=== 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: 5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8 ################# 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-hdf5 --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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/01 job id: 51922560 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/02 job id: 51922561 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/03 job id: 51922562 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/04 job id: 51922563 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/05 job id: 51922565 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/06 job id: 51922566 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/07 job id: 51922567 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/08 job id: 51922568 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/09 job id: 51922569 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/10 job id: 51922570 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/11 job id: 51922571 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/12 job id: 51922572 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/13 job id: 51922573 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/14 job id: 51922574 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/15 job id: 51922575 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/16 job id: 51922576 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/17 job id: 51922577 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/18 job id: 51922578 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/19 job id: 51922579 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/20 job id: 51922580 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/21 job id: 51922581 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/22 job id: 51922582 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/23 job id: 51922583 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/24 job id: 51922584 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/25 job id: 51922585 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/26 job id: 51922586 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/27 job id: 51922587 --- 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/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 0.028 0.034 137.984 137.985 farming_run 1 2.0 137.494 137.496 137.946 137.949 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.546572E+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 0.023 119.722 119.722 qs_energies 1 2.0 0.001 0.001 119.538 119.540 mp2_main 1 3.0 0.000 0.000 117.487 117.490 mp2_gpw_main 1 4.0 0.030 0.039 116.491 116.493 mp2_ri_gpw_compute_in 1 5.0 0.174 0.175 97.647 97.923 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.346 57.622 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.174 43.799 48.844 get_2c_integrals 1 6.0 0.008 0.009 39.599 40.126 integrate_v_rspace 273 8.0 0.443 0.456 25.575 30.123 fft_wrap_pw1pw2 5465 10.4 0.067 0.069 28.051 28.620 grid_integrate_task_list 273 9.0 20.827 26.314 20.827 26.314 fft_wrap_pw1pw2_100 2178 11.4 1.309 1.439 25.501 26.096 compute_2c_integrals 1 7.0 0.002 0.002 21.391 21.391 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.862 21.125 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.423 20.860 21.124 rpa_ri_compute_en 1 5.0 0.019 0.021 18.738 18.885 cp_fm_cholesky_decompose 12 8.2 18.142 18.682 18.142 18.682 cholesky_decomp 1 7.0 0.000 0.000 17.049 17.587 fft3d_s 5443 12.4 16.868 17.209 16.889 17.229 ao_to_mo_and_store_B_mult_1 272 7.0 10.763 15.316 10.763 15.316 calculate_wavefunction 272 8.0 5.456 5.600 12.901 13.785 calc_potential_gpw 544 9.5 0.004 0.004 10.783 11.580 rpa_num_int 1 6.0 0.000 0.006 10.670 10.671 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.634 10.662 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.798 10.506 potential_pw2rs 545 10.0 0.106 0.108 8.536 10.384 calc_mat_Q 8 8.0 0.000 0.000 9.498 9.612 contract_S_to_Q 8 9.0 0.000 0.000 8.920 9.034 collocate_single_gaussian 272 10.0 0.039 0.041 7.864 8.629 parallel_gemm_fm 14 9.1 0.000 0.000 8.488 8.611 parallel_gemm_fm_cosma 14 10.1 8.488 8.611 8.488 8.611 create_integ_mat 1 6.0 0.023 0.028 7.698 7.699 array2fm 1 7.0 0.000 0.000 6.647 7.141 pw_gather_s 2722 12.2 4.000 4.927 4.000 4.927 pw_poisson_solve 545 10.5 0.010 0.011 4.656 4.839 pw_scatter_s 2720 12.7 4.418 4.745 4.418 4.745 pw_poisson_set 548 11.5 0.018 0.019 3.331 3.538 array2fm_buffer_send 1 8.0 3.018 3.138 3.018 3.138 pw_copy 4911 11.6 2.631 2.764 2.631 2.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.492701, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2798.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 0.025 0.037 529.105 529.107 farming_run 1 2.0 528.266 528.272 529.053 529.057 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.274180E+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 3635357. 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 0.013 0.034 325.324 325.324 qs_energies 1 2.0 0.002 0.015 325.090 325.091 mp2_main 1 3.0 0.000 0.000 209.628 209.631 mp2_gpw_main 1 4.0 0.049 0.080 208.642 208.649 mp2_ri_gpw_compute_en 1 5.0 0.038 0.041 128.417 130.584 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.730 1.829 126.749 126.754 scf_env_do_scf 1 3.0 0.000 0.000 115.163 115.163 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.195 114.203 rebuild_ks_matrix 4 6.0 0.000 0.000 114.194 114.201 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 114.194 114.201 hfx_ks_matrix 4 8.0 0.001 0.001 113.838 113.842 integrate_four_center 4 9.0 0.154 0.466 113.838 113.841 mp2_ri_gpw_compute_en_expansio 172 7.0 0.460 0.500 111.033 111.969 local_gemm 172 8.0 110.573 111.504 110.573 111.504 integrate_four_center_main 4 10.0 0.112 0.503 101.894 105.576 integrate_four_center_bin 264 11.0 101.782 105.513 101.782 105.513 init_scf_loop 1 4.0 0.000 0.000 97.240 97.240 mp2_ri_gpw_compute_in 1 5.0 0.067 0.077 80.075 81.117 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.742 57.781 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 44.548 49.538 integrate_v_rspace 95 8.0 0.398 0.565 28.184 32.984 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.459 31.324 fft_wrap_pw1pw2_100 730 11.4 0.635 0.734 28.159 29.049 grid_integrate_task_list 95 9.0 23.479 28.471 23.479 28.471 ao_to_mo_and_store_B_mult_1 91 7.0 10.521 27.040 10.521 27.040 get_2c_integrals 1 6.0 0.000 0.000 23.233 23.271 compute_2c_integrals 1 7.0 0.002 0.003 22.220 22.231 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 21.731 22.096 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.824 21.730 22.095 fft3d_s 1823 12.4 19.199 20.095 19.213 20.108 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.920 17.920 calc_potential_gpw 182 9.5 0.002 0.002 12.440 13.265 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.750 11.467 calculate_wavefunction 91 8.0 2.028 2.072 10.050 10.841 potential_pw2rs 186 10.0 0.033 0.034 9.010 10.649 collocate_single_gaussian 91 10.0 0.017 0.024 8.270 9.018 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.522 7.884 8.969 mp_sendrecv_dm3 2068 8.0 5.913 7.023 5.913 7.023 integrate_four_center_load 4 10.0 0.000 0.000 6.754 6.759 hfx_load_balance 1 11.0 0.000 0.000 6.754 6.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.638552, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1527.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 447.246336E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 54.927 54.928 qs_mol_dyn_low 1 2.0 0.003 0.003 54.709 54.719 qs_forces 11 3.9 0.002 0.002 54.645 54.646 qs_energies 11 4.9 0.001 0.001 53.141 53.170 scf_env_do_scf 11 5.9 0.000 0.001 45.185 45.186 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.147 43.148 dbcsr_multiply_generic 2286 12.5 0.093 0.098 35.170 35.593 qs_scf_new_mos 108 7.5 0.000 0.001 33.356 33.654 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.356 33.654 ot_scf_mini 108 9.5 0.002 0.002 31.720 31.904 multiply_cannon 2286 13.5 0.184 0.189 27.895 29.199 multiply_cannon_loop 2286 14.5 1.825 1.935 27.255 28.585 velocity_verlet 10 3.0 0.001 0.001 25.764 25.764 ot_mini 108 10.5 0.001 0.001 19.088 19.349 qs_ot_get_derivative 108 11.5 0.001 0.001 16.136 16.317 mp_waitall_1 245248 16.5 8.099 14.491 8.099 14.491 multiply_cannon_multrec 54864 15.5 3.658 5.711 9.655 13.046 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 5.837 12.786 qs_ot_get_p 119 10.4 0.001 0.008 8.033 8.363 rebuild_ks_matrix 119 8.3 0.000 0.000 7.759 7.954 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.759 7.954 dbcsr_mm_accdrv_process 76910 16.1 1.843 2.917 5.911 7.602 mp_sum_l 7287 12.8 5.381 7.110 5.381 7.110 multiply_cannon_sync_h2d 54864 15.5 5.170 7.052 5.170 7.052 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.816 6.974 init_scf_run 11 5.9 0.000 0.001 6.665 6.665 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.665 6.665 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.549 5.991 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.310 5.450 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.260 5.316 calculate_dm_sparse 119 9.5 0.000 0.000 4.857 5.024 jit_kernel_multiply 13 15.8 4.005 4.621 4.005 4.621 calculate_first_density_matrix 1 7.0 0.000 0.000 4.457 4.473 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.426 4.426 sum_up_and_integrate 119 10.3 0.001 0.002 4.345 4.351 integrate_v_rspace 119 11.3 0.002 0.002 4.334 4.342 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.240 4.240 cp_fm_redistribute_end 50 14.0 2.163 4.217 2.169 4.220 cp_fm_diag_elpa_base 50 14.0 2.045 4.111 2.049 4.120 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.970 4.085 calculate_rho_elec 119 8.7 0.011 0.016 3.969 4.085 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.390 3.422 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.888 3.078 apply_single 119 13.6 0.000 0.000 2.888 3.078 acc_transpose_blocks 54864 15.5 0.220 0.241 2.212 2.749 ot_diis_step 108 11.5 0.006 0.006 2.656 2.657 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.640 2.652 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.412 2.495 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.277 2.278 wfi_extrapolate 11 7.9 0.001 0.001 2.129 2.129 grid_integrate_task_list 119 12.3 2.019 2.114 2.019 2.114 density_rs2pw 119 9.7 0.003 0.004 2.001 2.093 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.973 2.049 init_scf_loop 11 6.9 0.000 0.000 2.021 2.021 mp_sum_d 4139 12.0 1.315 1.925 1.315 1.925 potential_pw2rs 119 12.3 0.003 0.004 1.721 1.730 make_m2s 4572 13.5 0.053 0.055 1.581 1.631 make_images 4572 14.5 0.132 0.138 1.500 1.548 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.463 1.531 mp_alltoall_d11v 2130 13.8 1.282 1.466 1.282 1.466 acc_transpose_blocks_sync 164592 16.5 1.204 1.449 1.204 1.449 grid_collocate_task_list 119 9.7 1.344 1.411 1.344 1.411 transfer_rs2pw 487 10.6 0.006 0.007 1.309 1.406 mp_waitany 12084 13.8 1.231 1.392 1.231 1.392 transfer_pw2rs 487 13.2 0.005 0.006 1.297 1.306 fft3d_ps 1201 13.6 0.363 0.463 1.238 1.298 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.133 1.204 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.712 1.106 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.928000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=426.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 485.302272E+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 481194. 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 41.489 41.490 qs_mol_dyn_low 1 2.0 0.003 0.003 41.283 41.290 qs_forces 11 3.9 0.002 0.003 41.222 41.223 qs_energies 11 4.9 0.001 0.001 39.544 39.549 scf_env_do_scf 11 5.9 0.000 0.001 32.174 32.174 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.737 29.738 dbcsr_multiply_generic 2286 12.5 0.101 0.104 24.422 24.786 qs_scf_new_mos 108 7.5 0.001 0.001 20.944 21.167 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.943 21.166 multiply_cannon 2286 13.5 0.208 0.216 19.075 20.424 ot_scf_mini 108 9.5 0.002 0.003 20.017 20.173 multiply_cannon_loop 2286 14.5 1.185 1.236 17.879 19.351 velocity_verlet 10 3.0 0.001 0.002 18.605 18.606 ot_mini 108 10.5 0.001 0.001 12.354 12.579 multiply_cannon_multrec 27432 15.5 1.820 4.150 8.232 11.113 mp_waitall_1 200699 16.5 5.695 10.912 5.695 10.912 qs_ot_get_derivative 108 11.5 0.001 0.001 9.993 10.155 multiply_cannon_metrocomm3 27432 15.5 0.072 0.076 4.059 9.590 dbcsr_mm_accdrv_process 47894 16.0 4.347 7.755 6.333 8.454 rebuild_ks_matrix 119 8.3 0.000 0.000 6.889 7.027 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.888 7.027 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.069 6.195 init_scf_run 11 5.9 0.000 0.001 6.143 6.143 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.143 6.143 qs_ot_get_p 119 10.4 0.001 0.001 4.719 4.940 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.749 4.660 calculate_first_density_matrix 1 7.0 0.000 0.000 4.607 4.609 calculate_dm_sparse 119 9.5 0.000 0.000 4.072 4.155 mp_sum_l 7287 12.8 2.123 4.118 2.123 4.118 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.064 4.042 apply_single 119 13.6 0.000 0.000 3.064 4.042 sum_up_and_integrate 119 10.3 0.001 0.001 3.756 3.763 integrate_v_rspace 119 11.3 0.002 0.003 3.742 3.749 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.673 3.702 calculate_rho_elec 119 8.7 0.021 0.024 3.673 3.702 jit_kernel_multiply 8 16.3 1.927 3.693 1.927 3.693 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.261 3.269 make_m2s 4572 13.5 0.051 0.053 2.768 3.100 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.005 3.023 make_images 4572 14.5 0.205 0.242 2.678 3.014 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.601 2.602 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.377 2.462 init_scf_loop 11 6.9 0.000 0.000 2.414 2.414 ot_diis_step 108 11.5 0.011 0.011 2.313 2.314 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.252 2.252 cp_fm_redistribute_end 50 14.0 1.143 2.223 1.146 2.225 multiply_cannon_sync_h2d 27432 15.5 1.701 2.197 1.701 2.197 cp_fm_diag_elpa_base 50 14.0 1.048 2.127 1.076 2.