=== 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: 14576fcd0de922a1947a6e9f68f87a45f7ee4c54 ################# 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, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/01 job id: 49961626 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/02 job id: 49961627 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/03 job id: 49961628 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/04 job id: 49961629 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/05 job id: 49961630 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/06 job id: 49961631 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/07 job id: 49961632 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/08 job id: 49961633 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/09 job id: 49961634 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/10 job id: 49961635 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/11 job id: 49961636 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/12 job id: 49961637 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/13 job id: 49961638 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/14 job id: 49961639 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/15 job id: 49961640 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/16 job id: 49961642 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/17 job id: 49961643 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/18 job id: 49961644 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/19 job id: 49961645 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/20 job id: 49961646 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/21 job id: 49961647 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/22 job id: 49961649 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/23 job id: 49961651 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/24 job id: 49961652 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/25 job id: 49961653 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/26 job id: 49961654 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/27 job id: 49961656 --- 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/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.034 134.301 134.302 farming_run 1 2.0 133.789 133.789 134.270 134.273 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.491218E+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 0.007 0.027 115.160 115.160 qs_energies 1 2.0 0.000 0.000 114.957 114.961 mp2_main 1 3.0 0.000 0.000 112.742 112.745 mp2_gpw_main 1 4.0 0.027 0.034 111.585 111.588 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.296 92.801 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.627 55.134 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.031 46.219 get_2c_integrals 1 6.0 0.008 0.009 36.901 37.497 integrate_v_rspace 273 8.0 0.438 0.454 24.604 29.603 pw_transfer 6555 10.6 0.377 0.398 26.887 27.344 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 25.474 26.031 grid_integrate_task_list 273 9.0 20.512 25.976 20.512 25.976 fft_wrap_pw1pw2_100 2178 12.4 1.149 1.306 23.030 23.540 rpa_ri_compute_en 1 5.0 0.019 0.023 19.182 19.357 compute_2c_integrals 1 7.0 0.002 0.002 19.324 19.326 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.725 19.041 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.427 18.724 19.040 cp_fm_cholesky_decompose 12 8.2 17.611 18.215 17.611 18.215 cholesky_decomp 1 7.0 0.000 0.000 16.428 17.043 fft3d_s 5443 13.4 16.160 16.367 16.182 16.389 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.312 10.767 15.312 calculate_wavefunction 272 8.0 5.430 5.562 12.333 13.029 rpa_num_int 1 6.0 0.000 0.006 10.753 10.763 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.705 10.740 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.588 contract_S_to_Q 8 9.0 0.000 0.000 8.909 9.009 calc_potential_gpw 544 9.5 0.005 0.006 8.255 8.694 parallel_gemm_fm 14 9.1 0.000 0.000 8.492 8.617 parallel_gemm_fm_cosma 14 10.1 8.492 8.617 8.492 8.617 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.148 8.385 potential_pw2rs 545 10.0 0.107 0.109 7.542 8.127 create_integ_mat 1 6.0 0.014 0.029 7.830 7.840 collocate_single_gaussian 272 10.0 0.039 0.041 7.401 7.574 array2fm 1 7.0 0.000 0.000 6.752 7.256 pw_scatter_s 2720 13.7 4.319 4.399 4.319 4.399 pw_gather_s 2722 13.2 3.433 3.732 3.433 3.732 array2fm_buffer_send 1 8.0 3.001 3.159 3.001 3.159 pw_poisson_solve 545 10.5 1.129 1.183 2.247 2.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.588241, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2808.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.042 423.996 423.998 farming_run 1 2.0 423.256 423.261 423.931 423.933 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.236742E+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 23204152. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 218.118 218.119 qs_energies 1 2.0 0.000 0.000 217.873 217.881 scf_env_do_scf 1 3.0 0.000 0.000 115.336 115.337 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.227 114.235 rebuild_ks_matrix 4 6.0 0.000 0.000 114.225 114.233 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 114.225 114.233 hfx_ks_matrix 4 8.0 0.001 0.001 113.826 113.831 integrate_four_center 4 9.0 0.154 0.465 113.826 113.830 integrate_four_center_main 4 10.0 0.133 0.541 101.704 105.624 integrate_four_center_bin 268 11.0 101.572 105.581 101.572 105.581 mp2_main 1 3.0 0.000 0.000 102.228 102.236 mp2_gpw_main 1 4.0 0.047 0.072 101.218 101.228 init_scf_loop 1 4.0 0.000 0.000 97.381 97.382 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 74.074 75.075 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.775 54.774 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.157 41.348 46.474 integrate_v_rspace 95 8.0 0.395 0.567 27.871 32.826 pw_transfer 2240 10.6 0.145 0.169 29.497 29.967 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.498 28.978 mp2_ri_gpw_compute_en 1 5.0 0.056 0.064 26.993 28.531 grid_integrate_task_list 95 9.0 23.244 28.404 23.244 28.404 ao_to_mo_and_store_B_mult_1 91 7.0 10.753 27.818 10.753 27.818 fft_wrap_pw1pw2_100 730 12.4 1.276 1.452 26.239 26.753 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.893 25.331 25.341 get_2c_integrals 1 6.0 0.000 0.000 20.138 20.233 compute_2c_integrals 1 7.0 0.002 0.003 19.106 19.112 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.633 18.945 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.890 18.632 18.945 fft3d_s 1823 13.4 18.480 18.886 18.494 18.899 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.953 17.953 calculate_wavefunction 91 8.0 2.015 2.050 9.616 9.839 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.591 8.727 9.139 potential_pw2rs 186 10.0 0.033 0.035 8.429 9.086 mp2_ri_gpw_compute_en_comm 22 7.0 0.493 0.519 8.371 8.887 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.158 8.611 local_gemm 172 8.0 8.170 8.574 8.170 8.574 mp_sync 37 10.5 4.145 8.235 4.145 8.235 calc_potential_gpw 182 9.5 0.002 0.002 7.945 8.140 collocate_single_gaussian 91 10.0 0.017 0.020 7.747 8.022 mp_sendrecv_dm3 2068 8.0 6.425 6.946 6.425 6.946 integrate_four_center_load 4 10.0 0.000 0.000 6.760 6.764 hfx_load_balance 1 11.0 0.000 0.000 6.760 6.764 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.420 6.351 6.420 pw_gather_s 912 13.2 4.462 5.043 4.462 5.043 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.208584, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1497.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 452.214784E+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 1129132. 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 0.010 0.032 54.453 54.458 qs_mol_dyn_low 1 2.0 0.005 0.006 54.234 54.243 qs_forces 11 3.9 0.005 0.010 54.110 54.112 qs_energies 11 4.9 0.002 0.002 52.583 52.604 scf_env_do_scf 11 5.9 0.000 0.001 46.327 46.327 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.160 44.160 qs_scf_new_mos 108 7.5 0.000 0.001 33.965 34.282 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.964 34.281 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.787 34.163 ot_scf_mini 108 9.5 0.002 0.002 32.318 32.526 multiply_cannon 2286 13.5 0.189 0.197 26.361 27.782 multiply_cannon_loop 2286 14.5 1.821 1.919 25.640 27.072 velocity_verlet 10 3.0 0.002 0.002 26.799 26.801 ot_mini 108 10.5 0.001 0.001 19.249 19.501 qs_ot_get_derivative 108 11.5 0.001 0.001 16.209 16.400 mp_waitall_1 245248 16.5 8.417 14.369 8.417 14.369 multiply_cannon_metrocomm3 54864 15.5 0.072 0.079 5.976 12.590 multiply_cannon_multrec 54864 15.5 3.655 5.681 7.755 11.133 qs_ot_get_p 119 10.4 0.001 0.001 8.400 8.703 rebuild_ks_matrix 119 8.3 0.000 0.000 7.886 8.045 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.886 8.045 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.927 7.060 mp_sum_l 7287 12.8 5.341 7.034 5.341 7.034 multiply_cannon_sync_h2d 54864 15.5 5.191 6.280 5.191 6.280 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.865 6.275 dbcsr_mm_accdrv_process 76910 16.1 1.835 2.869 4.013 5.692 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.367 5.399 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.040 5.139 init_scf_run 11 5.9 0.000 0.001 4.959 4.959 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.959 4.959 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.512 4.512 sum_up_and_integrate 119 10.3 0.001 0.001 4.487 4.494 integrate_v_rspace 119 11.3 0.002 0.002 4.476 4.484 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.235 4.376 calculate_rho_elec 119 8.7 0.011 0.017 4.234 4.376 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.315 4.315 cp_fm_redistribute_end 50 14.0 2.203 4.288 2.210 4.291 cp_fm_diag_elpa_base 50 14.0 2.075 4.173 2.079 4.183 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.974 3.174 apply_single 119 13.6 0.000 0.000 2.973 3.173 calculate_dm_sparse 119 9.5 0.000 0.000 2.961 3.128 acc_transpose_blocks 54864 15.5 0.229 0.257 2.274 2.814 ot_diis_step 108 11.5 0.006 0.006 2.782 2.782 jit_kernel_multiply 13 15.8 2.115 2.754 2.115 2.754 calculate_first_density_matrix 1 7.0 0.000 0.001 2.728 2.735 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.711 2.658 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.456 2.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.365 2.367 density_rs2pw 119 9.7 0.004 0.004 2.171 2.316 init_scf_loop 11 6.9 0.000 0.001 2.149 2.150 wfi_extrapolate 11 7.9 0.001 0.001 2.146 2.146 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.070 2.136 grid_integrate_task_list 119 12.3 2.021 2.111 2.021 2.111 mp_sum_d 4135 12.0 1.327 1.926 1.327 1.926 potential_pw2rs 119 12.3 0.004 0.004 1.856 1.866 make_m2s 4572 13.5 0.053 0.055 1.776 1.822 make_images 4572 14.5 0.133 0.139 1.694 1.739 pw_transfer 1439 11.6 0.052 0.056 1.654 1.724 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.578 1.651 mp_alltoall_d11v 2130 13.8 1.412 1.572 1.412 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.492 1.529 transfer_rs2pw 487 10.6 0.005 0.006 1.384 1.493 acc_transpose_blocks_sync 164592 16.5 1.205 1.435 1.205 1.435 grid_collocate_task_list 119 9.7 1.355 1.417 1.355 1.417 mp_waitany 12084 13.8 1.257 1.413 1.257 1.413 transfer_pw2rs 487 13.2 0.006 0.006 1.369 1.379 fft3d_ps 1201 14.6 0.368 0.472 1.291 1.360 fft_wrap_pw1pw2_140 487 13.2 0.142 0.156 1.218 1.292 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.743 1.128 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.458000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.324544E+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 1413743. 