=== This is the CP2K Performance-Test === Updating 178857898..185b7b66b Fast-forward tools/dashboard/README.md | 115 ++++++++++++++++++++++++++++++++++ tools/dashboard/generate_dashboard.py | 2 +- tools/dashboard/screenshot.png | Bin 0 -> 146629 bytes tools/dashboard/status_failed.png | Bin 0 -> 1374 bytes tools/dashboard/status_failed_old.png | Bin 0 -> 1470 bytes tools/dashboard/status_ok.png | Bin 0 -> 1069 bytes tools/dashboard/status_ok_old.png | Bin 0 -> 1314 bytes tools/dashboard/status_outdated.png | Bin 0 -> 1643 bytes tools/dashboard/status_unknown.png | Bin 0 -> 1695 bytes 9 files changed, 116 insertions(+), 1 deletion(-) create mode 100644 tools/dashboard/README.md create mode 100644 tools/dashboard/screenshot.png create mode 100644 tools/dashboard/status_failed.png create mode 100644 tools/dashboard/status_failed_old.png create mode 100644 tools/dashboard/status_ok.png create mode 100644 tools/dashboard/status_ok_old.png create mode 100644 tools/dashboard/status_outdated.png create mode 100644 tools/dashboard/status_unknown.png Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 185b7b66b50ae386199c6584a49c01427632d034 ################# 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 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.8.2, 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: 28.08.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/01 job id: 48744261 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/02 job id: 48744262 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/03 job id: 48744263 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/04 job id: 48744264 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/05 job id: 48744265 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/06 job id: 48744266 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/07 job id: 48744268 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/08 job id: 48744269 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/09 job id: 48744270 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/10 job id: 48744271 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/11 job id: 48744272 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/12 job id: 48744273 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/13 job id: 48744275 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/14 job id: 48744276 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/15 job id: 48744277 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/16 job id: 48744278 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/17 job id: 48744279 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/18 job id: 48744280 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/19 job id: 48744281 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/20 job id: 48744283 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/21 job id: 48744284 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/22 job id: 48744285 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/23 job id: 48744286 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/24 job id: 48744287 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/25 job id: 48744288 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/26 job id: 48744289 --- 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/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/27 job id: 48744291 --- 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/185b7b66b50ae386199c6584a49c01427632d034_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 0.018 0.033 139.140 139.141 farming_run 1 2.0 138.531 138.533 139.101 139.102 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.487028E+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 0.026 116.269 116.269 qs_energies 1 2.0 0.000 0.000 115.961 115.970 mp2_main 1 3.0 0.008 0.008 113.475 113.483 mp2_gpw_main 1 4.0 0.076 0.083 112.265 112.274 mp2_ri_gpw_compute_in 1 5.0 0.180 0.190 92.809 93.136 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.850 55.177 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.086 46.081 get_2c_integrals 1 6.0 0.008 0.009 37.295 37.779 integrate_v_rspace 273 8.0 0.439 0.454 24.638 29.373 pw_transfer 6555 10.6 0.381 0.393 26.813 27.358 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.452 26.049 grid_integrate_task_list 273 9.0 20.531 25.716 20.531 25.716 fft_wrap_pw1pw2_100 2178 12.4 1.098 1.292 22.975 23.573 compute_2c_integrals 1 7.0 0.020 0.045 19.443 19.443 rpa_ri_compute_en 1 5.0 0.035 0.039 19.290 19.405 compute_2c_integrals_loop_lm 1 8.0 0.008 0.008 18.551 18.674 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.435 18.543 18.666 cp_fm_cholesky_decompose 12 8.2 17.853 18.313 17.853 18.313 cholesky_decomp 1 7.0 0.000 0.000 16.696 17.188 fft3d_s 5443 13.4 16.120 16.402 16.142 16.425 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.308 10.764 15.308 calculate_wavefunction 272 8.0 5.431 5.573 12.358 13.045 rpa_num_int 1 6.0 0.002 0.035 10.754 10.762 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.676 10.715 calc_mat_Q 8 8.0 0.000 0.000 9.490 9.580 contract_S_to_Q 8 9.0 0.000 0.000 8.914 9.001 calc_potential_gpw 544 9.5 0.005 0.006 8.224 8.689 parallel_gemm_fm 14 9.1 0.000 0.000 8.507 8.633 parallel_gemm_fm_cosma 14 10.1 8.507 8.633 8.507 8.633 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.040 8.279 create_integ_mat 1 6.0 0.023 0.027 8.098 8.098 potential_pw2rs 545 10.0 0.107 0.109 7.476 8.084 collocate_single_gaussian 272 10.0 0.040 0.042 7.302 7.507 array2fm 1 7.0 0.000 0.000 6.626 7.092 pw_scatter_s 2720 13.7 4.366 4.667 4.366 4.667 pw_gather_s 2722 13.2 3.436 3.814 3.436 3.814 array2fm_buffer_send 1 8.0 2.898 3.127 2.898 3.127 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.277 1.434 2.234 2.430 pw_poisson_solve 545 10.5 1.114 1.184 2.178 2.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.281554, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 0.027 0.040 426.157 426.159 farming_run 1 2.0 425.089 425.110 426.106 426.110 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.248170E+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 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 0.031 218.256 218.257 qs_energies 1 2.0 0.003 0.020 217.981 217.994 scf_env_do_scf 1 3.0 0.000 0.000 114.552 114.552 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.515 113.523 rebuild_ks_matrix 4 6.0 0.000 0.000 113.514 113.522 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 113.514 113.522 hfx_ks_matrix 4 8.0 0.001 0.001 113.114 113.119 integrate_four_center 4 9.0 0.153 0.464 113.113 113.118 integrate_four_center_main 4 10.0 0.126 0.609 101.879 105.033 integrate_four_center_bin 265 11.0 101.753 105.014 101.753 105.014 mp2_main 1 3.0 0.022 0.048 103.104 103.116 mp2_gpw_main 1 4.0 0.071 0.129 102.123 102.151 init_scf_loop 1 4.0 0.000 0.000 96.415 96.415 mp2_ri_gpw_compute_in 1 5.0 0.072 0.087 74.216 75.583 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.011 54.049 55.422 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 41.618 46.610 integrate_v_rspace 95 8.0 0.398 0.567 28.134 32.974 pw_transfer 2240 10.6 0.146 0.172 29.449 29.965 mp2_ri_gpw_compute_en 1 5.0 0.068 0.141 27.676 29.508 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.476 28.940 ao_to_mo_and_store_B_mult_1 91 7.0 10.756 28.869 10.756 28.869 grid_integrate_task_list 95 9.0 23.504 28.538 23.504 28.538 fft_wrap_pw1pw2_100 730 12.4 1.253 1.428 26.218 26.665 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.912 25.562 25.572 get_2c_integrals 1 6.0 0.004 0.036 20.038 20.088 compute_2c_integrals 1 7.0 0.003 0.007 19.004 19.039 fft3d_s 1823 13.4 18.470 18.885 18.484 18.899 compute_2c_integrals_loop_lm 1 8.0 0.007 0.013 18.541 18.860 mp2_eri_2c_integrate_gpw 1 9.0 1.757 2.063 18.534 18.859 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.134 18.134 mp2_ri_gpw_compute_en_expansio 172 7.0 0.560 0.595 9.115 9.936 calculate_wavefunction 91 8.0 2.012 2.050 9.624 9.858 local_gemm 172 8.0 8.556 9.358 8.556 9.358 potential_pw2rs 186 10.0 0.033 0.034 8.442 9.050 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.531 8.220 8.904 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.082 8.507 calc_potential_gpw 182 9.5 0.002 0.002 7.860 8.146 collocate_single_gaussian 91 10.0 0.037 0.259 7.724 8.055 mp_sendrecv_dm3 2068 8.0 6.244 6.946 6.244 6.946 integrate_four_center_load 4 10.0 0.000 0.000 6.744 6.748 hfx_load_balance 1 11.0 0.000 0.000 6.744 6.748 mp2_ri_gpw_compute_en_ener 172 7.0 6.336 6.425 6.336 6.425 mp_sync 37 10.5 3.229 5.721 3.229 5.721 pw_gather_s 912 13.2 4.489 5.038 4.489 5.038 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.147513, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 452.530176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 60.911 60.912 qs_mol_dyn_low 1 2.0 0.003 0.003 60.317 60.326 qs_forces 11 3.9 0.003 0.004 60.215 60.217 qs_energies 11 4.9 0.001 0.001 58.542 58.571 scf_env_do_scf 11 5.9 0.000 0.001 51.137 51.137 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 48.431 48.432 qs_scf_new_mos 108 7.5 0.000 0.001 35.861 36.151 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.860 36.150 dbcsr_multiply_generic 2286 12.5 0.096 0.101 35.547 36.149 ot_scf_mini 108 9.5 0.002 0.002 34.047 34.247 velocity_verlet 10 3.0 0.001 0.002 29.527 29.528 multiply_cannon 2286 13.5 0.188 0.199 27.268 28.903 multiply_cannon_loop 2286 14.5 1.825 1.973 26.176 27.603 ot_mini 108 10.5 0.001 0.001 19.895 20.160 qs_ot_get_derivative 108 11.5 0.001 0.001 16.823 17.014 mp_waitall_1 245248 16.5 9.195 15.289 9.195 15.289 multiply_cannon_metrocomm3 54864 15.5 0.076 0.086 6.001 13.692 multiply_cannon_multrec 54864 15.5 3.673 5.891 8.035 11.542 qs_ot_get_p 119 10.4 0.001 0.001 9.258 9.530 rebuild_ks_matrix 119 8.3 0.000 0.000 9.165 9.345 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 9.165 9.345 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.087 8.252 mp_sum_l 7287 12.8 5.110 7.097 5.110 7.097 multiply_cannon_sync_h2d 54864 15.5 5.112 6.397 5.112 6.397 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.919 6.386 dbcsr_mm_accdrv_process 76910 16.1 1.794 2.905 4.276 6.217 init_scf_run 11 5.9 0.000 0.001 5.995 5.995 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.995 5.995 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.925 5.976 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.719 5.877 calculate_rho_elec 119 8.7 0.011 0.016 5.719 5.876 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.423 5.532 sum_up_and_integrate 119 10.3 0.001 0.001 5.478 5.489 integrate_v_rspace 119 11.3 0.002 0.002 5.467 5.479 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.136 5.137 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.741 4.743 cp_fm_redistribute_end 50 14.0 2.413 4.700 2.421 4.706 cp_fm_diag_elpa_base 50 14.0 2.270 4.560 2.276 4.574 density_rs2pw 119 9.7 0.004 0.004 3.557 3.675 calculate_dm_sparse 119 9.5 0.000 0.000 3.348 3.519 calculate_first_density_matrix 1 7.0 0.000 0.001 3.480 3.485 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.013 3.243 apply_single 119 13.6 0.000 0.000 3.013 3.243 jit_kernel_multiply 13 15.8 2.419 3.223 2.419 3.223 acc_transpose_blocks 54864 15.5 0.236 0.265 2.501 3.095 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.984 2.987 make_m2s 4572 13.5 0.054 0.056 2.857 2.980 pw_transfer 1439 11.6 0.052 0.058 2.833 2.942 make_images 4572 14.5 0.136 0.142 2.774 2.896 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.758 2.869 ot_diis_step 108 11.5 0.006 0.006 2.821 2.821 multiply_cannon_metrocomm1 54864 15.5 0.059 0.065 1.764 2.799 potential_pw2rs 119 12.3 0.004 0.004 2.724 2.782 init_scf_loop 11 6.9 0.000 0.001 2.668 2.668 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.584 2.659 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.599 2.659 fft3d_ps 1201 14.6 0.375 0.485 2.478 2.585 wfi_extrapolate 11 7.9 0.001 0.001 2.415 2.415 transfer_rs2pw 487 10.6 0.005 0.006 2.164 2.327 fft_wrap_pw1pw2_140 487 13.2 0.135 0.149 2.069 2.179 mp_alltoall_d11v 2130 13.8 1.954 2.148 1.954 2.148 grid_integrate_task_list 119 12.3 2.013 2.128 2.013 2.128 mp_sum_d 4135 12.0 1.385 2.033 1.385 2.033 transfer_pw2rs 487 13.2 0.006 0.006 1.876 1.928 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.740 1.784 make_images_sizes 4572 15.5 0.004 0.005 1.347 1.676 mp_alltoall_i44 4572 16.5 1.343 1.672 1.343 1.672 mp_alltoall_z22v 1201 16.6 1.461 1.533 1.461 1.533 mp_waitany 12084 13.8 1.427 1.533 1.427 1.533 acc_transpose_blocks_sync 164592 16.5 1.181 1.457 1.181 1.457 prepare_preconditioner 11 7.9 0.000 0.000 1.401 1.434 make_preconditioner 11 8.9 0.000 0.000 1.401 1.434 grid_collocate_task_list 119 9.7 1.349 1.427 1.349 1.427 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.327 1.363 acc_transpose_blocks_kernels 54864 16.5 0.247 0.387 1.039 1.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.912000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 488.509440E+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 1220167. 