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.070 2.072 density_rs2pw 119 9.7 0.003 0.004 1.961 2.035 acc_transpose_blocks 27432 15.5 0.114 0.119 1.652 2.020 grid_integrate_task_list 119 12.3 1.838 1.913 1.838 1.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.858 1.897 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.731 1.765 make_images_data 4572 15.5 0.047 0.053 1.283 1.649 prepare_preconditioner 11 7.9 0.000 0.000 1.528 1.553 make_preconditioner 11 8.9 0.000 0.000 1.528 1.553 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.136 1.531 potential_pw2rs 119 12.3 0.006 0.006 1.488 1.498 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.428 1.486 fft3d_ps 1201 13.6 0.511 0.566 1.451 1.482 wfi_extrapolate 11 7.9 0.001 0.001 1.476 1.477 fft_wrap_pw1pw2_140 487 12.2 0.047 0.049 1.345 1.378 mp_alltoall_d11v 2130 13.8 1.262 1.370 1.262 1.370 grid_collocate_task_list 119 9.7 1.287 1.349 1.287 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.222 1.269 transfer_rs2pw 487 10.6 0.005 0.005 1.180 1.242 mp_allgather_i34 2286 14.5 0.620 1.036 0.620 1.036 mp_sum_d 4139 12.0 0.543 1.009 0.543 1.009 transfer_pw2rs 487 13.2 0.004 0.005 0.975 0.986 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.697 0.942 acc_transpose_blocks_sync 82296 16.5 0.816 0.938 0.816 0.938 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.930 0.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.888 make_images_sizes 4572 15.5 0.005 0.005 0.617 0.882 mp_alltoall_i44 4572 16.5 0.613 0.878 0.613 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.490000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.545455, yerr=1.876342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 516.452352E+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 0.015 0.033 34.617 34.623 qs_mol_dyn_low 1 2.0 0.014 0.043 34.329 34.338 qs_forces 11 3.9 0.003 0.004 34.124 34.127 qs_energies 11 4.9 0.004 0.026 32.538 32.551 scf_env_do_scf 11 5.9 0.000 0.001 25.673 25.673 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.089 23.090 dbcsr_multiply_generic 2286 12.5 0.096 0.098 18.997 19.089 multiply_cannon 2286 13.5 0.193 0.197 15.687 16.492 multiply_cannon_loop 2286 14.5 0.859 0.893 14.904 15.736 qs_scf_new_mos 108 7.5 0.001 0.001 15.535 15.547 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.534 15.547 velocity_verlet 10 3.0 0.023 0.064 15.336 15.344 ot_scf_mini 108 9.5 0.002 0.003 14.764 14.778 multiply_cannon_multrec 18288 15.5 1.872 2.884 8.918 9.190 ot_mini 108 10.5 0.001 0.001 9.149 9.167 dbcsr_mm_accdrv_process 38222 16.0 6.858 7.738 6.952 7.803 qs_ot_get_derivative 108 11.5 0.001 0.001 7.636 7.651 rebuild_ks_matrix 119 8.3 0.000 0.000 5.984 6.006 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 5.983 6.005 init_scf_run 11 5.9 0.000 0.001 5.489 5.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.488 5.489 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.247 5.268 calculate_first_density_matrix 1 7.0 0.001 0.002 4.252 4.253 mp_waitall_1 158411 16.6 2.687 3.779 2.687 3.779 calculate_dm_sparse 119 9.5 0.000 0.001 3.736 3.747 sum_up_and_integrate 119 10.3 0.001 0.002 3.518 3.528 integrate_v_rspace 119 11.3 0.003 0.003 3.505 3.517 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.855 3.449 qs_ot_get_p 119 10.4 0.001 0.002 3.372 3.406 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.158 3.166 calculate_rho_elec 119 8.7 0.031 0.032 3.158 3.165 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.061 3.066 init_scf_loop 11 6.9 0.001 0.004 2.566 2.567 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.522 2.470 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.011 2.361 apply_single 119 13.6 0.000 0.000 2.011 2.361 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.241 2.246 make_m2s 4572 13.5 0.043 0.045 1.823 1.966 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.944 1.944 grid_integrate_task_list 119 12.3 1.816 1.900 1.816 1.900 make_images 4572 14.5 0.191 0.204 1.738 1.880 density_rs2pw 119 9.7 0.004 0.010 1.722 1.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.004 1.817 1.819 prepare_preconditioner 11 7.9 0.000 0.000 1.796 1.800 make_preconditioner 11 8.9 0.000 0.002 1.796 1.800 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.757 1.766 acc_transpose_blocks 18288 15.5 0.079 0.081 1.639 1.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.644 1.732 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.695 1.702 cp_fm_diag_elpa_base 50 14.0 1.671 1.683 1.693 1.701 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.663 1.678 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.660 1.665 mp_sum_l 7287 12.8 1.117 1.566 1.117 1.566 ot_diis_step 108 11.5 0.011 0.012 1.495 1.496 fft3d_ps 1201 13.6 0.535 0.557 1.347 1.362 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.337 1.349 grid_collocate_task_list 119 9.7 1.234 1.348 1.234 1.348 potential_pw2rs 119 12.3 0.007 0.008 1.311 1.318 wfi_extrapolate 11 7.9 0.004 0.015 1.175 1.176 multiply_cannon_sync_h2d 18288 15.5 1.032 1.172 1.032 1.172 transfer_rs2pw 487 10.6 0.005 0.005 1.037 1.137 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.052 1.056 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.954 0.974 make_images_data 4572 15.5 0.047 0.050 0.789 0.943 hybrid_alltoall_any 4725 16.4 0.059 0.116 0.692 0.879 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.801 0.837 transfer_pw2rs 487 13.2 0.004 0.004 0.817 0.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.811 acc_transpose_blocks_sync 54864 16.5 0.740 0.805 0.740 0.805 mp_alltoall_d11v 2130 13.8 0.692 0.787 0.692 0.787 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.672 0.746 cp_fm_cholesky_invert 11 10.9 0.738 0.742 0.738 0.742 mp_alltoall_z22v 1201 15.6 0.665 0.728 0.665 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.623000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=491.090909, yerr=1.831767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.528768E+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 0.030 0.070 38.112 38.114 qs_mol_dyn_low 1 2.0 0.004 0.008 37.677 37.727 qs_forces 11 3.9 0.003 0.009 37.591 37.592 qs_energies 11 4.9 0.002 0.007 35.889 35.896 scf_env_do_scf 11 5.9 0.000 0.001 28.938 28.940 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.543 25.544 dbcsr_multiply_generic 2286 12.5 0.103 0.105 21.589 21.688 multiply_cannon 2286 13.5 0.218 0.227 17.950 18.358 velocity_verlet 10 3.0 0.003 0.010 18.269 18.272 qs_scf_new_mos 108 7.5 0.001 0.001 17.689 17.751 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.689 17.750 multiply_cannon_loop 2286 14.5 1.520 1.591 17.004 17.337 ot_scf_mini 108 9.5 0.002 0.003 16.658 16.715 multiply_cannon_multrec 27432 15.5 2.455 3.144 10.875 11.190 ot_mini 108 10.5 0.001 0.001 10.261 10.325 dbcsr_mm_accdrv_process 47916 15.9 7.831 9.583 8.317 9.646 qs_ot_get_derivative 108 11.5 0.001 0.001 8.392 8.448 rebuild_ks_matrix 119 8.3 0.000 0.000 6.243 6.301 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.032 6.243 6.301 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.534 5.586 init_scf_run 11 5.9 0.000 0.001 5.567 5.567 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.567 5.567 calculate_first_density_matrix 1 7.0 0.001 0.005 4.172 4.174 calculate_dm_sparse 119 9.5 0.000 0.000 3.976 4.031 qs_ot_get_p 119 10.4 0.002 0.007 3.561 3.639 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.117 3.514 sum_up_and_integrate 119 10.3 0.001 0.002 3.378 3.387 integrate_v_rspace 119 11.3 0.003 0.003 3.366 3.376 init_scf_loop 11 6.9 0.001 0.003 3.375 3.375 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.225 3.258 calculate_rho_elec 119 8.7 0.040 0.046 3.225 3.258 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.073 3.082 jit_kernel_multiply 6 16.0 0.419 2.593 0.419 2.593 acc_transpose_blocks 27432 15.5 0.117 0.121 2.395 2.528 prepare_preconditioner 11 7.9 0.000 0.000 2.515 2.523 make_preconditioner 11 8.9 0.000 0.001 2.515 2.523 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.123 2.444 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.322 2.351 make_m2s 4572 13.5 0.053 0.055 2.237 2.346 mp_waitall_1 137007 16.6 1.724 2.250 1.724 2.250 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.112 2.244 apply_single 119 13.6 0.000 0.000 2.112 2.244 make_images 4572 14.5 0.272 0.334 2.129 2.236 qs_ot_p2m_diag 50 11.0 0.016 0.026 2.160 2.170 grid_integrate_task_list 119 12.3 1.831 1.897 1.831 1.897 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.828 1.830 ot_diis_step 108 11.5 0.012 0.012 1.828 1.828 density_rs2pw 119 9.7 0.004 0.005 1.664 1.770 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.761 1.762 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.683 1.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.667 acc_transpose_blocks_sync 82296 16.5 1.421 1.556 1.421 1.556 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.519 1.528 cp_fm_diag_elpa_base 50 14.0 1.486 1.502 1.517 1.526 fft_wrap_pw1pw2_140 487 12.2 0.071 0.072 1.393 1.430 fft3d_ps 1201 13.6 0.576 0.639 1.345 1.368 wfi_extrapolate 11 7.9 0.003 0.022 1.341 1.341 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.765 1.333 grid_collocate_task_list 119 9.7 1.249 1.318 1.249 1.318 mp_sum_l 7287 12.8 1.020 1.304 1.020 1.304 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.284 1.302 potential_pw2rs 119 12.3 0.008 0.009 1.188 1.190 cp_fm_upper_to_full 72 14.2 0.821 1.163 0.821 1.163 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.114 1.115 dbcsr_complete_redistribute 329 12.2 0.121 0.161 0.770 1.046 transfer_rs2pw 487 10.6 0.005 0.005 0.904 1.020 make_images_data 4572 15.5 0.047 0.051 0.832 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.801 0.880 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.711 0.877 mp_alltoall_d11v 2130 13.8 0.731 0.871 0.731 0.871 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.829 0.843 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.799 0.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.114000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.272727, yerr=2.699862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 620.138496E+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 0.055 0.147 30.334 30.335 qs_mol_dyn_low 1 2.0 0.003 0.003 29.902 29.991 qs_forces 11 3.9 0.002 0.003 29.781 29.783 qs_energies 11 4.9 0.001 0.001 28.091 28.094 scf_env_do_scf 11 5.9 0.000 0.001 21.482 21.482 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.077 19.078 dbcsr_multiply_generic 2286 12.5 0.095 0.098 15.021 15.083 velocity_verlet 10 3.0 0.048 0.056 14.432 14.435 multiply_cannon 2286 13.5 0.225 0.237 12.478 12.959 multiply_cannon_loop 2286 14.5 0.644 0.661 11.565 11.742 qs_scf_new_mos 108 7.5 0.001 0.001 11.702 11.730 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.702 11.729 ot_scf_mini 108 9.5 0.002 0.002 10.998 11.023 multiply_cannon_multrec 9144 15.5 1.722 1.975 8.015 8.279 ot_mini 108 10.5 0.001 0.001 6.366 6.397 dbcsr_mm_accdrv_process 12550 15.8 5.318 6.231 6.182 6.288 rebuild_ks_matrix 119 8.3 0.000 0.000 5.645 5.669 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.645 5.669 init_scf_run 11 5.9 0.000 0.001 5.177 5.177 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.177 5.177 qs_ot_get_derivative 108 11.5 0.001 0.001 5.084 5.111 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.019 5.041 calculate_first_density_matrix 1 7.0 0.000 0.000 4.071 4.072 calculate_dm_sparse 119 9.5 0.000 0.000 3.612 3.628 jit_kernel_multiply 8 15.3 0.825 3.374 0.825 3.374 sum_up_and_integrate 119 10.3 0.001 0.001 3.315 3.320 integrate_v_rspace 119 11.3 0.003 0.003 3.305 3.310 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.159 3.165 calculate_rho_elec 119 8.7 0.060 0.061 3.158 3.164 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.003 3.009 qs_ot_get_p 119 10.4 0.001 0.002 2.759 2.796 init_scf_loop 11 6.9 0.000 0.000 2.385 2.386 grid_integrate_task_list 119 12.3 1.867 1.925 1.867 1.925 make_m2s 4572 13.5 0.033 0.035 1.753 1.918 make_images 4572 14.5 0.269 0.299 1.663 1.828 mp_waitall_1 115863 16.7 1.301 1.821 1.301 1.821 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.797 1.799 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.754 1.773 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.708 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.676 1.679 make_preconditioner 11 8.9 0.000 0.000 1.676 1.679 density_rs2pw 119 9.7 0.003 0.003 1.554 1.639 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.634 1.635 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.569 1.595 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.555 1.