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 0.009 0.030 39.858 39.859 qs_mol_dyn_low 1 2.0 0.021 0.027 39.628 39.636 qs_forces 11 3.9 0.007 0.010 39.229 39.247 qs_energies 11 4.9 0.002 0.003 37.556 37.583 scf_env_do_scf 11 5.9 0.000 0.001 32.069 32.069 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.607 29.608 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.462 22.829 qs_scf_new_mos 108 7.5 0.001 0.001 20.836 21.063 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.836 21.062 ot_scf_mini 108 9.5 0.002 0.003 19.907 20.070 velocity_verlet 10 3.0 0.001 0.002 18.854 18.862 multiply_cannon 2286 13.5 0.211 0.223 17.137 18.538 multiply_cannon_loop 2286 14.5 1.190 1.250 15.939 17.420 ot_mini 108 10.5 0.001 0.001 12.279 12.508 mp_waitall_1 200699 16.5 5.624 10.827 5.624 10.827 qs_ot_get_derivative 108 11.5 0.001 0.001 9.919 10.081 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.059 9.461 multiply_cannon_multrec 27432 15.5 1.838 4.262 6.376 9.348 rebuild_ks_matrix 119 8.3 0.000 0.000 6.831 6.975 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.830 6.975 dbcsr_mm_accdrv_process 47894 16.0 3.629 5.822 4.458 6.680 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.019 6.149 qs_ot_get_p 119 10.4 0.001 0.001 4.691 4.903 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.732 4.631 init_scf_run 11 5.9 0.000 0.001 4.270 4.271 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.270 4.270 mp_sum_l 7287 12.8 2.104 4.092 2.104 4.092 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.054 4.039 apply_single 119 13.6 0.000 0.000 3.054 4.039 sum_up_and_integrate 119 10.3 0.001 0.002 3.752 3.759 integrate_v_rspace 119 11.3 0.002 0.003 3.737 3.744 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.705 3.739 calculate_rho_elec 119 8.7 0.021 0.024 3.705 3.739 make_m2s 4572 13.5 0.051 0.053 2.772 3.102 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.015 3.034 make_images 4572 14.5 0.211 0.262 2.684 3.017 calculate_first_density_matrix 1 7.0 0.000 0.001 2.729 2.731 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.567 2.568 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.358 2.442 init_scf_loop 11 6.9 0.000 0.001 2.438 2.438 ot_diis_step 108 11.5 0.011 0.011 2.313 2.314 calculate_dm_sparse 119 9.5 0.000 0.001 2.188 2.264 multiply_cannon_sync_h2d 27432 15.5 1.678 2.251 1.678 2.251 jit_kernel_multiply 11 16.4 0.771 2.250 0.771 2.250 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.244 2.244 cp_fm_redistribute_end 50 14.0 1.138 2.219 1.141 2.222 cp_fm_diag_elpa_base 50 14.0 1.049 2.131 1.077 2.171 density_rs2pw 119 9.7 0.004 0.004 1.993 2.081 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.054 2.057 acc_transpose_blocks 27432 15.5 0.113 0.119 1.659 2.028 grid_integrate_task_list 119 12.3 1.841 1.939 1.841 1.939 pw_transfer 1439 11.6 0.065 0.070 1.882 1.915 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.852 1.893 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.790 1.826 make_images_data 4572 15.5 0.047 0.053 1.300 1.685 prepare_preconditioner 11 7.9 0.000 0.000 1.545 1.569 make_preconditioner 11 8.9 0.000 0.000 1.545 1.569 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.148 1.534 potential_pw2rs 119 12.3 0.006 0.007 1.498 1.505 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.443 1.502 wfi_extrapolate 11 7.9 0.001 0.001 1.478 1.478 fft3d_ps 1201 14.6 0.519 0.574 1.426 1.453 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.415 1.449 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.371 1.381 grid_collocate_task_list 119 9.7 1.289 1.345 1.289 1.345 mp_alltoall_d11v 2130 13.8 1.238 1.337 1.238 1.337 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.218 1.263 transfer_rs2pw 487 10.6 0.005 0.005 1.155 1.259 mp_allgather_i34 2286 14.5 0.606 1.030 0.606 1.030 mp_sum_d 4135 12.0 0.542 1.017 0.542 1.017 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.944 0.967 transfer_pw2rs 487 13.2 0.004 0.006 0.950 0.958 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.700 0.948 acc_transpose_blocks_sync 82296 16.5 0.820 0.945 0.820 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.870 0.884 make_images_sizes 4572 15.5 0.005 0.005 0.590 0.869 mp_alltoall_i44 4572 16.5 0.585 0.864 0.585 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.859000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.818182, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 523.264000E+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 0.008 0.028 33.062 33.064 qs_mol_dyn_low 1 2.0 0.003 0.003 32.857 32.865 qs_forces 11 3.9 0.002 0.003 32.796 32.796 qs_energies 11 4.9 0.002 0.002 31.235 31.238 scf_env_do_scf 11 5.9 0.000 0.001 26.537 26.537 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.994 23.994 dbcsr_multiply_generic 2286 12.5 0.096 0.097 17.519 17.577 qs_scf_new_mos 108 7.5 0.001 0.001 16.106 16.121 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.105 16.120 velocity_verlet 10 3.0 0.001 0.002 15.592 15.593 ot_scf_mini 108 9.5 0.002 0.002 15.331 15.340 multiply_cannon 2286 13.5 0.194 0.198 13.962 14.616 multiply_cannon_loop 2286 14.5 0.856 0.916 13.154 13.822 ot_mini 108 10.5 0.001 0.001 9.523 9.533 qs_ot_get_derivative 108 11.5 0.001 0.001 8.013 8.023 multiply_cannon_multrec 18288 15.5 1.875 2.828 7.012 7.259 rebuild_ks_matrix 119 8.3 0.000 0.000 6.091 6.109 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.091 6.109 dbcsr_mm_accdrv_process 38222 16.0 4.941 5.793 5.044 5.860 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.363 5.380 mp_waitall_1 158411 16.6 2.972 4.114 2.972 4.114 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.145 3.791 sum_up_and_integrate 119 10.3 0.001 0.001 3.650 3.654 integrate_v_rspace 119 11.3 0.002 0.003 3.637 3.643 qs_ot_get_p 119 10.4 0.001 0.001 3.512 3.531 init_scf_run 11 5.9 0.000 0.001 3.527 3.527 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.527 3.527 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.395 3.399 calculate_rho_elec 119 8.7 0.031 0.032 3.394 3.398 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.589 2.598 init_scf_loop 11 6.9 0.000 0.001 2.524 2.526 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.070 2.411 apply_single 119 13.6 0.000 0.000 2.070 2.411 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.315 2.323 calculate_first_density_matrix 1 7.0 0.000 0.000 2.261 2.262 make_m2s 4572 13.5 0.043 0.045 2.026 2.178 make_images 4572 14.5 0.194 0.211 1.942 2.092 density_rs2pw 119 9.7 0.004 0.004 1.939 2.064 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.998 1.999 grid_integrate_task_list 119 12.3 1.810 1.936 1.810 1.936 pw_transfer 1439 11.6 0.066 0.069 1.920 1.928 calculate_dm_sparse 119 9.5 0.000 0.001 1.877 1.887 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.827 1.837 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.760 1.766 acc_transpose_blocks 18288 15.5 0.080 0.083 1.637 1.762 prepare_preconditioner 11 7.9 0.000 0.000 1.759 1.762 make_preconditioner 11 8.9 0.000 0.000 1.759 1.762 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.728 1.749 cp_fm_diag_elpa_base 50 14.0 1.705 1.731 1.727 1.747 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.721 1.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.617 1.700 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.548 1.554 ot_diis_step 108 11.5 0.011 0.011 1.494 1.494 mp_sum_l 7287 12.8 1.141 1.465 1.141 1.465 fft_wrap_pw1pw2_140 487 13.2 0.212 0.220 1.436 1.447 potential_pw2rs 119 12.3 0.007 0.009 1.420 1.423 fft3d_ps 1201 14.6 0.528 0.544 1.395 1.403 grid_collocate_task_list 119 9.7 1.231 1.304 1.231 1.304 wfi_extrapolate 11 7.9 0.001 0.001 1.211 1.211 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.204 1.208 transfer_rs2pw 487 10.6 0.005 0.005 1.038 1.166 multiply_cannon_sync_h2d 18288 15.5 1.051 1.161 1.051 1.161 make_images_data 4572 15.5 0.047 0.052 0.871 1.063 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.976 0.994 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.751 0.983 jit_kernel_multiply 5 15.9 0.044 0.970 0.044 0.970 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.933 0.933 mp_alltoall_d11v 2130 13.8 0.743 0.870 0.743 0.870 acc_transpose_blocks_sync 54864 16.5 0.737 0.870 0.737 0.870 transfer_pw2rs 487 13.2 0.004 0.004 0.867 0.870 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.800 0.824 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.798 0.800 mp_alltoall_z22v 1201 16.6 0.707 0.774 0.707 0.774 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.486 0.758 cp_fm_cholesky_invert 11 10.9 0.750 0.755 0.750 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.670 0.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.064000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.545455, yerr=2.387986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.858816E+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 0.009 0.029 37.914 37.915 qs_mol_dyn_low 1 2.0 0.003 0.004 37.738 37.746 qs_forces 11 3.9 0.003 0.004 37.633 37.633 qs_energies 11 4.9 0.002 0.003 35.879 35.885 scf_env_do_scf 11 5.9 0.000 0.001 30.770 30.771 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.176 27.176 dbcsr_multiply_generic 2286 12.5 0.099 0.102 20.395 20.491 velocity_verlet 10 3.0 0.002 0.002 19.196 19.200 qs_scf_new_mos 108 7.5 0.001 0.001 18.742 18.794 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.741 18.793 ot_scf_mini 108 9.5 0.002 0.003 17.687 17.738 multiply_cannon 2286 13.5 0.219 0.226 16.413 16.868 multiply_cannon_loop 2286 14.5 1.533 1.603 15.389 15.793 ot_mini 108 10.5 0.001 0.001 10.809 10.873 multiply_cannon_multrec 27432 15.5 2.532 3.207 9.120 9.411 qs_ot_get_derivative 108 11.5 0.001 0.001 8.898 8.947 dbcsr_mm_accdrv_process 47916 15.9 5.849 7.548 6.485 7.951 rebuild_ks_matrix 119 8.3 0.000 0.000 6.522 6.567 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.521 6.567 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.763 5.803 qs_ot_get_p 119 10.4 0.001 0.001 3.970 4.047 init_scf_run 11 5.9 0.000 0.001 3.730 3.731 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.730 3.730 sum_up_and_integrate 119 10.3 0.001 0.001 3.671 3.678 integrate_v_rspace 119 11.3 0.002 0.003 3.659 3.667 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.591 3.624 calculate_rho_elec 119 8.7 0.040 0.046 3.591 3.624 init_scf_loop 11 6.9 0.000 0.001 3.571 3.572 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.177 3.566 mp_waitall_1 137007 16.6 2.133 2.834 2.133 2.834 prepare_preconditioner 11 7.9 0.000 0.000 2.725 2.733 make_preconditioner 11 8.9 0.000 0.000 2.725 2.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.321 2.650 make_m2s 4572 13.5 0.054 0.056 2.500 2.639 make_images 4572 14.5 0.278 0.340 2.393 2.529 acc_transpose_blocks 27432 15.5 0.121 0.127 2.405 2.505 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.426 2.453 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.438 2.448 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.139 2.262 apply_single 119 13.6 0.000 0.000 2.139 2.262 calculate_first_density_matrix 1 7.0 0.000 0.001 2.246 2.249 calculate_dm_sparse 119 9.5 0.000 0.000 2.170 2.216 density_rs2pw 119 9.7 0.004 0.004 2.019 2.100 pw_transfer 1439 11.6 0.065 0.069 2.056 2.092 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.009 2.010 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.963 2.002 grid_integrate_task_list 119 12.3 1.823 1.895 1.823 1.895 ot_diis_step 108 11.5 0.012 0.012 1.867 1.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.833 1.835 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.709 1.722 cp_fm_diag_elpa_base 50 14.0 1.674 1.690 1.707 1.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.641 1.653 fft_wrap_pw1pw2_140 487 13.2 0.245 0.260 1.566 1.608 acc_transpose_blocks_sync 82296 16.5 1.434 1.529 1.434 1.529 fft3d_ps 1201 14.6 0.556 0.607 1.483 1.507 potential_pw2rs 119 12.3 0.009 0.009 1.457 1.464 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.832 1.455 jit_kernel_multiply 8 16.2 0.571 1.447 0.571 1.447 wfi_extrapolate 11 7.9 0.001 0.001 1.429 1.429 mp_sum_l 7287 12.8 1.056 1.389 1.056 1.389 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.327 1.342 grid_collocate_task_list 119 9.7 1.249 1.319 1.249 1.319 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.260 cp_fm_upper_to_full 72 14.2 0.824 1.172 0.824 1.172 transfer_rs2pw 487 10.6 0.004 0.005 1.049 1.133 dbcsr_complete_redistribute 329 12.2 0.118 0.149 0.835 1.116 make_images_data 4572 15.5 0.047 0.052 0.962 1.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.099 1.100 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.828 1.018 mp_alltoall_d11v 2130 13.8 0.848 0.929 0.848 0.929 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.609 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.874 transfer_pw2rs 487 13.2 0.004 0.004 0.865 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.853 cp_fm_cholesky_invert 11 10.9 0.836 0.839 0.836 0.839 mp_alltoall_z22v 1201 16.6 0.808 0.834 0.808 0.834 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.821 0.834 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.915000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.272727, yerr=3.518546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 615.800832E+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 0.009 0.027 29.674 29.675 qs_mol_dyn_low 1 2.0 0.003 0.004 29.494 29.502 qs_forces 11 3.9 0.002 0.003 29.418 29.420 qs_energies 11 4.9 0.004 0.004 27.707 27.709 scf_env_do_scf 11 5.9 0.000 0.001 22.963 22.963 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 20.310 20.311 velocity_verlet 10 3.0 0.002 0.002 15.292 15.295 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.475 13.551 qs_scf_new_mos 108 7.5 0.001 0.001 12.470 12.501 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.470 12.500 ot_scf_mini 108 9.5 0.002 0.002 11.726 11.758 multiply_cannon 2286 13.5 0.224 0.230 10.622 11.129 multiply_cannon_loop 2286 14.5 0.646 0.672 9.662 9.860 ot_mini 108 10.5 0.001 0.001 6.694 6.733 multiply_cannon_multrec 9144 15.5 1.722 1.907 6.005 6.231 rebuild_ks_matrix 119 8.3 0.000 0.000 5.874 5.897 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.874 5.897 qs_ot_get_derivative 108 11.5 0.001 0.001 5.358 5.389 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.235 5.257 dbcsr_mm_accdrv_process 12550 15.8 3.371 3.787 4.173 4.246 sum_up_and_integrate 119 10.3 0.001 0.001 3.514 3.519 integrate_v_rspace 119 11.3 0.003 0.003 3.503 3.508 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.461 3.470 calculate_rho_elec 119 8.7 0.060 0.061 3.461 3.469 init_scf_run 11 5.9 0.000 0.001 3.281 3.281 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.281 3.281 qs_ot_get_p 119 10.4 0.001 0.001 3.090 3.137 init_scf_loop 11 6.9 0.001 0.001 2.631 2.632 make_m2s 4572 13.5 0.033 0.035 1.943 2.121 calculate_first_density_matrix 1 7.0 0.000 0.000 2.099 2.101 pw_transfer 