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 0.014 0.051 41.196 41.197 qs_mol_dyn_low 1 2.0 0.003 0.005 40.724 40.962 qs_forces 11 3.9 0.003 0.010 40.570 40.572 qs_energies 11 4.9 0.002 0.004 38.851 38.857 scf_env_do_scf 11 5.9 0.001 0.001 33.705 33.706 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.017 31.017 dbcsr_multiply_generic 2286 12.5 0.104 0.107 23.081 23.468 qs_scf_new_mos 108 7.5 0.001 0.001 22.009 22.238 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.008 22.238 ot_scf_mini 108 9.5 0.003 0.003 21.049 21.213 velocity_verlet 10 3.0 0.005 0.006 19.385 19.386 multiply_cannon 2286 13.5 0.210 0.219 17.418 18.991 multiply_cannon_loop 2286 14.5 1.196 1.264 16.146 17.959 ot_mini 108 10.5 0.001 0.002 13.073 13.301 mp_waitall_1 200699 16.5 6.165 11.354 6.165 11.354 qs_ot_get_derivative 108 11.5 0.001 0.002 10.651 10.813 multiply_cannon_metrocomm3 27432 15.5 0.075 0.078 4.187 9.902 multiply_cannon_multrec 27432 15.5 1.874 4.674 6.181 9.261 rebuild_ks_matrix 119 8.3 0.000 0.000 7.068 7.225 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 7.068 7.225 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.213 6.354 dbcsr_mm_accdrv_process 47894 16.0 3.548 5.988 4.228 6.325 qs_ot_get_p 119 10.4 0.001 0.001 4.968 5.189 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.868 4.780 mp_sum_l 7287 12.8 2.332 4.367 2.332 4.367 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.108 4.096 apply_single 119 13.6 0.000 0.000 3.108 4.095 sum_up_and_integrate 119 10.3 0.005 0.033 3.924 3.955 integrate_v_rspace 119 11.3 0.002 0.003 3.906 3.911 init_scf_run 11 5.9 0.000 0.001 3.907 3.907 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.906 3.907 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.812 3.855 calculate_rho_elec 119 8.7 0.021 0.024 3.811 3.854 make_m2s 4572 13.5 0.052 0.054 2.887 3.205 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.162 3.182 make_images 4572 14.5 0.208 0.247 2.798 3.118 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.744 2.744 init_scf_loop 11 6.9 0.000 0.001 2.665 2.665 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.462 2.546 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.384 2.384 ot_diis_step 108 11.5 0.011 0.015 2.364 2.365 cp_fm_redistribute_end 50 14.0 1.204 2.348 1.209 2.351 calculate_first_density_matrix 1 7.0 0.000 0.001 2.323 2.325 cp_fm_diag_elpa_base 50 14.0 1.107 2.242 1.137 2.292 density_rs2pw 119 9.7 0.004 0.004 2.104 2.223 multiply_cannon_sync_h2d 27432 15.5 1.692 2.196 1.692 2.196 calculate_dm_sparse 119 9.5 0.001 0.001 2.104 2.179 pw_transfer 1439 11.6 0.065 0.068 1.943 1.974 acc_transpose_blocks 27432 15.5 0.118 0.123 1.618 1.959 grid_integrate_task_list 119 12.3 1.839 1.927 1.839 1.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.900 1.903 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.852 1.884 make_images_data 4572 15.5 0.049 0.056 1.325 1.733 prepare_preconditioner 11 7.9 0.000 0.000 1.705 1.731 make_preconditioner 11 8.9 0.000 0.000 1.705 1.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.670 1.711 potential_pw2rs 119 12.3 0.006 0.007 1.664 1.673 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.596 1.655 hybrid_alltoall_any 4725 16.4 0.055 0.117 1.179 1.544 jit_kernel_multiply 8 16.4 0.622 1.531 0.622 1.531 wfi_extrapolate 11 7.9 0.001 0.003 1.516 1.516 fft3d_ps 1201 14.6 0.516 0.568 1.487 1.514 fft_wrap_pw1pw2_140 487 13.2 0.159 0.167 1.414 1.447 transfer_rs2pw 487 10.6 0.005 0.005 1.246 1.396 mp_alltoall_d11v 2130 13.8 1.297 1.393 1.297 1.393 grid_collocate_task_list 119 9.7 1.288 1.349 1.288 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.248 1.293 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.264 1.272 mp_allgather_i34 2286 14.5 0.690 1.126 0.690 1.126 mp_sum_d 4135 12.0 0.599 1.083 0.599 1.083 transfer_pw2rs 487 13.2 0.005 0.005 1.073 1.082 make_images_sizes 4572 15.5 0.005 0.005 0.685 0.970 mp_alltoall_i44 4572 16.5 0.680 0.965 0.680 0.965 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.944 0.945 acc_transpose_blocks_sync 82296 16.5 0.814 0.943 0.814 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.930 multiply_cannon_metrocomm1 27432 15.5 0.034 0.038 0.530 0.917 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.661 0.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.197000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.909091, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 521.183232E+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 0.055 0.082 38.218 38.219 qs_mol_dyn_low 1 2.0 0.003 0.004 37.835 37.844 qs_forces 11 3.9 0.002 0.006 37.769 37.774 qs_energies 11 4.9 0.002 0.007 36.053 36.061 scf_env_do_scf 11 5.9 0.001 0.001 30.737 30.738 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.506 27.508 dbcsr_multiply_generic 2286 12.5 0.098 0.101 18.619 18.713 velocity_verlet 10 3.0 0.002 0.004 18.327 18.328 qs_scf_new_mos 108 7.5 0.001 0.001 17.765 17.783 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.765 17.782 ot_scf_mini 108 9.5 0.002 0.003 16.918 16.931 multiply_cannon 2286 13.5 0.196 0.202 14.190 14.812 multiply_cannon_loop 2286 14.5 0.860 0.909 13.211 13.904 ot_mini 108 10.5 0.001 0.001 10.124 10.138 qs_ot_get_derivative 108 11.5 0.001 0.001 8.548 8.558 multiply_cannon_multrec 18288 15.5 1.871 2.912 7.006 7.323 rebuild_ks_matrix 119 8.3 0.000 0.000 7.273 7.295 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.032 7.272 7.294 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.429 6.449 dbcsr_mm_accdrv_process 38222 16.0 4.951 5.861 5.043 5.925 mp_waitall_1 158411 16.6 4.009 5.151 4.009 5.151 sum_up_and_integrate 119 10.3 0.001 0.001 4.706 4.713 integrate_v_rspace 119 11.3 0.003 0.003 4.693 4.702 qs_ot_get_p 119 10.4 0.001 0.001 4.370 4.394 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.353 4.356 calculate_rho_elec 119 8.7 0.031 0.031 4.352 4.356 init_scf_run 11 5.9 0.000 0.001 4.016 4.016 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.015 4.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.218 3.843 init_scf_loop 11 6.9 0.023 0.049 3.203 3.204 density_rs2pw 119 9.7 0.004 0.004 2.867 2.981 make_m2s 4572 13.5 0.044 0.045 2.742 2.920 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.856 2.865 make_images 4572 14.5 0.195 0.207 2.657 2.834 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.511 2.511 calculate_first_density_matrix 1 7.0 0.000 0.002 2.475 2.477 potential_pw2rs 119 12.3 0.008 0.009 2.438 2.462 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.089 2.417 apply_single 119 13.6 0.000 0.000 2.088 2.417 multiply_cannon_metrocomm3 18288 15.5 0.049 0.050 1.468 2.411 pw_transfer 1439 11.6 0.065 0.069 2.326 2.351 prepare_preconditioner 11 7.9 0.000 0.001 2.273 2.306 make_preconditioner 11 8.9 0.000 0.002 2.273 2.306 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.234 2.259 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.100 2.230 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.141 2.154 cp_fm_diag_elpa_base 50 14.0 2.112 2.124 2.139 2.151 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.015 2.023 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.968 1.970 calculate_dm_sparse 119 9.5 0.000 0.001 1.961 1.969 transfer_rs2pw 487 10.6 0.005 0.005 1.816 1.945 grid_integrate_task_list 119 12.3 1.806 1.925 1.806 1.925 acc_transpose_blocks 18288 15.5 0.082 0.084 1.779 1.853 fft3d_ps 1201 14.6 0.533 0.551 1.805 1.833 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.743 1.749 transfer_pw2rs 487 13.2 0.004 0.004 1.689 1.714 mp_sum_l 7287 12.8 1.305 1.626 1.305 1.626 fft_wrap_pw1pw2_140 487 13.2 0.209 0.216 1.583 1.608 ot_diis_step 108 11.5 0.012 0.012 1.543 1.544 make_images_data 4572 15.5 0.049 0.054 1.197 1.509 wfi_extrapolate 11 7.9 0.001 0.001 1.473 1.473 hybrid_alltoall_any 4725 16.4 0.059 0.118 1.007 1.339 grid_collocate_task_list 119 9.7 1.235 1.316 1.235 1.316 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.240 1.245 multiply_cannon_sync_h2d 18288 15.5 1.034 1.164 1.034 1.164 mp_alltoall_z22v 1201 16.6 1.070 1.133 1.070 1.133 mp_alltoall_d11v 2130 13.8 0.974 1.128 0.974 1.128 make_images_sizes 4572 15.5 0.005 0.005 0.788 1.127 mp_alltoall_i44 4572 16.5 0.783 1.122 0.783 1.122 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.058 1.075 mp_sendrecv_dv 11067 12.7 1.026 1.054 1.026 1.054 cp_fm_cholesky_invert 11 10.9 1.029 1.034 1.029 1.034 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.980 0.985 acc_transpose_blocks_kernels 18288 16.5 0.218 0.228 0.948 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.920 mp_allgather_i34 2286 14.5 0.423 0.846 0.423 0.846 acc_transpose_blocks_sync 54864 16.5 0.731 0.811 0.731 0.811 transfer_pw2rs_50 119 14.3 0.145 0.152 0.707 0.787 dbcsr_complete_redistribute 329 12.2 0.094 0.101 0.667 0.782 transfer_rs2pw_50 119 11.7 0.156 0.159 0.720 0.767 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.219000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 557.088768E+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 0.030 44.531 44.533 qs_mol_dyn_low 1 2.0 0.003 0.004 44.156 44.164 qs_forces 11 3.9 0.009 0.011 43.776 43.779 qs_energies 11 4.9 0.001 0.001 41.905 41.915 scf_env_do_scf 11 5.9 0.000 0.001 35.060 35.062 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.784 30.785 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.837 22.945 velocity_verlet 10 3.0 0.242 0.323 22.320 22.322 qs_scf_new_mos 108 7.5 0.001 0.001 20.853 20.920 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.853 20.920 ot_scf_mini 108 9.5 0.002 0.003 19.769 19.825 multiply_cannon 2286 13.5 0.220 0.229 17.828 18.355 multiply_cannon_loop 2286 14.5 1.557 1.612 16.658 16.999 ot_mini 108 10.5 0.001 0.001 11.847 11.907 multiply_cannon_multrec 27432 15.5 2.610 3.356 9.926 10.251 qs_ot_get_derivative 108 11.5 0.001 0.001 9.910 9.967 dbcsr_mm_accdrv_process 47916 15.9 6.711 8.615 7.214 8.678 rebuild_ks_matrix 119 8.3 0.000 0.000 7.365 7.416 qs_ks_build_kohn_sham_matrix 119 9.3 0.031 0.039 7.365 7.415 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.540 6.585 init_scf_run 11 5.9 0.000 0.001 5.375 5.376 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.375 5.375 qs_ot_get_p 119 10.4 0.001 0.001 4.804 4.883 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.651 4.699 calculate_rho_elec 119 8.7 0.040 0.046 4.651 4.698 sum_up_and_integrate 119 10.3 0.001 0.001 4.393 4.403 integrate_v_rspace 119 11.3 0.003 0.003 4.381 4.392 init_scf_loop 11 6.9 0.033 0.045 4.247 4.248 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.535 4.014 mp_waitall_1 137007 16.6 3.113 3.667 3.113 3.667 calculate_first_density_matrix 1 7.0 0.000 0.000 3.472 3.473 make_m2s 4572 13.5 0.055 0.057 3.268 3.370 prepare_preconditioner 11 7.9 0.000 0.000 3.237 3.297 make_preconditioner 11 8.9 0.000 0.000 3.237 3.297 make_images 4572 14.5 0.276 0.337 3.159 3.258 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.830 3.149 density_rs2pw 119 9.7 0.004 0.004 3.040 3.138 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.972 2.987 acc_transpose_blocks 27432 15.5 0.123 0.127 2.783 2.883 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.722 2.753 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.243 2.700 apply_single 119 13.6 0.000 0.000 2.243 2.700 calculate_dm_sparse 119 9.5 0.000 0.000 2.552 2.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.589 2.592 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.494 2.495 pw_transfer 1439 11.6 0.065 0.069 2.412 2.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.351 2.363 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.320 2.361 potential_pw2rs 119 12.3 0.009 0.010 2.140 2.151 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.138 2.147 cp_fm_diag_elpa_base 50 14.0 2.097 2.117 2.135 2.144 transfer_rs2pw 487 10.6 0.004 0.005 1.900 2.009 mp_sum_l 7287 12.8 1.311 1.955 1.311 1.955 jit_kernel_multiply 6 16.5 0.437 1.919 0.437 1.919 ot_diis_step 108 11.5 0.012 0.012 1.892 1.893 grid_integrate_task_list 119 12.3 1.818 1.880 1.818 1.880 fft3d_ps 1201 14.6 0.556 0.607 1.841 1.870 wfi_extrapolate 11 7.9 0.001 0.001 1.836 1.836 fft_wrap_pw1pw2_140 487 13.2 0.244 0.258 1.767 1.808 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.615 1.625 acc_transpose_blocks_sync 82296 16.5 1.467 1.564 1.467 1.564 make_images_data 4572 15.5 0.049 0.053 1.340 1.564 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.453 1.469 transfer_pw2rs 487 13.2 0.004 0.004 1.424 1.432 hybrid_alltoall_any 4725 16.4 0.067 0.156 1.097 1.371 grid_collocate_task_list 119 9.7 1.251 1.367 1.251 1.367 multiply_cannon_metrocomm3 27432 15.5 0.042 0.043 0.796 1.322 dbcsr_complete_redistribute 329 12.2 0.115 0.141 0.981 1.253 mp_alltoall_d11v 2130 13.8 1.100 1.251 1.100 1.251 mp_alltoall_z22v 1201 16.6 1.167 1.198 1.167 1.198 cp_fm_upper_to_full 72 14.2 0.843 1.198 0.843 1.198 mp_sendrecv_dv 8211 12.7 1.171 1.194 1.171 1.194 acc_transpose_blocks_kernels 27432 16.5 0.273 0.282 1.164 1.182 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.149 1.157 cp_fm_cholesky_invert 11 10.9 1.119 1.123 1.119 1.123 make_images_sizes 4572 15.5 0.005 0.005 0.780 1.102 mp_alltoall_i44 4572 16.5 0.775 1.097 0.775 1.097 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.667 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.916 0.921 jit_kernel_transpose 5 15.6 0.892 0.907 0.892 0.