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.480 1.489 acc_transpose_blocks 9144 15.5 0.042 0.044 1.461 1.484 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.421 1.434 grid_collocate_task_list 119 9.7 1.297 1.376 1.297 1.376 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.326 1.339 fft3d_ps 1201 13.6 0.635 0.650 1.320 1.331 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.274 1.283 cp_fm_diag_elpa_base 50 14.0 1.248 1.264 1.273 1.281 ot_diis_step 108 11.5 0.013 0.013 1.271 1.271 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.210 1.246 apply_single 119 13.6 0.000 0.000 1.210 1.246 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.221 1.223 potential_pw2rs 119 12.3 0.010 0.010 1.130 1.131 wfi_extrapolate 11 7.9 0.001 0.001 1.057 1.057 hybrid_alltoall_any 4725 16.4 0.066 0.179 0.753 0.982 make_images_data 4572 15.5 0.042 0.047 0.757 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.927 cp_fm_cholesky_invert 11 10.9 0.840 0.842 0.840 0.842 mp_alltoall_d11v 2130 13.8 0.747 0.825 0.747 0.825 transfer_rs2pw 487 10.6 0.004 0.005 0.748 0.820 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.811 0.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.745 0.748 acc_transpose_blocks_sync 27432 16.5 0.720 0.742 0.720 0.742 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.667 0.719 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.684 0.690 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.346 0.680 transfer_pw2rs 487 13.2 0.003 0.004 0.647 0.649 mp_allgather_i34 2286 14.5 0.230 0.638 0.230 0.638 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.335000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.818182, yerr=5.490789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 768.081920E+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 0.016 0.033 44.125 44.125 qs_mol_dyn_low 1 2.0 0.003 0.003 43.920 43.930 qs_forces 11 3.9 0.002 0.003 43.850 43.851 qs_energies 11 4.9 0.001 0.001 41.887 41.891 scf_env_do_scf 11 5.9 0.001 0.001 33.957 33.957 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.230 26.232 velocity_verlet 10 3.0 0.013 0.014 23.556 23.562 dbcsr_multiply_generic 2286 12.5 0.105 0.106 21.013 21.261 multiply_cannon 2286 13.5 0.295 0.297 17.177 18.125 qs_scf_new_mos 108 7.5 0.001 0.001 17.151 17.247 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.150 17.246 multiply_cannon_loop 2286 14.5 0.864 0.875 15.941 16.920 ot_scf_mini 108 9.5 0.002 0.002 15.992 16.091 multiply_cannon_multrec 9144 15.5 3.356 4.713 10.953 11.049 ot_mini 108 10.5 0.001 0.001 9.774 9.891 dbcsr_mm_accdrv_process 12550 15.8 6.737 8.700 7.458 8.732 qs_ot_get_derivative 108 11.5 0.001 0.001 7.737 7.838 init_scf_loop 11 6.9 0.000 0.000 7.701 7.704 rebuild_ks_matrix 119 8.3 0.000 0.000 6.902 7.042 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.901 7.042 prepare_preconditioner 11 7.9 0.000 0.000 6.771 6.785 make_preconditioner 11 8.9 0.000 0.000 6.771 6.785 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.266 6.665 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.218 6.346 init_scf_run 11 5.9 0.000 0.001 5.881 5.881 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.881 5.881 cp_fm_upper_to_full 72 14.2 3.207 4.659 3.207 4.659 calculate_first_density_matrix 1 7.0 0.000 0.000 4.468 4.469 calculate_dm_sparse 119 9.5 0.000 0.000 4.363 4.382 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.845 3.850 calculate_rho_elec 119 8.7 0.118 0.121 3.844 3.849 sum_up_and_integrate 119 10.3 0.001 0.001 3.634 3.642 integrate_v_rspace 119 11.3 0.004 0.004 3.624 3.631 qs_ot_get_p 119 10.4 0.001 0.002 3.312 3.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.363 3.370 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.816 3.212 mp_waitall_1 94719 16.7 2.188 3.164 2.188 3.164 dbcsr_complete_redistribute 329 12.2 0.285 0.289 2.013 2.861 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.728 2.571 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.180 2.502 apply_single 119 13.6 0.000 0.000 2.180 2.502 make_m2s 4572 13.5 0.037 0.037 2.270 2.413 mp_alltoall_i22 627 13.8 1.479 2.354 1.479 2.354 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.342 2.350 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.500 2.336 make_images 4572 14.5 0.354 0.389 2.150 2.293 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.313 2.220 grid_integrate_task_list 119 12.3 2.067 2.099 2.067 2.099 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.040 2.091 ot_diis_step 108 11.5 0.014 0.015 2.010 2.010 fft_wrap_pw1pw2_140 487 12.2 0.179 0.179 1.997 2.005 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.954 1.956 density_rs2pw 119 9.7 0.003 0.003 1.891 1.912 acc_transpose_blocks 9144 15.5 0.044 0.045 1.832 1.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.826 1.826 fft3d_ps 1201 13.6 0.859 0.875 1.794 1.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.773 1.774 mp_sum_l 7287 12.8 1.019 1.750 1.019 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.585 1.629 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.602 1.602 grid_collocate_task_list 119 9.7 1.512 1.529 1.512 1.529 cp_fm_cholesky_invert 11 10.9 1.362 1.365 1.362 1.365 wfi_extrapolate 11 7.9 0.001 0.001 1.356 1.356 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.326 1.326 cp_fm_diag_elpa_base 50 14.0 1.180 1.234 1.324 1.324 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.302 1.323 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.004 1.202 potential_pw2rs 119 12.3 0.013 0.014 1.163 1.165 acc_transpose_blocks_sync 27432 16.5 1.101 1.127 1.101 1.127 mp_alltoall_d11v 2130 13.8 1.106 1.123 1.106 1.123 qs_env_update_s_mstruct 11 6.9 0.008 0.017 1.099 1.116 make_images_data 4572 15.5 0.046 0.049 0.942 1.114 jit_kernel_multiply 6 15.5 0.692 1.055 0.692 1.055 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.034 qs_create_task_list 11 7.9 0.000 0.001 0.930 0.943 generate_qs_task_list 11 8.9 0.369 0.390 0.929 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.892 0.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.125000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=724.363636, yerr=11.818881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 498.581504E+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 1147888. 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 0.014 0.026 83.056 83.057 qs_mol_dyn_low 1 2.0 0.003 0.003 82.768 82.778 qs_forces 11 3.9 0.003 0.003 82.699 82.700 qs_energies 11 4.9 0.001 0.002 79.839 79.855 scf_env_do_scf 11 5.9 0.000 0.001 69.160 69.163 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.576 63.577 dbcsr_multiply_generic 2055 12.4 0.105 0.107 52.564 52.828 qs_scf_new_mos 99 7.5 0.000 0.001 46.766 46.881 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.766 46.880 multiply_cannon 2055 13.4 0.182 0.188 44.320 45.210 ot_scf_mini 99 9.5 0.002 0.002 44.409 44.518 multiply_cannon_loop 2055 14.4 1.791 1.836 43.358 44.289 velocity_verlet 10 3.0 0.001 0.002 42.963 42.964 ot_mini 99 10.5 0.001 0.001 26.361 26.474 multiply_cannon_multrec 49320 15.4 11.411 12.035 19.356 19.868 qs_ot_get_derivative 99 11.5 0.001 0.001 19.556 19.667 rebuild_ks_matrix 110 8.3 0.000 0.000 14.462 14.611 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.461 14.610 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.692 12.825 mp_waitall_1 220248 16.4 10.523 11.368 10.523 11.368 multiply_cannon_sync_h2d 49320 15.4 9.633 10.238 9.633 10.238 qs_ot_get_p 110 10.4 0.001 0.002 9.622 9.726 init_scf_run 11 5.9 0.000 0.002 8.658 8.659 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.658 8.659 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.649 8.177 dbcsr_mm_accdrv_process 87628 16.1 3.069 3.162 7.815 8.074 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.423 7.781 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.272 7.735 apply_single 110 13.6 0.000 0.001 7.272 7.734 sum_up_and_integrate 110 10.3 0.001 0.002 6.880 6.892 integrate_v_rspace 110 11.3 0.003 0.003 6.855 6.874 ot_diis_step 99 11.5 0.006 0.006 6.601 6.601 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.548 6.581 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.093 6.227 calculate_rho_elec 110 8.6 0.020 0.024 6.092 6.226 calculate_dm_sparse 110 9.5 0.001 0.001 5.611 5.722 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.664 5.665 init_scf_loop 11 6.9 0.000 0.000 5.556 5.557 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.225 5.298 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.100 5.123 cp_fm_diag_elpa_base 48 14.0 5.088 5.111 5.099 5.120 mp_sum_l 6594 12.7 3.920 4.746 3.920 4.746 calculate_first_density_matrix 1 7.0 0.000 0.000 4.505 4.512 jit_kernel_multiply 13 15.9 4.468 4.489 4.468 4.489 wfi_extrapolate 11 7.9 0.001 0.001 4.054 4.054 make_m2s 4110 13.4 0.060 0.065 3.867 3.961 make_images 4110 14.4 0.177 0.190 3.772 3.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.572 3.576 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.532 3.562 grid_integrate_task_list 110 12.3 3.256 3.455 3.256 3.455 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 2.290 3.416 prepare_preconditioner 11 7.9 0.000 0.000 3.382 3.401 make_preconditioner 11 8.9 0.000 0.000 3.382 3.401 density_rs2pw 110 9.6 0.004 0.004 3.143 3.341 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.233 3.280 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.130 3.178 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.129 3.168 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 2.645 2.709 acc_transpose_blocks 49320 15.4 0.206 0.214 2.558 2.636 mp_alltoall_d11v 2046 13.8 2.100 2.593 2.100 2.593 potential_pw2rs 110 12.3 0.005 0.006 2.467 2.483 fft_wrap_pw1pw2_140 451 12.1 0.092 0.097 2.235 2.298 grid_collocate_task_list 110 9.6 2.156 2.279 2.156 2.279 fft3d_ps 1111 13.6 0.731 0.817 2.185 2.236 mp_waitany 14300 13.8 1.814 2.090 1.814 2.090 transfer_rs2pw 451 10.6 0.006 0.006 1.865 2.063 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.908 1.927 make_images_data 4110 15.4 0.043 0.046 1.733 1.841 mp_sum_d 3893 11.9 1.275 1.809 1.275 1.809 cp_fm_cholesky_invert 11 10.9 1.802 1.805 1.802 1.805 hybrid_alltoall_any 4261 16.3 0.084 0.487 1.501 1.777 transfer_pw2rs 451 13.1 0.006 0.007 1.747 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.057000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.909091, yerr=3.232071 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 592.076800E+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 1258009. 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 0.017 0.035 70.932 70.934 qs_mol_dyn_low 1 2.0 0.007 0.021 70.596 70.609 qs_forces 11 3.9 0.003 0.004 70.468 70.470 qs_energies 11 4.9 0.001 0.003 67.137 67.142 scf_env_do_scf 11 5.9 0.000 0.001 56.248 56.252 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.729 48.729 dbcsr_multiply_generic 2055 12.4 0.115 0.118 40.341 40.563 velocity_verlet 10 3.0 0.006 0.020 35.926 35.931 multiply_cannon 2055 13.4 0.223 0.242 33.766 35.030 qs_scf_new_mos 99 7.5 0.001 0.001 33.089 33.235 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.089 33.235 multiply_cannon_loop 2055 14.4 1.158 1.179 32.482 33.169 ot_scf_mini 99 9.5 0.003 0.003 31.421 31.574 ot_mini 99 10.5 0.001 0.001 18.347 18.497 multiply_cannon_multrec 24660 15.4 6.963 8.791 16.408 18.196 rebuild_ks_matrix 110 8.3 0.000 0.000 13.517 13.705 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 13.517 13.705 qs_ot_get_derivative 99 11.5 0.001 0.001 12.537 12.683 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.891 12.056 mp_waitall_1 176588 16.5 7.547 9.851 7.547 9.851 dbcsr_mm_accdrv_process 52282 16.1 7.308 8.924 9.274 9.566 init_scf_run 11 5.9 0.000 0.002 8.505 8.506 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.505 8.506 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.152 7.705 multiply_cannon_sync_h2d 24660 15.4 6.417 7.610 6.417 7.610 init_scf_loop 11 6.9 0.000 0.000 7.483 7.484 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.500 7.136 apply_single 110 13.6 0.000 0.001 6.500 7.136 qs_ot_get_p 110 10.4 0.001 0.001 6.393 6.588 sum_up_and_integrate 110 10.3 0.002 0.003 6.225 6.235 integrate_v_rspace 110 11.3 0.002 0.003 6.197 6.209 ot_diis_step 99 11.5 0.010 0.010 5.764 5.764 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.667 5.675 calculate_rho_elec 110 8.6 0.039 0.047 5.666 5.675 prepare_preconditioner 11 7.9 0.000 0.000 5.470 5.493 make_preconditioner 11 8.9 0.000 0.000 5.470 5.493 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.714 5.418 calculate_dm_sparse 110 9.5 0.001 0.001 5.294 5.328 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.066 5.223 calculate_first_density_matrix 1 7.0 0.000 0.001 4.911 4.913 make_m2s 4110 13.4 0.057 0.060 4.189 4.677 make_images 4110 14.4 0.405 0.462 4.077 4.560 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.509 4.530 jit_kernel_multiply 9 16.4 1.600 4.170 1.600 4.170 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.048 4.049 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.860 3.870 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.489 3.502 wfi_extrapolate 11 7.9 0.001 0.003 3.501 3.502 cp_fm_diag_elpa_base 48 14.0 3.443 3.466 3.486 3.