1439 11.6 0.066 0.068 2.083 2.095 mp_waitall_1 115863 16.7 1.595 2.089 1.595 2.089 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.035 2.038 make_images 4572 14.5 0.272 0.304 1.855 2.031 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.989 2.002 grid_integrate_task_list 119 12.3 1.869 1.927 1.869 1.927 density_rs2pw 119 9.7 0.003 0.004 1.830 1.911 prepare_preconditioner 11 7.9 0.000 0.000 1.882 1.886 make_preconditioner 11 8.9 0.000 0.000 1.882 1.886 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.845 1.868 calculate_dm_sparse 119 9.5 0.000 0.000 1.811 1.835 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.766 1.794 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.776 1.776 fft_wrap_pw1pw2_140 487 13.2 0.322 0.331 1.617 1.632 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.517 1.518 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.485 1.499 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.463 1.471 cp_fm_diag_elpa_base 50 14.0 1.435 1.451 1.461 1.469 acc_transpose_blocks 9144 15.5 0.042 0.043 1.427 1.462 fft3d_ps 1201 14.6 0.561 0.575 1.403 1.417 grid_collocate_task_list 119 9.7 1.299 1.373 1.299 1.373 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.352 1.360 jit_kernel_multiply 8 15.5 0.762 1.338 0.762 1.338 ot_diis_step 108 11.5 0.013 0.013 1.319 1.320 potential_pw2rs 119 12.3 0.010 0.011 1.307 1.311 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.258 1.289 apply_single 119 13.6 0.000 0.000 1.258 1.289 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.230 1.231 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.170 1.175 wfi_extrapolate 11 7.9 0.001 0.001 1.123 1.123 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.840 1.079 make_images_data 4572 15.5 0.042 0.046 0.845 1.061 cp_fm_cholesky_invert 11 10.9 0.994 0.997 0.994 0.997 mp_alltoall_d11v 2130 13.8 0.882 0.953 0.882 0.953 transfer_rs2pw 487 10.6 0.004 0.004 0.859 0.940 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.927 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.876 0.884 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.764 0.767 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.355 0.761 acc_transpose_blocks_sync 27432 16.5 0.727 0.759 0.727 0.759 mp_alltoall_z22v 1201 16.6 0.714 0.750 0.714 0.750 mp_allgather_i34 2286 14.5 0.275 0.727 0.275 0.727 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.676 0.715 transfer_pw2rs 487 13.2 0.003 0.004 0.700 0.703 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.643 0.648 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.675000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.090909, yerr=3.476010 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 767.389696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 44.679 44.680 qs_mol_dyn_low 1 2.0 0.003 0.003 44.477 44.485 qs_forces 11 3.9 0.002 0.003 43.970 43.970 qs_energies 11 4.9 0.002 0.002 41.945 41.950 scf_env_do_scf 11 5.9 0.001 0.001 36.114 36.115 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.007 28.010 velocity_verlet 10 3.0 0.002 0.006 25.433 25.494 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.716 19.976 qs_scf_new_mos 108 7.5 0.001 0.001 18.196 18.312 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.195 18.312 ot_scf_mini 108 9.5 0.002 0.002 16.989 17.108 multiply_cannon 2286 13.5 0.299 0.309 15.316 16.301 multiply_cannon_loop 2286 14.5 0.868 0.883 14.026 14.967 ot_mini 108 10.5 0.001 0.001 10.276 10.417 multiply_cannon_multrec 9144 15.5 3.354 4.663 8.790 8.956 qs_ot_get_derivative 108 11.5 0.001 0.001 8.049 8.162 init_scf_loop 11 6.9 0.001 0.001 8.078 8.082 rebuild_ks_matrix 119 8.3 0.000 0.001 7.364 7.547 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.364 7.547 prepare_preconditioner 11 7.9 0.000 0.000 7.086 7.105 make_preconditioner 11 8.9 0.000 0.000 7.086 7.105 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.607 6.979 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.633 6.798 dbcsr_mm_accdrv_process 12550 15.8 4.722 6.588 5.298 6.621 cp_fm_upper_to_full 72 14.2 3.194 4.616 3.194 4.616 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.260 4.266 calculate_rho_elec 119 8.7 0.118 0.121 4.259 4.265 sum_up_and_integrate 119 10.3 0.001 0.001 3.926 3.933 integrate_v_rspace 119 11.3 0.003 0.004 3.916 3.923 qs_ot_get_p 119 10.4 0.001 0.001 3.673 3.828 init_scf_run 11 5.9 0.000 0.001 3.760 3.760 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.760 3.760 mp_waitall_1 94719 16.7 2.760 3.677 2.760 3.677 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.923 3.320 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.059 2.906 make_m2s 4572 13.5 0.037 0.037 2.625 2.796 pw_transfer 1439 11.6 0.069 0.069 2.740 2.744 make_images 4572 14.5 0.355 0.389 2.505 2.676 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.641 2.647 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.303 2.582 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.743 2.582 apply_single 119 13.6 0.000 0.000 2.303 2.582 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.588 2.471 mp_alltoall_i22 627 13.8 1.511 2.351 1.511 2.351 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.474 2.304 calculate_dm_sparse 119 9.5 0.000 0.000 2.278 2.296 density_rs2pw 119 9.7 0.003 0.004 2.238 2.270 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.184 2.245 fft_wrap_pw1pw2_140 487 13.2 0.575 0.580 2.224 2.231 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.203 2.206 calculate_first_density_matrix 1 7.0 0.000 0.000 2.193 2.195 ot_diis_step 108 11.5 0.014 0.015 2.187 2.188 grid_integrate_task_list 119 12.3 2.114 2.150 2.114 2.150 mp_sum_l 7287 12.8 1.230 2.063 1.230 2.063 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.972 1.975 acc_transpose_blocks 9144 15.5 0.044 0.045 1.807 1.836 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.829 1.831 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.780 1.780 fft3d_ps 1201 14.6 0.594 0.607 1.737 1.740 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.643 1.645 grid_collocate_task_list 119 9.7 1.546 1.591 1.546 1.591 cp_fm_cholesky_invert 11 10.9 1.581 1.585 1.581 1.585 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.524 1.524 cp_fm_diag_elpa_base 50 14.0 1.370 1.424 1.522 1.523 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.240 1.512 wfi_extrapolate 11 7.9 0.001 0.001 1.501 1.501 make_images_data 4572 15.5 0.046 0.049 1.200 1.427 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.369 1.394 potential_pw2rs 119 12.3 0.014 0.014 1.351 1.354 mp_alltoall_d11v 2130 13.8 1.285 1.323 1.285 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.248 acc_transpose_blocks_sync 27432 16.5 1.106 1.133 1.106 1.133 qs_env_update_s_mstruct 11 6.9 0.010 0.012 1.110 1.126 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.987 1.030 mp_alltoall_z22v 1201 16.6 1.006 1.024 1.006 1.024 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.949 0.965 qs_create_task_list 11 7.9 0.001 0.001 0.940 0.952 generate_qs_task_list 11 8.9 0.368 0.387 0.939 0.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.680000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=723.272727, yerr=12.106279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 503.967744E+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 982542. 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 0.015 0.034 82.440 82.442 qs_mol_dyn_low 1 2.0 0.004 0.006 82.074 82.104 qs_forces 11 3.9 0.004 0.009 81.834 81.835 qs_energies 11 4.9 0.003 0.010 78.960 78.976 scf_env_do_scf 11 5.9 0.001 0.002 70.051 70.054 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 64.461 64.463 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.108 51.463 qs_scf_new_mos 99 7.5 0.000 0.001 47.333 47.445 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.333 47.445 ot_scf_mini 99 9.5 0.002 0.003 44.914 45.020 velocity_verlet 10 3.0 0.001 0.002 43.786 43.788 multiply_cannon 2055 13.4 0.186 0.191 42.517 43.281 multiply_cannon_loop 2055 14.4 1.780 1.820 41.562 42.345 ot_mini 99 10.5 0.001 0.001 26.687 26.796 qs_ot_get_derivative 99 11.5 0.001 0.001 19.818 19.941 multiply_cannon_multrec 49320 15.4 11.396 12.153 17.349 18.003 rebuild_ks_matrix 110 8.3 0.000 0.000 14.601 14.730 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 14.601 14.730 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.794 12.909 mp_waitall_1 220248 16.4 10.832 11.639 10.832 11.639 multiply_cannon_sync_h2d 49320 15.4 9.627 10.339 9.627 10.339 qs_ot_get_p 110 10.4 0.001 0.001 9.701 9.818 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.703 8.236 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.268 7.850 apply_single 110 13.6 0.000 0.000 7.268 7.849 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.474 7.320 sum_up_and_integrate 110 10.3 0.002 0.003 7.064 7.081 integrate_v_rspace 110 11.3 0.003 0.003 7.039 7.062 init_scf_run 11 5.9 0.000 0.001 6.807 6.807 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.806 6.807 ot_diis_step 99 11.5 0.005 0.006 6.633 6.633 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.348 6.501 calculate_rho_elec 110 8.6 0.021 0.025 6.348 6.501 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.428 6.443 dbcsr_mm_accdrv_process 87628 16.1 3.010 3.129 5.822 6.124 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.619 5.619 init_scf_loop 11 6.9 0.001 0.004 5.561 5.564 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.321 5.379 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.054 5.074 cp_fm_diag_elpa_base 48 14.0 5.041 5.062 5.052 5.073 mp_sum_l 6594 12.7 4.150 5.069 4.150 5.069 wfi_extrapolate 11 7.9 0.001 0.001 4.104 4.104 make_m2s 4110 13.4 0.060 0.065 3.987 4.089 make_images 4110 14.4 0.176 0.189 3.892 3.997 calculate_dm_sparse 110 9.5 0.001 0.001 3.785 3.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.552 3.556 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.470 3.515 density_rs2pw 110 9.6 0.004 0.004 3.341 3.503 grid_integrate_task_list 110 12.3 3.259 3.424 3.259 3.424 prepare_preconditioner 11 7.9 0.000 0.001 3.359 3.377 make_preconditioner 11 8.9 0.000 0.002 3.359 3.377 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.232 3.278 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.207 3.251 pw_transfer 1331 11.6 0.055 0.068 3.112 3.187 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.135 3.177 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.023 3.100 fft_wrap_pw1pw2_140 451 13.1 0.391 0.430 2.572 2.657 potential_pw2rs 110 12.3 0.005 0.006 2.624 2.640 acc_transpose_blocks 49320 15.4 0.212 0.220 2.554 2.627 calculate_first_density_matrix 1 7.0 0.001 0.004 2.599 2.601 mp_alltoall_d11v 2046 13.8 2.165 2.568 2.165 2.568 jit_kernel_multiply 13 15.9 2.536 2.549 2.536 2.549 fft3d_ps 1111 14.6 0.793 0.885 2.317 2.372 grid_collocate_task_list 110 9.6 2.161 2.282 2.161 2.282 mp_waitany 14300 13.8 1.872 2.051 1.872 2.051 transfer_rs2pw 451 10.6 0.005 0.006 1.900 2.041 mp_sum_d 3889 11.9 1.443 1.990 1.443 1.990 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.942 1.959 make_images_data 4110 15.4 0.043 0.047 1.813 1.954 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.580 1.844 cp_fm_cholesky_invert 11 10.9 1.816 1.820 1.816 1.820 transfer_pw2rs 451 13.1 0.006 0.007 1.769 1.777 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.646 1.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.442000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=3.106910 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 592.994304E+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 2226166. 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 0.013 0.036 68.886 68.891 qs_mol_dyn_low 1 2.0 0.004 0.004 68.578 68.590 qs_forces 11 3.9 0.003 0.003 68.503 68.504 qs_energies 11 4.9 0.003 0.011 65.178 65.180 scf_env_do_scf 11 5.9 0.001 0.002 56.668 56.672 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.094 49.094 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.850 38.028 velocity_verlet 10 3.0 0.002 0.002 36.022 36.024 qs_scf_new_mos 99 7.5 0.001 0.001 33.233 33.355 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.232 33.354 multiply_cannon 2055 13.4 0.225 0.245 31.231 32.237 ot_scf_mini 99 9.5 0.003 0.004 31.564 31.686 multiply_cannon_loop 2055 14.4 1.163 1.192 29.987 30.931 ot_mini 99 10.5 0.001 0.001 18.344 18.470 multiply_cannon_multrec 24660 15.4 6.978 8.859 14.014 16.015 rebuild_ks_matrix 110 8.3 0.000 0.000 13.542 13.658 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.542 13.658 qs_ot_get_derivative 99 11.5 0.001 0.001 12.566 12.688 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.907 12.012 mp_waitall_1 176588 16.5 7.496 9.997 7.496 9.997 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.082 7.727 init_scf_loop 11 6.9 0.001 0.005 7.536 7.537 multiply_cannon_sync_h2d 24660 15.4 6.389 7.530 6.389 7.530 dbcsr_mm_accdrv_process 52282 16.1 5.464 6.292 6.865 7.210 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.448 7.098 apply_single 110 13.6 0.000 0.001 6.448 7.098 qs_ot_get_p 110 10.4 0.001 0.001 6.558 6.738 sum_up_and_integrate 110 10.3 0.002 0.003 6.393 6.406 integrate_v_rspace 110 11.3 0.002 0.003 6.365 6.382 init_scf_run 11 5.9 0.000 0.001 6.083 6.084 scf_env_initial_rho_setup 11 6.9 0.002 0.006 6.083 6.084 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.897 5.905 calculate_rho_elec 110 8.6 0.039 0.047 5.897 5.905 ot_diis_step 99 11.5 0.010 0.012 5.731 5.731 prepare_preconditioner 11 7.9 0.000 0.000 5.515 5.539 make_preconditioner 11 8.9 0.000 0.003 5.515 5.539 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.710 5.447 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.082 5.238 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.662 4.682 make_m2s 4110 13.4 0.056 0.059 4.258 4.680 make_images 4110 14.4 0.411 0.470 4.149 4.568 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.186 4.186 pw_transfer 1331 11.6 0.066 0.074 3.557 3.695 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.596 3.610 cp_fm_diag_elpa_base 48 14.0 3.549 3.563 3.593 3.607 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.449 3.589 wfi_extrapolate 11 7.9 0.001 0.001 3.503 3.503 density_rs2pw 110 9.6 0.004 0.005 3.145 3.