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=44.533000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.727273, yerr=3.466487 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 624.529408E+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 0.010 0.039 35.127 35.128 qs_mol_dyn_low 1 2.0 0.003 0.003 34.925 34.938 qs_forces 11 3.9 0.002 0.003 34.813 34.815 qs_energies 11 4.9 0.001 0.001 32.971 32.977 scf_env_do_scf 11 5.9 0.000 0.001 27.767 27.767 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.422 24.424 velocity_verlet 10 3.0 0.006 0.008 18.212 18.218 dbcsr_multiply_generic 2286 12.5 0.096 0.100 15.510 15.723 qs_scf_new_mos 108 7.5 0.001 0.001 15.118 15.166 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.118 15.165 ot_scf_mini 108 9.5 0.002 0.002 14.254 14.292 multiply_cannon 2286 13.5 0.227 0.236 11.540 12.152 multiply_cannon_loop 2286 14.5 0.641 0.675 10.298 10.824 ot_mini 108 10.5 0.001 0.001 8.017 8.072 rebuild_ks_matrix 119 8.3 0.000 0.000 6.792 6.859 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.792 6.859 qs_ot_get_derivative 108 11.5 0.001 0.001 6.538 6.579 multiply_cannon_multrec 9144 15.5 1.762 1.956 6.126 6.344 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.064 6.123 dbcsr_mm_accdrv_process 12550 15.8 3.629 4.309 4.254 4.352 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.297 4.318 calculate_rho_elec 119 8.7 0.060 0.061 4.297 4.318 sum_up_and_integrate 119 10.3 0.001 0.001 4.274 4.280 integrate_v_rspace 119 11.3 0.003 0.003 4.264 4.270 mp_waitall_1 115863 16.7 3.076 4.092 3.076 4.092 qs_ot_get_p 119 10.4 0.001 0.001 3.988 4.036 init_scf_run 11 5.9 0.000 0.001 3.666 3.666 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.665 3.665 init_scf_loop 11 6.9 0.001 0.001 3.316 3.318 make_m2s 4572 13.5 0.034 0.035 2.765 3.092 make_images 4572 14.5 0.270 0.301 2.676 3.002 density_rs2pw 119 9.7 0.004 0.004 2.568 2.619 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.572 2.576 prepare_preconditioner 11 7.9 0.000 0.000 2.447 2.455 make_preconditioner 11 8.9 0.000 0.000 2.447 2.455 pw_transfer 1439 11.6 0.066 0.068 2.419 2.435 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.311 2.354 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.325 2.344 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.273 2.295 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.271 2.272 calculate_first_density_matrix 1 7.0 0.000 0.000 2.096 2.098 grid_integrate_task_list 119 12.3 1.930 2.013 1.930 2.013 calculate_dm_sparse 119 9.5 0.000 0.000 1.934 1.967 potential_pw2rs 119 12.3 0.010 0.011 1.944 1.955 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.899 1.921 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.880 1.888 cp_fm_diag_elpa_base 50 14.0 1.846 1.868 1.878 1.886 fft_wrap_pw1pw2_140 487 13.2 0.322 0.331 1.789 1.808 make_images_data 4572 15.5 0.043 0.047 1.321 1.798 fft3d_ps 1201 14.6 0.564 0.576 1.739 1.751 hybrid_alltoall_any 4725 16.4 0.066 0.177 1.220 1.689 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.666 1.669 wfi_extrapolate 11 7.9 0.001 0.001 1.503 1.503 transfer_rs2pw 487 10.6 0.004 0.004 1.425 1.488 acc_transpose_blocks 9144 15.5 0.042 0.043 1.425 1.481 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.421 1.446 jit_kernel_multiply 8 15.2 0.586 1.433 0.586 1.433 ot_diis_step 108 11.5 0.013 0.013 1.425 1.426 grid_collocate_task_list 119 9.7 1.346 1.392 1.346 1.392 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.336 1.390 apply_single 119 13.6 0.000 0.000 1.336 1.390 cp_fm_cholesky_invert 11 10.9 1.312 1.317 1.312 1.317 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.438 1.291 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.254 1.256 transfer_pw2rs 487 13.2 0.003 0.003 1.219 1.229 mp_alltoall_d11v 2130 13.8 1.094 1.220 1.094 1.220 mp_allgather_i34 2286 14.5 0.550 1.214 0.550 1.214 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.188 1.198 multiply_cannon_metrocomm1 9144 15.5 0.024 0.025 0.817 1.138 mp_alltoall_z22v 1201 16.6 1.049 1.086 1.049 1.086 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.061 1.085 make_images_sizes 4572 15.5 0.005 0.005 0.657 1.046 mp_alltoall_i44 4572 16.5 0.652 1.041 0.652 1.041 mp_sum_l 7287 12.8 0.760 0.983 0.760 0.983 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.875 0.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.853 0.860 acc_transpose_blocks_sync 27432 16.5 0.713 0.766 0.713 0.766 mp_sendrecv_dv 5355 12.7 0.751 0.760 0.751 0.760 dbcsr_complete_redistribute 329 12.2 0.163 0.172 0.715 0.755 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.700 0.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=35.128000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.000000, yerr=4.898979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 788.307968E+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 0.012 0.031 49.535 49.536 qs_mol_dyn_low 1 2.0 0.003 0.004 49.319 49.327 qs_forces 11 3.9 0.002 0.002 49.260 49.261 qs_energies 11 4.9 0.001 0.001 47.069 47.077 scf_env_do_scf 11 5.9 0.001 0.001 40.274 40.275 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 31.288 31.290 velocity_verlet 10 3.0 0.002 0.002 27.483 27.488 dbcsr_multiply_generic 2286 12.5 0.105 0.106 21.997 22.420 qs_scf_new_mos 108 7.5 0.001 0.001 20.377 20.457 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.376 20.456 ot_scf_mini 108 9.5 0.002 0.002 19.085 19.173 multiply_cannon 2286 13.5 0.303 0.309 16.843 18.264 multiply_cannon_loop 2286 14.5 0.877 0.894 15.316 16.849 ot_mini 108 10.5 0.001 0.001 11.466 11.604 qs_ot_get_derivative 108 11.5 0.001 0.001 9.217 9.303 multiply_cannon_multrec 9144 15.5 3.671 5.097 9.073 9.186 init_scf_loop 11 6.9 0.001 0.001 8.950 8.953 rebuild_ks_matrix 119 8.3 0.000 0.000 8.045 8.209 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 8.045 8.209 prepare_preconditioner 11 7.9 0.000 0.000 7.874 7.890 make_preconditioner 11 8.9 0.000 0.000 7.874 7.890 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.420 7.746 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.184 7.332 dbcsr_mm_accdrv_process 12550 15.8 4.752 6.539 5.264 6.572 mp_waitall_1 94719 16.7 3.421 5.068 3.421 5.068 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.888 4.901 calculate_rho_elec 119 8.7 0.118 0.121 4.887 4.901 init_scf_run 11 5.9 0.000 0.001 4.662 4.662 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.662 4.662 cp_fm_upper_to_full 72 14.2 3.164 4.538 3.164 4.538 sum_up_and_integrate 119 10.3 0.001 0.001 4.517 4.523 integrate_v_rspace 119 11.3 0.003 0.003 4.506 4.513 qs_ot_get_p 119 10.4 0.001 0.001 4.347 4.512 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.178 3.661 make_m2s 4572 13.5 0.038 0.038 3.063 3.338 make_images 4572 14.5 0.354 0.387 2.943 3.218 pw_transfer 1439 11.6 0.068 0.069 3.175 3.191 multiply_cannon_metrocomm3 9144 15.5 0.022 0.023 1.739 3.158 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.077 3.093 dbcsr_complete_redistribute 329 12.2 0.282 0.285 2.151 2.965 calculate_first_density_matrix 1 7.0 0.000 0.000 2.926 2.927 density_rs2pw 119 9.7 0.004 0.004 2.763 2.773 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.714 2.716 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.655 2.659 acc_transpose_blocks 9144 15.5 0.045 0.046 2.637 2.656 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.367 2.569 apply_single 119 13.6 0.000 0.000 2.367 2.569 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.754 2.568 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.469 2.520 mp_sum_l 7287 12.8 1.546 2.518 1.546 2.518 fft_wrap_pw1pw2_140 487 13.2 0.575 0.580 2.485 2.500 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.466 calculate_dm_sparse 119 9.5 0.000 0.000 2.399 2.417 mp_alltoall_i22 627 13.8 1.477 2.364 1.477 2.364 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.256 2.257 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.449 2.247 grid_integrate_task_list 119 12.3 2.173 2.192 2.173 2.192 fft3d_ps 1201 14.6 0.597 0.607 2.174 2.191 ot_diis_step 108 11.5 0.014 0.014 2.173 2.174 cp_fm_cholesky_invert 11 10.9 2.079 2.084 2.079 2.084 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.877 1.877 cp_fm_diag_elpa_base 50 14.0 1.723 1.776 1.875 1.875 potential_pw2rs 119 12.3 0.014 0.014 1.846 1.852 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.788 1.789 make_images_data 4572 15.5 0.047 0.051 1.380 1.774 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.378 1.718 wfi_extrapolate 11 7.9 0.001 0.001 1.663 1.663 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.562 1.612 grid_collocate_task_list 119 9.7 1.587 1.593 1.587 1.593 mp_alltoall_d11v 2130 13.8 1.498 1.528 1.498 1.528 mp_alltoall_z22v 1201 16.6 1.438 1.470 1.438 1.470 acc_transpose_blocks_kernels 9144 16.5 0.121 0.122 1.462 1.467 jit_kernel_transpose 5 15.6 1.342 1.346 1.342 1.346 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.309 transfer_rs2pw 487 10.6 0.005 0.005 1.149 1.170 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.118 1.137 acc_transpose_blocks_sync 27432 16.5 1.113 1.133 1.113 1.133 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.073 1.091 cp_fm_cholesky_decompose 22 10.9 1.031 1.038 1.031 1.038 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.000 1.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=49.536000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=736.727273, yerr=15.920775 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 501.436416E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.025 89.604 89.605 qs_mol_dyn_low 1 2.0 0.003 0.013 88.908 88.918 qs_forces 11 3.9 0.003 0.003 88.840 88.841 qs_energies 11 4.9 0.001 0.002 85.851 85.864 scf_env_do_scf 11 5.9 0.000 0.001 76.604 76.606 scf_env_do_scf_inner_loop 99 6.5 0.048 0.189 70.058 70.059 dbcsr_multiply_generic 2055 12.4 0.110 0.112 52.673 53.025 qs_scf_new_mos 99 7.5 0.000 0.001 50.895 51.092 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.894 51.092 ot_scf_mini 99 9.5 0.002 0.002 48.404 48.590 velocity_verlet 10 3.0 0.001 0.002 46.443 46.444 multiply_cannon 2055 13.4 0.184 0.188 42.945 43.857 multiply_cannon_loop 2055 14.4 1.807 1.843 41.754 42.665 ot_mini 99 10.5 0.001 0.001 27.666 27.829 qs_ot_get_derivative 99 11.5 0.001 0.001 20.804 20.963 multiply_cannon_multrec 49320 15.4 11.281 11.951 17.276 18.160 rebuild_ks_matrix 110 8.3 0.000 0.000 15.722 15.913 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.721 15.913 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.816 13.981 mp_waitall_1 220248 16.4 11.986 13.202 11.986 13.202 qs_ot_get_p 110 10.4 0.001 0.001 12.047 12.185 multiply_cannon_sync_h2d 49320 15.4 9.444 10.323 9.444 10.323 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.943 8.515 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.173 8.206 sum_up_and_integrate 110 10.3 0.001 0.003 7.960 7.972 integrate_v_rspace 110 11.3 0.003 0.003 7.935 7.953 multiply_cannon_metrocomm3 49320 15.4 0.085 0.090 6.593 7.895 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.252 7.721 apply_single 110 13.6 0.000 0.000 7.252 7.721 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.533 7.663 calculate_rho_elec 110 8.6 0.021 0.024 7.532 7.662 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.219 7.220 init_scf_run 11 5.9 0.000 0.001 7.138 7.138 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.138 7.138 ot_diis_step 99 11.5 0.006 0.006 6.649 6.649 init_scf_loop 11 6.9 0.001 0.004 6.511 6.513 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.423 6.445 cp_fm_diag_elpa_base 48 14.0 6.404 6.427 6.421 6.443 dbcsr_mm_accdrv_process 87628 16.1 3.032 3.109 5.866 6.096 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.796 5.847 mp_sum_l 6594 12.7 4.488 5.344 4.488 5.344 make_m2s 4110 13.4 0.061 0.065 4.776 4.907 make_images 4110 14.4 0.180 0.191 4.679 4.814 density_rs2pw 110 9.6 0.004 0.005 4.531 4.703 wfi_extrapolate 11 7.9 0.001 0.001 4.372 4.372 pw_transfer 1331 11.6 0.056 0.067 4.052 4.210 prepare_preconditioner 11 7.9 0.000 0.000 4.143 4.164 make_preconditioner 11 8.9 0.000 0.000 4.143 4.164 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.963 4.125 calculate_dm_sparse 110 9.5 0.001 0.001 3.872 3.963 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.891 3.958 multiply_cannon_metrocomm1 49320 15.4 0.067 0.069 2.725 3.747 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.602 3.608 potential_pw2rs 110 12.3 0.006 0.006 3.492 3.562 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.403 3.463 grid_integrate_task_list 110 12.3 3.253 3.433 3.253 3.433 fft_wrap_pw1pw2_140 451 13.1 0.385 0.425 3.239 3.406 fft3d_ps 1111 14.6 0.792 0.890 3.262 3.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.214 3.276 mp_alltoall_d11v 2046 13.8 2.316 2.817 2.316 2.817 transfer_rs2pw 451 10.6 0.005 0.006 2.590 2.777 calculate_first_density_matrix 1 7.0 0.000 0.000 2.643 2.647 acc_transpose_blocks 49320 15.4 0.222 0.229 2.589 2.645 jit_kernel_multiply 13 15.9 2.556 2.577 2.556 2.577 transfer_pw2rs 451 13.1 0.006 0.007 2.391 2.458 cp_fm_cholesky_invert 11 10.9 2.298 2.304 2.298 2.304 make_images_data 4110 15.4 0.045 0.048 1.990 2.284 mp_alltoall_z22v 1111 16.6 2.065 2.279 2.065 2.279 grid_collocate_task_list 110 9.6 2.160 2.277 2.160 2.277 mp_waitany 14300 13.8 1.906 2.156 1.906 2.156 hybrid_alltoall_any 4261 16.3 0.085 0.482 1.685 2.097 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.036 2.063 mp_sum_d 3889 11.9 1.433 1.890 1.433 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=89.605000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.962576 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 590.307328E+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 3063405. 