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.350 3.352 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.348 grid_integrate_task_list 110 12.3 3.169 3.291 3.169 3.291 fft_wrap_pw1pw2 1111 11.6 0.015 0.016 3.073 3.215 density_rs2pw 110 9.6 0.004 0.004 2.903 3.106 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.008 3.069 hybrid_alltoall_any 4261 16.3 0.106 0.458 2.029 2.835 make_images_data 4110 15.4 0.048 0.052 2.300 2.808 fft_wrap_pw1pw2_140 451 12.1 0.110 0.116 2.617 2.758 mp_sum_l 6594 12.7 1.767 2.640 1.767 2.640 fft3d_ps 1111 13.6 1.071 1.279 2.451 2.585 cp_fm_cholesky_invert 11 10.9 2.527 2.533 2.527 2.533 grid_collocate_task_list 110 9.6 2.172 2.330 2.172 2.330 potential_pw2rs 110 12.3 0.008 0.008 2.101 2.114 mp_alltoall_d11v 2046 13.8 1.745 2.000 1.745 2.000 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.967 1.996 acc_transpose_blocks 24660 15.4 0.115 0.118 1.973 1.994 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.876 1.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.807 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.769 1.748 mp_allgather_i34 2055 14.4 0.596 1.640 0.596 1.640 transfer_rs2pw 451 10.6 0.007 0.007 1.425 1.631 mp_irecv_dv 57340 16.2 0.638 1.519 0.638 1.519 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.934000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.454545, yerr=7.228273 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.394624E+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 0.009 0.027 61.047 61.048 qs_mol_dyn_low 1 2.0 0.011 0.066 60.786 60.797 qs_forces 11 3.9 0.003 0.003 60.619 60.620 qs_energies 11 4.9 0.001 0.001 57.447 57.451 scf_env_do_scf 11 5.9 0.000 0.001 47.566 47.566 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.121 39.122 velocity_verlet 10 3.0 0.007 0.047 32.284 32.293 dbcsr_multiply_generic 2055 12.4 0.111 0.116 30.415 30.653 multiply_cannon 2055 13.4 0.211 0.222 24.376 25.553 qs_scf_new_mos 99 7.5 0.001 0.001 24.843 24.931 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.843 24.930 multiply_cannon_loop 2055 14.4 0.815 0.838 23.232 24.315 ot_scf_mini 99 9.5 0.002 0.003 23.597 23.704 ot_mini 99 10.5 0.001 0.001 13.570 13.680 multiply_cannon_multrec 16440 15.4 3.720 4.612 11.895 12.728 rebuild_ks_matrix 110 8.3 0.000 0.000 12.010 12.129 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.009 12.128 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.566 10.673 mp_waitall_1 139946 16.5 6.618 9.688 6.618 9.688 qs_ot_get_derivative 99 11.5 0.001 0.001 9.184 9.290 init_scf_loop 11 6.9 0.000 0.000 8.410 8.411 dbcsr_mm_accdrv_process 34862 16.1 7.306 7.698 8.019 8.158 init_scf_run 11 5.9 0.000 0.002 7.391 7.391 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.391 7.391 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.200 7.118 prepare_preconditioner 11 7.9 0.000 0.000 6.694 6.713 make_preconditioner 11 8.9 0.000 0.000 6.694 6.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.016 6.367 sum_up_and_integrate 110 10.3 0.001 0.002 6.105 6.120 integrate_v_rspace 110 11.3 0.003 0.003 6.079 6.094 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.454 5.464 calculate_rho_elec 110 8.6 0.058 0.058 5.454 5.463 qs_ot_get_p 110 10.4 0.001 0.002 5.247 5.395 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.868 5.356 apply_single 110 13.6 0.000 0.000 4.868 5.356 calculate_dm_sparse 110 9.5 0.001 0.001 4.399 4.433 make_m2s 4110 13.4 0.049 0.051 4.022 4.400 calculate_first_density_matrix 1 7.0 0.000 0.000 4.376 4.377 ot_diis_step 99 11.5 0.011 0.011 4.358 4.358 make_images 4110 14.4 0.397 0.517 3.906 4.285 multiply_cannon_sync_h2d 16440 15.4 3.262 3.983 3.262 3.983 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.171 3.777 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.654 3.658 grid_integrate_task_list 110 12.3 3.186 3.399 3.186 3.399 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.338 3.350 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.290 3.290 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.020 3.028 wfi_extrapolate 11 7.9 0.001 0.001 2.926 2.926 density_rs2pw 110 9.6 0.004 0.004 2.708 2.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.848 2.850 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.768 2.777 cp_fm_diag_elpa_base 48 14.0 2.703 2.728 2.767 2.775 make_images_data 4110 15.4 0.045 0.049 2.241 2.700 hybrid_alltoall_any 4261 16.3 0.109 0.380 2.014 2.674 fft_wrap_pw1pw2_140 451 12.1 0.124 0.128 2.599 2.614 cp_fm_cholesky_invert 11 10.9 2.607 2.613 2.607 2.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.568 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.413 2.480 grid_collocate_task_list 110 9.6 2.225 2.377 2.225 2.377 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.888 2.366 fft3d_ps 1111 13.6 1.089 1.107 2.337 2.356 mp_irecv_dv 48980 15.7 0.812 2.230 0.812 2.230 mp_sum_l 6594 12.7 1.397 2.064 1.397 2.064 qs_energies_init_hamiltonians 11 5.9 0.008 0.018 2.054 2.055 mp_alltoall_d11v 2046 13.8 1.726 1.980 1.726 1.980 potential_pw2rs 110 12.3 0.010 0.010 1.969 1.977 dbcsr_complete_redistribute 325 12.2 0.338 0.369 1.424 1.892 cp_fm_upper_to_full 70 14.2 1.405 1.778 1.405 1.778 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.656 acc_transpose_blocks 16440 15.4 0.076 0.079 1.588 1.640 cp_fm_cholesky_decompose 22 10.9 1.558 1.579 1.558 1.579 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.376 1.474 mp_allgather_i34 2055 14.4 0.462 1.467 0.462 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.969 1.423 transfer_rs2pw 451 10.6 0.005 0.006 1.254 1.399 mp_waitany 17072 13.8 1.137 1.314 1.137 1.314 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.263 1.272 qs_env_update_s_mstruct 11 6.9 0.069 0.116 1.107 1.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.048000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.909091, yerr=8.151459 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 742.043648E+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 0.009 0.028 67.449 67.450 qs_mol_dyn_low 1 2.0 0.003 0.003 67.212 67.224 qs_forces 11 3.9 0.003 0.003 67.137 67.138 qs_energies 11 4.9 0.001 0.002 63.733 63.738 scf_env_do_scf 11 5.9 0.000 0.001 53.594 53.597 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.924 41.924 velocity_verlet 10 3.0 0.003 0.014 37.209 37.211 dbcsr_multiply_generic 2055 12.4 0.117 0.123 32.700 32.951 qs_scf_new_mos 99 7.5 0.001 0.001 27.579 27.692 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.579 27.691 multiply_cannon 2055 13.4 0.239 0.257 25.586 26.394 ot_scf_mini 99 9.5 0.002 0.003 25.889 25.981 multiply_cannon_loop 2055 14.4 1.402 1.476 24.183 24.835 multiply_cannon_multrec 24660 15.4 4.124 6.838 15.007 16.110 ot_mini 99 10.5 0.001 0.001 14.977 15.092 rebuild_ks_matrix 110 8.3 0.000 0.000 11.927 12.021 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.927 12.021 init_scf_loop 11 6.9 0.000 0.000 11.630 11.631 dbcsr_mm_accdrv_process 52304 16.0 9.702 11.111 10.724 11.623 qs_ot_get_derivative 99 11.5 0.001 0.001 10.835 10.938 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.535 10.620 prepare_preconditioner 11 7.9 0.000 0.000 9.899 9.914 make_preconditioner 11 8.9 0.000 0.000 9.899 9.914 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.066 9.537 init_scf_run 11 5.9 0.000 0.001 7.412 7.412 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.412 7.412 mp_waitall_1 121746 16.5 4.351 6.133 4.351 6.133 sum_up_and_integrate 110 10.3 0.001 0.002 6.102 6.114 integrate_v_rspace 110 11.3 0.003 0.003 6.076 6.088 qs_ot_get_p 110 10.4 0.001 0.002 5.768 5.913 make_m2s 4110 13.4 0.059 0.061 5.478 5.832 make_images 4110 14.4 0.572 0.692 5.337 5.687 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.581 5.592 calculate_rho_elec 110 8.6 0.077 0.081 5.580 5.591 calculate_dm_sparse 110 9.5 0.001 0.001 4.881 4.917 cp_fm_upper_to_full 70 14.2 3.388 4.904 3.388 4.904 calculate_first_density_matrix 1 7.0 0.000 0.000 4.334 4.337 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.037 4.108 apply_single 110 13.6 0.000 0.000 4.037 4.108 ot_diis_step 99 11.5 0.011 0.011 4.105 4.105 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.899 3.917 dbcsr_complete_redistribute 325 12.2 0.419 0.461 2.685 3.817 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.530 3.577 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.435 3.455 grid_integrate_task_list 110 12.3 3.270 3.435 3.270 3.435 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.331 3.384 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.339 3.340 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.174 3.305 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.433 3.241 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.184 3.210 hybrid_alltoall_any 4261 16.3 0.123 0.465 2.336 3.147 make_images_data 4110 15.4 0.048 0.051 2.750 3.109 wfi_extrapolate 11 7.9 0.001 0.001 2.986 2.986 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.824 2.942 density_rs2pw 110 9.6 0.004 0.004 2.710 2.898 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.839 2.840 mp_alltoall_i22 605 13.7 1.658 2.831 1.658 2.831 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.809 2.820 cp_fm_diag_elpa_base 48 14.0 2.663 2.722 2.807 2.818 fft_wrap_pw1pw2_140 451 12.1 0.142 0.145 2.717 2.745 cp_fm_cholesky_invert 11 10.9 2.627 2.635 2.627 2.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.473 2.505 acc_transpose_blocks 24660 15.4 0.110 0.112 2.401 2.499 fft3d_ps 1111 13.6 1.138 1.188 2.458 2.487 multiply_cannon_sync_h2d 24660 15.4 2.381 2.481 2.381 2.481 grid_collocate_task_list 110 9.6 2.261 2.392 2.261 2.392 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.231 2.232 mp_alltoall_d11v 2046 13.8 1.809 1.955 1.809 1.955 potential_pw2rs 110 12.3 0.012 0.012 1.916 1.920 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.797 1.831 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.718 cp_fm_cholesky_decompose 22 10.9 1.646 1.696 1.646 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.649 mp_sum_l 6594 12.7 0.947 1.553 0.947 1.553 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.835 1.542 acc_transpose_blocks_sync 73980 16.4 1.389 1.495 1.389 1.495 jit_kernel_multiply 8 15.7 0.677 1.485 0.677 1.485 mp_irecv_dv 62702 16.1 0.732 1.460 0.732 1.460 transfer_rs2pw 451 10.6 0.005 0.006 1.173 1.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.450000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.272727, yerr=11.386522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 860.155904E+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 0.010 0.027 55.923 55.924 qs_mol_dyn_low 1 2.0 0.003 0.003 55.699 55.709 qs_forces 11 3.9 0.003 0.003 55.592 55.593 qs_energies 11 4.9 0.001 0.001 51.974 51.978 scf_env_do_scf 11 5.9 0.000 0.001 41.873 41.874 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.340 34.341 velocity_verlet 10 3.0 0.044 0.047 30.526 30.529 dbcsr_multiply_generic 2055 12.4 0.106 0.109 25.016 25.146 multiply_cannon 2055 13.4 0.237 0.249 19.581 20.715 qs_scf_new_mos 99 7.5 0.001 0.001 20.281 20.331 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.280 20.331 ot_scf_mini 99 9.5 0.002 0.002 19.027 19.058 multiply_cannon_loop 2055 14.4 0.603 0.616 18.328 18.625 rebuild_ks_matrix 110 8.3 0.000 0.000 11.424 11.443 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.424 11.443 ot_mini 99 10.5 0.001 0.001 10.635 10.657 multiply_cannon_multrec 8220 15.4 3.187 4.466 9.557 10.492 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.165 10.182 mp_waitall_1 103326 16.6 5.760 7.492 5.760 7.492 init_scf_loop 11 6.9 0.000 0.000 7.489 7.490 dbcsr_mm_accdrv_process 17442 15.9 4.803 5.980 6.229 7.135 qs_ot_get_derivative 99 11.5 0.001 0.001 6.965 6.993 init_scf_run 11 5.9 0.000 0.001 6.961 6.961 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.961 6.961 sum_up_and_integrate 110 10.3 0.001 0.002 5.988 5.999 integrate_v_rspace 110 11.3 0.003 0.003 5.961 5.973 prepare_preconditioner 11 7.9 0.000 0.000 5.858 5.865 make_preconditioner 11 8.9 0.000 0.000 5.858 5.865 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.617 5.631 calculate_rho_elec 110 8.6 0.115 0.116 5.616 5.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.440 5.513 qs_ot_get_p 110 10.4 0.001 0.002 4.468 4.483 make_m2s 4110 13.4 0.037 0.039 4.142 4.434 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.994 4.425 calculate_dm_sparse 110 9.5 0.001 0.001 4.345 4.377 make_images 4110 14.4 0.645 0.703 4.013 4.303 calculate_first_density_matrix 1 7.0 0.000 0.000 4.229 4.231 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.712 3.764 apply_single 110 13.6 0.000 0.000 3.712 3.764 ot_diis_step 99 11.5 0.012 0.012 3.647 3.647 grid_integrate_task_list 110 12.3 3.385 3.528 3.385 3.528 jit_kernel_multiply 8 15.5 1.111 3.341 1.111 3.