=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 443e0b30de9442b60cbf758811016979827d8a6a ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/01 job id: 50227617 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/02 job id: 50227619 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/03 job id: 50227620 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/04 job id: 50227623 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/05 job id: 50227626 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/06 job id: 50227627 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/07 job id: 50227629 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/08 job id: 50227630 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/09 job id: 50227631 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/10 job id: 50227634 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/11 job id: 50227636 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/12 job id: 50227637 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/13 job id: 50227638 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/14 job id: 50227640 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/15 job id: 50227641 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/16 job id: 50227643 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/17 job id: 50227645 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/18 job id: 50227647 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/19 job id: 50227648 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/20 job id: 50227649 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/21 job id: 50227651 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/22 job id: 50227652 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/23 job id: 50227655 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/24 job id: 50227656 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/25 job id: 50227657 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/26 job id: 50227658 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/27 job id: 50227660 --- 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/443e0b30de9442b60cbf758811016979827d8a6a_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 0.026 0.042 132.447 132.448 farming_run 1 2.0 131.844 131.845 132.402 132.405 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.498788E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.026 113.842 113.842 qs_energies 1 2.0 0.000 0.000 113.622 113.624 mp2_main 1 3.0 0.000 0.000 111.568 111.570 mp2_gpw_main 1 4.0 0.027 0.033 110.579 110.581 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 91.873 91.979 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.513 54.619 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.161 40.927 45.721 get_2c_integrals 1 6.0 0.008 0.009 36.465 37.187 integrate_v_rspace 273 8.0 0.436 0.451 24.534 28.990 pw_transfer 6555 10.6 0.378 0.386 26.706 27.041 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.334 25.628 grid_integrate_task_list 273 9.0 20.458 25.329 20.458 25.329 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.873 23.162 compute_2c_integrals 1 7.0 0.002 0.002 18.936 18.938 rpa_ri_compute_en 1 5.0 0.019 0.020 18.601 18.690 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.527 18.674 mp2_eri_2c_integrate_gpw 1 9.0 2.387 2.426 18.526 18.673 cp_fm_cholesky_decompose 12 8.2 17.461 18.172 17.461 18.172 cholesky_decomp 1 7.0 0.000 0.000 16.370 17.076 fft3d_s 5443 13.4 16.081 16.327 16.103 16.349 ao_to_mo_and_store_B_mult_1 272 7.0 10.765 15.311 10.765 15.311 calculate_wavefunction 272 8.0 5.406 5.488 12.304 13.075 rpa_num_int 1 6.0 0.001 0.010 10.650 10.650 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.583 10.600 calc_mat_Q 8 8.0 0.000 0.000 9.452 9.547 contract_S_to_Q 8 9.0 0.000 0.000 8.875 8.969 calc_potential_gpw 544 9.5 0.004 0.004 8.245 8.695 parallel_gemm_fm 14 9.1 0.000 0.000 8.468 8.569 parallel_gemm_fm_cosma 14 10.1 8.468 8.569 8.468 8.569 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.043 8.269 potential_pw2rs 545 10.0 0.106 0.109 7.424 8.067 create_integ_mat 1 6.0 0.032 0.037 7.752 7.752 collocate_single_gaussian 272 10.0 0.038 0.041 7.290 7.494 array2fm 1 7.0 0.000 0.000 6.696 7.194 pw_scatter_s 2720 13.7 4.355 4.521 4.355 4.521 pw_gather_s 2722 13.2 3.398 3.829 3.398 3.829 array2fm_buffer_send 1 8.0 2.971 3.151 2.971 3.151 pw_poisson_solve 545 10.5 1.118 1.173 2.195 2.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.578581, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2815.