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 0.010 0.028 79.709 79.710 qs_mol_dyn_low 1 2.0 0.003 0.006 79.108 79.117 qs_forces 11 3.9 0.004 0.011 78.996 78.998 qs_energies 11 4.9 0.002 0.005 75.098 75.114 scf_env_do_scf 11 5.9 0.001 0.001 64.994 64.997 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 55.709 55.709 dbcsr_multiply_generic 2055 12.4 0.122 0.127 41.951 42.583 velocity_verlet 10 3.0 0.001 0.002 40.737 40.745 qs_scf_new_mos 99 7.5 0.001 0.001 38.085 38.259 qs_scf_loop_do_ot 99 8.5 0.001 0.001 38.085 38.258 ot_scf_mini 99 9.5 0.003 0.003 36.240 36.399 multiply_cannon 2055 13.4 0.225 0.246 33.301 35.428 multiply_cannon_loop 2055 14.4 1.175 1.204 31.623 33.693 ot_mini 99 10.5 0.001 0.002 20.783 20.909 multiply_cannon_multrec 24660 15.4 6.906 8.709 14.921 16.686 rebuild_ks_matrix 110 8.3 0.000 0.000 14.924 14.998 qs_ks_build_kohn_sham_matrix 110 9.3 0.064 0.423 14.924 14.997 qs_ot_get_derivative 99 11.5 0.001 0.002 14.816 14.997 mp_waitall_1 176588 16.5 9.059 13.765 9.059 13.765 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.750 12.819 multiply_cannon_metrocomm3 24660 15.4 0.075 0.078 5.719 10.594 init_scf_loop 11 6.9 0.001 0.001 9.242 9.243 qs_ot_get_p 110 10.4 0.001 0.001 8.317 8.452 dbcsr_mm_accdrv_process 52282 16.1 6.188 7.447 7.849 8.126 init_scf_run 11 5.9 0.000 0.001 7.594 7.594 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.594 7.594 multiply_cannon_sync_h2d 24660 15.4 6.323 7.459 6.323 7.459 sum_up_and_integrate 110 10.3 0.002 0.002 7.212 7.224 integrate_v_rspace 110 11.3 0.003 0.004 7.185 7.196 prepare_preconditioner 11 7.9 0.000 0.000 7.111 7.138 make_preconditioner 11 8.9 0.000 0.000 7.110 7.138 qs_rho_update_rho_low 110 7.6 0.001 0.002 7.026 7.061 calculate_rho_elec 110 8.6 0.039 0.047 7.025 7.061 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.591 7.054 apply_single 110 13.6 0.000 0.001 6.590 7.054 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.618 6.807 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.273 6.016 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.889 5.927 ot_diis_step 99 11.5 0.011 0.013 5.807 5.808 make_m2s 4110 13.4 0.057 0.060 5.100 5.572 make_images 4110 14.4 0.410 0.461 4.989 5.457 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.308 5.309 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.532 4.546 cp_fm_diag_elpa_base 48 14.0 4.470 4.495 4.529 4.543 density_rs2pw 110 9.6 0.004 0.005 4.143 4.389 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.340 4.342 pw_transfer 1331 11.6 0.067 0.072 4.139 4.315 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.031 4.211 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.136 4.208 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.945 3.991 wfi_extrapolate 11 7.9 0.001 0.001 3.943 3.943 mp_sum_l 6594 12.7 2.911 3.690 2.911 3.690 cp_fm_cholesky_invert 11 10.9 3.563 3.576 3.563 3.576 calculate_first_density_matrix 1 7.0 0.000 0.002 3.516 3.518 grid_integrate_task_list 110 12.3 3.137 3.331 3.137 3.331 fft3d_ps 1111 14.6 1.114 1.341 3.132 3.327 fft_wrap_pw1pw2_140 451 13.1 0.456 0.474 3.125 3.306 make_images_data 4110 15.4 0.050 0.054 2.695 3.245 calculate_dm_sparse 110 9.5 0.001 0.001 3.177 3.208 hybrid_alltoall_any 4261 16.3 0.108 0.456 2.296 3.207 potential_pw2rs 110 12.3 0.009 0.009 3.003 3.044 jit_kernel_multiply 10 16.2 1.298 2.849 1.298 2.849 mp_alltoall_d11v 2046 13.8 2.206 2.691 2.206 2.691 transfer_rs2pw 451 10.6 0.006 0.007 2.184 2.430 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.342 2.392 grid_collocate_task_list 110 9.6 2.169 2.346 2.169 2.346 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.332 2.341 transfer_pw2rs 451 13.1 0.005 0.006 2.135 2.219 mp_allgather_i34 2055 14.4 0.980 2.124 0.980 2.124 acc_transpose_blocks 24660 15.4 0.119 0.122 2.050 2.076 mp_waitany 10164 13.8 1.694 2.032 1.694 2.032 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.919 1.922 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.768 1.836 cp_fm_cholesky_decompose 22 10.9 1.826 1.834 1.826 1.834 mp_alltoall_z22v 1111 16.6 1.709 1.821 1.709 1.821 dbcsr_complete_redistribute 325 12.2 0.231 0.305 1.393 1.695 mp_irecv_dv 57340 16.2 0.640 1.612 0.640 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.583 1.606 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=79.710000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.454545, yerr=5.805384 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 667.361280E+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 0.169 0.523 68.683 68.684 qs_mol_dyn_low 1 2.0 0.005 0.023 67.460 67.805 qs_forces 11 3.9 0.013 0.077 67.296 67.297 qs_energies 11 4.9 0.007 0.028 63.895 63.899 scf_env_do_scf 11 5.9 0.001 0.001 55.182 55.183 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.807 44.808 velocity_verlet 10 3.0 0.003 0.012 36.585 36.590 dbcsr_multiply_generic 2055 12.4 0.110 0.119 30.844 31.114 qs_scf_new_mos 99 7.5 0.001 0.001 28.644 28.743 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.644 28.742 ot_scf_mini 99 9.5 0.002 0.003 27.321 27.435 multiply_cannon 2055 13.4 0.213 0.224 23.094 24.690 multiply_cannon_loop 2055 14.4 0.822 0.847 21.706 23.152 ot_mini 99 10.5 0.001 0.002 15.372 15.484 rebuild_ks_matrix 110 8.3 0.000 0.000 13.132 13.261 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 13.132 13.261 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.576 11.687 mp_waitall_1 139946 16.5 8.232 11.351 8.232 11.351 qs_ot_get_derivative 99 11.5 0.001 0.002 10.887 11.001 multiply_cannon_multrec 16440 15.4 3.684 4.721 9.939 10.894 init_scf_loop 11 6.9 0.002 0.006 10.329 10.332 prepare_preconditioner 11 7.9 0.000 0.000 8.395 8.410 make_preconditioner 11 8.9 0.001 0.004 8.394 8.410 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.570 7.980 multiply_cannon_metrocomm3 16440 15.4 0.048 0.049 4.441 7.666 sum_up_and_integrate 110 10.3 0.002 0.003 7.085 7.101 integrate_v_rspace 110 11.3 0.003 0.004 7.059 7.076 qs_ot_get_p 110 10.4 0.001 0.001 6.862 7.001 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.636 6.655 calculate_rho_elec 110 8.6 0.059 0.060 6.635 6.652 dbcsr_mm_accdrv_process 34862 16.1 5.387 5.829 6.102 6.245 init_scf_run 11 5.9 0.000 0.001 6.034 6.034 scf_env_initial_rho_setup 11 6.9 0.003 0.015 6.034 6.034 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.152 5.710 apply_single 110 13.6 0.000 0.000 5.152 5.710 make_m2s 4110 13.4 0.050 0.051 5.207 5.601 make_images 4110 14.4 0.400 0.524 5.091 5.486 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.871 4.877 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.705 4.489 ot_diis_step 99 11.5 0.011 0.014 4.446 4.447 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.436 4.437 pw_transfer 1331 11.6 0.065 0.074 4.085 4.100 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.978 3.996 density_rs2pw 110 9.6 0.004 0.005 3.791 3.955 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.768 3.775 cp_fm_diag_elpa_base 48 14.0 3.691 3.731 3.766 3.774 multiply_cannon_sync_h2d 16440 15.4 3.187 3.704 3.187 3.704 cp_fm_cholesky_invert 11 10.9 3.599 3.607 3.599 3.607 make_images_data 4110 15.4 0.047 0.052 2.882 3.503 wfi_extrapolate 11 7.9 0.007 0.048 3.453 3.453 hybrid_alltoall_any 4261 16.3 0.110 0.380 2.494 3.412 grid_integrate_task_list 110 12.3 3.175 3.378 3.175 3.378 fft_wrap_pw1pw2_140 451 13.1 0.576 0.586 3.222 3.239 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.964 2.969 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.881 2.937 fft3d_ps 1111 14.6 1.092 1.105 2.924 2.936 potential_pw2rs 110 12.3 0.011 0.012 2.823 2.834 mp_sum_l 6594 12.7 1.937 2.732 1.937 2.732 calculate_dm_sparse 110 9.5 0.001 0.001 2.610 2.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.521 2.567 mp_alltoall_d11v 2046 13.8 2.257 2.478 2.257 2.478 calculate_first_density_matrix 1 7.0 0.001 0.003 2.468 2.472 grid_collocate_task_list 110 9.6 2.225 2.425 2.225 2.425 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.888 2.386 dbcsr_complete_redistribute 325 12.2 0.324 0.363 1.814 2.310 mp_irecv_dv 48980 15.7 0.813 2.251 0.813 2.251 cp_fm_cholesky_decompose 22 10.9 2.111 2.137 2.111 2.137 transfer_rs2pw 451 10.6 0.005 0.006 1.875 2.049 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.017 2.018 acc_transpose_blocks 16440 15.4 0.079 0.080 1.615 1.893 mp_allgather_i34 2055 14.4 0.694 1.858 0.694 1.858 cp_fm_upper_to_full 70 14.2 1.426 1.854 1.426 1.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.756 1.776 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.261 1.739 transfer_pw2rs 451 13.1 0.005 0.005 1.687 1.695 mp_alltoall_z22v 1111 16.6 1.615 1.635 1.615 1.635 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.495 1.504 make_images_sizes 4110 15.4 0.005 0.005 0.880 1.495 mp_alltoall_i44 4110 16.4 0.875 1.491 0.875 1.491 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 1.380 1.481 mp_waitany 17072 13.8 1.298 1.459 1.298 1.459 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.424 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=68.684000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.636364, yerr=8.678843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 734.765056E+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 0.008 0.030 68.735 68.736 qs_mol_dyn_low 1 2.0 0.003 0.003 68.201 68.523 qs_forces 11 3.9 0.003 0.003 68.118 68.119 qs_energies 11 4.9 0.001 0.001 64.706 64.710 scf_env_do_scf 11 5.9 0.000 0.001 55.798 55.800 scf_env_do_scf_inner_loop 99 6.5 0.041 0.058 43.764 43.765 velocity_verlet 10 3.0 0.002 0.002 38.094 38.096 dbcsr_multiply_generic 2055 12.4 0.116 0.122 31.468 31.697 qs_scf_new_mos 99 7.5 0.001 0.001 28.862 28.990 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.862 28.989 ot_scf_mini 99 9.5 0.003 0.003 27.129 27.261 multiply_cannon 2055 13.4 0.238 0.257 23.920 25.162 multiply_cannon_loop 2055 14.4 1.408 1.478 22.413 22.953 ot_mini 99 10.5 0.001 0.001 15.437 15.578 multiply_cannon_multrec 24660 15.4 4.172 7.147 13.297 14.648 rebuild_ks_matrix 110 8.3 0.000 0.000 12.093 12.193 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.092 12.192 init_scf_loop 11 6.9 0.000 0.001 11.990 11.991 qs_ot_get_derivative 99 11.5 0.001 0.001 11.213 11.348 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.694 10.784 prepare_preconditioner 11 7.9 0.000 0.000 10.196 10.210 make_preconditioner 11 8.9 0.000 0.000 10.195 10.210 dbcsr_mm_accdrv_process 52304 16.0 7.806 9.335 8.967 9.948 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.415 9.852 qs_ot_get_p 110 10.4 0.001 0.001 6.503 6.643 mp_waitall_1 121746 16.5 4.411 6.381 4.411 6.381 sum_up_and_integrate 110 10.3 0.001 0.002 6.357 6.371 integrate_v_rspace 110 11.3 0.003 0.003 6.331 6.344 make_m2s 4110 13.4 0.060 0.062 5.736 6.101 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.001 6.009 calculate_rho_elec 110 8.6 0.078 0.081 6.000 6.009 make_images 4110 14.4 0.578 0.698 5.593 5.953 init_scf_run 11 5.9 0.000 0.001 5.746 5.747 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.746 5.747 cp_fm_upper_to_full 70 14.2 3.335 4.827 3.335 4.827 qs_ot_p2m_diag 48 11.0 0.055 0.065 4.458 4.475 ot_diis_step 99 11.5 0.011 0.012 4.178 4.178 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.080 4.167 apply_single 110 13.6 0.000 0.000 4.080 4.167 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.873 3.874 dbcsr_complete_redistribute 325 12.2 0.415 0.457 2.715 3.826 pw_transfer 1331 11.6 0.066 0.073 3.755 3.790 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.649 3.686 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.583 3.666 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.560 3.615 grid_integrate_task_list 110 12.3 3.262 3.449 3.262 3.449 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.274 3.284 cp_fm_diag_elpa_base 48 14.0 3.115 3.176 3.271 3.282 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.163 3.262 density_rs2pw 110 9.6 0.004 0.004 3.113 3.247 make_images_data 4110 15.4 0.050 0.054 2.834 3.218 multiply_cannon_metrocomm3 24660 15.4 0.039 0.040 1.415 3.164 hybrid_alltoall_any 4261 16.3 0.125 0.463 2.400 3.136 fft_wrap_pw1pw2_140 451 13.1 0.611 0.632 3.089 3.130 wfi_extrapolate 11 7.9 0.001 0.001 3.119 3.119 calculate_dm_sparse 110 9.5 0.001 0.001 3.043 3.081 cp_fm_cholesky_invert 11 10.9 2.888 2.896 2.888 2.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.876 2.876 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.772 2.853 mp_alltoall_i22 605 13.7 1.643 2.815 1.643 2.815 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.644 2.645 acc_transpose_blocks 24660 15.4 0.116 0.119 2.412 2.550 fft3d_ps 1111 14.6 1.087 1.115 2.529 2.548 calculate_first_density_matrix 1 7.0 0.000 0.000 2.529 2.533 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.483 2.512 multiply_cannon_sync_h2d 24660 15.4 2.367 2.510 2.367 2.510 grid_collocate_task_list 110 9.6 2.267 2.428 2.267 2.428 potential_pw2rs 110 12.3 0.013 0.014 2.146 2.152 mp_alltoall_d11v 2046 13.8 1.909 2.053 1.909 2.053 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.868 1.902 cp_fm_cholesky_decompose 22 10.9 1.797 1.838 1.797 1.838 transfer_rs2pw 451 10.6 0.005 0.006 1.372 1.786 mp_sum_l 6594 12.7 1.118 1.786 1.118 1.786 jit_kernel_multiply 8 16.0 0.814 1.747 0.814 1.747 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.718 mp_allgather_i34 2055 14.4 0.589 1.686 0.589 1.686 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.586 1.678 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.644 1.655 mp_waitany 13376 13.8 1.128 1.634 1.128 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.559 1.573 acc_transpose_blocks_sync 73980 16.4 1.388 1.511 1.388 1.511 multiply_cannon_metrocomm4 20550 15.4 0.064 0.068 0.845 1.480 mp_irecv_dv 62702 16.1 0.739 1.396 0.739 1.396 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.736000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.727273, yerr=9.808076 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 861.253632E+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 0.017 0.039 63.959 63.961 qs_mol_dyn_low 1 2.0 0.003 0.003 63.328 63.354 qs_forces 11 3.9 0.003 0.003 63.248 63.249 qs_energies 11 4.9 0.001 0.002 59.495 59.500 scf_env_do_scf 11 5.9 0.000 0.001 49.433 49.434 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.809 39.810 velocity_verlet 10 3.0 0.015 0.019 35.108 35.111 dbcsr_multiply_generic 2055 12.4 0.108 0.113 26.453 26.627 qs_scf_new_mos 99 7.5 0.001 0.001 24.163 24.240 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.162 24.240 ot_scf_mini 99 9.5 0.002 