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.295 3.303 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.281 3.290 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.148 3.152 fft_wrap_pw1pw2_140 451 12.1 0.193 0.195 2.838 2.856 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.814 2.815 cp_fm_cholesky_invert 11 10.9 2.809 2.813 2.809 2.813 density_rs2pw 110 9.6 0.003 0.004 2.632 2.763 hybrid_alltoall_any 4261 16.3 0.205 0.889 2.189 2.712 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.689 2.690 wfi_extrapolate 11 7.9 0.001 0.001 2.641 2.641 make_images_data 4110 15.4 0.041 0.046 2.230 2.612 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.578 2.580 multiply_cannon_sync_h2d 8220 15.4 2.386 2.524 2.386 2.524 grid_collocate_task_list 110 9.6 2.367 2.468 2.367 2.468 fft3d_ps 1111 13.6 1.273 1.292 2.439 2.450 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.320 2.328 cp_fm_diag_elpa_base 48 14.0 2.266 2.288 2.319 2.326 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.226 2.233 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.983 2.015 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.793 2.001 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.875 1.888 potential_pw2rs 110 12.3 0.015 0.015 1.796 1.799 mp_alltoall_d11v 2046 13.8 1.568 1.796 1.568 1.796 cp_fm_cholesky_decompose 22 10.9 1.617 1.632 1.617 1.632 qs_env_update_s_mstruct 11 6.9 0.008 0.014 1.511 1.632 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.587 1.590 dbcsr_complete_redistribute 325 12.2 0.571 0.588 1.408 1.493 mp_allgather_i34 2055 14.4 0.429 1.476 0.429 1.476 acc_transpose_blocks 8220 15.4 0.040 0.040 1.370 1.394 qs_create_task_list 11 7.9 0.002 0.002 1.206 1.305 generate_qs_task_list 11 8.9 0.373 0.440 1.205 1.303 transfer_rs2pw 451 10.6 0.005 0.005 1.093 1.238 mp_waitany 9240 13.8 1.043 1.194 1.043 1.194 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.737 1.167 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.106 1.121 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.924000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.090909, yerr=9.755693 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 1.393500E+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 0.014 0.032 87.327 87.328 qs_mol_dyn_low 1 2.0 0.003 0.003 87.034 87.044 qs_forces 11 3.9 0.003 0.003 86.964 86.964 qs_energies 11 4.9 0.001 0.002 82.877 82.878 scf_env_do_scf 11 5.9 0.000 0.001 71.173 71.174 velocity_verlet 10 3.0 0.002 0.002 54.794 54.801 scf_env_do_scf_inner_loop 99 6.5 0.015 0.019 43.245 43.247 dbcsr_multiply_generic 2055 12.4 0.132 0.137 31.020 31.123 init_scf_loop 11 6.9 0.000 0.000 27.858 27.862 qs_scf_new_mos 99 7.5 0.001 0.001 26.464 26.499 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.464 26.498 prepare_preconditioner 11 7.9 0.000 0.000 25.905 25.913 make_preconditioner 11 8.9 0.000 0.000 25.905 25.913 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.259 25.379 multiply_cannon 2055 13.4 0.333 0.353 24.084 24.811 ot_scf_mini 99 9.5 0.002 0.002 24.673 24.699 multiply_cannon_loop 2055 14.4 0.824 0.838 22.316 22.603 cp_fm_upper_to_full 70 14.2 12.810 18.166 12.810 18.166 ot_mini 99 10.5 0.001 0.001 13.973 14.003 rebuild_ks_matrix 110 8.3 0.000 0.000 13.313 13.348 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.312 13.348 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.075 12.106 multiply_cannon_multrec 8220 15.4 4.065 4.239 11.530 11.688 dbcsr_complete_redistribute 325 12.2 1.031 1.057 7.193 10.242 qs_ot_get_derivative 99 11.5 0.001 0.001 9.414 9.442 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.181 9.228 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.631 8.659 mp_alltoall_i22 605 13.7 5.244 8.283 5.244 8.283 mp_waitall_1 84994 16.7 7.285 8.156 7.285 8.156 dbcsr_mm_accdrv_process 11614 15.7 3.872 4.134 7.316 7.527 init_scf_run 11 5.9 0.000 0.001 7.439 7.439 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.438 7.438 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.773 6.808 calculate_rho_elec 110 8.6 0.227 0.227 6.772 6.808 sum_up_and_integrate 110 10.3 0.002 0.002 6.523 6.536 integrate_v_rspace 110 11.3 0.004 0.004 6.495 6.508 make_m2s 4110 13.4 0.043 0.043 5.218 5.704 make_images 4110 14.4 0.880 0.938 5.030 5.515 qs_ot_get_p 110 10.4 0.001 0.002 5.440 5.472 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.731 5.203 apply_single 110 13.6 0.000 0.000 4.731 5.203 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.746 5.079 cp_fm_cholesky_invert 11 10.9 5.020 5.024 5.020 5.024 calculate_dm_sparse 110 9.5 0.001 0.001 4.878 4.904 ot_diis_step 99 11.5 0.015 0.015 4.542 4.542 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.424 4.429 calculate_first_density_matrix 1 7.0 0.000 0.000 4.078 4.079 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.909 3.917 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.875 3.884 grid_integrate_task_list 110 12.3 3.671 3.736 3.671 3.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.663 3.664 jit_kernel_multiply 10 15.2 3.250 3.501 3.250 3.501 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.995 3.479 hybrid_alltoall_any 4261 16.3 0.266 0.575 2.745 3.451 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.383 3.400 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.395 3.396 make_images_data 4110 15.4 0.044 0.047 2.725 3.312 fft3d_ps 1111 13.6 1.828 1.843 3.284 3.293 density_rs2pw 110 9.6 0.004 0.004 3.253 3.273 wfi_extrapolate 11 7.9 0.001 0.001 3.251 3.251 multiply_cannon_sync_h2d 8220 15.4 3.131 3.154 3.131 3.154 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.939 2.940 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.826 2.827 cp_fm_diag_elpa_base 48 14.0 2.285 2.472 2.824 2.824 grid_collocate_task_list 110 9.6 2.662 2.675 2.662 2.675 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.607 2.622 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.448 2.461 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.180 2.236 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.184 potential_pw2rs 110 12.3 0.021 0.021 2.078 2.080 mp_alltoall_d11v 2046 13.8 1.954 1.999 1.954 1.999 cp_fm_cholesky_decompose 22 10.9 1.921 1.952 1.921 1.952 qs_create_task_list 11 7.9 0.001 0.001 1.878 1.924 generate_qs_task_list 11 8.9 0.733 0.786 1.877 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.842 acc_transpose_blocks 8220 15.4 0.041 0.042 1.728 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.328000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1239.818182, yerr=56.500713 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.476800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 57647. MP_Allreduce 11104 795. MP_Sync 87 MP_Alltoall 2226 1622406. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.040 207.685 207.687 qs_mol_dyn_low 1 2.0 0.004 0.010 207.123 207.136 qs_forces 11 3.9 0.005 0.013 207.028 207.030 qs_energies 11 4.9 0.002 0.004 201.437 201.456 scf_env_do_scf 11 5.9 0.001 0.001 183.159 183.163 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.786 162.788 dbcsr_multiply_generic 2507 12.6 0.177 0.181 126.972 127.404 qs_scf_new_mos 117 7.6 0.001 0.001 123.852 124.070 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.851 124.069 velocity_verlet 10 3.0 0.002 0.006 123.057 123.059 ot_scf_mini 117 9.6 0.003 0.003 117.150 117.367 multiply_cannon 2507 13.6 0.239 0.249 103.061 104.103 multiply_cannon_loop 2507 14.6 2.382 2.446 100.942 102.082 ot_mini 117 10.6 0.001 0.001 65.911 66.129 multiply_cannon_multrec 60168 15.6 31.863 33.796 43.318 45.225 qs_ot_get_derivative 117 11.6 0.001 0.001 41.103 41.327 rebuild_ks_matrix 128 8.3 0.001 0.001 33.592 33.869 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.592 33.869 mp_waitall_1 267128 16.5 28.571 31.655 28.571 31.655 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.209 30.442 qs_ot_get_p 128 10.4 0.001 0.002 29.155 29.469 multiply_cannon_sync_h2d 60168 15.6 26.471 28.171 26.471 28.171 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.373 25.003 apply_single 128 13.6 0.001 0.001 24.373 25.003 ot_diis_step 117 11.6 0.007 0.008 24.546 24.547 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.517 22.599 init_scf_loop 11 6.9 0.000 0.000 20.297 20.298 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.614 19.805 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.617 19.617 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 15.948 18.045 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.686 16.710 cp_fm_diag_elpa_base 83 14.4 16.617 16.649 16.682 16.707 prepare_preconditioner 11 7.9 0.000 0.000 15.664 15.697 make_preconditioner 11 8.9 0.000 0.000 15.664 15.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.918 15.091 make_m2s 5014 13.6 0.104 0.113 13.858 14.268 init_scf_run 11 5.9 0.000 0.001 14.139 14.140 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.139 14.140 make_images 5014 14.6 0.401 0.420 13.674 14.094 sum_up_and_integrate 128 10.3 0.002 0.005 13.715 13.730 integrate_v_rspace 128 11.3 0.004 0.005 13.657 13.674 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.597 12.719 calculate_rho_elec 128 8.7 0.045 0.064 12.596 12.718 dbcsr_mm_accdrv_process 124484 16.2 4.723 4.878 11.015 11.474 calculate_dm_sparse 128 9.5 0.001 0.001 10.196 10.300 mp_sum_l 7950 12.9 8.799 9.817 8.799 9.817 wfi_extrapolate 11 7.9 0.001 0.004 8.971 8.972 cp_fm_cholesky_invert 11 10.9 8.913 8.921 8.913 8.921 multiply_cannon_metrocomm1 60168 15.6 0.092 0.097 6.140 8.214 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.046 8.133 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.998 8.116 make_images_data 5014 15.6 0.065 0.071 6.772 7.700 grid_integrate_task_list 128 12.3 7.068 7.455 7.068 7.455 hybrid_alltoall_any 5200 16.5 0.293 2.258 5.942 7.140 density_rs2pw 128 9.7 0.005 0.006 6.261 7.044 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.444 6.740 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.680 6.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.801 5.880 fft_wrap_pw1pw2_140 523 12.2 0.233 0.243 5.587 5.754 fft3d_ps 1291 13.7 2.093 2.596 5.271 5.518 grid_collocate_task_list 128 9.7 4.839 5.349 4.839 5.349 mp_alltoall_d11v 2415 14.1 4.240 5.274 4.240 5.274 calculate_first_density_matrix 1 7.0 0.002 0.013 4.993 4.998 cp_fm_cholesky_decompose 22 10.9 4.594 4.606 4.594 4.606 jit_kernel_multiply 13 15.9 4.221 4.279 4.221 4.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.205 4.250 potential_pw2rs 128 12.3 0.009 0.010 4.217 4.238 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.687000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.727273, yerr=7.174334 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.166473E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 842.928128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069299E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842592400 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58313. MP_Allreduce 11066 961. MP_Sync 86 MP_Alltoall 1955 5932926. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 192.010 192.010 qs_mol_dyn_low 1 2.0 0.003 0.003 191.644 191.661 qs_forces 11 3.9 0.004 0.005 191.473 191.474 qs_energies 11 4.9 0.001 0.002 184.663 184.674 scf_env_do_scf 11 5.9 0.001 0.001 166.541 166.551 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.109 133.111 velocity_verlet 10 3.0 0.001 0.002 119.358 119.361 dbcsr_multiply_generic 2485 12.5 0.186 0.192 98.739 99.934 qs_scf_new_mos 116 7.6 0.001 0.001 94.807 95.405 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.807 95.404 ot_scf_mini 116 9.6 0.004 0.004 90.019 90.749 multiply_cannon 2485 13.5 0.502 0.557 78.381 83.394 multiply_cannon_loop 2485 14.5 1.548 1.607 74.903 77.442 ot_mini 116 10.6 0.001 0.001 50.198 50.865 mp_waitall_1 212858 16.6 23.990 38.799 23.990 38.799 multiply_cannon_multrec 29820 15.5 20.827 25.157 32.925 37.666 init_scf_loop 11 6.9 0.000 0.000 33.335 33.336 rebuild_ks_matrix 127 8.3 0.001 0.001 32.477 33.230 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 32.476 33.229 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.199 29.873 qs_ot_get_derivative 116 11.6 0.001 0.002 28.381 29.094 prepare_preconditioner 11 7.9 0.000 0.000 28.943 29.027 make_preconditioner 11 8.9 0.000 0.000 28.943 29.027 multiply_cannon_metrocomm3 29820 15.5 0.095 0.101 15.487 28.477 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.550 28.086 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.903 23.058 apply_single 127 13.6 0.001 0.001 21.903 23.057 qs_ot_get_p 127 10.4 0.001 0.002 21.897 22.658 ot_diis_step 116 11.6 0.014 0.015 21.631 21.633 multiply_cannon_sync_h2d 29820 15.5 17.897 21.027 17.897 21.027 qs_ot_p2m_diag 82 11.4 0.186 0.214 16.898 16.932 cp_fm_cholesky_invert 11 10.9 16.776 16.789 16.776 16.789 make_m2s 4970 13.5 0.086 0.092 14.320 16.045 make_images 4970 14.5 1.155 1.353 14.105 15.831 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.685 15.686 sum_up_and_integrate 127 10.3 0.002 0.004 13.825 13.852 integrate_v_rspace 127 11.3 0.003 0.004 13.766 13.795 init_scf_run 11 5.9 0.000 0.002 13.298 13.300 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.298 13.300 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.972 12.999 calculate_rho_elec 127 8.7 0.087 0.105 12.971 12.998 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.482 12.514 cp_fm_diag_elpa_base 82 14.4 12.220 12.327 12.475 12.508 dbcsr_mm_accdrv_process 61726 16.2 6.730 7.734 11.547 12.136 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.220 11.730 multiply_cannon_metrocomm4 27335 15.5 0.106 0.119 3.840 10.609 make_images_data 4970 15.5 0.065 0.074 8.422 10.435 mp_irecv_dv 68888 16.3 3.632 10.216 3.632 10.216 hybrid_alltoall_any 5155 16.4 0.349 1.521 7.253 9.856 wfi_extrapolate 11 7.9 0.001 0.001 8.436 8.436 calculate_dm_sparse 127 9.5 0.001 0.001 7.960 8.120 fft_wrap_pw1pw2 1281 11.7 0.019 0.025 7.525 7.602 grid_integrate_task_list 127 12.3 7.132 7.427 7.132 7.427 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.333 7.128 cp_fm_cholesky_decompose 22 10.9 7.026 7.122 7.026 7.122 density_rs2pw 127 9.7 0.005 0.006 6.680 7.043 fft_wrap_pw1pw2_140 519 12.2 0.246 0.254 6.526 6.598 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.245 6.257 mp_sum_l 7884 12.9 4.252 6.174 4.252 6.174 fft3d_ps 1281 13.7 2.741 2.913 5.991 6.056 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.297 5.517 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.369 5.468 grid_collocate_task_list 127 9.7 4.965 5.310 4.965 5.310 mp_allgather_i34 2485 14.5 2.010 5.122 2.010 5.122 mp_alltoall_d11v 2401 14.1 4.391 4.832 4.391 4.832 calculate_first_density_matrix 1 7.0 0.000 0.000 4.674 4.679 potential_pw2rs 127 12.3 0.014 0.017 4.328 4.340 dbcsr_complete_redistribute 393 12.7 0.787 0.892 3.416 4.256 mp_sum_d 4452 12.1 2.700 4.055 2.700 4.055 jit_kernel_multiply 12 15.9 1.782 3.885 1.782 3.