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.288 3.364 grid_integrate_task_list 110 12.3 3.159 3.329 3.159 3.329 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.230 3.231 fft_wrap_pw1pw2_140 451 13.1 0.460 0.476 2.943 3.081 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.001 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.890 2.938 make_images_data 4110 15.4 0.048 0.053 2.303 2.736 fft3d_ps 1111 14.6 1.115 1.341 2.546 2.702 hybrid_alltoall_any 4261 16.3 0.106 0.459 1.967 2.700 cp_fm_cholesky_invert 11 10.9 2.580 2.587 2.580 2.587 calculate_first_density_matrix 1 7.0 0.001 0.006 2.481 2.484 mp_sum_l 6594 12.7 1.737 2.333 1.737 2.333 grid_collocate_task_list 110 9.6 2.166 2.292 2.166 2.292 potential_pw2rs 110 12.3 0.008 0.009 2.269 2.285 jit_kernel_multiply 11 16.3 1.040 2.074 1.040 2.074 acc_transpose_blocks 24660 15.4 0.115 0.118 1.975 2.006 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.963 1.985 mp_alltoall_d11v 2046 13.8 1.746 1.924 1.746 1.924 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.898 1.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.803 transfer_rs2pw 451 10.6 0.006 0.007 1.474 1.689 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.556 mp_allgather_i34 2055 14.4 0.551 1.548 0.551 1.548 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.776 1.527 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.383 1.488 mp_waitany 10164 13.8 1.217 1.432 1.217 1.432 dbcsr_complete_redistribute 325 12.2 0.243 0.307 1.146 1.413 mp_irecv_dv 57340 16.2 0.649 1.403 0.649 1.403 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.891000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.818182, yerr=7.196188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 667.836416E+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 0.021 0.045 60.099 60.101 qs_mol_dyn_low 1 2.0 0.004 0.004 59.699 59.708 qs_forces 11 3.9 0.003 0.003 59.484 59.487 qs_energies 11 4.9 0.002 0.007 56.285 56.290 scf_env_do_scf 11 5.9 0.001 0.001 48.353 48.354 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.751 39.752 velocity_verlet 10 3.0 0.001 0.002 32.752 32.758 dbcsr_multiply_generic 2055 12.4 0.107 0.111 28.706 28.958 qs_scf_new_mos 99 7.5 0.001 0.001 25.101 25.201 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.100 25.200 ot_scf_mini 99 9.5 0.002 0.003 23.840 23.960 multiply_cannon 2055 13.4 0.214 0.224 22.511 23.669 multiply_cannon_loop 2055 14.4 0.814 0.838 21.365 22.325 ot_mini 99 10.5 0.001 0.001 13.690 13.811 rebuild_ks_matrix 110 8.3 0.000 0.000 12.123 12.284 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.123 12.284 multiply_cannon_multrec 16440 15.4 3.644 4.866 9.833 11.009 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.674 10.817 mp_waitall_1 139946 16.5 6.968 10.039 6.968 10.039 qs_ot_get_derivative 99 11.5 0.001 0.001 9.269 9.387 init_scf_loop 11 6.9 0.001 0.003 8.567 8.567 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.373 7.241 prepare_preconditioner 11 7.9 0.000 0.000 6.827 6.841 make_preconditioner 11 8.9 0.000 0.002 6.827 6.841 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.141 6.490 sum_up_and_integrate 110 10.3 0.001 0.002 6.291 6.306 integrate_v_rspace 110 11.3 0.003 0.004 6.265 6.280 dbcsr_mm_accdrv_process 34862 16.1 5.312 5.690 6.032 6.154 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.771 5.777 calculate_rho_elec 110 8.6 0.059 0.059 5.770 5.777 init_scf_run 11 5.9 0.000 0.001 5.490 5.490 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.489 5.490 qs_ot_get_p 110 10.4 0.001 0.001 5.318 5.488 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.849 5.386 apply_single 110 13.6 0.000 0.000 4.849 5.385 make_m2s 4110 13.4 0.049 0.051 4.185 4.534 make_images 4110 14.4 0.401 0.532 4.071 4.420 ot_diis_step 99 11.5 0.011 0.011 4.393 4.393 multiply_cannon_sync_h2d 16440 15.4 3.254 3.898 3.254 3.898 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.243 3.827 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.683 3.686 pw_transfer 1331 11.6 0.066 0.073 3.580 3.588 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.472 3.481 grid_integrate_task_list 110 12.3 3.199 3.444 3.199 3.444 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.302 3.303 density_rs2pw 110 9.6 0.004 0.005 2.999 3.287 fft_wrap_pw1pw2_140 451 13.1 0.581 0.587 2.995 3.006 wfi_extrapolate 11 7.9 0.001 0.001 2.980 2.980 make_images_data 4110 15.4 0.045 0.050 2.340 2.791 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.767 2.779 cp_fm_diag_elpa_base 48 14.0 2.702 2.733 2.766 2.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.765 2.766 hybrid_alltoall_any 4261 16.3 0.109 0.386 2.096 2.710 cp_fm_cholesky_invert 11 10.9 2.682 2.688 2.682 2.688 calculate_dm_sparse 110 9.5 0.001 0.001 2.563 2.594 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.438 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.433 2.482 multiply_cannon_metrocomm4 14385 15.4 0.049 0.052 0.877 2.446 fft3d_ps 1111 14.6 1.091 1.101 2.412 2.424 calculate_first_density_matrix 1 7.0 0.001 0.004 2.418 2.421 grid_collocate_task_list 110 9.6 2.223 2.413 2.223 2.413 mp_irecv_dv 48980 15.7 0.802 2.309 0.802 2.309 mp_alltoall_d11v 2046 13.8 1.795 2.180 1.795 2.180 potential_pw2rs 110 12.3 0.011 0.012 2.125 2.129 mp_sum_l 6594 12.7 1.395 1.993 1.395 1.993 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.987 1.990 dbcsr_complete_redistribute 325 12.2 0.329 0.351 1.443 1.900 cp_fm_upper_to_full 70 14.2 1.409 1.783 1.409 1.783 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.659 acc_transpose_blocks 16440 15.4 0.077 0.080 1.581 1.620 cp_fm_cholesky_decompose 22 10.9 1.595 1.619 1.595 1.619 transfer_rs2pw 451 10.6 0.005 0.006 1.323 1.618 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.497 1.512 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.488 mp_allgather_i34 2055 14.4 0.458 1.456 0.458 1.456 mp_waitany 17072 13.8 1.175 1.445 1.175 1.445 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.981 1.432 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.301 1.315 rs_gather_matrices 110 12.3 0.138 0.151 0.889 1.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.101000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.090909, yerr=8.712108 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 744.792064E+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 0.024 0.044 66.893 66.895 qs_mol_dyn_low 1 2.0 0.084 0.097 66.333 66.343 qs_forces 11 3.9 0.066 0.074 66.174 66.188 qs_energies 11 4.9 0.002 0.002 62.714 62.723 scf_env_do_scf 11 5.9 0.000 0.001 54.402 54.405 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.592 42.593 velocity_verlet 10 3.0 0.002 0.002 37.608 37.611 dbcsr_multiply_generic 2055 12.4 0.113 0.120 30.875 31.140 qs_scf_new_mos 99 7.5 0.001 0.001 27.940 28.044 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.940 28.043 ot_scf_mini 99 9.5 0.002 0.003 26.248 26.340 multiply_cannon 2055 13.4 0.252 0.269 23.782 24.675 multiply_cannon_loop 2055 14.4 1.417 1.491 22.240 22.925 ot_mini 99 10.5 0.001 0.001 15.087 15.210 multiply_cannon_multrec 24660 15.4 4.085 6.706 12.989 14.153 rebuild_ks_matrix 110 8.3 0.000 0.000 11.974 12.081 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.973 12.081 init_scf_loop 11 6.9 0.000 0.001 11.770 11.771 qs_ot_get_derivative 99 11.5 0.001 0.001 10.911 11.009 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.582 10.678 prepare_preconditioner 11 7.9 0.000 0.000 10.036 10.050 make_preconditioner 11 8.9 0.000 0.000 10.036 10.050 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.262 9.726 dbcsr_mm_accdrv_process 52304 16.0 7.724 9.069 8.745 9.654 mp_waitall_1 121746 16.5 4.300 6.346 4.300 6.346 sum_up_and_integrate 110 10.3 0.001 0.002 6.249 6.261 integrate_v_rspace 110 11.3 0.003 0.003 6.223 6.236 qs_ot_get_p 110 10.4 0.001 0.001 6.007 6.138 make_m2s 4110 13.4 0.059 0.061 5.433 5.885 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.865 5.879 calculate_rho_elec 110 8.6 0.078 0.081 5.864 5.878 make_images 4110 14.4 0.581 0.708 5.293 5.740 init_scf_run 11 5.9 0.000 0.001 5.547 5.548 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.547 5.548 cp_fm_upper_to_full 70 14.2 3.366 4.882 3.366 4.882 ot_diis_step 99 11.5 0.011 0.011 4.135 4.135 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.039 4.121 apply_single 110 13.6 0.000 0.000 4.039 4.121 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.067 4.083 dbcsr_complete_redistribute 325 12.2 0.438 0.482 2.675 3.816 pw_transfer 1331 11.6 0.066 0.075 3.672 3.707 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.565 3.603 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.522 3.571 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.505 3.505 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.395 3.447 grid_integrate_task_list 110 12.3 3.288 3.403 3.288 3.403 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.152 3.280 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.355 3.154 density_rs2pw 110 9.6 0.004 0.004 2.986 3.144 make_images_data 4110 15.4 0.048 0.052 2.655 3.126 fft_wrap_pw1pw2_140 451 13.1 0.611 0.635 3.047 3.085 wfi_extrapolate 11 7.9 0.001 0.001 3.057 3.057 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.464 3.045 calculate_dm_sparse 110 9.5 0.001 0.001 2.987 3.025 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.950 2.959 cp_fm_diag_elpa_base 48 14.0 2.798 2.859 2.948 2.957 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.765 2.878 mp_alltoall_i22 605 13.7 1.658 2.856 1.658 2.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.844 2.846 cp_fm_cholesky_invert 11 10.9 2.769 2.777 2.769 2.777 multiply_cannon_sync_h2d 24660 15.4 2.360 2.522 2.360 2.522 acc_transpose_blocks 24660 15.4 0.114 0.119 2.422 2.518 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.488 fft3d_ps 1111 14.6 1.085 1.119 2.445 2.461 grid_collocate_task_list 110 9.6 2.264 2.413 2.264 2.413 calculate_first_density_matrix 1 7.0 0.000 0.002 2.393 2.397 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.271 2.277 potential_pw2rs 110 12.3 0.012 0.013 2.044 2.051 mp_alltoall_d11v 2046 13.8 1.798 1.996 1.798 1.996 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.809 1.844 cp_fm_cholesky_decompose 22 10.9 1.704 1.747 1.704 1.747 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.715 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.631 1.644 mp_sum_l 6594 12.7 0.973 1.631 0.973 1.631 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.558 acc_transpose_blocks_sync 73980 16.4 1.410 1.501 1.410 1.501 multiply_cannon_metrocomm4 20550 15.4 0.063 0.066 0.868 1.497 jit_kernel_multiply 8 16.0 0.678 1.437 0.678 1.437 mp_irecv_dv 62702 16.1 0.763 1.412 0.763 1.412 transfer_rs2pw 451 10.6 0.005 0.006 1.231 1.397 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.895000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.636364, yerr=13.472347 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 858.898432E+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 0.028 0.045 57.464 57.466 qs_mol_dyn_low 1 2.0 0.036 0.037 56.839 56.849 qs_forces 11 3.9 0.062 0.063 55.467 55.468 qs_energies 11 4.9 0.003 0.004 51.759 51.765 scf_env_do_scf 11 5.9 0.000 0.001 43.502 43.502 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.637 35.638 velocity_verlet 10 3.0 0.020 0.029 32.665 32.747 dbcsr_multiply_generic 2055 12.4 0.106 0.111 23.441 23.614 qs_scf_new_mos 99 7.5 0.001 0.001 20.983 21.028 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.982 21.028 ot_scf_mini 99 9.5 0.002 0.002 19.715 19.735 multiply_cannon 2055 13.4 0.235 0.246 17.689 19.210 multiply_cannon_loop 2055 14.4 0.604 0.623 16.398 16.792 rebuild_ks_matrix 110 8.3 0.000 0.000 11.702 11.718 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.701 11.718 ot_mini 99 10.5 0.001 0.001 10.897 10.