0.002 22.819 22.865 multiply_cannon 2055 13.4 0.238 0.249 19.692 21.203 multiply_cannon_loop 2055 14.4 0.609 0.632 18.275 18.708 rebuild_ks_matrix 110 8.3 0.000 0.000 12.350 12.408 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.350 12.408 ot_mini 99 10.5 0.001 0.001 12.243 12.273 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.978 11.028 multiply_cannon_multrec 8220 15.4 3.294 4.537 8.864 9.871 init_scf_loop 11 6.9 0.001 0.001 9.562 9.566 mp_waitall_1 103326 16.6 7.053 8.642 7.053 8.642 qs_ot_get_derivative 99 11.5 0.001 0.001 8.470 8.512 prepare_preconditioner 11 7.9 0.000 0.000 7.765 7.770 make_preconditioner 11 8.9 0.000 0.000 7.765 7.770 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.262 7.371 init_scf_run 11 5.9 0.000 0.001 6.815 6.816 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.815 6.815 sum_up_and_integrate 110 10.3 0.001 0.002 6.785 6.798 integrate_v_rspace 110 11.3 0.003 0.003 6.759 6.771 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.654 6.675 calculate_rho_elec 110 8.6 0.114 0.114 6.654 6.675 dbcsr_mm_accdrv_process 17442 15.9 3.314 4.459 5.432 6.541 qs_ot_get_p 110 10.4 0.001 0.001 6.372 6.401 make_m2s 4110 13.4 0.038 0.039 5.066 5.372 make_images 4110 14.4 0.648 0.709 4.935 5.240 qs_ot_p2m_diag 48 11.0 0.082 0.086 4.646 4.656 pw_transfer 1331 11.6 0.066 0.071 4.361 4.382 multiply_cannon_metrocomm3 8220 15.4 0.020 0.020 2.859 4.299 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.253 4.276 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.222 4.222 cp_fm_cholesky_invert 11 10.9 3.953 3.957 3.953 3.957 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.772 3.830 apply_single 110 13.6 0.000 0.000 3.772 3.830 ot_diis_step 99 11.5 0.012 0.012 3.730 3.731 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.696 3.698 density_rs2pw 110 9.6 0.004 0.004 3.535 3.658 fft_wrap_pw1pw2_140 451 13.1 0.775 0.785 3.529 3.566 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.530 3.538 cp_fm_diag_elpa_base 48 14.0 3.461 3.493 3.528 3.535 grid_integrate_task_list 110 12.3 3.376 3.480 3.376 3.480 calculate_first_density_matrix 1 7.0 0.000 0.000 3.443 3.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.274 3.299 wfi_extrapolate 11 7.9 0.001 0.001 3.255 3.255 make_images_data 4110 15.4 0.042 0.048 2.716 3.239 hybrid_alltoall_any 4261 16.3 0.202 0.857 2.630 3.194 fft3d_ps 1111 14.6 1.140 1.195 2.920 2.956 calculate_dm_sparse 110 9.5 0.001 0.001 2.692 2.738 qs_energies_init_hamiltonians 11 5.9 0.007 0.010 2.701 2.702 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.470 2.495 multiply_cannon_sync_h2d 8220 15.4 2.363 2.491 2.363 2.491 potential_pw2rs 110 12.3 0.015 0.016 2.484 2.489 grid_collocate_task_list 110 9.6 2.373 2.488 2.373 2.488 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.406 2.443 mp_alltoall_d11v 2046 13.8 2.114 2.279 2.114 2.279 cp_fm_cholesky_decompose 22 10.9 2.238 2.246 2.238 2.246 jit_kernel_multiply 6 15.5 1.799 2.199 1.799 2.199 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.985 multiply_cannon_metrocomm1 8220 15.4 0.023 0.024 1.376 1.881 dbcsr_complete_redistribute 325 12.2 0.549 0.589 1.742 1.868 mp_allgather_i34 2055 14.4 0.579 1.786 0.579 1.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.707 1.713 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.521 1.637 transfer_rs2pw 451 10.6 0.005 0.005 1.472 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.575 1.589 mp_alltoall_z22v 1111 16.6 1.542 1.561 1.542 1.561 acc_transpose_blocks 8220 15.4 0.040 0.041 1.510 1.541 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.372 1.393 make_images_sizes 4110 15.4 0.005 0.005 0.743 1.353 mp_alltoall_i44 4110 16.4 0.738 1.348 0.738 1.348 mp_waitany 9240 13.8 1.195 1.335 1.195 1.335 qs_create_task_list 11 7.9 0.001 0.002 1.226 1.326 generate_qs_task_list 11 8.9 0.372 0.439 1.225 1.324 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.479 1.307 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=63.961000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.090909, yerr=11.919842 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 1.383072E+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 0.041 0.060 88.781 88.784 qs_mol_dyn_low 1 2.0 0.003 0.003 88.020 88.029 qs_forces 11 3.9 0.003 0.003 87.948 87.949 qs_energies 11 4.9 0.001 0.002 83.825 83.827 scf_env_do_scf 11 5.9 0.000 0.001 73.721 73.721 velocity_verlet 10 3.0 0.002 0.002 55.980 55.987 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.253 45.255 dbcsr_multiply_generic 2055 12.4 0.121 0.127 30.078 30.193 init_scf_loop 11 6.9 0.001 0.001 28.392 28.395 qs_scf_new_mos 99 7.5 0.001 0.001 27.779 27.830 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.779 27.829 prepare_preconditioner 11 7.9 0.000 0.000 26.404 26.413 make_preconditioner 11 8.9 0.000 0.000 26.404 26.413 ot_scf_mini 99 9.5 0.002 0.002 25.961 26.004 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.646 25.893 multiply_cannon 2055 13.4 0.332 0.346 22.816 23.527 multiply_cannon_loop 2055 14.4 0.823 0.832 21.008 21.281 cp_fm_upper_to_full 70 14.2 12.631 18.132 12.631 18.132 ot_mini 99 10.5 0.001 0.001 14.570 14.610 rebuild_ks_matrix 110 8.3 0.000 0.000 13.720 13.768 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.719 13.767 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.451 12.494 dbcsr_complete_redistribute 325 12.2 1.013 1.039 7.371 10.610 multiply_cannon_multrec 8220 15.4 4.113 4.275 9.852 10.023 qs_ot_get_derivative 99 11.5 0.001 0.001 9.922 9.966 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.331 9.579 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.743 8.959 mp_waitall_1 84994 16.7 7.820 8.726 7.820 8.726 mp_alltoall_i22 605 13.7 5.395 8.669 5.395 8.669 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.152 7.191 calculate_rho_elec 110 8.6 0.225 0.225 7.152 7.190 sum_up_and_integrate 110 10.3 0.002 0.002 6.867 6.881 integrate_v_rspace 110 11.3 0.004 0.004 6.839 6.853 qs_ot_get_p 110 10.4 0.001 0.001 6.018 6.072 make_m2s 4110 13.4 0.044 0.044 5.448 5.940 dbcsr_mm_accdrv_process 11614 15.7 3.922 4.213 5.591 5.843 init_scf_run 11 5.9 0.000 0.001 5.830 5.830 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.829 5.830 make_images 4110 14.4 0.889 0.946 5.259 5.751 multiply_cannon_metrocomm3 8220 15.4 0.020 0.021 5.121 5.493 cp_fm_cholesky_invert 11 10.9 5.428 5.432 5.428 5.432 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.777 5.248 apply_single 110 13.6 0.000 0.000 4.777 5.248 pw_transfer 1331 11.6 0.075 0.075 4.911 4.916 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.794 4.799 ot_diis_step 99 11.5 0.015 0.015 4.625 4.626 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.376 4.383 fft_wrap_pw1pw2_140 451 13.1 1.280 1.283 4.165 4.170 grid_integrate_task_list 110 12.3 3.736 3.857 3.736 3.857 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.844 3.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.645 3.646 density_rs2pw 110 9.6 0.004 0.004 3.574 3.631 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.166 3.625 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.893 3.546 make_images_data 4110 15.4 0.047 0.050 2.821 3.481 wfi_extrapolate 11 7.9 0.001 0.001 3.378 3.378 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.230 3.230 cp_fm_diag_elpa_base 48 14.0 2.685 2.886 3.228 3.229 calculate_dm_sparse 110 9.5 0.001 0.001 3.190 3.220 multiply_cannon_sync_h2d 8220 15.4 3.132 3.150 3.132 3.150 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.956 2.957 fft3d_ps 1111 14.6 1.294 1.303 2.845 2.850 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.761 2.785 grid_collocate_task_list 110 9.6 2.715 2.762 2.715 2.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.427 2.455 calculate_first_density_matrix 1 7.0 0.000 0.000 2.335 2.337 potential_pw2rs 110 12.3 0.022 0.022 2.279 2.283 mp_alltoall_d11v 2046 13.8 2.101 2.220 2.101 2.220 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.166 2.218 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.180 cp_fm_cholesky_decompose 22 10.9 2.074 2.091 2.074 2.091 qs_create_task_list 11 7.9 0.001 0.001 1.881 1.925 generate_qs_task_list 11 8.9 0.730 0.782 1.881 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.871 1.875 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.797 1.871 jit_kernel_multiply 10 15.3 1.470 1.850 1.470 1.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.784000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1239.636364, yerr=53.184538 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 637.423616E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.012 0.031 223.197 223.198 qs_mol_dyn_low 1 2.0 0.003 0.004 222.716 222.768 qs_forces 11 3.9 0.005 0.005 222.624 222.626 qs_energies 11 4.9 0.001 0.002 216.843 216.878 scf_env_do_scf 11 5.9 0.001 0.001 199.253 199.257 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 175.739 175.742 qs_scf_new_mos 117 7.6 0.001 0.001 132.802 133.134 qs_scf_loop_do_ot 117 8.6 0.001 0.001 132.801 133.134 velocity_verlet 10 3.0 0.001 0.002 132.926 132.929 dbcsr_multiply_generic 2507 12.6 0.183 0.187 128.825 129.552 ot_scf_mini 117 9.6 0.003 0.003 126.119 126.426 multiply_cannon 2507 13.6 0.243 0.254 103.070 105.107 multiply_cannon_loop 2507 14.6 2.390 2.484 100.545 102.633 ot_mini 117 10.6 0.001 0.001 68.674 68.939 qs_ot_get_derivative 117 11.6 0.001 0.001 43.772 44.048 multiply_cannon_multrec 60168 15.6 31.825 33.696 41.966 43.607 rebuild_ks_matrix 128 8.3 0.001 0.001 35.376 35.729 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 35.376 35.728 qs_ot_get_p 128 10.4 0.001 0.001 34.846 35.242 mp_waitall_1 267128 16.5 31.341 35.153 31.341 35.153 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.813 32.129 multiply_cannon_sync_h2d 60168 15.6 26.407 28.022 26.407 28.022 qs_ot_p2m_diag 83 11.4 0.079 0.092 27.574 27.664 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.406 24.975 apply_single 128 13.6 0.001 0.001 24.406 24.975 ot_diis_step 117 11.6 0.008 0.008 24.630 24.631 cp_dbcsr_syevd 83 12.4 0.005 0.005 24.624 24.627 init_scf_loop 11 6.9 0.001 0.001 23.428 23.430 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.955 20.985 cp_fm_diag_elpa_base 83 14.4 20.833 20.873 20.950 20.980 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.448 20.697 prepare_preconditioner 11 7.9 0.000 0.000 18.661 18.706 make_preconditioner 11 8.9 0.000 0.000 18.661 18.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.838 18.016 multiply_cannon_metrocomm3 60168 15.6 0.121 0.128 16.262 17.943 make_m2s 5014 13.6 0.104 0.112 15.380 15.845 make_images 5014 14.6 0.402 0.422 15.198 15.670 sum_up_and_integrate 128 10.3 0.002 0.004 15.570 15.588 integrate_v_rspace 128 11.3 0.004 0.004 15.511 15.532 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.282 15.413 calculate_rho_elec 128 8.7 0.045 0.064 15.282 15.412 init_scf_run 11 5.9 0.000 0.001 13.267 13.268 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.267 13.268 cp_fm_cholesky_invert 11 10.9 10.796 10.805 10.796 10.805 mp_sum_l 7950 12.9 9.112 10.416 9.112 10.416 dbcsr_mm_accdrv_process 124484 16.2 4.541 4.811 9.709 10.316 wfi_extrapolate 11 7.9 0.001 0.001 9.701 9.701 density_rs2pw 128 9.7 0.006 0.007 8.803 9.389 pw_transfer 1547 11.6 0.076 0.092 8.660 8.915 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.778 8.893 calculate_dm_sparse 128 9.5 0.001 0.001 8.653 8.805 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.456 8.711 multiply_cannon_metrocomm1 60168 15.6 0.098 0.102 6.806 8.684 make_images_data 5014 15.6 0.068 0.075 7.534 8.539 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.316 8.407 hybrid_alltoall_any 5200 16.5 0.295 2.253 6.470 7.843 grid_integrate_task_list 128 12.3 7.003 7.515 7.003 7.515 fft_wrap_pw1pw2_140 523 13.2 1.109 1.163 6.873 7.122 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.922 6.937 fft3d_ps 1291 14.7 2.190 2.823 6.475 6.872 mp_alltoall_d11v 2415 14.1 4.983 6.379 4.983 6.379 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.910 6.096 potential_pw2rs 128 12.3 0.009 0.010 5.872 5.972 cp_fm_cholesky_decompose 22 10.9 5.577 5.590 5.577 5.590 grid_collocate_task_list 128 9.7 4.838 5.220 4.838 5.220 transfer_rs2pw 523 10.6 0.008 0.009 4.458 5.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=223.198000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.454545, yerr=7.981383 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 844.980224E+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 4617890. 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 0.025 0.055 200.161 200.167 qs_mol_dyn_low 1 2.0 0.006 0.029 199.680 199.694 qs_forces 11 3.9 0.011 0.056 199.559 199.560 qs_energies 11 4.9 0.004 0.020 192.633 192.645 scf_env_do_scf 11 5.9 0.008 0.056 175.494 175.504 scf_env_do_scf_inner_loop 117 6.6 0.048 0.075 139.694 139.703 velocity_verlet 10 3.0 0.002 0.003 125.206 125.209 dbcsr_multiply_generic 2507 12.6 0.192 0.198 99.069 100.414 qs_scf_new_mos 117 7.6 0.001 0.001 99.141 99.725 qs_scf_loop_do_ot 117 8.6 0.001 0.001 99.140 99.724 ot_scf_mini 117 9.6 0.004 0.005 94.295 94.929 multiply_cannon 2507 13.6 0.505 0.560 77.648 82.043 multiply_cannon_loop 2507 14.6 1.581 1.657 74.096 76.667 ot_mini 117 10.6 0.001 0.001 51.358 51.947 mp_waitall_1 214728 16.6 25.602 39.216 25.602 39.216 init_scf_loop 11 6.9 0.001 0.006 35.696 35.697 multiply_cannon_multrec 30084 15.6 20.777 24.498 31.524 35.651 rebuild_ks_matrix 128 8.3 0.001 0.001 33.471 34.136 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 33.470 34.136 prepare_preconditioner 11 7.9 0.000 0.000 31.053 31.122 make_preconditioner 11 8.9 0.000 0.001 31.053 31.122 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.114 30.