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.010000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.454545, yerr=2.425757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/17/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 2.950886E+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 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 987.787264E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 176.578 176.580 qs_mol_dyn_low 1 2.0 0.011 0.035 176.127 176.143 qs_forces 11 3.9 0.007 0.031 175.995 176.001 qs_energies 11 4.9 0.028 0.054 169.448 169.459 scf_env_do_scf 11 5.9 0.001 0.001 152.150 152.157 scf_env_do_scf_inner_loop 118 6.6 0.006 0.025 117.019 117.020 velocity_verlet 10 3.0 0.005 0.016 112.544 112.553 dbcsr_multiply_generic 2527 12.6 0.185 0.189 84.034 85.021 qs_scf_new_mos 118 7.6 0.001 0.001 80.052 80.425 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.052 80.425 ot_scf_mini 118 9.6 0.004 0.004 75.766 76.187 multiply_cannon 2527 13.6 0.510 0.527 64.189 68.955 multiply_cannon_loop 2527 14.6 1.136 1.195 61.227 63.440 ot_mini 118 10.6 0.001 0.001 42.791 43.193 init_scf_loop 11 6.9 0.002 0.014 35.033 35.034 mp_waitall_1 171878 16.6 25.040 35.006 25.040 35.006 prepare_preconditioner 11 7.9 0.000 0.000 30.982 31.028 make_preconditioner 11 8.9 0.003 0.024 30.982 31.028 rebuild_ks_matrix 129 8.3 0.001 0.001 30.524 31.015 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 30.524 31.014 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.426 29.937 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.574 28.027 multiply_cannon_multrec 20216 15.6 13.135 16.179 24.096 27.169 multiply_cannon_metrocomm3 20216 15.6 0.061 0.064 15.621 24.595 qs_ot_get_derivative 118 11.6 0.001 0.002 22.798 23.217 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.086 21.114 apply_single 129 13.6 0.001 0.001 20.086 21.114 ot_diis_step 118 11.6 0.018 0.018 19.877 19.877 qs_ot_get_p 129 10.4 0.002 0.002 18.171 18.674 make_m2s 5054 13.6 0.078 0.083 14.841 16.056 make_images 5054 14.6 1.149 1.236 14.604 15.808 multiply_cannon_sync_h2d 20216 15.6 13.683 15.466 13.683 15.466 cp_fm_cholesky_invert 11 10.9 14.220 14.229 14.220 14.229 qs_ot_p2m_diag 83 11.4 0.266 0.278 13.969 13.977 sum_up_and_integrate 129 10.3 0.002 0.003 13.689 13.717 integrate_v_rspace 129 11.3 0.004 0.005 13.628 13.651 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.095 13.129 calculate_rho_elec 129 8.7 0.133 0.147 13.095 13.128 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.917 12.919 init_scf_run 11 5.9 0.001 0.005 12.039 12.039 scf_env_initial_rho_setup 11 6.9 0.003 0.004 12.038 12.039 make_images_data 5054 15.6 0.062 0.072 8.985 10.612 dbcsr_mm_accdrv_process 41830 16.2 7.238 7.582 10.416 10.586 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.766 9.787 cp_fm_diag_elpa_base 83 14.4 9.358 9.510 9.763 9.784 hybrid_alltoall_any 5240 16.5 0.453 2.065 7.764 9.751 multiply_cannon_metrocomm4 17689 15.6 0.067 0.076 3.545 9.684 mp_irecv_dv 50625 16.2 3.413 9.416 3.413 9.416 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.930 9.223 grid_integrate_task_list 129 12.3 7.374 7.782 7.374 7.782 fft_wrap_pw1pw2 1301 11.7 0.019 0.025 7.522 7.629 calculate_dm_sparse 129 9.5 0.001 0.001 7.388 7.479 cp_fm_upper_to_full 105 14.8 5.817 7.460 5.817 7.460 cp_fm_cholesky_decompose 22 10.9 7.316 7.406 7.316 7.406 wfi_extrapolate 11 7.9 0.002 0.006 7.367 7.370 fft_wrap_pw1pw2_140 527 12.2 0.285 0.293 6.647 6.777 density_rs2pw 129 9.7 0.006 0.014 6.399 6.763 dbcsr_complete_redistribute 395 12.7 1.160 1.194 4.770 6.550 fft3d_ps 1301 13.7 2.776 3.007 5.820 5.892 grid_collocate_task_list 129 9.7 5.197 5.689 5.197 5.689 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.539 5.552 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.626 5.402 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.621 5.302 mp_allgather_i34 2527 14.6 1.467 4.813 1.467 4.813 mp_alltoall_d11v 2423 14.1 4.359 4.783 4.359 4.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.614 4.760 calculate_first_density_matrix 1 7.0 0.006 0.031 4.471 4.475 mp_sum_l 8010 12.9 3.165 4.456 3.165 4.456 transfer_fm_to_dbcsr 11 9.9 0.017 0.022 2.534 4.296 potential_pw2rs 129 12.3 0.019 0.021 4.070 4.083 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.977 4.006 mp_alltoall_i22 718 14.1 2.102 3.977 2.102 3.977 qs_energies_init_hamiltonians 11 5.9 0.020 0.026 3.929 3.937 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.654 3.687 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.580000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=935.818182, yerr=13.663609 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/18/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 4.353791E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.152356E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.039 188.603 188.604 qs_mol_dyn_low 1 2.0 0.003 0.003 188.174 188.189 qs_forces 11 3.9 0.004 0.005 188.047 188.059 qs_energies 11 4.9 0.012 0.014 180.954 180.965 scf_env_do_scf 11 5.9 0.001 0.001 162.004 162.017 velocity_verlet 10 3.0 0.002 0.002 123.799 123.802 scf_env_do_scf_inner_loop 117 6.6 0.031 0.046 115.555 115.556 dbcsr_multiply_generic 2507 12.6 0.190 0.195 82.148 82.985 qs_scf_new_mos 117 7.6 0.001 0.001 79.896 80.248 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.895 80.247 ot_scf_mini 117 9.6 0.003 0.003 75.327 75.662 multiply_cannon 2507 13.6 0.554 0.580 57.696 60.613 multiply_cannon_loop 2507 14.6 1.830 1.906 54.121 56.183 init_scf_loop 11 6.9 0.000 0.000 46.326 46.327 ot_mini 117 10.6 0.001 0.001 42.641 42.987 prepare_preconditioner 11 7.9 0.000 0.000 42.247 42.268 make_preconditioner 11 8.9 0.000 0.000 42.247 42.268 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.755 40.886 multiply_cannon_multrec 30084 15.6 13.580 18.672 28.456 33.226 rebuild_ks_matrix 128 8.3 0.001 0.001 29.214 29.534 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.213 29.533 mp_waitall_1 147882 16.7 17.011 26.756 17.011 26.756 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.371 26.653 qs_ot_get_derivative 117 11.6 0.001 0.002 23.050 23.396 make_m2s 5014 13.6 0.092 0.098 20.100 21.077 make_images 5014 14.6 1.940 2.286 19.792 20.777 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.009 19.696 apply_single 128 13.6 0.001 0.001 19.009 19.696 ot_diis_step 117 11.6 0.017 0.018 19.463 19.464 qs_ot_get_p 128 10.4 0.016 0.019 18.863 19.204 cp_fm_upper_to_full 105 14.8 11.671 17.208 11.671 17.208 cp_fm_cholesky_invert 11 10.9 15.956 15.965 15.956 15.965 dbcsr_mm_accdrv_process 62264 16.2 10.064 11.147 14.443 14.980 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.427 14.922 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.740 14.792 sum_up_and_integrate 128 10.3 0.002 0.003 13.615 13.643 integrate_v_rspace 128 11.3 0.003 0.004 13.556 13.585 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.345 13.346 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.102 13.147 calculate_rho_elec 128 8.7 0.174 0.190 13.102 13.146 init_scf_run 11 5.9 0.000 0.002 13.033 13.034 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.033 13.034 dbcsr_complete_redistribute 395 12.7 1.491 1.640 9.229 13.025 make_images_data 5014 15.6 0.064 0.070 10.766 12.513 hybrid_alltoall_any 5200 16.5 0.526 2.204 9.693 11.730 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.881 11.705 multiply_cannon_sync_h2d 30084 15.6 10.556 11.278 10.556 11.278 transfer_fm_to_dbcsr 11 9.9 0.000 0.004 6.469 10.226 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.155 10.173 cp_fm_diag_elpa_base 83 14.4 9.188 9.516 10.149 10.165 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.678 9.917 mp_alltoall_i22 716 14.1 5.710 9.487 5.710 9.487 calculate_dm_sparse 128 9.5 0.001 0.001 8.162 8.251 grid_integrate_task_list 128 12.3 7.573 7.885 7.573 7.885 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.654 7.725 wfi_extrapolate 11 7.9 0.001 0.001 7.695 7.695 cp_fm_cholesky_decompose 22 10.9 7.435 7.538 7.435 7.538 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.783 7.376 mp_irecv_dv 76098 16.2 2.630 7.105 2.630 7.105 fft_wrap_pw1pw2_140 523 12.2 0.323 0.333 6.781 6.858 density_rs2pw 128 9.7 0.005 0.005 6.148 6.540 fft3d_ps 1291 13.7 2.934 3.025 5.851 5.913 grid_collocate_task_list 128 9.7 5.286 5.665 5.286 5.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.590 5.654 mp_alltoall_d11v 2415 14.1 4.873 5.300 4.873 5.300 calculate_first_density_matrix 1 7.0 0.000 0.000 5.142 5.148 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.442 4.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.467 4.546 qs_energies_init_hamiltonians 11 5.9 0.006 0.014 4.407 4.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.261 4.293 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.231 4.286 potential_pw2rs 128 12.3 0.022 0.022 3.894 3.902 jit_kernel_multiply 10 15.7 1.607 3.815 1.607 3.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.604000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1088.545455, yerr=15.640063 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.557103E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58130. MP_Allreduce 11095 1166. MP_Sync 86 MP_Alltoall 1700 18828153. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.038 171.235 171.236 qs_mol_dyn_low 1 2.0 0.003 0.003 170.697 170.713 qs_forces 11 3.9 0.014 0.025 170.577 170.580 qs_energies 11 4.9 0.010 0.010 163.160 163.177 scf_env_do_scf 11 5.9 0.001 0.001 143.470 143.478 velocity_verlet 10 3.0 0.002 0.002 110.319 110.324 scf_env_do_scf_inner_loop 116 6.6 0.015 0.022 107.475 107.476 dbcsr_multiply_generic 2485 12.5 0.214 0.220 73.914 74.394 qs_scf_new_mos 116 7.6 0.001 0.001 71.486 71.638 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.486 71.637 ot_scf_mini 116 9.6 0.003 0.004 67.123 67.295 multiply_cannon 2485 13.5 0.553 0.572 55.120 58.292 multiply_cannon_loop 2485 14.5 0.804 0.838 52.145 53.229 ot_mini 116 10.6 0.001 0.001 37.446 37.628 init_scf_loop 11 6.9 0.000 0.000 35.841 35.843 prepare_preconditioner 11 7.9 0.000 0.000 31.537 31.572 make_preconditioner 11 8.9 0.000 0.000 31.537 31.572 mp_waitall_1 124680 16.7 25.026 31.550 25.026 31.550 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.373 29.635 rebuild_ks_matrix 127 8.3 0.001 0.001 28.951 29.122 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.950 29.122 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.391 26.551 multiply_cannon_multrec 9940 15.5 10.160 14.636 19.791 24.416 qs_ot_get_derivative 116 11.6 0.001 0.002 20.427 20.608 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 12.932 20.606 cp_fm_cholesky_invert 11 10.9 18.006 18.013 18.006 18.013 qs_ot_get_p 127 10.4 0.085 0.123 17.363 17.564 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.065 17.272 apply_single 127 13.6 0.001 0.001 17.064 17.272 ot_diis_step 116 11.6 0.020 0.020 16.948 16.948 make_m2s 4970 13.5 0.064 0.070 14.872 15.733 make_images 4970 14.5 2.135 2.560 14.566 15.432 sum_up_and_integrate 127 10.3 0.002 0.002 13.682 13.728 integrate_v_rspace 127 11.3 0.004 0.004 13.623 13.671 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.356 13.404 calculate_rho_elec 127 8.7 0.255 0.266 13.356 13.403 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.256 13.273 init_scf_run 11 5.9 0.001 0.011 12.784 12.784 scf_env_initial_rho_setup 11 6.9 0.002 0.011 12.783 12.784 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.102 12.102 multiply_cannon_sync_h2d 9940 15.5 10.575 11.016 10.575 11.016 dbcsr_mm_accdrv_process 20590 16.1 3.201 4.021 9.287 9.984 make_images_data 4970 15.5 0.054 0.065 8.609 9.826 hybrid_alltoall_any 5155 16.4 0.838 3.748 8.400 9.777 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.252 9.264 cp_fm_diag_elpa_base 82 14.4 9.019 9.091 9.249 9.261 calculate_dm_sparse 127 9.5 0.001 0.001 8.067 8.155 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.944 8.069 fft_wrap_pw1pw2 1281 11.7 0.018 0.020 8.019 8.053 grid_integrate_task_list 127 12.3 7.723 8.036 7.723 8.036 cp_fm_cholesky_decompose 22 10.9 7.794 7.917 7.794 7.917 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.236 7.819 wfi_extrapolate 11 7.9 0.001 0.001 7.297 7.297 fft_wrap_pw1pw2_140 519 12.2 0.430 0.437 7.048 7.099 density_rs2pw 127 9.7 0.005 0.005 6.098 6.481 fft3d_ps 1281 13.7 3.092 3.211 6.010 6.035 grid_collocate_task_list 127 9.7 5.517 5.766 5.517 5.766 dbcsr_complete_redistribute 393 12.7 2.124 2.239 5.306 5.697 multiply_cannon_metrocomm4 7455 15.5 0.025 0.027 1.856 5.633 mp_irecv_dv 28618 15.9 1.816 5.536 1.816 5.536 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.340 5.351 calculate_first_density_matrix 1 7.0 0.000 0.000 5.265 5.268 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.245 5.255 mp_alltoall_d11v 2401 14.1 4.640 5.119 4.640 5.119 mp_allgather_i34 2485 14.5 1.185 4.682 1.185 4.682 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.392 4.