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 0.023 0.038 420.524 420.525 farming_run 1 2.0 419.809 419.815 420.490 420.492 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.245880E+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 11728562. 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 0.011 0.031 216.945 216.945 qs_energies 1 2.0 0.000 0.000 216.708 216.726 scf_env_do_scf 1 3.0 0.000 0.000 114.815 114.816 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.892 113.898 rebuild_ks_matrix 4 6.0 0.000 0.000 113.891 113.897 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.891 113.897 hfx_ks_matrix 4 8.0 0.001 0.001 113.544 113.548 integrate_four_center 4 9.0 0.152 0.459 113.543 113.548 integrate_four_center_main 4 10.0 0.114 0.542 101.798 105.214 integrate_four_center_bin 264 11.0 101.683 105.183 101.683 105.183 mp2_main 1 3.0 0.000 0.000 101.603 101.621 mp2_gpw_main 1 4.0 0.051 0.077 100.654 100.676 init_scf_loop 1 4.0 0.000 0.000 96.828 96.828 mp2_ri_gpw_compute_in 1 5.0 0.065 0.076 73.849 74.942 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.721 54.813 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.405 46.461 integrate_v_rspace 95 8.0 0.395 0.558 27.894 32.824 pw_transfer 2240 10.6 0.146 0.165 29.462 29.807 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.485 28.848 mp2_ri_gpw_compute_en 1 5.0 0.057 0.077 26.653 28.724 grid_integrate_task_list 95 9.0 23.315 28.463 23.315 28.463 ao_to_mo_and_store_B_mult_1 91 7.0 10.641 27.740 10.641 27.740 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.245 26.626 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.896 24.899 24.910 get_2c_integrals 1 6.0 0.000 0.000 20.041 20.064 compute_2c_integrals 1 7.0 0.002 0.003 19.027 19.031 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.566 18.895 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.860 18.565 18.894 fft3d_s 1823 13.4 18.480 18.774 18.493 18.787 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.986 17.986 calculate_wavefunction 91 8.0 2.008 2.041 9.608 9.818 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.586 8.749 9.387 potential_pw2rs 186 10.0 0.033 0.035 8.431 8.961 local_gemm 172 8.0 8.193 8.825 8.193 8.825 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.123 8.561 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.527 7.926 8.395 calc_potential_gpw 182 9.5 0.002 0.002 7.858 8.073 collocate_single_gaussian 91 10.0 0.016 0.018 7.720 8.005 mp_sync 37 10.5 3.725 6.932 3.725 6.932 integrate_four_center_load 4 10.0 0.000 0.000 6.751 6.756 hfx_load_balance 1 11.0 0.000 0.000 6.751 6.756 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.452 6.341 6.452 mp_sendrecv_dm3 2068 8.0 5.957 6.401 5.957 6.401 pw_gather_s 912 13.2 4.501 4.923 4.501 4.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.647171, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 453.476352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.078 53.934 53.936 qs_mol_dyn_low 1 2.0 0.003 0.005 53.606 53.615 qs_forces 11 3.9 0.002 0.002 53.538 53.539 qs_energies 11 4.9 0.001 0.002 52.067 52.083 scf_env_do_scf 11 5.9 0.000 0.001 45.910 45.910 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.909 43.909 qs_scf_new_mos 108 7.5 0.000 0.001 34.039 34.331 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.038 34.331 dbcsr_multiply_generic 2286 12.5 0.092 0.096 33.728 34.130 ot_scf_mini 108 9.5 0.002 0.002 32.361 32.557 multiply_cannon 2286 13.5 0.194 0.208 26.126 27.589 multiply_cannon_loop 2286 14.5 1.816 1.922 25.485 26.979 velocity_verlet 10 3.0 0.001 0.002 25.692 25.693 ot_mini 108 10.5 0.001 0.001 19.361 19.623 qs_ot_get_derivative 108 11.5 0.001 0.001 16.370 16.590 mp_waitall_1 245248 16.5 8.384 14.596 8.384 14.596 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.014 12.777 multiply_cannon_multrec 54864 15.5 3.667 5.779 7.636 10.981 qs_ot_get_p 119 10.4 0.001 0.001 8.267 8.564 rebuild_ks_matrix 119 8.3 0.000 0.000 7.740 7.897 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.740 7.896 mp_sum_l 7287 12.8 5.692 7.443 5.692 7.443 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.814 6.951 multiply_cannon_sync_h2d 54864 15.5 5.177 6.551 5.177 6.551 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.502 5.989 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.645 5.754 dbcsr_mm_accdrv_process 76910 16.1 1.831 2.883 3.882 5.566 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.441 5.500 init_scf_run 11 5.9 0.000 0.001 4.881 4.881 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.881 4.881 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.580 