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.614 30.217 qs_ot_get_derivative 117 11.6 0.001 0.002 29.340 29.965 multiply_cannon_metrocomm3 30084 15.6 0.100 0.105 15.862 29.176 qs_ot_get_p 128 10.4 0.001 0.001 24.828 25.627 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.134 23.375 apply_single 128 13.6 0.001 0.001 22.134 23.375 ot_diis_step 117 11.6 0.014 0.015 21.840 21.842 multiply_cannon_sync_h2d 30084 15.6 17.815 20.334 17.815 20.334 qs_ot_p2m_diag 83 11.4 0.188 0.216 19.475 19.518 cp_dbcsr_syevd 83 12.4 0.006 0.006 18.161 18.162 cp_fm_cholesky_invert 11 10.9 18.038 18.053 18.038 18.053 make_m2s 5014 13.6 0.090 0.094 15.204 16.871 make_images 5014 14.6 1.180 1.379 14.993 16.659 sum_up_and_integrate 128 10.3 0.002 0.004 15.113 15.143 integrate_v_rspace 128 11.3 0.003 0.005 15.053 15.087 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.589 14.618 cp_fm_diag_elpa_base 83 14.4 14.294 14.414 14.583 14.609 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.420 14.454 calculate_rho_elec 128 8.7 0.087 0.104 14.420 14.453 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.822 12.270 init_scf_run 11 5.9 0.000 0.001 12.194 12.197 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.194 12.196 multiply_cannon_metrocomm4 27577 15.6 0.106 0.119 3.926 10.847 make_images_data 5014 15.6 0.069 0.078 8.839 10.845 dbcsr_mm_accdrv_process 62242 16.2 5.633 6.444 10.200 10.738 mp_irecv_dv 69486 16.3 3.720 10.444 3.720 10.444 hybrid_alltoall_any 5200 16.5 0.354 1.544 7.394 10.171 pw_transfer 1547 11.6 0.085 0.097 9.023 9.092 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.800 8.872 wfi_extrapolate 11 7.9 0.001 0.001 8.662 8.662 density_rs2pw 128 9.7 0.006 0.007 8.104 8.510 cp_fm_cholesky_decompose 22 10.9 7.969 8.065 7.969 8.065 grid_integrate_task_list 128 12.3 7.179 7.641 7.179 7.641 fft_wrap_pw1pw2_140 523 13.2 1.207 1.230 7.517 7.607 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.432 7.189 calculate_dm_sparse 128 9.5 0.001 0.003 6.786 6.927 fft3d_ps 1291 14.7 2.844 3.030 6.367 6.422 mp_sum_l 7950 12.9 4.438 6.420 4.438 6.420 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.219 6.231 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.504 5.578 potential_pw2rs 128 12.3 0.016 0.019 5.544 5.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.257 5.417 grid_collocate_task_list 128 9.7 5.008 5.311 5.008 5.311 mp_allgather_i34 2507 14.6 2.062 5.189 2.062 5.189 mp_alltoall_d11v 2415 14.1 4.452 5.025 4.452 5.025 dbcsr_complete_redistribute 395 12.7 0.783 0.878 3.501 4.419 transfer_rs2pw 523 10.6 0.007 0.008 3.770 4.312 mp_sum_d 4465 12.1 2.691 4.004 2.691 4.004 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=200.167000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.818182, yerr=3.638636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 981.581824E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58094. MP_Allreduce 11102 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.513 192.798 192.799 qs_mol_dyn_low 1 2.0 0.003 0.003 191.459 191.473 qs_forces 11 3.9 0.010 0.054 191.038 191.040 qs_energies 11 4.9 0.005 0.026 184.316 184.327 scf_env_do_scf 11 5.9 0.001 0.001 167.392 167.393 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 128.529 128.530 velocity_verlet 10 3.0 0.008 0.012 121.323 121.336 qs_scf_new_mos 117 7.6 0.001 0.001 89.632 89.980 qs_scf_loop_do_ot 117 8.6 0.001 0.001 89.631 89.979 dbcsr_multiply_generic 2507 12.6 0.184 0.188 85.752 87.000 ot_scf_mini 117 9.6 0.004 0.004 85.324 85.709 multiply_cannon 2507 13.6 0.512 0.529 64.369 68.937 multiply_cannon_loop 2507 14.6 1.135 1.200 60.676 64.028 ot_mini 117 10.6 0.001 0.001 45.147 45.547 init_scf_loop 11 6.9 0.025 0.194 38.754 38.755 mp_waitall_1 170520 16.6 26.270 37.285 26.270 37.285 prepare_preconditioner 11 7.9 0.000 0.000 34.464 34.515 make_preconditioner 11 8.9 0.000 0.002 34.464 34.515 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.007 33.453 rebuild_ks_matrix 128 8.3 0.001 0.001 31.127 31.558 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.127 31.558 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.080 28.461 multiply_cannon_metrocomm3 20056 15.6 0.064 0.069 15.877 26.611 multiply_cannon_multrec 20056 15.6 13.011 16.103 23.203 26.353 qs_ot_get_derivative 117 11.6 0.002 0.002 25.395 25.789 qs_ot_get_p 128 10.4 0.001 0.001 25.160 25.633 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.925 20.957 apply_single 128 13.6 0.001 0.001 19.924 20.956 qs_ot_p2m_diag 83 11.4 0.310 0.628 20.463 20.519 ot_diis_step 117 11.6 0.018 0.018 19.623 19.624 cp_dbcsr_syevd 83 12.4 0.006 0.007 19.125 19.173 make_m2s 5014 13.6 0.080 0.083 15.370 16.780 cp_fm_cholesky_invert 11 10.9 16.755 16.771 16.755 16.771 make_images 5014 14.6 1.162 1.271 15.137 16.547 multiply_cannon_sync_h2d 20056 15.6 13.517 15.088 13.517 15.088 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.975 14.992 cp_fm_diag_elpa_base 83 14.4 14.519 14.695 14.970 14.988 sum_up_and_integrate 128 10.3 0.002 0.003 14.779 14.801 integrate_v_rspace 128 11.3 0.003 0.004 14.719 14.743 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.653 14.687 calculate_rho_elec 128 8.7 0.130 0.145 14.652 14.687 init_scf_run 11 5.9 0.000 0.001 11.693 11.693 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.693 11.693 make_images_data 5014 15.6 0.064 0.074 8.995 10.954 hybrid_alltoall_any 5200 16.5 0.451 2.034 7.857 10.508 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.162 10.430 dbcsr_mm_accdrv_process 41502 16.2 6.442 6.869 9.663 9.844 multiply_cannon_metrocomm4 17549 15.6 0.067 0.077 3.481 9.408 pw_transfer 1547 11.6 0.086 0.103 9.133 9.262 mp_irecv_dv 50230 16.2 3.353 9.150 3.353 9.150 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.908 9.045 cp_fm_cholesky_decompose 22 10.9 8.447 8.510 8.447 8.510 density_rs2pw 128 9.7 0.006 0.006 7.987 8.392 grid_integrate_task_list 128 12.3 7.285 7.879 7.285 7.879 wfi_extrapolate 11 7.9 0.001 0.001 7.849 7.850 fft_wrap_pw1pw2_140 523 13.2 1.299 1.324 7.667 7.823 cp_fm_upper_to_full 105 14.8 5.810 7.359 5.810 7.359 dbcsr_complete_redistribute 395 12.7 1.174 1.201 5.118 6.945 calculate_dm_sparse 128 9.5 0.001 0.002 6.636 6.718 fft3d_ps 1291 14.7 2.758 2.989 6.288 6.388 mp_sum_l 7950 12.9 4.439 6.039 4.439 6.039 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.019 5.813 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.845 5.675 grid_collocate_task_list 128 9.7 5.163 5.646 5.163 5.646 mp_alltoall_d11v 2415 14.1 4.666 5.643 4.666 5.643 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.615 5.622 potential_pw2rs 128 12.3 0.021 0.023 5.168 5.200 mp_allgather_i34 2507 14.6 2.174 5.098 2.174 5.098 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.614 4.739 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.226 4.275 mp_alltoall_i22 716 14.1 2.248 4.212 2.248 4.212 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.437 4.187 transfer_rs2pw 523 10.6 0.007 0.007 3.540 3.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=192.799000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=930.272727, yerr=11.521450 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.148822E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57809. MP_Allreduce 11152 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.064 191.579 191.580 qs_mol_dyn_low 1 2.0 0.003 0.003 190.894 190.907 qs_forces 11 3.9 0.004 0.004 190.781 190.787 qs_energies 11 4.9 0.003 0.012 183.080 183.089 scf_env_do_scf 11 5.9 0.001 0.001 164.759 164.774 velocity_verlet 10 3.0 0.002 0.002 125.879 125.882 scf_env_do_scf_inner_loop 118 6.6 0.005 0.021 118.508 118.510 dbcsr_multiply_generic 2527 12.6 0.209 0.215 83.176 83.987 qs_scf_new_mos 118 7.6 0.001 0.001 81.894 82.217 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.894 82.217 ot_scf_mini 118 9.6 0.004 0.004 77.180 77.513 multiply_cannon 2527 13.6 0.558 0.594 57.309 61.117 multiply_cannon_loop 2527 14.6 1.874 1.973 53.557 55.573 init_scf_loop 11 6.9 0.001 0.006 46.126 46.129 ot_mini 118 10.6 0.001 0.001 43.964 44.289 prepare_preconditioner 11 7.9 0.000 0.000 42.026 42.054 make_preconditioner 11 8.9 0.002 0.014 42.026 42.054 make_full_inverse_cholesky 11 9.9 0.013 0.023 35.711 40.664 multiply_cannon_multrec 30324 15.6 13.790 19.634 27.578 32.992 rebuild_ks_matrix 129 8.3 0.001 0.001 30.008 30.406 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 30.007 30.406 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.611 26.972 mp_waitall_1 149060 16.7 17.482 26.857 17.482 26.857 qs_ot_get_derivative 118 11.6 0.002 0.002 24.058 24.389 make_m2s 5054 13.6 0.097 0.100 21.231 22.180 make_images 5054 14.6 1.991 2.476 20.924 21.875 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.297 19.846 apply_single 129 13.6 0.001 0.001 19.297 19.846 ot_diis_step 118 11.6 0.018 0.020 19.781 19.782 qs_ot_get_p 129 10.4 0.001 0.002 19.154 19.522 cp_fm_upper_to_full 105 14.8 11.378 16.710 11.378 16.710 cp_fm_cholesky_invert 11 10.9 16.051 16.060 16.051 16.060 multiply_cannon_metrocomm3 30324 15.6 0.051 0.054 6.260 15.247 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.679 14.729 sum_up_and_integrate 129 10.3 0.002 0.003 14.522 14.621 integrate_v_rspace 129 11.3 0.003 0.004 14.462 14.562 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.888 13.941 calculate_rho_elec 129 8.7 0.174 0.190 13.887 13.941 dbcsr_mm_accdrv_process 62756 16.2 8.948 10.099 13.350 13.920 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.276 13.278 make_images_data 5054 15.6 0.069 0.076 11.329 12.977 dbcsr_complete_redistribute 395 12.7 1.533 1.642 8.942 12.614 hybrid_alltoall_any 5240 16.5 0.538 2.229 10.108 12.238 init_scf_run 11 5.9 0.000 0.001 11.912 11.914 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.912 11.914 multiply_cannon_sync_h2d 30324 15.6 10.480 11.724 10.480 11.724 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.592 11.250 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.086 10.334 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.204 10.215 cp_fm_diag_elpa_base 83 14.4 9.243 9.562 10.198 10.210 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.291 9.901 pw_transfer 1559 11.6 0.086 0.105 9.446 9.592 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.219 9.369 mp_alltoall_i22 718 14.1 5.553 9.265 5.553 9.265 fft_wrap_pw1pw2_140 527 13.2 1.459 1.494 8.271 8.430 grid_integrate_task_list 129 12.3 7.567 7.974 7.567 7.974 wfi_extrapolate 11 7.9 0.001 0.001 7.762 7.762 cp_fm_cholesky_decompose 22 10.9 7.518 7.621 7.518 7.621 multiply_cannon_metrocomm4 25270 15.6 0.087 0.096 2.838 7.363 density_rs2pw 129 9.7 0.006 0.006 6.784 7.193 calculate_dm_sparse 129 9.5 0.001 0.001 7.083 7.186 mp_irecv_dv 76685 16.2 2.679 7.081 2.679 7.081 fft3d_ps 1301 14.7 2.866 2.975 6.372 6.500 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.904 5.964 grid_collocate_task_list 129 9.7 5.328 5.741 5.328 5.741 mp_alltoall_d11v 2423 14.1 5.089 5.523 5.089 5.523 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.760 4.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.730 4.818 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.814 4.815 potential_pw2rs 129 12.3 0.023 0.024 4.724 4.805 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.326 4.399 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 4.054 4.086 calculate_first_density_matrix 1 7.0 0.000 0.003 3.952 3.954 mp_sum_l 8010 12.9 2.665 3.864 2.665 3.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.580000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1088.545455, yerr=16.272473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/19/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 5.910120E+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 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 1.533207E+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 5.526748E+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 726592466352 ------------------------------------------------------------------------------- - - - 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 4045 57596. MP_Allreduce 11189 1162. MP_Sync 88 MP_Alltoall 1724 18848034. 