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.409 10.427 multiply_cannon_multrec 8220 15.4 3.167 4.318 7.553 8.493 init_scf_loop 11 6.9 0.000 0.001 7.816 7.818 mp_waitall_1 103326 16.6 6.013 7.543 6.013 7.543 qs_ot_get_derivative 99 11.5 0.001 0.001 7.196 7.215 sum_up_and_integrate 110 10.3 0.001 0.002 6.256 6.268 integrate_v_rspace 110 11.3 0.003 0.003 6.230 6.241 prepare_preconditioner 11 7.9 0.000 0.000 6.146 6.149 make_preconditioner 11 8.9 0.000 0.000 6.146 6.149 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.023 6.036 calculate_rho_elec 110 8.6 0.113 0.115 6.023 6.035 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.714 5.795 dbcsr_mm_accdrv_process 17442 15.9 3.071 3.984 4.247 5.176 init_scf_run 11 5.9 0.000 0.001 5.111 5.111 scf_env_initial_rho_setup 11 6.9 0.010 0.011 5.110 5.111 qs_ot_get_p 110 10.4 0.001 0.001 4.842 4.857 make_m2s 4110 13.4 0.038 0.039 4.357 4.651 make_images 4110 14.4 0.653 0.716 4.228 4.522 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.968 4.243 pw_transfer 1331 11.6 0.066 0.072 3.942 3.963 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.834 3.857 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.728 3.767 apply_single 110 13.6 0.000 0.000 3.728 3.767 ot_diis_step 99 11.5 0.012 0.012 3.672 3.672 grid_integrate_task_list 110 12.3 3.371 3.491 3.371 3.491 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.414 3.419 fft_wrap_pw1pw2_140 451 13.1 0.776 0.789 3.314 3.353 density_rs2pw 110 9.6 0.004 0.004 2.998 3.104 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.061 3.061 cp_fm_cholesky_invert 11 10.9 2.935 2.939 2.935 2.939 hybrid_alltoall_any 4261 16.3 0.200 0.855 2.289 2.829 make_images_data 4110 15.4 0.040 0.047 2.338 2.758 wfi_extrapolate 11 7.9 0.001 0.001 2.754 2.754 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.675 2.676 calculate_dm_sparse 110 9.5 0.001 0.001 2.501 2.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.534 2.535 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.526 2.534 cp_fm_diag_elpa_base 48 14.0 2.470 2.495 2.525 2.532 grid_collocate_task_list 110 9.6 2.372 2.532 2.372 2.532 fft3d_ps 1111 14.6 1.132 1.168 2.500 2.512 multiply_cannon_sync_h2d 8220 15.4 2.378 2.449 2.378 2.449 calculate_first_density_matrix 1 7.0 0.000 0.001 2.251 2.253 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.138 2.154 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.044 2.069 potential_pw2rs 110 12.3 0.015 0.015 2.027 2.030 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.974 1.990 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.988 mp_alltoall_d11v 2046 13.8 1.758 1.895 1.758 1.895 cp_fm_cholesky_decompose 22 10.9 1.738 1.753 1.738 1.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.632 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.504 1.618 dbcsr_complete_redistribute 325 12.2 0.562 0.598 1.490 1.596 mp_allgather_i34 2055 14.4 0.461 1.588 0.461 1.588 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.451 acc_transpose_blocks 8220 15.4 0.040 0.041 1.365 1.389 mp_sum_dm 438 4.9 1.248 1.341 1.248 1.341 md_write_output 11 3.9 0.080 1.274 0.084 1.324 qs_create_task_list 11 7.9 0.000 0.001 1.216 1.316 generate_qs_task_list 11 8.9 0.375 0.445 1.215 1.315 update_particle_set 20 4.0 0.000 0.000 1.220 1.309 transfer_rs2pw 451 10.6 0.005 0.005 1.151 1.283 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.196 1.219 mp_waitany 9240 13.8 1.078 1.217 1.078 1.217 mp_alltoall_z22v 1111 16.6 1.135 1.156 1.135 1.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.466000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.727273, yerr=10.913635 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 1.421431E+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 0.015 0.036 88.595 88.596 qs_mol_dyn_low 1 2.0 0.003 0.003 88.244 88.256 qs_forces 11 3.9 0.003 0.004 88.068 88.068 qs_energies 11 4.9 0.002 0.002 83.923 83.924 scf_env_do_scf 11 5.9 0.001 0.001 73.818 73.818 velocity_verlet 10 3.0 0.002 0.002 56.637 56.657 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.179 45.180 dbcsr_multiply_generic 2055 12.4 0.119 0.125 30.214 30.311 init_scf_loop 11 6.9 0.001 0.001 28.565 28.568 qs_scf_new_mos 99 7.5 0.001 0.001 27.679 27.758 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.678 27.757 prepare_preconditioner 11 7.9 0.000 0.000 26.548 26.558 make_preconditioner 11 8.9 0.000 0.000 26.548 26.558 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.812 26.024 ot_scf_mini 99 9.5 0.002 0.002 25.843 25.913 multiply_cannon 2055 13.4 0.334 0.348 22.792 23.583 multiply_cannon_loop 2055 14.4 0.824 0.838 20.923 21.210 cp_fm_upper_to_full 70 14.2 12.725 18.184 12.725 18.184 ot_mini 99 10.5 0.001 0.001 14.629 14.689 rebuild_ks_matrix 110 8.3 0.000 0.001 13.754 13.829 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.753 13.828 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.478 12.548 dbcsr_complete_redistribute 325 12.2 1.028 1.059 7.373 10.598 multiply_cannon_multrec 8220 15.4 4.060 4.184 9.740 9.866 qs_ot_get_derivative 99 11.5 0.001 0.001 9.762 9.829 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.311 9.532 mp_waitall_1 84994 16.7 7.991 8.972 7.991 8.972 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.720 8.921 mp_alltoall_i22 605 13.7 5.365 8.568 5.365 8.568 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.179 7.216 calculate_rho_elec 110 8.6 0.223 0.224 7.178 7.216 sum_up_and_integrate 110 10.3 0.002 0.002 6.839 6.855 integrate_v_rspace 110 11.3 0.003 0.004 6.811 6.827 make_m2s 4110 13.4 0.043 0.044 5.554 6.082 make_images 4110 14.4 0.887 0.939 5.365 5.892 qs_ot_get_p 110 10.4 0.001 0.001 5.820 5.891 init_scf_run 11 5.9 0.000 0.001 5.824 5.824 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.824 5.824 dbcsr_mm_accdrv_process 11614 15.7 3.845 4.151 5.533 5.755 cp_fm_cholesky_invert 11 10.9 5.454 5.459 5.454 5.459 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.149 5.446 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.955 5.436 apply_single 110 13.6 0.000 0.000 4.954 5.436 pw_transfer 1331 11.6 0.074 0.075 4.980 4.986 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.863 4.869 ot_diis_step 99 11.5 0.015 0.016 4.848 4.848 fft_wrap_pw1pw2_140 451 13.1 1.280 1.285 4.232 4.238 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.143 4.151 grid_integrate_task_list 110 12.3 3.704 3.802 3.704 3.802 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.659 3.660 hybrid_alltoall_any 4261 16.3 0.262 0.562 2.917 3.658 density_rs2pw 110 9.6 0.004 0.004 3.605 3.649 make_images_data 4110 15.4 0.045 0.048 2.961 3.649 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.621 3.622 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.119 3.571 wfi_extrapolate 11 7.9 0.001 0.001 3.414 3.414 calculate_dm_sparse 110 9.5 0.001 0.001 3.200 3.231 multiply_cannon_sync_h2d 8220 15.4 3.129 3.139 3.129 3.139 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.027 3.027 cp_fm_diag_elpa_base 48 14.0 2.486 2.672 3.025 3.025 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.987 2.989 fft3d_ps 1111 14.6 1.293 1.300 2.915 2.927 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.755 2.794 grid_collocate_task_list 110 9.6 2.689 2.733 2.689 2.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.458 potential_pw2rs 110 12.3 0.021 0.021 2.295 2.302 calculate_first_density_matrix 1 7.0 0.000 0.000 2.286 2.286 qs_env_update_s_mstruct 11 6.9 0.002 0.002 2.181 2.240 mp_alltoall_d11v 2046 13.8 2.142 2.218 2.142 2.218 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.180 cp_fm_cholesky_decompose 22 10.9 2.099 2.121 2.099 2.121 qs_create_task_list 11 7.9 0.001 0.001 1.890 1.935 generate_qs_task_list 11 8.9 0.733 0.790 1.889 1.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.876 1.880 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.802 1.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.596000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1261.727273, yerr=68.093640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.616064E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2529110. 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 0.019 0.040 207.072 207.074 qs_mol_dyn_low 1 2.0 0.004 0.004 206.625 206.640 qs_forces 11 3.9 0.005 0.005 206.252 206.253 qs_energies 11 4.9 0.002 0.003 200.683 200.696 scf_env_do_scf 11 5.9 0.001 0.001 184.332 184.336 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 163.894 163.896 dbcsr_multiply_generic 2507 12.6 0.180 0.183 125.884 126.430 qs_scf_new_mos 117 7.6 0.001 0.001 124.608 124.883 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.607 124.882 velocity_verlet 10 3.0 0.001 0.002 123.513 123.517 ot_scf_mini 117 9.6 0.003 0.003 117.933 118.196 multiply_cannon 2507 13.6 0.240 0.249 101.655 103.587 multiply_cannon_loop 2507 14.6 2.406 2.450 99.525 101.077 ot_mini 117 10.6 0.001 0.001 66.692 66.932 multiply_cannon_multrec 60168 15.6 31.803 33.809 41.535 43.589 qs_ot_get_derivative 117 11.6 0.001 0.001 41.802 42.047 rebuild_ks_matrix 128 8.3 0.001 0.001 33.491 33.823 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.490 33.823 mp_waitall_1 267128 16.5 28.953 32.218 28.953 32.218 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.076 30.373 qs_ot_get_p 128 10.4 0.001 0.001 28.995 29.312 multiply_cannon_sync_h2d 60168 15.6 26.309 28.355 26.309 28.355 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.353 25.216 apply_single 128 13.6 0.001 0.001 24.353 25.216 ot_diis_step 117 11.6 0.008 0.008 24.610 24.611 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.396 22.480 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.257 20.449 init_scf_loop 11 6.9 0.001 0.002 20.365 20.366 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.691 19.692 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 16.213 18.240 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.724 16.754 cp_fm_diag_elpa_base 83 14.4 16.655 16.694 16.720 16.751 prepare_preconditioner 11 7.9 0.000 0.000 15.781 15.826 make_preconditioner 11 8.9 0.000 0.000 15.781 15.826 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.023 15.217 make_m2s 5014 13.6 0.104 0.112 13.844 14.290 make_images 5014 14.6 0.396 0.415 13.661 14.118 sum_up_and_integrate 128 10.3 0.002 0.004 13.909 13.924 integrate_v_rspace 128 11.3 0.004 0.004 13.851 13.870 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.041 13.197 calculate_rho_elec 128 8.7 0.045 0.064 13.041 13.196 init_scf_run 11 5.9 0.000 0.001 12.228 12.229 scf_env_initial_rho_setup 11 6.9 0.006 0.008 12.228 12.229 mp_sum_l 7950 12.9 9.124 10.552 9.124 10.552 dbcsr_mm_accdrv_process 124484 16.2 4.721 4.867 9.301 9.833 wfi_extrapolate 11 7.9 0.001 0.001 9.032 9.033 cp_fm_cholesky_invert 11 10.9 8.967 8.975 8.967 8.975 calculate_dm_sparse 128 9.5 0.001 0.001 8.482 8.585 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.127 8.265 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.148 8.257 multiply_cannon_metrocomm1 60168 15.6 0.097 0.101 6.329 8.048 pw_transfer 1547 11.6 0.074 0.091 7.543 7.718 make_images_data 5014 15.6 0.066 0.071 6.784 7.691 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.340 7.513 grid_integrate_task_list 128 12.3 7.056 7.415 7.056 7.415 density_rs2pw 128 9.7 0.006 0.007 6.671 7.249 hybrid_alltoall_any 5200 16.5 0.294 2.263 5.956 7.189 fft_wrap_pw1pw2_140 523 13.2 1.129 1.174 6.464 6.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.620 6.633 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.748 5.863 fft3d_ps 1291 14.7 2.189 2.848 5.343 5.680 mp_alltoall_d11v 2415 14.1 4.349 5.608 4.349 5.608 grid_collocate_task_list 128 9.7 4.837 5.208 4.837 5.208 cp_fm_cholesky_decompose 22 10.9 4.629 4.642 4.629 4.642 potential_pw2rs 128 12.3 0.009 0.010 4.382 4.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.074000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.636364, yerr=6.328474 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.946240E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5046142. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.036 188.939 188.941 qs_mol_dyn_low 1 2.0 0.004 0.005 188.580 188.595 qs_forces 11 3.9 0.004 0.005 188.468 188.469 qs_energies 11 4.9 0.002 0.003 181.712 181.724 scf_env_do_scf 11 5.9 0.001 0.001 165.357 165.367 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.492 132.495 velocity_verlet 10 3.0 0.002 0.002 119.199 119.201 dbcsr_multiply_generic 2507 12.6 0.190 0.196 97.595 98.841 qs_scf_new_mos 117 7.6 0.001 0.001 93.934 94.536 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.933 94.535 ot_scf_mini 117 9.6 0.004 0.004 89.132 89.772 multiply_cannon 2507 13.6 0.504 0.556 77.392 81.637 multiply_cannon_loop 2507 14.6 1.575 1.649 73.908 76.817 ot_mini 117 10.6 0.001 0.001 50.147 50.731 mp_waitall_1 214728 16.6 23.838 39.304 23.838 39.304 multiply_cannon_multrec 30084 15.6 21.077 25.958 31.650 37.001 rebuild_ks_matrix 128 8.3 0.001 0.001 32.346 33.187 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.346 33.187 init_scf_loop 11 6.9 0.000 0.001 32.775 32.776 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.082 29.860 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.548 29.239 qs_ot_get_derivative 117 11.6 0.001 0.001 28.215 28.823 prepare_preconditioner 11 7.9 0.000 0.000 28.451 28.526 make_preconditioner 11 8.9 0.000 0.000 28.451 28.