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.169 4.229 jit_kernel_multiply 11 15.4 3.544 4.002 3.544 4.002 potential_pw2rs 127 12.3 0.026 0.027 3.912 3.919 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.593 3.895 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.503 3.825 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.642 3.718 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.605 3.641 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.401 3.532 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.458 3.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.236000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1458.909091, yerr=46.241372 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 3.029635E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59121. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.063 291.433 291.433 qs_mol_dyn_low 1 2.0 0.003 0.003 290.617 290.635 qs_forces 11 3.9 0.004 0.005 290.517 290.520 qs_energies 11 4.9 0.002 0.002 281.747 281.751 scf_env_do_scf 11 5.9 0.001 0.001 256.958 256.965 velocity_verlet 10 3.0 0.003 0.005 208.802 208.810 scf_env_do_scf_inner_loop 116 6.6 0.032 0.036 131.755 131.757 init_scf_loop 11 6.9 0.000 0.000 124.928 124.932 prepare_preconditioner 11 7.9 0.000 0.000 120.037 120.065 make_preconditioner 11 8.9 0.000 0.000 120.037 120.065 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.226 117.174 qs_scf_new_mos 116 7.6 0.001 0.001 89.532 89.637 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.531 89.637 dbcsr_multiply_generic 2485 12.5 0.234 0.243 84.531 85.306 ot_scf_mini 116 9.6 0.004 0.004 84.679 84.753 cp_fm_upper_to_full 104 14.8 53.139 76.951 53.139 76.951 multiply_cannon 2485 13.5 0.682 0.750 61.261 62.399 multiply_cannon_loop 2485 14.5 1.040 1.059 57.070 58.930 ot_mini 116 10.6 0.001 0.001 44.633 44.726 dbcsr_complete_redistribute 393 12.7 3.979 4.029 30.656 44.002 copy_fm_to_dbcsr 208 11.6 0.001 0.001 27.201 40.562 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.769 37.974 mp_alltoall_i22 712 14.1 22.554 36.114 22.554 36.114 rebuild_ks_matrix 127 8.3 0.001 0.001 33.471 33.599 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.471 33.599 mp_waitall_1 102768 16.8 28.447 33.551 28.447 33.551 cp_fm_cholesky_invert 11 10.9 33.109 33.115 33.109 33.115 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.112 31.218 qs_ot_get_p 127 10.4 0.026 0.026 25.058 25.149 qs_ot_get_derivative 116 11.6 0.002 0.002 24.540 24.616 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 19.193 21.068 qs_ot_p2m_diag 82 11.4 0.869 0.883 20.908 20.935 make_m2s 4970 13.5 0.073 0.076 18.546 20.223 multiply_cannon_multrec 9940 15.5 10.185 11.943 20.069 20.171 ot_diis_step 116 11.6 0.022 0.022 20.049 20.049 make_images 4970 14.5 3.037 3.230 18.068 19.749 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.308 19.611 apply_single 127 13.6 0.001 0.001 19.308 19.610 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.141 19.142 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.027 16.028 cp_fm_diag_elpa_base 82 14.4 11.641 13.248 16.023 16.023 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.582 15.595 calculate_rho_elec 127 8.7 0.478 0.480 15.581 15.594 sum_up_and_integrate 127 10.3 0.002 0.002 15.471 15.563 integrate_v_rspace 127 11.3 0.005 0.005 15.409 15.501 init_scf_run 11 5.9 0.000 0.001 14.491 14.491 scf_env_initial_rho_setup 11 6.9 0.001 0.002 14.491 14.491 multiply_cannon_sync_h2d 9940 15.5 14.200 14.215 14.200 14.215 hybrid_alltoall_any 5155 16.4 1.301 3.042 10.798 13.308 make_images_data 4970 15.5 0.061 0.066 10.546 12.932 dbcsr_mm_accdrv_process 20590 16.0 4.273 6.029 9.634 11.496 fft_wrap_pw1pw2 1281 11.7 0.021 0.021 9.815 9.821 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.602 9.662 calculate_dm_sparse 127 9.5 0.001 0.001 8.796 8.845 wfi_extrapolate 11 7.9 0.001 0.001 8.845 8.845 cp_fm_cholesky_decompose 22 10.9 8.793 8.812 8.793 8.812 fft_wrap_pw1pw2_140 519 12.2 0.817 0.819 8.704 8.717 grid_integrate_task_list 127 12.3 8.495 8.692 8.495 8.692 qs_energies_init_hamiltonians 11 5.9 0.017 0.019 8.158 8.161 fft3d_ps 1281 13.7 3.951 3.966 7.141 7.157 mp_alltoall_d11v 2401 14.1 6.695 6.846 6.695 6.846 density_rs2pw 127 9.7 0.005 0.005 6.714 6.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.457 6.487 grid_collocate_task_list 127 9.7 6.363 6.443 6.363 6.443 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.092 6.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.433000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2707.000000, yerr=161.289689 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 1.265504E+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 0.047 0.087 85.569 85.571 qs_energies 1 2.0 0.000 0.000 84.785 84.793 ls_scf 1 3.0 0.000 0.000 83.562 83.571 dbcsr_multiply_generic 111 6.7 0.014 0.014 72.380 72.547 multiply_cannon 111 7.7 0.017 0.020 55.780 56.863 multiply_cannon_loop 111 8.7 0.228 0.250 52.415 53.754 ls_scf_main 1 4.0 0.000 0.000 52.269 52.269 density_matrix_trs4 2 5.0 0.002 0.003 46.679 46.757 ls_scf_init_scf 1 4.0 0.000 0.000 28.257 28.265 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.119 27.160 mp_waitall_1 11031 10.9 22.249 25.844 22.249 25.844 matrix_sqrt_Newton_Schulz 2 6.5 0.011 0.027 24.908 24.922 multiply_cannon_multrec 2664 9.7 8.160 8.996 15.559 17.385 multiply_cannon_sync_h2d 2664 9.7 13.653 15.778 13.653 15.778 make_m2s 222 7.7 0.008 0.011 13.044 13.509 make_images 222 8.7 0.099 0.108 13.022 13.489 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.625 12.260 make_images_data 222 9.7 0.004 0.005 7.616 8.169 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.712 7.013 7.972 hybrid_alltoall_any 227 10.6 0.216 1.843 6.601 7.878 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.355 7.869 dbcsr_mm_accdrv_process_sort 4760 11.4 6.222 7.106 6.222 7.106 calculate_norms 4752 9.8 5.531 6.277 5.531 6.277 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.036 5.180 mp_sum_l 887 5.1 3.040 4.405 3.040 4.405 make_images_sizes 222 9.7 0.000 0.000 0.668 3.713 mp_alltoall_i44 222 10.7 0.667 3.712 0.667 3.712 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.038 3.471 mp_irecv_dv 6231 10.9 2.022 3.447 2.022 3.447 arnoldi_extremal 4 6.8 0.000 0.000 3.271 3.303 arnoldi_normal_ev 4 7.8 0.023 0.029 3.271 3.303 build_subspace 16 8.4 0.009 0.012 3.147 3.149 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.236 3.135 ls_scf_post 1 4.0 0.000 0.000 3.037 3.044 ls_scf_store_result 1 5.0 0.000 0.000 2.858 2.897 dbcsr_special_finalize 555 9.7 0.005 0.006 2.402 2.748 dbcsr_merge_single_wm 555 10.7 0.460 0.583 2.393 2.739 make_images_pack 222 9.7 2.211 2.631 2.213 2.633 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.323 2.568 dbcsr_sort_data 658 11.4 2.193 2.489 2.193 2.489 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.462 2.071 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.332 2.430 buffer_matrices_ensure_size 222 8.7 1.751 2.108 1.751 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.810 rebuild_ks_matrix 3 7.3 0.000 0.000 1.799 1.800 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.799 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.571000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 2.168136E+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 0.068 0.279 91.025 91.028 qs_energies 1 2.0 0.000 0.000 89.594 89.614 ls_scf 1 3.0 0.000 0.000 88.041 88.045 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.133 74.420 multiply_cannon 111 7.7 0.027 0.040 52.484 57.470 ls_scf_main 1 4.0 0.000 0.000 54.449 54.453 multiply_cannon_loop 111 8.7 0.135 0.145 49.695 53.240 density_matrix_trs4 2 5.0 0.002 0.003 48.756 48.981 mp_waitall_1 9105 10.9 20.504 29.936 20.504 29.936 ls_scf_init_scf 1 4.0 0.000 0.000 29.871 29.873 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.667 28.774 multiply_cannon_multrec 1332 9.7 13.338 16.572 22.632 26.694 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.012 26.285 26.300 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.299 21.658 make_m2s 222 7.7 0.006 0.007 15.070 15.579 make_images 222 8.7 1.370 1.699 15.039 15.549 dbcsr_mm_accdrv_process 4041 10.4 0.340 0.525 8.890 10.442 dbcsr_mm_accdrv_process_sort 4041 11.4 8.407 9.919 8.407 9.919 hybrid_alltoall_any 227 10.6 0.543 2.540 8.066 9.620 make_images_data 222 9.7 0.004 0.004 8.715 9.555 mp_sum_l 887 5.1 5.333 8.388 5.333 8.388 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.194 7.641 mp_irecv_dv 3311 11.0 3.173 7.610 3.173 7.610 calculate_norms 2376 9.8 6.062 6.742 6.062 6.742 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.848 6.473 multiply_cannon_sync_h2d 1332 9.7 4.881 6.187 4.881 6.187 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.928 5.120 arnoldi_extremal 4 6.8 0.000 0.000 4.618 4.639 arnoldi_normal_ev 4 7.8 0.001 0.004 4.618 4.639 build_subspace 16 8.4 0.014 0.020 4.367 4.370 ls_scf_post 1 4.0 0.000 0.000 3.721 3.727 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.546 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.133 3.362 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.225 2.746 3.227 mp_allgather_i34 111 8.7 0.784 3.006 0.784 3.006 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.603 2.678 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.136 2.528 dbcsr_data_new 4174 10.1 2.113 2.405 2.113 2.405 make_images_pack 222 9.7 1.819 2.124 1.822 2.126 dbcsr_sort_data 436 11.2 1.837 2.122 1.837 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.890 1.892 rebuild_ks_matrix 3 7.3 0.000 0.000 1.877 1.879 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.877 1.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.028000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 2.944709E+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 0.024 0.044 93.457 93.458 qs_energies 1 2.0 0.000 0.000 92.142 92.147 ls_scf 1 3.0 0.000 0.000 90.664 90.683 dbcsr_multiply_generic 111 6.7 0.016 0.018 75.394 75.725 ls_scf_main 1 4.0 0.000 0.000 56.340 56.344 multiply_cannon 111 7.7 0.029 0.078 52.029 55.721 multiply_cannon_loop 111 8.7 0.116 0.129 49.270 53.498 density_matrix_trs4 2 5.0 0.002 0.003 50.457 50.626 mp_waitall_1 7281 11.0 23.550 33.603 23.550 33.603 ls_scf_init_scf 1 4.0 0.000 0.000 30.654 30.676 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.443 29.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.040 27.053 multiply_cannon_multrec 888 9.7 12.663 15.353 21.250 24.720 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.916 23.451 make_m2s 222 7.7 0.006 0.007 16.454 17.197 make_images 222 8.7 1.586 1.850 16.416 17.157 make_images_data 222 9.7 0.004 0.004 9.498 10.693 hybrid_alltoall_any 227 10.6 0.639 2.943 9.045 10.346 dbcsr_mm_accdrv_process 3754 10.4 0.316 0.490 8.102 9.343 dbcsr_mm_accdrv_process_sort 3754 11.4 7.668 8.852 7.668 8.852 mp_sum_l 887 5.1 5.249 8.401 5.249 8.401 multiply_cannon_sync_h2d 888 9.7 5.967 7.446 5.967 7.446 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.466 7.315 mp_irecv_dv 2335 11.1 2.449 7.253 2.449 7.253 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.779 7.186 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.054 6.726 arnoldi_extremal 4 6.8 0.000 0.000 5.092 5.108 arnoldi_normal_ev 4 7.8 0.001 0.005 5.092 5.108 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.726 4.952 calculate_norms 1584 9.8 4.415 4.838 4.415 4.838 build_subspace 16 8.4 0.014 0.020 4.789 4.794 mp_allgather_i34 111 8.7 0.878 3.809 0.878 3.809 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.423 3.758 ls_scf_post 1 4.0 0.000 0.000 3.670 3.676 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.599 3.021 3.601 ls_scf_store_result 1 5.0 0.000 0.000 3.424 3.493 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.849 2.970 dbcsr_data_new 4116 9.9 2.116 2.477 2.116 2.477 make_images_sizes 222 9.7 0.000 0.000 1.124 2.356 mp_alltoall_i44 222 10.7 1.124 2.355 1.124 2.355 dbcsr_sort_data 325 11.1 1.875 2.115 1.875 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.917 1.919 rebuild_ks_matrix 3 7.3 0.000 0.000 1.898 1.901 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.898 1.901 make_images_pack 222 9.7 1.622 1.880 1.625 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.458000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2302.