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 0.014 0.032 178.382 178.383 qs_mol_dyn_low 1 2.0 0.003 0.004 177.815 177.828 qs_forces 11 3.9 0.004 0.004 177.697 177.699 qs_energies 11 4.9 0.002 0.003 170.234 170.251 scf_env_do_scf 11 5.9 0.001 0.001 151.428 151.437 velocity_verlet 10 3.0 0.002 0.003 116.039 116.044 scf_env_do_scf_inner_loop 118 6.6 0.010 0.044 114.239 114.241 dbcsr_multiply_generic 2529 12.6 0.184 0.196 76.808 77.447 qs_scf_new_mos 118 7.6 0.001 0.001 76.401 76.553 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.400 76.552 ot_scf_mini 118 9.6 0.004 0.004 71.820 71.999 multiply_cannon 2529 13.6 0.564 0.601 56.194 59.410 multiply_cannon_loop 2529 14.6 0.820 0.858 52.907 54.777 ot_mini 118 10.6 0.001 0.001 39.983 40.164 init_scf_loop 11 6.9 0.001 0.001 37.037 37.038 mp_waitall_1 126876 16.7 27.366 34.198 27.366 34.198 prepare_preconditioner 11 7.9 0.000 0.000 33.125 33.150 make_preconditioner 11 8.9 0.000 0.000 33.125 33.150 make_full_inverse_cholesky 11 9.9 0.016 0.027 30.989 31.332 rebuild_ks_matrix 129 8.3 0.001 0.001 29.672 29.828 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.672 29.828 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.046 27.192 multiply_cannon_multrec 10116 15.6 10.346 14.892 18.521 22.901 qs_ot_get_derivative 118 11.6 0.002 0.002 22.295 22.464 multiply_cannon_metrocomm3 10116 15.6 0.027 0.028 13.644 21.534 cp_fm_cholesky_invert 11 10.9 19.043 19.050 19.043 19.050 qs_ot_get_p 129 10.4 0.001 0.001 18.615 18.804 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.173 18.482 apply_single 129 13.6 0.001 0.001 18.172 18.482 ot_diis_step 118 11.6 0.020 0.021 17.595 17.596 make_m2s 5058 13.6 0.067 0.071 15.680 16.734 make_images 5058 14.6 2.188 2.620 15.372 16.430 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.549 14.600 calculate_rho_elec 129 8.7 0.258 0.269 14.549 14.599 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.517 14.536 sum_up_and_integrate 129 10.3 0.002 0.002 14.418 14.471 integrate_v_rspace 129 11.3 0.004 0.004 14.357 14.412 cp_dbcsr_syevd 84 12.4 0.006 0.006 13.266 13.267 init_scf_run 11 5.9 0.000 0.001 11.982 11.982 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.982 11.982 multiply_cannon_sync_h2d 10116 15.6 10.714 11.472 10.714 11.472 make_images_data 5058 15.6 0.057 0.067 9.157 10.680 hybrid_alltoall_any 5245 16.5 0.849 3.800 8.854 10.329 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.206 10.218 cp_fm_diag_elpa_base 84 14.4 9.965 10.042 10.202 10.215 multiply_cannon_metrocomm1 10116 15.6 0.031 0.032 5.287 9.796 pw_transfer 1559 11.6 0.086 0.095 9.524 9.561 dbcsr_mm_accdrv_process 20934 16.1 3.187 4.128 7.826 9.483 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.300 9.343 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.757 8.876 cp_fm_cholesky_decompose 22 10.9 8.317 8.446 8.317 8.446 fft_wrap_pw1pw2_140 527 13.2 1.817 1.853 8.184 8.224 grid_integrate_task_list 129 12.3 7.811 8.220 7.811 8.220 wfi_extrapolate 11 7.9 0.001 0.001 7.568 7.568 density_rs2pw 129 9.7 0.006 0.006 7.019 7.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.525 6.547 calculate_dm_sparse 129 9.5 0.001 0.001 6.246 6.337 fft3d_ps 1301 14.7 2.783 2.860 6.005 6.045 grid_collocate_task_list 129 9.7 5.614 5.914 5.614 5.914 dbcsr_complete_redistribute 397 12.7 2.116 2.172 5.360 5.635 multiply_cannon_metrocomm4 7587 15.6 0.027 0.029 1.887 5.620 mp_irecv_dv 29102 15.9 1.845 5.521 1.845 5.521 mp_alltoall_d11v 2429 14.1 5.017 5.493 5.017 5.493 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.364 5.445 mp_allgather_i34 2529 14.6 1.451 5.263 1.451 5.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.227 5.228 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.846 4.456 potential_pw2rs 129 12.3 0.027 0.028 4.392 4.399 calculate_first_density_matrix 1 7.0 0.000 0.000 4.163 4.165 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.578 3.876 mp_sum_l 8016 12.9 2.660 3.874 2.660 3.874 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.764 3.810 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.443 3.762 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.568 3.656 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.383000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1457.454545, yerr=7.703171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.171361E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 59207. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.092 292.009 292.018 qs_mol_dyn_low 1 2.0 0.003 0.003 291.060 291.089 qs_forces 11 3.9 0.005 0.005 289.467 289.472 qs_energies 11 4.9 0.002 0.002 280.663 280.674 scf_env_do_scf 11 5.9 0.001 0.001 256.940 256.948 velocity_verlet 10 3.0 0.277 0.278 210.620 210.781 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.419 132.420 init_scf_loop 11 6.9 0.001 0.001 124.237 124.240 prepare_preconditioner 11 7.9 0.000 0.000 119.241 119.262 make_preconditioner 11 8.9 0.000 0.000 119.241 119.262 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.303 116.407 qs_scf_new_mos 116 7.6 0.001 0.001 89.090 89.153 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.089 89.152 ot_scf_mini 116 9.6 0.004 0.004 84.267 84.279 dbcsr_multiply_generic 2485 12.5 0.211 0.222 81.893 82.280 cp_fm_upper_to_full 104 14.8 52.356 75.395 52.356 75.395 multiply_cannon 2485 13.5 0.670 0.705 58.990 60.663 multiply_cannon_loop 2485 14.5 1.039 1.082 54.983 56.129 ot_mini 116 10.6 0.001 0.001 43.540 43.554 dbcsr_complete_redistribute 393 12.7 4.016 4.108 30.084 42.823 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.477 39.190 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.895 36.377 mp_alltoall_i22 712 14.1 21.672 34.483 21.672 34.483 rebuild_ks_matrix 127 8.3 0.001 0.001 33.891 33.906 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.890 33.906 cp_fm_cholesky_invert 11 10.9 33.594 33.601 33.594 33.601 mp_waitall_1 102768 16.8 27.626 32.077 27.626 32.077 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.410 31.437 qs_ot_get_p 127 10.4 0.001 0.001 25.723 25.761 qs_ot_get_derivative 116 11.6 0.002 0.002 24.379 24.389 qs_ot_p2m_diag 82 11.4 0.869 0.874 21.635 21.663 make_m2s 4970 13.5 0.076 0.077 18.447 19.920 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.855 19.858 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 18.498 19.652 make_images 4970 14.5 3.044 3.235 17.970 19.443 ot_diis_step 116 11.6 0.021 0.022 19.135 19.135 multiply_cannon_multrec 9940 15.5 10.243 12.018 18.614 18.702 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.479 18.561 apply_single 127 13.6 0.001 0.001 18.478 18.560 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.693 16.693 cp_fm_diag_elpa_base 82 14.4 12.271 13.959 16.689 16.689 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.518 16.532 calculate_rho_elec 127 8.7 0.477 0.477 16.517 16.531 sum_up_and_integrate 127 10.3 0.002 0.002 16.139 16.229 integrate_v_rspace 127 11.3 0.005 0.005 16.077 16.167 multiply_cannon_sync_h2d 9940 15.5 14.200 14.225 14.200 14.225 init_scf_run 11 5.9 0.000 0.001 13.132 13.132 scf_env_initial_rho_setup 11 6.9 0.009 0.012 13.131 13.132 hybrid_alltoall_any 5155 16.4 1.311 3.067 10.669 12.920 make_images_data 4970 15.5 0.065 0.071 10.380 12.513 pw_transfer 1535 11.6 0.094 0.095 11.463 11.471 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 11.228 11.236 fft_wrap_pw1pw2_140 519 13.2 3.023 3.076 9.990 9.998 dbcsr_mm_accdrv_process 20590 16.0 4.271 5.891 8.124 9.978 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.590 9.593 wfi_extrapolate 11 7.9 0.001 0.001 9.214 9.214 cp_fm_cholesky_decompose 22 10.9 9.162 9.180 9.162 9.180 grid_integrate_task_list 127 12.3 8.496 8.686 8.496 8.686 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.357 8.359 density_rs2pw 127 9.7 0.005 0.005 7.552 7.604 calculate_dm_sparse 127 9.5 0.001 0.001 7.357 7.438 mp_alltoall_d11v 2401 14.1 7.180 7.324 7.180 7.324 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.639 6.693 fft3d_ps 1281 14.7 2.834 2.850 6.512 6.550 grid_collocate_task_list 127 9.7 6.366 6.441 6.366 6.441 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.313 6.401 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.018000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2864.090909, yerr=148.499437 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 1.259667E+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 25315686. 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 0.072 0.196 86.385 86.395 qs_energies 1 2.0 0.000 0.000 85.154 85.197 ls_scf 1 3.0 0.000 0.001 83.883 83.905 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.547 72.741 multiply_cannon 111 7.7 0.017 0.020 55.904 57.099 multiply_cannon_loop 111 8.7 0.227 0.240 52.470 53.978 ls_scf_main 1 4.0 0.000 0.006 52.233 52.242 density_matrix_trs4 2 5.0 0.002 0.004 46.660 46.764 ls_scf_init_scf 1 4.0 0.000 0.001 28.605 28.620 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.170 27.222 mp_waitall_1 11031 10.9 22.415 25.688 22.415 25.688 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.066 25.082 multiply_cannon_multrec 2664 9.7 8.150 8.917 15.632 17.425 multiply_cannon_sync_h2d 2664 9.7 13.431 15.801 13.431 15.801 make_m2s 222 7.7 0.009 0.012 12.989 13.565 make_images 222 8.7 0.099 0.110 12.966 13.546 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.750 11.886 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.427 8.471 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.708 7.098 8.140 make_images_data 222 9.7 0.004 0.005 7.558 8.069 hybrid_alltoall_any 227 10.6 0.216 1.832 6.499 7.751 dbcsr_mm_accdrv_process_sort 4760 11.4 6.303 7.262 6.303 7.262 calculate_norms 4752 9.8 5.498 6.178 5.498 6.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.054 5.168 mp_sum_l 887 5.1 3.147 4.823 3.147 4.823 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.044 3.965 mp_irecv_dv 6231 10.9 2.026 3.934 2.026 3.934 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.347 3.702 make_images_sizes 222 9.7 0.000 0.000 0.681 3.301 mp_alltoall_i44 222 10.7 0.681 3.300 0.681 3.300 arnoldi_extremal 4 6.8 0.000 0.000 3.179 3.203 arnoldi_normal_ev 4 7.8 0.001 0.003 3.179 3.202 build_subspace 16 8.4 0.009 0.012 3.087 3.090 ls_scf_post 1 4.0 0.003 0.025 3.045 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.394 2.836 dbcsr_merge_single_wm 555 10.7 0.463 0.601 2.385 2.827 make_images_pack 222 9.7 2.208 2.639 2.209 2.640 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.555 dbcsr_sort_data 658 11.4 2.179 2.552 2.179 2.552 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.456 2.068 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.300 2.347 buffer_matrices_ensure_size 222 8.7 1.753 2.084 1.753 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.839 1.853 rebuild_ks_matrix 3 7.3 0.000 0.000 1.828 1.843 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.024 1.828 1.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.395000, 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/185b7b66b50ae386199c6584a49c01427632d034_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 2.191532E+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 0.022 0.040 90.305 90.306 qs_energies 1 2.0 0.000 0.000 89.659 89.663 ls_scf 1 3.0 0.000 0.000 88.323 88.327 dbcsr_multiply_generic 111 6.7 0.015 0.017 74.237 74.521 multiply_cannon 111 7.7 0.027 0.036 52.843 57.091 ls_scf_main 1 4.0 0.000 0.000 54.772 54.840 multiply_cannon_loop 111 8.7 0.135 0.146 49.973 53.780 density_matrix_trs4 2 5.0 0.002 0.003 49.067 49.279 mp_waitall_1 9105 10.9 20.672 30.324 20.672 30.324 ls_scf_init_scf 1 4.0 0.000 0.000 29.877 29.900 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.606 28.692 multiply_cannon_multrec 1332 9.7 13.385 16.848 22.773 27.386 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.247 26.256 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.498 21.450 make_m2s 222 7.7 0.006 0.007 14.945 15.558 make_images 222 8.7 1.372 1.693 14.914 15.528 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.506 8.988 10.534 dbcsr_mm_accdrv_process_sort 4041 11.4 8.508 10.007 8.508 10.007 make_images_data 222 9.7 0.004 0.005 8.683 9.588 hybrid_alltoall_any 227 10.6 0.542 2.542 8.036 9.375 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.187 7.870 mp_irecv_dv 3311 11.0 3.166 7.815 3.166 7.815 mp_sum_l 887 5.1 4.981 7.803 4.981 7.803 calculate_norms 2376 9.8 6.047 6.803 6.047 6.803 multiply_cannon_sync_h2d 1332 9.7 4.842 6.188 4.842 6.188 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.698 6.114 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.931 5.166 arnoldi_extremal 4 6.8 0.000 0.000 4.830 4.847 arnoldi_normal_ev 4 7.8 0.001 0.005 4.830 4.847 build_subspace 16 8.4 0.014 0.021 4.578 4.580 ls_scf_post 1 4.0 0.000 0.000 3.674 3.678 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.498 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.256 3.442 dbcsr_matrix_vector_mult_local 304 10.0 2.781 3.268 2.783 3.270 mp_allgather_i34 111 8.7 0.855 3.051 0.855 3.051 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.606 2.678 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.134 2.531 dbcsr_data_new 4174 10.1 2.114 2.409 2.114 2.409 make_images_pack 222 9.7 1.815 2.117 1.817 2.120 dbcsr_sort_data 436 11.2 1.812 2.000 1.812 2.000 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.938 1.962 rebuild_ks_matrix 3 7.3 0.000 0.000 1.926 1.950 qs_ks_build_kohn_sham_matrix 3 8.3 0.062 0.085 1.926 1.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.306000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 2.922562E+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 0.074 0.136 94.214 94.216 qs_energies 1 2.0 0.000 0.001 93.315 93.317 ls_scf 1 3.0 0.003 0.029 91.869 91.874 dbcsr_multiply_generic 111 6.7 0.027 0.040 76.000 76.354 ls_scf_main 1 4.0 0.008 0.064 57.019 57.023 multiply_cannon 111 7.7 0.035 0.074 52.045 56.572 multiply_cannon_loop 111 8.7 0.118 0.131 49.250 53.465 density_matrix_trs4 2 5.0 0.003 0.004 50.993 51.204 mp_waitall_1 7281 11.0 23.535 33.921 23.535 33.921 ls_scf_init_scf 1 4.0 0.000 0.002 31.149 31.156 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.843 29.932 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.203 27.223 multiply_cannon_multrec 888 9.7 12.729 15.333 21.405 24.612 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.665 23.407 make_m2s 222 7.7 0.006 0.008 16.888 17.618 make_images 222 8.7 1.586 1.860 16.850 17.579 make_images_data 222 9.7 0.004 0.005 9.711 10.867 hybrid_alltoall_any 227 10.6 0.643 2.965 9.171 10.473 dbcsr_mm_accdrv_process 3754 10.4 0.305 0.489 8.206 9.437 mp_sum_l 887 5.1 5.375 9.379 5.375 9.379 dbcsr_mm_accdrv_process_sort 3754 11.4 7.771 8.949 7.771 8.949 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.209 7.543 multiply_cannon_sync_h2d 888 9.7 5.957 7.229 5.957 7.229 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.453 6.961 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.787 6.952 mp_irecv_dv 2335 11.1 2.437 6.916 2.437 6.916 arnoldi_extremal 4 6.8 0.000 0.000 5.357 5.379 arnoldi_normal_ev 4 7.8 0.001 0.005 5.356 5.379 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.018 5.235 build_subspace 16 8.4 0.014 0.020 5.052 5.057 calculate_norms 1584 9.8 4.331 4.757 4.331 4.757 mp_allgather_i34 111 8.7 0.904 4.065 0.904 4.065 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.601 3.895 ls_scf_post 1 4.0 0.000 0.001 3.698 3.702 dbcsr_matrix_vector_mult_local 304 10.0 3.061 3.639 3.063 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.429 3.520 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.956 3.062 dbcsr_data_new 4116 9.9 2.110 2.462 2.110 2.462 make_images_sizes 222 9.7 0.000 0.000 1.208 2.385 mp_alltoall_i44 222 10.7 1.208 2.385 1.208 2.385 make_images_pack 222 9.7 1.782 2.131 1.785 2.133 dbcsr_sort_data 325 11.1 1.862 2.130 1.862 2.130 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.968 1.970 rebuild_ks_matrix 3 7.3 0.000 0.000 1.948 1.950 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.948 1.950 dbcsr_finalize 304 7.8 0.026 0.032 1.613 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.216000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2290.