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.143 27.696 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.070 23.115 apply_single 128 13.6 0.001 0.001 22.070 23.115 ot_diis_step 117 11.6 0.014 0.015 21.762 21.764 qs_ot_get_p 128 10.4 0.001 0.001 20.921 21.640 multiply_cannon_sync_h2d 30084 15.6 18.073 21.525 18.073 21.525 cp_fm_cholesky_invert 11 10.9 16.633 16.645 16.633 16.645 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.960 15.994 make_m2s 5014 13.6 0.089 0.096 14.143 15.915 make_images 5014 14.6 1.179 1.384 13.931 15.703 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.763 14.764 sum_up_and_integrate 128 10.3 0.002 0.004 13.947 13.974 integrate_v_rspace 128 11.3 0.003 0.004 13.887 13.918 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.329 13.360 calculate_rho_elec 128 8.7 0.087 0.104 13.328 13.360 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.192 11.659 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.597 11.632 cp_fm_diag_elpa_base 83 14.4 11.342 11.438 11.591 11.622 init_scf_run 11 5.9 0.000 0.001 11.571 11.573 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.571 11.573 multiply_cannon_metrocomm4 27577 15.6 0.106 0.119 3.835 10.957 dbcsr_mm_accdrv_process 62242 16.2 5.415 6.369 10.022 10.599 mp_irecv_dv 69486 16.3 3.629 10.556 3.629 10.556 make_images_data 5014 15.6 0.066 0.076 8.303 10.401 hybrid_alltoall_any 5200 16.5 0.352 1.487 7.087 10.135 pw_transfer 1547 11.6 0.085 0.099 8.572 8.631 wfi_extrapolate 11 7.9 0.001 0.001 8.422 8.422 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.348 8.410 density_rs2pw 128 9.7 0.006 0.007 7.044 7.591 grid_integrate_task_list 128 12.3 7.201 7.558 7.201 7.558 fft_wrap_pw1pw2_140 523 13.2 1.212 1.234 7.383 7.459 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.281 7.094 cp_fm_cholesky_decompose 22 10.9 6.908 6.991 6.908 6.991 calculate_dm_sparse 128 9.5 0.001 0.001 6.459 6.600 mp_sum_l 7950 12.9 4.259 6.199 4.259 6.199 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.181 6.192 fft3d_ps 1291 14.7 2.835 3.016 5.910 5.949 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.255 5.474 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.347 5.415 grid_collocate_task_list 128 9.7 5.003 5.328 5.003 5.328 mp_allgather_i34 2507 14.6 1.999 5.137 1.999 5.137 mp_alltoall_d11v 2415 14.1 4.091 4.610 4.091 4.610 potential_pw2rs 128 12.3 0.016 0.019 4.439 4.451 dbcsr_complete_redistribute 395 12.7 0.779 0.868 3.146 4.006 mp_sum_d 4465 12.1 2.603 3.957 2.603 3.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.941000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.454545, yerr=1.437399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 960.561152E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. 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 0.034 0.062 175.105 175.105 qs_mol_dyn_low 1 2.0 0.003 0.004 174.695 174.708 qs_forces 11 3.9 0.004 0.005 174.606 174.606 qs_energies 11 4.9 0.002 0.002 168.076 168.087 scf_env_do_scf 11 5.9 0.001 0.001 152.620 152.621 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.958 117.958 velocity_verlet 10 3.0 0.001 0.002 112.861 112.863 dbcsr_multiply_generic 2529 12.6 0.186 0.195 82.301 83.599 qs_scf_new_mos 118 7.6 0.001 0.001 80.692 81.016 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.691 81.015 ot_scf_mini 118 9.6 0.003 0.004 76.419 76.847 multiply_cannon 2529 13.6 0.512 0.532 62.414 67.479 multiply_cannon_loop 2529 14.6 1.135 1.199 59.481 62.439 ot_mini 118 10.6 0.001 0.001 42.886 43.271 mp_waitall_1 172006 16.6 24.692 34.987 24.692 34.987 init_scf_loop 11 6.9 0.000 0.001 34.565 34.565 rebuild_ks_matrix 129 8.3 0.001 0.001 30.358 30.846 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 30.358 30.846 prepare_preconditioner 11 7.9 0.000 0.000 30.565 30.611 make_preconditioner 11 8.9 0.000 0.000 30.565 30.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.203 29.639 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.384 27.827 multiply_cannon_multrec 20232 15.6 13.311 16.611 22.672 26.021 multiply_cannon_metrocomm3 20232 15.6 0.063 0.066 15.146 24.807 qs_ot_get_derivative 118 11.6 0.002 0.002 23.221 23.619 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.651 20.827 apply_single 129 13.6 0.001 0.001 19.651 20.827 ot_diis_step 118 11.6 0.018 0.018 19.559 19.559 qs_ot_get_p 129 10.4 0.001 0.001 18.861 19.361 make_m2s 5058 13.6 0.080 0.088 14.795 15.987 make_images 5058 14.6 1.162 1.252 14.560 15.749 multiply_cannon_sync_h2d 20232 15.6 13.711 15.631 13.711 15.631 qs_ot_p2m_diag 84 11.4 0.268 0.274 14.494 14.504 cp_fm_cholesky_invert 11 10.9 14.140 14.149 14.140 14.149 sum_up_and_integrate 129 10.3 0.002 0.003 14.048 14.073 integrate_v_rspace 129 11.3 0.003 0.004 13.988 14.013 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.581 13.627 calculate_rho_elec 129 8.7 0.131 0.145 13.581 13.627 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.426 13.427 make_images_data 5058 15.6 0.063 0.074 8.860 10.510 init_scf_run 11 5.9 0.000 0.001 10.454 10.454 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.454 10.454 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.226 10.242 cp_fm_diag_elpa_base 84 14.4 9.813 9.973 10.222 10.239 hybrid_alltoall_any 5245 16.5 0.451 2.039 7.683 10.042 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.223 9.538 multiply_cannon_metrocomm4 17703 15.6 0.068 0.077 3.455 9.401 mp_irecv_dv 50659 16.2 3.325 9.138 3.325 9.138 dbcsr_mm_accdrv_process 41846 16.2 5.609 5.963 8.812 8.946 pw_transfer 1559 11.6 0.086 0.104 8.723 8.831 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.498 8.611 grid_integrate_task_list 129 12.3 7.368 7.778 7.368 7.778 fft_wrap_pw1pw2_140 527 13.2 1.313 1.342 7.540 7.663 wfi_extrapolate 11 7.9 0.001 0.001 7.416 7.416 density_rs2pw 129 9.7 0.006 0.006 6.894 7.299 cp_fm_upper_to_full 106 14.8 5.778 7.284 5.778 7.284 cp_fm_cholesky_decompose 22 10.9 7.226 7.248 7.226 7.248 dbcsr_complete_redistribute 397 12.7 1.172 1.199 4.562 6.313 calculate_dm_sparse 129 9.5 0.001 0.001 5.867 5.986 fft3d_ps 1301 14.7 2.781 3.010 5.852 5.946 grid_collocate_task_list 129 9.7 5.189 5.717 5.189 5.717 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.436 5.440 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.651 5.288 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.410 5.157 mp_alltoall_d11v 2429 14.1 4.328 4.938 4.328 4.938 mp_sum_l 8016 12.9 3.236 4.714 3.236 4.714 mp_allgather_i34 2529 14.6 1.413 4.693 1.413 4.693 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.533 4.688 potential_pw2rs 129 12.3 0.020 0.022 4.393 4.403 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.342 4.050 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.927 3.970 mp_alltoall_i22 720 14.1 1.965 3.877 1.965 3.877 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.728 3.729 mp_sum_d 4507 12.2 2.277 3.538 2.277 3.538 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.472 3.518 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.105000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.181818, yerr=15.236429 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.153782E+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 1.770319E+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 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503084. 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 0.010 0.029 189.141 189.141 qs_mol_dyn_low 1 2.0 0.003 0.003 188.776 188.791 qs_forces 11 3.9 0.004 0.004 188.650 188.659 qs_energies 11 4.9 0.002 0.003 181.599 181.614 scf_env_do_scf 11 5.9 0.001 0.001 164.823 164.834 velocity_verlet 10 3.0 0.002 0.002 125.567 125.570 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.202 118.206 dbcsr_multiply_generic 2507 12.6 0.187 0.193 81.477 82.304 qs_scf_new_mos 117 7.6 0.001 0.001 81.827 82.182 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.826 82.181 ot_scf_mini 117 9.6 0.003 0.004 77.071 77.421 multiply_cannon 2507 13.6 0.556 0.592 56.605 59.521 multiply_cannon_loop 2507 14.6 1.823 1.913 52.196 54.051 init_scf_loop 11 6.9 0.001 0.001 46.493 46.501 ot_mini 117 10.6 0.001 0.001 43.598 43.946 prepare_preconditioner 11 7.9 0.000 0.000 42.394 42.416 make_preconditioner 11 8.9 0.000 0.000 42.394 42.416 make_full_inverse_cholesky 11 9.9 0.012 0.022 35.935 40.921 multiply_cannon_multrec 30084 15.6 13.499 19.109 26.381 31.672 rebuild_ks_matrix 128 8.3 0.001 0.001 29.189 29.550 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.189 29.549 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.348 26.674 mp_waitall_1 147882 16.7 17.179 26.117 17.179 26.117 qs_ot_get_derivative 117 11.6 0.002 0.002 23.886 24.241 make_m2s 5014 13.6 0.097 0.101 20.354 21.157 make_images 5014 14.6 1.981 2.498 20.044 20.848 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.078 19.702 apply_single 128 13.6 0.001 0.001 19.078 19.702 qs_ot_get_p 128 10.4 0.001 0.001 19.334 19.702 ot_diis_step 117 11.6 0.018 0.019 19.581 19.582 cp_fm_upper_to_full 105 14.8 11.426 16.813 11.426 16.813 cp_fm_cholesky_invert 11 10.9 16.153 16.163 16.153 16.163 multiply_cannon_metrocomm3 30084 15.6 0.050 0.052 6.426 15.109 qs_ot_p2m_diag 83 11.4 0.342 0.390 15.027 15.079 sum_up_and_integrate 128 10.3 0.002 0.003 13.990 14.019 integrate_v_rspace 128 11.3 0.003 0.004 13.930 13.960 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.924 13.955 calculate_rho_elec 128 8.7 0.172 0.189 13.923 13.955 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.616 13.617 dbcsr_mm_accdrv_process 62264 16.2 8.483 9.255 12.455 12.954 dbcsr_complete_redistribute 395 12.7 1.498 1.628 9.210 12.929 make_images_data 5014 15.6 0.066 0.073 10.845 12.593 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.840 11.555 multiply_cannon_sync_h2d 30084 15.6 10.473 11.290 10.473 11.290 hybrid_alltoall_any 5200 16.5 0.536 2.222 9.740 11.109 init_scf_run 11 5.9 0.000 0.001 10.813 10.814 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.813 10.814 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.179 10.433 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.366 10.381 cp_fm_diag_elpa_base 83 14.4 9.407 9.729 10.359 10.373 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 6.437 10.024 mp_alltoall_i22 716 14.1 5.606 9.329 5.606 9.329 pw_transfer 1547 11.6 0.085 0.102 9.030 9.106 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.805 8.886 grid_integrate_task_list 128 12.3 7.567 7.939 7.567 7.939 fft_wrap_pw1pw2_140 523 13.2 1.444 1.481 7.834 7.925 cp_fm_cholesky_decompose 22 10.9 7.606 7.709 7.606 7.709 wfi_extrapolate 11 7.9 0.001 0.001 7.679 7.679 multiply_cannon_metrocomm4 25070 15.6 0.086 0.096 2.888 7.206 density_rs2pw 128 9.7 0.006 0.006 6.794 7.181 mp_irecv_dv 76098 16.2 2.731 6.935 2.731 6.935 calculate_dm_sparse 128 9.5 0.001 0.001 6.450 6.543 fft3d_ps 1291 14.7 2.857 2.924 5.984 6.042 grid_collocate_task_list 128 9.7 5.287 5.852 5.287 5.852 mp_alltoall_d11v 2415 14.1 5.387 5.754 5.387 5.754 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.458 5.511 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.604 4.707 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.592 4.642 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.358 4.439 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.434 4.434 potential_pw2rs 128 12.3 0.023 0.023 4.243 4.251 mp_sum_l 7950 12.9 2.583 3.858 2.583 3.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.141000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1092.272727, yerr=12.628186 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.527411E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526746E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848081. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.046 172.584 172.587 qs_mol_dyn_low 1 2.0 0.007 0.012 172.097 172.111 qs_forces 11 3.9 0.004 0.004 171.553 171.558 qs_energies 11 4.9 0.002 0.002 164.127 164.136 scf_env_do_scf 11 5.9 0.001 0.001 146.800 146.818 velocity_verlet 10 3.0 0.002 0.003 113.839 113.870 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 110.630 110.632 dbcsr_multiply_generic 2529 12.6 0.183 0.195 73.168 73.647 qs_scf_new_mos 118 7.6 0.001 0.001 73.466 73.589 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.465 73.588 ot_scf_mini 118 9.6 0.003 0.004 68.962 69.102 multiply_cannon 2529 13.6 0.567 0.589 54.225 57.456 multiply_cannon_loop 2529 14.6 0.818 0.854 51.046 52.187 ot_mini 118 10.6 0.001 0.001 38.381 38.489 init_scf_loop 11 6.9 0.001 0.001 36.017 36.019 mp_waitall_1 126876 16.7 25.399 32.201 25.399 32.201 prepare_preconditioner 11 7.9 0.000 0.000 32.141 32.167 make_preconditioner 11 8.9 0.000 0.000 32.141 32.167 make_full_inverse_cholesky 11 9.9 0.015 0.025 30.119 30.390 rebuild_ks_matrix 129 8.3 0.001 0.001 29.295 29.470 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.295 29.470 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.750 26.910 multiply_cannon_multrec 10116 15.6 10.438 16.200 18.017 22.628 qs_ot_get_derivative 118 11.6 0.001 0.002 20.870 20.992 multiply_cannon_metrocomm3 10116 15.6 0.026 0.027 13.159 20.847 cp_fm_cholesky_invert 11 10.9 18.524 18.530 18.524 18.530 qs_ot_get_p 129 10.4 0.001 0.001 17.689 17.847 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.526 17.696 apply_single 129 13.6 0.001 0.001 17.526 17.696 ot_diis_step 118 11.6 0.020 0.021 17.438 17.439 make_m2s 5058 13.6 0.066 0.070 14.975 15.926 make_images 5058 14.6 2.200 2.641 14.665 15.607 sum_up_and_integrate 129 10.3 0.002 0.002 14.298 14.340 integrate_v_rspace 129 11.3 0.004 0.004 14.237 14.280 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.194 14.225 calculate_rho_elec 129 8.7 0.257 0.268 14.194 14.224 qs_ot_p2m_diag 84 11.4 0.501 0.507 13.844 