2pw 445 10.6 0.007 0.008 8.252 8.797 integrate_v_rspace 60 11.3 0.002 0.002 8.777 8.793 fft3d_pb 915 16.0 2.394 2.676 8.441 8.711 xc_rho_set_and_dset_create 110 12.4 0.075 0.094 7.320 7.586 make_m2s 2890 13.2 0.078 0.087 6.672 7.246 make_images 2890 14.2 0.238 0.260 6.564 7.138 make_full_single_inverse 7 9.6 0.001 0.003 7.017 7.057 xc_vxc_pw_create 60 11.3 0.038 0.049 6.968 7.015 multiply_cannon_metrocomm1 69360 15.2 0.094 0.100 4.548 6.867 acc_transpose_blocks 69360 15.2 0.354 0.370 6.139 6.508 xc_pw_derive 510 13.4 0.005 0.007 6.172 6.256 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 5.870 5.887 mp_alltoall_z22v 2340 17.7 5.488 5.747 5.488 5.747 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.429 5.485 calculate_first_density_matrix 1 7.0 0.000 0.005 5.454 5.479 mp_waitany 7680 13.5 4.608 5.257 4.608 5.257 multiply_cannon_metrocomm4 67915 15.2 0.187 0.204 2.029 5.162 potential_pw2rs 60 12.3 0.002 0.003 5.016 5.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=235.028000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=552.800000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 577b676415f3d2d1924bda4bef9a238f96821cb1 Summary: empty Status: OK