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 3.335758E+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 0.035 0.060 97.467 97.469 qs_energies 1 2.0 0.000 0.000 96.803 96.810 ls_scf 1 3.0 0.000 0.000 94.819 94.826 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.402 78.679 ls_scf_main 1 4.0 0.000 0.000 58.901 58.902 multiply_cannon 111 7.7 0.038 0.088 51.656 55.792 density_matrix_trs4 2 5.0 0.002 0.003 52.782 52.895 multiply_cannon_loop 111 8.7 0.152 0.170 46.681 49.785 ls_scf_init_scf 1 4.0 0.000 0.000 32.672 32.674 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.467 31.526 mp_waitall_1 6369 11.0 22.824 29.262 22.824 29.262 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.995 29.008 multiply_cannon_multrec 1332 9.7 14.164 17.407 22.177 25.612 make_m2s 222 7.7 0.007 0.008 21.261 22.588 make_images 222 8.7 3.135 3.608 21.210 22.540 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.142 16.337 make_images_data 222 9.7 0.004 0.004 11.907 13.527 hybrid_alltoall_any 227 10.6 0.799 3.776 11.261 12.899 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.473 7.643 9.166 dbcsr_mm_accdrv_process_sort 3641 11.4 7.203 8.670 7.203 8.670 mp_sum_l 887 5.1 3.906 7.457 3.906 7.457 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.086 5.998 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.034 5.929 mp_irecv_dv 3229 10.9 2.060 5.913 2.060 5.913 multiply_cannon_sync_h2d 1332 9.7 5.394 5.887 5.394 5.887 arnoldi_extremal 4 6.8 0.000 0.000 5.400 5.413 arnoldi_normal_ev 4 7.8 0.001 0.005 5.400 5.413 build_subspace 16 8.4 0.014 0.021 5.067 5.073 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.601 5.049 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.513 4.726 calculate_norms 2376 9.8 4.184 4.542 4.184 4.542 mp_allgather_i34 111 8.7 2.092 4.182 2.092 4.182 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.709 3.993 dbcsr_matrix_vector_mult_local 304 10.0 3.238 3.731 3.240 3.733 dbcsr_sort_data 658 11.4 3.065 3.378 3.065 3.378 ls_scf_post 1 4.0 0.000 0.000 3.247 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.811 3.194 dbcsr_merge_single_wm 555 10.7 0.536 0.661 2.802 3.186 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.038 3.080 ls_scf_store_result 1 5.0 0.000 0.000 2.992 3.051 dbcsr_data_release 10477 10.7 1.586 2.360 1.586 2.360 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.975 1.977 dbcsr_finalize 304 7.8 0.049 0.061 1.804 1.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.469000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2701.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 4.765585E+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 0.038 0.058 92.400 92.401 qs_energies 1 2.0 0.000 0.000 91.483 91.489 ls_scf 1 3.0 0.000 0.000 89.339 89.344 dbcsr_multiply_generic 111 6.7 0.018 0.019 70.708 70.907 ls_scf_main 1 4.0 0.000 0.000 56.516 56.517 multiply_cannon 111 7.7 0.088 0.188 52.763 56.400 multiply_cannon_loop 111 8.7 0.088 0.093 50.176 51.487 density_matrix_trs4 2 5.0 0.002 0.003 49.526 49.587 ls_scf_init_scf 1 4.0 0.000 0.000 29.442 29.446 mp_waitall_1 5436 11.0 24.452 28.604 24.452 28.604 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.182 28.214 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.072 26.080 multiply_cannon_multrec 444 9.7 13.673 16.141 20.844 22.274 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.616 15.734 make_m2s 222 7.7 0.005 0.005 13.366 14.319 make_images 222 8.7 2.041 2.477 13.298 14.249 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.509 14.059 hybrid_alltoall_any 227 10.6 0.802 3.824 7.989 9.605 make_images_data 222 9.7 0.003 0.004 8.144 9.467 multiply_cannon_sync_h2d 444 9.7 6.727 8.792 6.727 8.792 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.407 6.869 8.074 dbcsr_mm_accdrv_process_sort 3003 11.4 6.507 7.667 6.507 7.667 arnoldi_extremal 4 6.8 0.000 0.000 6.039 6.049 arnoldi_normal_ev 4 7.8 0.002 0.005 6.039 6.049 build_subspace 16 8.4 0.015 0.020 5.643 5.652 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.459 4.616 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.277 4.465 dbcsr_matrix_vector_mult_local 304 10.0 3.768 4.253 3.770 4.256 mp_sum_l 887 5.1 2.588 4.217 2.588 4.217 calculate_norms 792 9.8 3.619 3.750 3.619 3.750 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.520 3.608 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.521 3.548 mp_allgather_i34 111 8.7 1.159 3.545 1.159 3.545 mp_irecv_dv 1241 11.2 1.510 3.532 1.510 3.532 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.792 3.427 ls_scf_post 1 4.0 0.000 0.000 3.381 3.386 make_images_sizes 222 9.7 0.000 0.000 0.829 3.230 mp_alltoall_i44 222 10.7 0.829 3.230 0.829 3.230 ls_scf_store_result 1 5.0 0.000 0.000 3.176 3.207 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.278 dbcsr_data_new 4608 9.7 1.788 2.244 1.788 2.244 dbcsr_merge_all 275 8.9 0.479 0.527 2.052 2.139 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.128 2.128 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.022 2.022 rebuild_ks_matrix 3 7.3 0.000 0.000 1.990 1.990 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.990 1.990 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.401000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 8.834490E+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 0.144 0.174 106.893 106.895 qs_energies 1 2.0 0.000 0.000 104.956 104.964 ls_scf 1 3.0 0.000 0.000 102.018 102.026 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.774 74.947 ls_scf_main 1 4.0 0.000 0.000 64.043 64.044 density_matrix_trs4 2 5.0 0.002 0.003 54.918 54.982 multiply_cannon 111 7.7 0.143 0.218 48.089 49.982 multiply_cannon_loop 111 8.7 0.098 0.099 45.085 45.915 ls_scf_init_scf 1 4.0 0.000 0.000 34.193 34.194 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.485 32.513 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.790 29.797 mp_waitall_1 4527 11.1 21.581 25.767 21.581 25.767 make_m2s 222 7.7 0.005 0.005 22.829 23.914 make_images 222 8.7 3.572 3.873 22.722 23.804 multiply_cannon_multrec 444 9.7 17.798 18.378 22.500 23.113 hybrid_alltoall_any 227 10.6 1.653 3.627 12.938 15.802 make_images_data 222 9.7 0.003 0.004 13.148 15.371 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 9.832 10.346 multiply_cannon_sync_h2d 444 9.7 8.793 8.843 8.793 8.843 arnoldi_extremal 4 6.8 0.000 0.000 7.728 7.734 arnoldi_normal_ev 4 7.8 0.003 0.010 7.728 7.734 build_subspace 16 8.4 0.026 0.036 7.171 7.185 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.651 5.818 dbcsr_matrix_vector_mult_local 304 10.0 5.140 5.462 5.143 5.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.184 5.274 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.910 5.168 dbcsr_mm_accdrv_process 1814 10.4 0.290 0.357 4.516 4.645 dbcsr_mm_accdrv_process_sort 1814 11.4 4.179 4.317 4.179 4.317 make_images_sizes 222 9.7 0.000 0.000 1.555 3.808 mp_alltoall_i44 222 10.7 1.555 3.808 1.555 3.808 ls_scf_post 1 4.0 0.000 0.000 3.782 3.791 ls_scf_store_result 1 5.0 0.000 0.000 3.500 3.534 mp_allgather_i34 111 8.7 1.111 3.463 1.111 3.463 calculate_norms 792 9.8 3.227 3.268 3.227 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.068 3.136 dbcsr_complete_redistribute 5 7.6 1.432 1.473 2.778 2.919 dbcsr_merge_all 275 8.9 0.887 0.926 2.851 2.913 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.907 2.907 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.545 dbcsr_sort_data 325 11.1 2.435 2.494 2.435 2.494 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.435 2.436 rebuild_ks_matrix 3 7.3 0.000 0.000 2.368 2.370 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.368 2.370 dbcsr_new_transposed 4 7.5 0.241 0.252 2.354 2.363 dbcsr_data_new 6591 9.6 1.834 2.298 1.834 2.298 mp_alltoall_d11v 48 9.2 2.171 2.241 2.171 2.241 dbcsr_frobenius_norm 74 6.6 2.058 2.139 2.209 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.130 2.212 dbcsr_add_anytype 103 7.2 0.859 0.890 2.130 2.211 dbcsr_data_release 12724 10.6 1.969 2.188 1.969 2.188 mp_sum_l 887 5.1 1.469 2.169 1.469 2.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.895000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7078.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/185b7b66b50ae386199c6584a49c01427632d034_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 592.687104E+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 4806001. 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 0.040 0.124 235.385 235.415 qs_mol_dyn_low 1 2.0 0.006 0.060 234.194 234.225 qs_forces 5 3.8 0.004 0.007 234.021 234.030 qs_energies 5 4.8 0.018 0.056 230.914 230.942 scf_env_do_scf 5 5.8 0.000 0.002 214.515 214.521 scf_env_do_scf_inner_loop 105 6.6 0.004 0.095 186.503 186.593 qs_scf_new_mos 105 7.6 0.000 0.001 145.417 145.607 qs_scf_loop_do_ot 105 8.6 0.001 0.001 145.417 145.606 dbcsr_multiply_generic 1445 12.2 0.128 0.135 135.325 135.695 ot_scf_mini 105 9.6 0.003 0.004 135.556 135.674 multiply_cannon 1445 13.2 0.275 0.286 115.877 118.107 multiply_cannon_loop 1445 14.2 2.865 3.017 113.988 115.231 velocity_verlet 4 3.0 0.008 0.022 108.062 108.066 ot_mini 105 10.6 0.001 0.001 60.633 60.740 mp_waitall_1 488190 16.1 36.095 45.725 36.095 45.725 qs_ot_get_p 112 10.4 0.001 0.001 44.767 45.048 multiply_cannon_multrec 69360 15.2 29.605 34.618 39.862 44.688 qs_ot_get_derivative 55 11.6 0.001 0.002 38.877 38.988 multiply_cannon_metrocomm3 69360 15.2 0.208 0.217 26.232 35.930 qs_ot_p2m_diag 40 11.0 0.020 0.030 33.630 33.700 multiply_cannon_sync_h2d 69360 15.2 29.061 33.065 29.061 33.065 cp_dbcsr_syevd 40 12.0 0.002 0.003 30.223 30.225 rebuild_ks_matrix 110 8.4 0.000 0.000 29.620 29.799 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.015 29.619 29.799 init_scf_loop 7 6.6 0.001 0.009 27.973 27.976 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.128 27.291 cp_fm_syevd 40 13.0 0.000 0.000 25.070 25.201 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.096 23.355 apply_single 62 13.6 0.000 0.000 23.096 23.355 prepare_preconditioner 7 7.6 0.000 0.001 22.868 22.903 make_preconditioner 7 8.6 0.000 0.002 22.868 22.903 ot_new_cg_direction 55 11.6 0.001 0.001 21.088 21.089 cp_fm_redistribute_end 40 14.0 10.162 20.275 10.166 20.277 cp_fm_syevd_base 40 14.0 10.103 20.224 10.103 20.224 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.566 17.965 calculate_rho_elec 110 8.6 0.030 0.036 17.566 17.964 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.384 15.442 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.881 15.015 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.992 14.091 init_scf_run 5 5.8 0.000 0.002 13.235 13.236 scf_env_initial_rho_setup 5 6.8 0.002 0.003 13.235 13.236 mp_sum_l 4764 12.2 12.257 12.968 12.257 12.968 density_rs2pw 110 9.6 0.006 0.007 11.898 12.382 pw_transfer 1645 12.4 0.080 0.100 12.052 12.295 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.914 12.159 calculate_dm_sparse 110 9.5 0.000 0.001 11.833 12.097 dbcsr_mm_accdrv_process 154766 15.8 6.139 6.322 10.125 11.229 cp_fm_cholesky_invert 7 10.6 10.721 10.729 10.721 10.729 qs_vxc_create 110 10.4 0.003 0.024 10.642 10.689 fft_wrap_pw1pw2_240 915 15.0 1.141 1.237 10.442 10.644 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.518 10.611 check_diag 80 13.5 8.685 8.934 9.532 9.662 multiply_cannon_metrocomm1 69360 15.2 0.098 0.106 5.005 9.472 transfer_rs2pw 445 10.6 0.007 0.008 8.531 9.039 sum_up_and_integrate 60 10.3 0.001 0.002 8.925 8.940 fft3d_pb 915 16.0 2.396 2.685 8.645 8.929 integrate_v_rspace 60 11.3 0.002 0.002 8.908 8.923 acc_transpose_blocks 69360 15.2 0.364 0.385 7.844 8.860 calculate_first_density_matrix 1 7.0 0.000 0.006 8.296 8.315 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.255 8.270 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.783 7.835 xc_rho_set_and_dset_create 110 12.4 0.080 0.108 7.422 7.689 make_m2s 2890 13.2 0.079 0.089 6.757 7.287 make_full_single_inverse 7 9.6 0.002 0.053 7.174 7.209 xc_vxc_pw_create 60 11.3 0.039 0.053 7.161 7.209 make_images 2890 14.2 0.240 0.261 6.650 7.181 xc_pw_derive 510 13.4 0.006 0.007 6.374 6.463 acc_transpose_blocks_kernels 69360 16.2 0.852 0.893 5.082 5.995 mp_alltoall_z22v 2340 17.7 5.694 5.950 5.694 5.950 mp_waitany 7680 13.5 4.494 5.257 4.494 5.257 potential_pw2rs 60 12.3 0.003 0.003 5.116 5.130 jit_kernel_transpose 5 15.0 4.230 5.121 4.230 5.121 multiply_cannon_metrocomm4 67915 15.2 0.189 0.203 2.024 4.874 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=235.415000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.429286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 185b7b66b50ae386199c6584a49c01427632d034 Summary: empty Status: OK