13.861 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.643 12.644 multiply_cannon_sync_h2d 10116 15.6 10.784 11.263 10.784 11.263 init_scf_run 11 5.9 0.000 0.001 10.435 10.435 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.435 10.435 make_images_data 5058 15.6 0.056 0.065 8.598 10.047 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.699 9.710 cp_fm_diag_elpa_base 84 14.4 9.456 9.542 9.695 9.706 hybrid_alltoall_any 5245 16.5 0.849 3.797 8.409 9.556 pw_transfer 1559 11.6 0.086 0.094 9.414 9.438 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.190 9.219 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.278 8.370 grid_integrate_task_list 129 12.3 7.825 8.181 7.825 8.181 fft_wrap_pw1pw2_140 527 13.2 1.823 1.850 8.121 8.155 cp_fm_cholesky_decompose 22 10.9 8.004 8.155 8.004 8.155 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.397 7.961 dbcsr_mm_accdrv_process 20934 16.1 3.414 4.337 7.227 7.878 wfi_extrapolate 11 7.9 0.001 0.001 7.409 7.409 density_rs2pw 129 9.7 0.006 0.006 6.816 7.120 calculate_dm_sparse 129 9.5 0.001 0.001 6.166 6.248 fft3d_ps 1301 14.7 2.795 2.873 5.886 5.927 multiply_cannon_metrocomm4 7587 15.6 0.027 0.031 1.890 5.911 grid_collocate_task_list 129 9.7 5.617 5.893 5.617 5.893 mp_irecv_dv 29102 15.9 1.848 5.817 1.848 5.817 dbcsr_complete_redistribute 397 12.7 2.117 2.179 5.193 5.472 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.273 5.278 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.154 5.178 mp_alltoall_d11v 2429 14.1 4.720 5.152 4.720 5.152 mp_allgather_i34 2529 14.6 1.351 4.777 1.351 4.777 potential_pw2rs 129 12.3 0.026 0.027 4.332 4.339 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.043 4.094 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.579 3.920 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.706 3.764 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.635 3.716 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.331 3.645 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.477 3.633 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.452 3.467 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.587000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1443.363636, yerr=15.709133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.235643E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58530. MP_Allreduce 11189 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.056 292.210 292.211 qs_mol_dyn_low 1 2.0 0.003 0.004 291.661 291.677 qs_forces 11 3.9 0.005 0.005 291.565 291.566 qs_energies 11 4.9 0.002 0.002 282.740 282.748 scf_env_do_scf 11 5.9 0.001 0.001 260.108 260.117 velocity_verlet 10 3.0 0.002 0.002 211.653 211.661 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 135.654 135.655 init_scf_loop 11 6.9 0.001 0.001 124.178 124.180 prepare_preconditioner 11 7.9 0.000 0.000 119.346 119.366 make_preconditioner 11 8.9 0.000 0.000 119.345 119.366 make_full_inverse_cholesky 11 9.9 0.036 0.039 95.542 116.490 qs_scf_new_mos 118 7.6 0.001 0.001 91.752 91.860 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.751 91.859 ot_scf_mini 118 9.6 0.004 0.004 86.791 86.838 dbcsr_multiply_generic 2529 12.6 0.215 0.228 83.309 83.920 cp_fm_upper_to_full 106 14.8 52.605 75.366 52.605 75.366 multiply_cannon 2529 13.6 0.685 0.738 59.609 60.907 multiply_cannon_loop 2529 14.6 1.048 1.063 55.498 57.013 ot_mini 118 10.6 0.001 0.001 45.302 45.379 dbcsr_complete_redistribute 397 12.7 4.012 4.053 29.879 42.714 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.285 39.045 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.760 36.260 mp_alltoall_i22 720 14.1 21.524 34.388 21.524 34.388 rebuild_ks_matrix 129 8.3 0.001 0.001 34.282 34.335 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 34.281 34.334 cp_fm_cholesky_invert 11 10.9 33.751 33.758 33.751 33.758 mp_waitall_1 104580 16.8 28.621 32.770 28.621 32.770 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.817 31.879 qs_ot_get_p 129 10.4 0.001 0.001 26.380 26.514 qs_ot_get_derivative 118 11.6 0.002 0.002 25.033 25.080 qs_ot_p2m_diag 84 11.4 0.890 0.896 22.219 22.251 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 19.104 20.578 make_m2s 5058 13.6 0.076 0.079 18.994 20.514 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.382 20.383 ot_diis_step 118 11.6 0.022 0.023 20.227 20.228 make_images 5058 14.6 3.099 3.314 18.509 20.026 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.520 19.687 apply_single 129 13.6 0.001 0.001 19.519 19.686 multiply_cannon_multrec 10116 15.6 10.417 12.268 18.320 18.386 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.134 17.135 cp_fm_diag_elpa_base 84 14.4 12.690 14.304 17.130 17.130 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.703 16.719 calculate_rho_elec 129 8.7 0.484 0.485 16.702 16.718 sum_up_and_integrate 129 10.3 0.002 0.002 16.172 16.261 integrate_v_rspace 129 11.3 0.004 0.004 16.109 16.197 multiply_cannon_sync_h2d 10116 15.6 14.427 14.457 14.427 14.457 hybrid_alltoall_any 5245 16.5 1.319 3.071 10.959 13.114 make_images_data 5058 15.6 0.065 0.069 10.796 13.027 init_scf_run 11 5.9 0.000 0.001 12.306 12.308 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.306 12.308 pw_transfer 1559 11.6 0.096 0.098 11.532 11.540 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.292 11.300 fft_wrap_pw1pw2_140 527 13.2 3.084 3.148 10.058 10.082 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.949 9.987 dbcsr_mm_accdrv_process 20934 16.1 4.224 6.044 7.651 9.560 cp_fm_cholesky_decompose 22 10.9 9.076 9.161 9.076 9.161 wfi_extrapolate 11 7.9 0.001 0.001 9.081 9.081 grid_integrate_task_list 129 12.3 8.617 8.811 8.617 8.811 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.126 8.126 density_rs2pw 129 9.7 0.005 0.006 7.613 7.664 mp_alltoall_d11v 2429 14.1 7.057 7.178 7.057 7.178 calculate_dm_sparse 129 9.5 0.001 0.001 6.928 6.967 grid_collocate_task_list 129 9.7 6.468 6.541 6.468 6.541 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.439 6.524 fft3d_ps 1301 14.7 2.879 2.919 6.485 6.523 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.277 6.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.211000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2960.454545, yerr=153.307987 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 1.260028E+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 29859882. 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 0.012 0.033 84.882 84.883 qs_energies 1 2.0 0.000 0.000 84.419 84.433 ls_scf 1 3.0 0.000 0.000 83.520 83.535 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.419 72.561 multiply_cannon 111 7.7 0.018 0.020 55.698 56.974 multiply_cannon_loop 111 8.7 0.227 0.245 52.332 53.734 ls_scf_main 1 4.0 0.000 0.000 52.258 52.269 density_matrix_trs4 2 5.0 0.002 0.003 46.682 46.759 ls_scf_init_scf 1 4.0 0.000 0.000 28.175 28.177 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.031 27.089 mp_waitall_1 11031 10.9 22.361 25.141 22.361 25.141 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.954 24.969 multiply_cannon_multrec 2664 9.7 8.150 8.910 15.512 17.243 multiply_cannon_sync_h2d 2664 9.7 13.483 15.303 13.483 15.303 make_m2s 222 7.7 0.009 0.011 13.017 13.438 make_images 222 8.7 0.098 0.108 12.995 13.414 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.654 12.453 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.457 8.323 make_images_data 222 9.7 0.004 0.005 7.588 8.118 dbcsr_mm_accdrv_process 4760 10.4 0.583 0.691 6.978 7.978 hybrid_alltoall_any 227 10.6 0.216 1.851 6.551 7.662 dbcsr_mm_accdrv_process_sort 4760 11.4 6.196 7.121 6.196 7.121 calculate_norms 4752 9.8 5.512 6.275 5.512 6.275 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.047 5.273 mp_sum_l 887 5.1 3.162 4.786 3.162 4.786 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.377 3.734 make_images_sizes 222 9.7 0.000 0.000 0.657 3.557 mp_alltoall_i44 222 10.7 0.657 3.557 0.657 3.557 arnoldi_extremal 4 6.8 0.000 0.000 3.286 3.313 arnoldi_normal_ev 4 7.8 0.001 0.002 3.286 3.313 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.063 3.252 mp_irecv_dv 6231 10.9 2.046 3.222 2.046 3.222 build_subspace 16 8.4 0.009 0.012 3.183 3.187 ls_scf_post 1 4.0 0.000 0.000 3.087 3.101 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.911 dbcsr_special_finalize 555 9.7 0.005 0.006 2.419 2.740 dbcsr_merge_single_wm 555 10.7 0.461 0.598 2.410 2.731 make_images_pack 222 9.7 2.207 2.635 2.209 2.637 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.335 2.581 dbcsr_sort_data 658 11.4 2.209 2.477 2.209 2.477 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.460 2.066 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.328 2.416 buffer_matrices_ensure_size 222 8.7 1.750 2.061 1.750 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.795 1.796 rebuild_ks_matrix 3 7.3 0.000 0.000 1.785 1.787 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.021 1.785 1.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.883000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 2.172568E+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 0.025 0.046 90.459 90.461 qs_energies 1 2.0 0.000 0.000 89.985 89.989 ls_scf 1 3.0 0.000 0.000 88.638 88.641 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.686 75.038 multiply_cannon 111 7.7 0.027 0.041 52.719 57.880 ls_scf_main 1 4.0 0.000 0.000 54.939 54.946 multiply_cannon_loop 111 8.7 0.136 0.148 49.914 53.593 density_matrix_trs4 2 5.0 0.002 0.003 49.190 49.418 mp_waitall_1 9105 10.9 20.963 30.081 20.963 30.081 ls_scf_init_scf 1 4.0 0.000 0.000 30.021 30.023 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.816 28.923 multiply_cannon_multrec 1332 9.7 13.265 17.148 22.520 27.476 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.460 26.468 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.567 20.868 make_m2s 222 7.7 0.006 0.007 15.226 15.772 make_images 222 8.7 1.368 1.704 15.195 15.742 dbcsr_mm_accdrv_process 4041 10.4 0.311 0.531 8.853 10.399 dbcsr_mm_accdrv_process_sort 4041 11.4 8.368 9.869 8.368 9.869 make_images_data 222 9.7 0.004 0.004 8.861 9.735 hybrid_alltoall_any 227 10.6 0.543 2.580 8.148 9.694 mp_sum_l 887 5.1 5.259 8.390 5.259 8.390 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.185 7.734 mp_irecv_dv 3311 11.0 3.165 7.679 3.165 7.679 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.019 6.711 calculate_norms 2376 9.8 6.071 6.648 6.071 6.648 multiply_cannon_sync_h2d 1332 9.7 4.884 6.308 4.884 6.308 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.965 5.187 arnoldi_extremal 4 6.8 0.000 0.000 4.690 4.714 arnoldi_normal_ev 4 7.8 0.001 0.005 4.689 4.714 build_subspace 16 8.4 0.014 0.021 4.440 4.443 ls_scf_post 1 4.0 0.000 0.000 3.677 3.682 ls_scf_store_result 1 5.0 0.000 0.000 3.388 3.502 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.180 3.404 dbcsr_matrix_vector_mult_local 304 10.0 2.760 3.256 2.762 3.257 mp_allgather_i34 111 8.7 0.791 3.082 0.791 3.082 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.180 2.727 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.630 2.712 dbcsr_data_new 4174 10.1 2.116 2.408 2.116 2.408 make_images_pack 222 9.7 1.820 2.135 1.822 2.137 dbcsr_sort_data 436 11.2 1.851 2.098 1.851 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.875 1.877 rebuild_ks_matrix 3 7.3 0.000 0.000 1.862 1.864 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.862 1.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.461000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 2.874262E+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 0.029 0.046 92.083 92.085 qs_energies 1 2.0 0.000 0.000 91.531 91.535 ls_scf 1 3.0 0.000 0.000 90.116 90.120 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.702 75.028 ls_scf_main 1 4.0 0.000 0.000 56.010 56.015 multiply_cannon 111 7.7 0.030 0.066 51.770 55.696 multiply_cannon_loop 111 8.7 0.117 0.130 49.048 53.143 density_matrix_trs4 2 5.0 0.002 0.003 50.079 50.230 mp_waitall_1 7281 11.0 23.386 33.449 23.386 33.449 ls_scf_init_scf 1 4.0 0.000 0.000 30.403 30.406 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.193 29.274 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.840 26.854 multiply_cannon_multrec 888 9.7 12.603 15.478 21.143 24.649 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.006 22.834 make_m2s 222 7.7 0.006 0.007 16.298 17.045 make_images 222 8.7 1.584 1.857 16.260 17.005 make_images_data 222 9.7 0.004 0.004 9.493 10.583 hybrid_alltoall_any 227 10.6 0.639 2.933 9.109 10.335 dbcsr_mm_accdrv_process 3754 10.4 0.298 0.491 8.074 9.323 dbcsr_mm_accdrv_process_sort 3754 11.4 7.640 8.833 7.640 8.833 mp_sum_l 887 5.1 4.958 8.139 4.958 8.139 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.491 7.519 mp_irecv_dv 2335 11.1 2.475 7.458 2.475 7.458 multiply_cannon_sync_h2d 888 9.7 6.077 7.006 6.077 7.006 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.528 6.808 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.768 6.402 arnoldi_extremal 4 6.8 0.000 0.000 5.149 5.168 arnoldi_normal_ev 4 7.8 0.001 0.005 5.149 5.167 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.753 4.997 build_subspace 16 8.4 0.014 0.020 4.843 4.849 calculate_norms 1584 9.8 4.337 4.761 4.337 4.761 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.474 3.801 ls_scf_post 1 4.0 0.000 0.000 3.703 3.707 mp_allgather_i34 111 8.7 0.845 3.689 0.845 3.689 dbcsr_matrix_vector_mult_local 304 10.0 3.040 3.629 3.043 3.631 ls_scf_store_result 1 5.0 0.000 0.000 3.442 3.531 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.869 2.982 dbcsr_data_new 4116 9.9 2.108 2.460 2.108 2.460 dbcsr_sort_data 325 11.1 1.880 2.094 1.880 2.094 make_images_sizes 222 9.7 0.000 0.000 0.972 2.035 mp_alltoall_i44 222 10.7 0.971 2.035 0.971 2.035 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.898 1.899 make_images_pack 222 9.7 1.621 1.884 1.624 1.887 dbcsr_finalize 304 7.8 0.026 0.032 1.603 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.879 1.881 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.879 1.