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 3.354931E+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 0.081 0.083 97.126 97.126 qs_energies 1 2.0 0.000 0.000 96.299 96.309 ls_scf 1 3.0 0.000 0.000 94.593 94.601 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.305 78.535 ls_scf_main 1 4.0 0.000 0.000 58.761 58.763 multiply_cannon 111 7.7 0.039 0.098 51.649 55.621 density_matrix_trs4 2 5.0 0.002 0.003 52.659 52.800 multiply_cannon_loop 111 8.7 0.151 0.168 46.652 49.425 ls_scf_init_scf 1 4.0 0.000 0.000 32.631 32.634 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.415 31.508 mp_waitall_1 6369 11.0 22.700 30.160 22.700 30.160 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.011 28.883 28.894 multiply_cannon_multrec 1332 9.7 14.185 17.115 22.115 25.574 make_m2s 222 7.7 0.007 0.008 21.100 22.488 make_images 222 8.7 3.141 3.609 21.050 22.440 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.254 16.844 make_images_data 222 9.7 0.004 0.004 11.797 13.413 hybrid_alltoall_any 227 10.6 0.796 3.766 11.209 12.879 dbcsr_mm_accdrv_process 3641 10.4 0.305 0.490 7.558 9.089 dbcsr_mm_accdrv_process_sort 3641 11.4 7.115 8.593 7.115 8.593 mp_sum_l 887 5.1 3.971 6.888 3.971 6.888 multiply_cannon_sync_h2d 1332 9.7 5.400 6.026 5.400 6.026 multiply_cannon_metrocomm4 1110 9.7 0.005 0.006 2.090 5.967 mp_irecv_dv 3229 10.9 2.065 5.893 2.065 5.893 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.037 5.414 arnoldi_extremal 4 6.8 0.000 0.000 5.177 5.194 arnoldi_normal_ev 4 7.8 0.004 0.009 5.177 5.194 build_subspace 16 8.4 0.014 0.020 4.839 4.846 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.480 4.690 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.508 4.671 calculate_norms 2376 9.8 4.214 4.589 4.214 4.589 mp_allgather_i34 111 8.7 2.117 4.388 2.117 4.388 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.576 3.871 dbcsr_matrix_vector_mult_local 304 10.0 3.191 3.691 3.193 3.693 dbcsr_sort_data 658 11.4 3.046 3.464 3.046 3.464 ls_scf_post 1 4.0 0.000 0.000 3.201 3.211 dbcsr_special_finalize 555 9.7 0.006 0.007 2.789 3.192 dbcsr_merge_single_wm 555 10.7 0.532 0.659 2.780 3.184 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.003 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.963 3.005 dbcsr_data_release 10477 10.7 1.591 2.443 1.591 2.443 dbcsr_finalize 304 7.8 0.049 0.061 1.802 2.028 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.126000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 4.800373E+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 0.083 0.088 93.111 93.112 qs_energies 1 2.0 0.000 0.000 92.048 92.054 ls_scf 1 3.0 0.000 0.000 90.050 90.056 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.586 71.827 ls_scf_main 1 4.0 0.000 0.000 57.242 57.243 multiply_cannon 111 7.7 0.091 0.127 52.946 55.937 multiply_cannon_loop 111 8.7 0.088 0.093 50.332 52.199 density_matrix_trs4 2 5.0 0.002 0.003 50.305 50.370 ls_scf_init_scf 1 4.0 0.000 0.000 29.399 29.404 mp_waitall_1 5436 11.0 25.026 29.200 25.026 29.200 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.161 28.206 matrix_sqrt_Newton_Schulz 2 6.5 0.010 0.020 26.065 26.078 multiply_cannon_multrec 444 9.7 13.680 16.213 20.764 21.959 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.857 16.354 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.496 15.370 make_m2s 222 7.7 0.005 0.005 13.710 14.680 make_images 222 8.7 2.038 2.480 13.642 14.612 hybrid_alltoall_any 227 10.6 0.801 3.813 8.290 9.972 make_images_data 222 9.7 0.003 0.004 8.489 9.809 multiply_cannon_sync_h2d 444 9.7 6.716 8.354 6.716 8.354 dbcsr_mm_accdrv_process 3003 10.4 0.340 0.397 6.776 7.960 dbcsr_mm_accdrv_process_sort 3003 11.4 6.414 7.563 6.414 7.563 arnoldi_extremal 4 6.8 0.000 0.000 5.911 5.920 arnoldi_normal_ev 4 7.8 0.056 0.072 5.911 5.920 build_subspace 16 8.4 0.015 0.019 5.440 5.450 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.436 4.684 mp_sum_l 887 5.1 2.919 4.512 2.919 4.512 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.183 4.390 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.185 3.708 4.187 mp_allgather_i34 111 8.7 1.191 3.835 1.191 3.835 calculate_norms 792 9.8 3.607 3.728 3.607 3.728 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.547 3.696 mp_irecv_dv 1241 11.2 1.532 3.649 1.532 3.649 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.471 3.544 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.063 3.532 ls_scf_post 1 4.0 0.000 0.000 3.410 3.415 make_images_sizes 222 9.7 0.000 0.000 0.836 3.329 mp_alltoall_i44 222 10.7 0.836 3.329 0.836 3.329 ls_scf_store_result 1 5.0 0.000 0.000 3.201 3.235 dbcsr_finalize 304 7.8 0.062 0.076 2.207 2.323 dbcsr_data_new 4608 9.7 1.784 2.319 1.784 2.319 dbcsr_merge_all 275 8.9 0.481 0.526 2.064 2.159 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.035 2.036 rebuild_ks_matrix 3 7.3 0.000 0.000 2.002 2.003 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.002 2.003 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.983 1.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.112000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3751.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 8.999743E+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 0.132 0.134 106.764 106.765 qs_energies 1 2.0 0.000 0.000 105.047 105.057 ls_scf 1 3.0 0.000 0.000 102.111 102.121 dbcsr_multiply_generic 111 6.7 0.024 0.026 76.177 76.329 ls_scf_main 1 4.0 0.000 0.000 63.304 63.304 density_matrix_trs4 2 5.0 0.002 0.003 54.522 54.583 multiply_cannon 111 7.7 0.120 0.250 48.657 50.513 multiply_cannon_loop 111 8.7 0.098 0.099 45.662 47.455 ls_scf_init_scf 1 4.0 0.000 0.000 35.044 35.045 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.631 33.655 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.935 30.941 mp_waitall_1 4527 11.1 21.537 25.344 21.537 25.344 make_m2s 222 7.7 0.005 0.005 22.329 23.427 make_images 222 8.7 3.569 3.859 22.222 23.318 multiply_cannon_multrec 444 9.7 17.860 18.424 22.497 23.103 hybrid_alltoall_any 227 10.6 1.650 3.608 12.560 15.267 make_images_data 222 9.7 0.003 0.004 12.731 14.872 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.219 10.579 multiply_cannon_sync_h2d 444 9.7 8.782 8.836 8.782 8.836 arnoldi_extremal 4 6.8 0.000 0.000 7.396 7.401 arnoldi_normal_ev 4 7.8 0.003 0.009 7.396 7.400 build_subspace 16 8.4 0.026 0.037 6.855 6.867 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.501 5.658 dbcsr_matrix_vector_mult_local 304 10.0 5.058 5.375 5.061 5.378 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.113 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.908 5.005 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.350 4.462 4.583 dbcsr_mm_accdrv_process_sort 1814 11.4 4.126 4.249 4.126 4.249 ls_scf_post 1 4.0 0.000 0.000 3.763 3.774 mp_sum_l 887 5.1 2.813 3.661 2.813 3.661 make_images_sizes 222 9.7 0.000 0.000 1.469 3.564 mp_alltoall_i44 222 10.7 1.468 3.564 1.468 3.564 mp_allgather_i34 111 8.7 1.116 3.542 1.116 3.542 ls_scf_store_result 1 5.0 0.000 0.000 3.485 3.521 calculate_norms 792 9.8 3.239 3.307 3.239 3.307 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.152 dbcsr_merge_all 275 8.9 0.883 0.918 2.860 2.932 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.905 2.905 dbcsr_complete_redistribute 5 7.6 1.425 1.468 2.743 2.853 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.209 2.669 matrix_ls_to_qs 2 6.0 0.000 0.000 2.396 2.525 dbcsr_sort_data 325 11.1 2.438 2.483 2.438 2.483 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.473 2.479 dbcsr_data_new 6591 9.6 1.876 2.372 1.876 2.372 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.280 2.282 dbcsr_new_transposed 4 7.5 0.244 0.255 2.259 2.270 rebuild_ks_matrix 3 7.3 0.000 0.000 2.215 2.216 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.215 2.216 dbcsr_frobenius_norm 74 6.6 2.054 2.124 2.183 2.214 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.200 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.199 dbcsr_data_release 12724 10.6 1.982 2.197 1.982 2.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.765000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8_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 590.053376E+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 4222701. 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 0.071 0.206 232.907 232.909 qs_mol_dyn_low 1 2.0 0.013 0.123 231.301 231.324 qs_forces 5 3.8 0.004 0.012 231.216 231.227 qs_energies 5 4.8 0.007 0.011 228.123 228.154 scf_env_do_scf 5 5.8 0.000 0.002 212.468 212.470 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 184.733 184.734 qs_scf_new_mos 105 7.6 0.000 0.001 143.554 143.715 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.553 143.714 dbcsr_multiply_generic 1445 12.2 0.129 0.137 134.551 134.908 ot_scf_mini 105 9.6 0.003 0.004 133.712 133.872 multiply_cannon 1445 13.2 0.275 0.285 115.283 117.160 multiply_cannon_loop 1445 14.2 2.851 2.997 113.481 114.727 velocity_verlet 4 3.0 0.001 0.007 106.383 106.384 ot_mini 105 10.6 0.001 0.002 60.303 60.436 multiply_cannon_multrec 69360 15.2 29.775 34.558 41.211 45.957 qs_ot_get_p 112 10.4 0.001 0.001 43.401 43.723 mp_waitall_1 488190 16.1 35.164 42.026 35.164 42.026 qs_ot_get_derivative 55 11.6 0.001 0.001 38.575 38.727 multiply_cannon_sync_h2d 69360 15.2 29.192 33.521 29.192 33.521 multiply_cannon_metrocomm3 69360 15.2 0.203 0.215 25.700 32.445 qs_ot_p2m_diag 40 11.0 0.020 0.031 32.303 32.376 rebuild_ks_matrix 110 8.4 0.000 0.000 29.866 30.018 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.017 29.865 30.017 cp_dbcsr_syevd 40 12.0 0.002 0.004 28.925 28.927 init_scf_loop 7 6.6 0.000 0.005 27.691 27.692 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.377 27.509 cp_fm_syevd 40 13.0 0.000 0.002 23.390 23.522 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.098 23.343 apply_single 62 13.6 0.000 0.000 23.098 23.343 prepare_preconditioner 7 7.6 0.000 0.000 22.604 22.636 make_preconditioner 7 8.6 0.000 0.007 22.604 22.636 ot_new_cg_direction 55 11.6 0.001 0.001 21.060 21.061 cp_fm_redistribute_end 40 14.0 9.276 18.499 9.282 18.501 cp_fm_syevd_base 40 14.0 9.209 18.432 9.209 18.432 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.354 17.743 calculate_rho_elec 110 8.6 0.029 0.035 17.353 17.742 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.914 14.977 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.797 14.969 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.779 13.890 mp_sum_l 4764 12.2 12.220 13.030 12.220 13.030 calculate_dm_sparse 110 9.5 0.000 0.001 12.806 12.985 init_scf_run 5 5.8 0.000 0.003 12.901 12.902 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.901 12.902 fft_wrap_pw1pw2 1425 12.5 0.019 0.027 12.072 12.385 dbcsr_mm_accdrv_process 154766 15.8 6.259 6.456 11.302 12.286 density_rs2pw 110 9.6 0.004 0.009 11.436 11.886 fft_wrap_pw1pw2_240 915 14.0 0.196 0.225 10.657 10.841 qs_vxc_create 110 10.4 0.002 0.003 10.752 10.776 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.396 10.466 cp_fm_cholesky_invert 7 10.6 10.278 10.288 10.278 10.288 check_diag 80 13.5 8.683 8.935 9.873 10.009 fft3d_pb 915 15.0 2.230 2.430 9.652 9.820 multiply_cannon_metrocomm1 69360 15.2 0.095 0.101 4.818 8.835 sum_up_and_integrate 60 10.3 0.001 0.003 8.745 8.755 integrate_v_rspace 60 11.3 0.002 0.003 8.727 8.738 transfer_rs2pw 445 10.6 0.008 0.009 8.091 8.560 calculate_first_density_matrix 1 7.0 0.000 0.012 8.013 8.027 xc_rho_set_and_dset_create 110 12.4 0.071 0.077 7.507 7.749 acc_transpose_blocks 69360 15.2 0.348 0.368 6.634 7.495 make_full_single_inverse 7 9.6 0.001 0.005 7.387 7.419 xc_vxc_pw_create 60 11.3 0.036 0.040 7.288 7.317 make_m2s 2890 13.2 0.078 0.088 6.648 7.259 make_images 2890 14.2 0.239 0.259 6.540 7.150 mp_alltoall_z22v 2340 16.7 6.888 7.091 6.888 7.091 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.987 7.003 xc_pw_derive 510 13.4 0.005 0.006 6.563 6.637 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.514 6.564 mp_waitany 7680 13.5 4.497 5.013 4.497 5.013 potential_pw2rs 60 12.3 0.002 0.003 4.852 4.891 multiply_cannon_metrocomm4 67915 15.2 0.180 0.197 2.003 4.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=232.909000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5417965e490c4f1d24a8566dbeb5fce0f3c2e8f8 Summary: empty Status: OK