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.085000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2246.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 3.327685E+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 0.028 0.044 97.266 97.267 qs_energies 1 2.0 0.000 0.000 96.672 96.700 ls_scf 1 3.0 0.000 0.000 95.018 95.048 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.688 78.941 ls_scf_main 1 4.0 0.000 0.000 59.131 59.133 multiply_cannon 111 7.7 0.047 0.102 51.747 56.001 density_matrix_trs4 2 5.0 0.002 0.005 52.994 53.117 multiply_cannon_loop 111 8.7 0.153 0.169 46.652 49.550 ls_scf_init_scf 1 4.0 0.000 0.000 32.628 32.631 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.413 31.479 mp_waitall_1 6369 11.0 22.864 29.069 22.864 29.069 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.976 28.991 multiply_cannon_multrec 1332 9.7 14.139 17.294 22.010 24.636 make_m2s 222 7.7 0.006 0.008 21.130 22.615 make_images 222 8.7 3.142 3.608 21.080 22.568 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.376 17.647 make_images_data 222 9.7 0.004 0.004 11.822 13.464 hybrid_alltoall_any 227 10.6 0.797 3.762 11.173 12.994 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.475 7.518 9.045 dbcsr_mm_accdrv_process_sort 3641 11.4 7.079 8.546 7.079 8.546 mp_sum_l 887 5.1 4.209 8.024 4.209 8.024 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.248 6.310 multiply_cannon_sync_h2d 1332 9.7 5.424 6.052 5.424 6.052 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.080 6.001 mp_irecv_dv 3229 10.9 2.056 5.932 2.056 5.932 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.497 5.403 arnoldi_extremal 4 6.8 0.000 0.000 5.300 5.323 arnoldi_normal_ev 4 7.8 0.001 0.005 5.300 5.323 build_subspace 16 8.4 0.014 0.021 4.966 4.972 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.576 4.757 mp_allgather_i34 111 8.7 2.210 4.727 2.210 4.727 calculate_norms 2376 9.8 4.161 4.426 4.161 4.426 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.630 3.963 dbcsr_matrix_vector_mult_local 304 10.0 3.223 3.732 3.225 3.734 dbcsr_sort_data 658 11.4 3.104 3.421 3.104 3.421 ls_scf_post 1 4.0 0.000 0.000 3.259 3.287 dbcsr_special_finalize 555 9.7 0.006 0.007 2.842 3.180 dbcsr_merge_single_wm 555 10.7 0.538 0.652 2.834 3.172 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.061 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.015 3.055 dbcsr_data_release 10477 10.7 1.572 2.398 1.572 2.398 dbcsr_finalize 304 7.8 0.049 0.061 1.811 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.267000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 4.760441E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.047 92.415 92.416 qs_energies 1 2.0 0.000 0.000 91.688 91.693 ls_scf 1 3.0 0.000 0.000 89.751 89.755 dbcsr_multiply_generic 111 6.7 0.018 0.019 71.363 71.579 ls_scf_main 1 4.0 0.000 0.000 57.047 57.047 multiply_cannon 111 7.7 0.088 0.202 52.883 55.942 multiply_cannon_loop 111 8.7 0.088 0.093 50.242 51.850 density_matrix_trs4 2 5.0 0.002 0.003 50.094 50.184 ls_scf_init_scf 1 4.0 0.000 0.000 29.286 29.287 mp_waitall_1 5436 11.0 25.003 29.104 25.003 29.104 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.049 28.080 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.950 25.968 multiply_cannon_multrec 444 9.7 13.629 16.211 20.684 21.959 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.217 15.829 make_m2s 222 7.7 0.005 0.005 13.723 14.708 make_images 222 8.7 2.041 2.478 13.655 14.638 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.099 13.847 hybrid_alltoall_any 227 10.6 0.801 3.817 8.279 9.953 make_images_data 222 9.7 0.003 0.004 8.505 9.886 multiply_cannon_sync_h2d 444 9.7 6.749 8.420 6.749 8.420 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.402 6.750 7.851 dbcsr_mm_accdrv_process_sort 3003 11.4 6.386 7.449 6.386 7.449 arnoldi_extremal 4 6.8 0.000 0.000 5.830 5.847 arnoldi_normal_ev 4 7.8 0.001 0.005 5.830 5.847 build_subspace 16 8.4 0.015 0.020 5.427 5.436 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.454 4.683 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.171 4.386 mp_sum_l 887 5.1 2.761 4.279 2.761 4.279 dbcsr_matrix_vector_mult_local 304 10.0 3.723 4.180 3.725 4.182 mp_allgather_i34 111 8.7 1.215 3.880 1.215 3.880 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.548 3.769 mp_irecv_dv 1241 11.2 1.532 3.720 1.532 3.720 calculate_norms 792 9.8 3.604 3.703 3.604 3.703 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.513 3.575 ls_scf_post 1 4.0 0.000 0.000 3.418 3.423 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.949 3.324 make_images_sizes 222 9.7 0.000 0.000 0.837 3.314 mp_alltoall_i44 222 10.7 0.837 3.314 0.837 3.314 ls_scf_store_result 1 5.0 0.000 0.000 3.209 3.243 dbcsr_finalize 304 7.8 0.062 0.076 2.193 2.263 dbcsr_data_new 4608 9.7 1.775 2.155 1.775 2.155 dbcsr_merge_all 275 8.9 0.478 0.521 2.051 2.110 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.010 2.011 rebuild_ks_matrix 3 7.3 0.000 0.000 1.978 1.978 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.978 1.978 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.921 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.416000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 8.878617E+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 0.072 0.091 105.180 105.182 qs_energies 1 2.0 0.000 0.000 103.872 103.878 ls_scf 1 3.0 0.000 0.000 100.939 100.945 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.637 74.774 ls_scf_main 1 4.0 0.000 0.000 63.537 63.538 density_matrix_trs4 2 5.0 0.002 0.003 54.555 54.607 multiply_cannon 111 7.7 0.122 0.235 48.474 50.556 multiply_cannon_loop 111 8.7 0.098 0.102 45.434 46.028 ls_scf_init_scf 1 4.0 0.000 0.000 33.624 33.628 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.124 32.148 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.410 29.418 mp_waitall_1 4527 11.1 21.605 25.518 21.605 25.518 make_m2s 222 7.7 0.005 0.005 22.480 23.592 make_images 222 8.7 3.583 3.888 22.372 23.483 multiply_cannon_multrec 444 9.7 17.894 18.429 22.532 23.096 hybrid_alltoall_any 227 10.6 1.656 3.609 12.598 15.429 make_images_data 222 9.7 0.003 0.004 12.841 15.017 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.176 10.575 multiply_cannon_sync_h2d 444 9.7 8.793 8.850 8.793 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.442 7.452 arnoldi_normal_ev 4 7.8 0.002 0.008 7.442 7.452 build_subspace 16 8.4 0.026 0.036 6.894 6.908 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.531 5.664 dbcsr_matrix_vector_mult_local 304 10.0 5.099 5.361 5.101 5.363 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.086 5.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.915 5.162 dbcsr_mm_accdrv_process 1814 10.4 0.314 0.357 4.451 4.589 dbcsr_mm_accdrv_process_sort 1814 11.4 4.113 4.254 4.113 4.254 ls_scf_post 1 4.0 0.000 0.000 3.778 3.785 make_images_sizes 222 9.7 0.000 0.000 1.501 3.712 mp_alltoall_i44 222 10.7 1.501 3.711 1.501 3.711 mp_allgather_i34 111 8.7 1.131 3.563 1.131 3.563 ls_scf_store_result 1 5.0 0.000 0.000 3.501 3.536 calculate_norms 792 9.8 3.227 3.267 3.227 3.267 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.182 dbcsr_merge_all 275 8.9 0.884 0.914 2.868 2.962 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.900 2.900 dbcsr_complete_redistribute 5 7.6 1.440 1.498 2.745 2.868 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.524 dbcsr_sort_data 325 11.1 2.445 2.512 2.445 2.512 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.387 2.393 dbcsr_data_new 6591 9.6 1.889 2.386 1.889 2.386 rebuild_ks_matrix 3 7.3 0.000 0.000 2.321 2.327 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 2.321 2.327 dbcsr_new_transposed 4 7.5 0.241 0.252 2.293 2.309 dbcsr_frobenius_norm 74 6.6 2.056 2.139 2.203 2.234 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.201 dbcsr_add_anytype 103 7.2 0.859 0.891 2.124 2.200 dbcsr_data_release 12724 10.6 1.976 2.176 1.976 2.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.182000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7009.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/14576fcd0de922a1947a6e9f68f87a45f7ee4c54_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 589.877248E+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 1170800662. 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 0.039 0.113 232.735 232.736 qs_mol_dyn_low 1 2.0 0.004 0.004 231.405 231.429 qs_forces 5 3.8 0.006 0.007 231.336 231.338 qs_energies 5 4.8 0.004 0.065 228.316 228.349 scf_env_do_scf 5 5.8 0.000 0.002 213.112 213.114 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 185.753 185.754 qs_scf_new_mos 105 7.6 0.000 0.001 144.804 144.953 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.804 144.953 ot_scf_mini 105 9.6 0.003 0.004 134.934 135.067 dbcsr_multiply_generic 1445 12.2 0.126 0.134 134.580 135.004 multiply_cannon 1445 13.2 0.276 0.288 115.395 117.475 multiply_cannon_loop 1445 14.2 2.842 2.980 113.658 115.422 velocity_verlet 4 3.0 0.004 0.017 107.746 107.746 ot_mini 105 10.6 0.001 0.001 60.533 60.652 multiply_cannon_multrec 69360 15.2 29.688 34.740 39.542 44.682 qs_ot_get_p 112 10.4 0.001 0.001 44.280 44.532 mp_waitall_1 488190 16.1 35.212 43.480 35.212 43.480 qs_ot_get_derivative 55 11.6 0.001 0.001 38.813 38.950 multiply_cannon_metrocomm3 69360 15.2 0.202 0.216 25.968 34.465 qs_ot_p2m_diag 40 11.0 0.020 0.030 33.233 33.337 multiply_cannon_sync_h2d 69360 15.2 29.202 33.073 29.202 33.073 cp_dbcsr_syevd 40 12.0 0.002 0.002 29.919 29.919 rebuild_ks_matrix 110 8.4 0.000 0.000 29.593 29.769 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 29.593 29.769 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.195 27.356 init_scf_loop 7 6.6 0.000 0.005 27.327 27.328 cp_fm_syevd 40 13.0 0.000 0.001 24.772 24.902 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.002 23.293 apply_single 62 13.6 0.000 0.000 23.002 23.293 prepare_preconditioner 7 7.6 0.000 0.000 22.469 22.503 make_preconditioner 7 8.6 0.000 0.003 22.468 22.503 ot_new_cg_direction 55 11.6 0.001 0.001 21.018 21.018 cp_fm_redistribute_end 40 14.0 10.029 20.014 10.034 20.015 cp_fm_syevd_base 40 14.0 9.974 19.962 9.974 19.962 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.173 17.519 calculate_rho_elec 110 8.6 0.030 0.032 17.173 17.518 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.094 15.172 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.914 15.054 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.995 14.070 mp_sum_l 4764 12.2 12.239 13.220 12.239 13.220 init_scf_run 5 5.8 0.000 0.001 12.586 12.586 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.585 12.586 pw_transfer 1645 12.4 0.079 0.095 12.370 12.565 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.232 12.426 density_rs2pw 110 9.6 0.005 0.007 11.479 11.859 calculate_dm_sparse 110 9.5 0.000 0.001 11.484 11.688 fft_wrap_pw1pw2_240 915 15.0 1.150 1.222 10.729 10.920 qs_vxc_create 110 10.4 0.002 0.009 10.769 10.804 cp_fm_cholesky_invert 7 10.6 10.556 10.564 10.556 10.564 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.474 10.548 dbcsr_mm_accdrv_process 154766 15.8 6.175 6.369 9.722 10.510 check_diag 80 13.5 8.686 8.950 9.512 9.666 fft3d_pb 915 16.0 2.377 2.563 8.934 9.125 sum_up_and_integrate 60 10.3 0.001 0.002 8.756 8.767 integrate_v_rspace 60 11.3 0.001 0.002 8.738 8.749 acc_transpose_blocks 69360 15.2 0.355 0.372 8.291 8.723 transfer_rs2pw 445 10.6 0.007 0.009 7.987 8.420 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.140 8.156 xc_rho_set_and_dset_create 110 12.4 0.078 0.106 7.524 7.772 calculate_first_density_matrix 1 7.0 0.000 0.003 7.700 7.716 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.665 7.715 xc_vxc_pw_create 60 11.3 0.039 0.051 7.264 7.298 multiply_cannon_metrocomm1 69360 15.2 0.099 0.106 4.719 7.294 make_full_single_inverse 7 9.6 0.001 0.004 7.084 7.119 make_m2s 2890 13.2 0.079 0.087 6.490 7.064 make_images 2890 14.2 0.239 0.259 6.383 6.955 xc_pw_derive 510 13.4 0.005 0.007 6.530 6.603 mp_alltoall_z22v 2340 17.7 5.979 6.305 5.979 6.305 acc_transpose_blocks_kernels 69360 16.2 0.847 0.892 5.535 5.860 multiply_cannon_metrocomm4 67915 15.2 0.186 0.205 2.018 4.995 jit_kernel_transpose 5 15.0 4.688 4.990 4.688 4.990 potential_pw2rs 60 12.3 0.003 0.003 4.933 4.965 mp_waitany 7680 13.5 4.378 4.918 4.378 4.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=232.736000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.400000, yerr=2.059126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 14576fcd0de922a1947a6e9f68f87a45f7ee4c54 Summary: empty Status: OK