4.580 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.389 4.390 sum_up_and_integrate 119 10.3 0.001 0.002 4.372 4.379 integrate_v_rspace 119 11.3 0.002 0.002 4.362 4.370 cp_fm_redistribute_end 50 14.0 2.238 4.366 2.244 4.369 cp_fm_diag_elpa_base 50 14.0 2.119 4.269 2.123 4.275 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.045 4.162 calculate_rho_elec 119 8.7 0.011 0.016 4.045 4.162 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.892 3.099 apply_single 119 13.6 0.000 0.000 2.892 3.099 calculate_dm_sparse 119 9.5 0.000 0.000 2.951 3.091 acc_transpose_blocks 54864 15.5 0.223 0.245 2.197 2.718 calculate_first_density_matrix 1 7.0 0.000 0.000 2.701 2.704 ot_diis_step 108 11.5 0.006 0.006 2.704 2.704 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.737 2.654 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.527 2.609 jit_kernel_multiply 13 15.8 1.988 2.608 1.988 2.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.333 2.334 density_rs2pw 119 9.7 0.004 0.004 2.076 2.171 grid_integrate_task_list 119 12.3 2.013 2.139 2.013 2.139 wfi_extrapolate 11 7.9 0.001 0.001 2.112 2.112 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.033 2.089 init_scf_loop 11 6.9 0.000 0.000 1.984 1.985 mp_sum_d 4137 12.0 1.343 1.903 1.343 1.903 potential_pw2rs 119 12.3 0.004 0.004 1.757 1.767 pw_transfer 1439 11.6 0.051 0.056 1.601 1.672 make_m2s 4572 13.5 0.053 0.055 1.619 1.663 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.526 1.599 make_images 4572 14.5 0.133 0.139 1.536 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.009 0.041 1.435 1.461 acc_transpose_blocks_sync 164592 16.5 1.196 1.434 1.196 1.434 mp_waitany 12084 13.8 1.248 1.430 1.248 1.430 grid_collocate_task_list 119 9.7 1.352 1.424 1.352 1.424 mp_alltoall_d11v 2130 13.8 1.272 1.403 1.272 1.403 transfer_rs2pw 487 10.6 0.005 0.006 1.321 1.400 fft3d_ps 1201 14.6 0.371 0.477 1.240 1.307 transfer_pw2rs 487 13.2 0.006 0.006 1.290 1.298 fft_wrap_pw1pw2_140 487 13.2 0.107 0.114 1.184 1.254 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.739 1.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.936000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.406464E+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 1032117. 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 0.008 0.028 41.121 41.122 qs_mol_dyn_low 1 2.0 0.003 0.003 40.899 40.907 qs_forces 11 3.9 0.002 0.003 40.833 40.834 qs_energies 11 4.9 0.001 0.001 39.138 39.143 scf_env_do_scf 11 5.9 0.000 0.001 32.422 32.422 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 29.867 29.868 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.421 22.802 qs_scf_new_mos 108 7.5 0.001 0.001 21.152 21.381 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.152 21.380 ot_scf_mini 108 9.5 0.002 0.003 20.203 20.361 velocity_verlet 10 3.0 0.001 0.002 18.804 18.805 multiply_cannon 2286 13.5 0.210 0.218 17.038 18.486 multiply_cannon_loop 2286 14.5 1.187 1.244 15.826 17.339 ot_mini 108 10.5 0.001 0.001 12.403 12.630 mp_waitall_1 200699 16.5 5.720 10.836 5.720 10.836 qs_ot_get_derivative 108 11.5 0.001 0.001 10.021 10.182 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.149 9.506 multiply_cannon_multrec 27432 15.5 1.831 4.335 6.160 9.050 rebuild_ks_matrix 119 8.3 0.000 0.000 6.831 6.980 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.830 6.980 dbcsr_mm_accdrv_process 47894 16.0 3.477 5.699 4.248 6.322 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.000 6.135 init_scf_run 11 5.9 0.000 0.001 5.494 5.494 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.494 5.494 qs_ot_get_p 119 10.4 0.001 0.001 4.867 5.076 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.822 4.640 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.983 4.076 apply_single 119 13.6 0.000 0.000 2.982 4.076 mp_sum_l 7287 12.8 2.114 3.961 2.114 3.961 calculate_first_density_matrix 1 7.0 0.000 0.000 3.902 3.905 sum_up_and_integrate 119 10.3 0.001 0.001 3.743 3.749 integrate_v_rspace 119 11.3 0.002 0.002 3.729 3.734 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.689 3.730 calculate_rho_elec 119 8.7 0.021 0.024 3.688 3.729 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.133 3.154 make_m2s 4572 13.5 0.052 0.053 2.817 3.122 make_images 4572 14.5 0.204 0.242 2.727 3.035 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.726 2.726 init_scf_loop 11 6.9 0.000 0.000 2.525 2.526 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.409 2.494 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.379 2.379 cp_fm_redistribute_end 50 14.0 1.202 2.347 1.206 2.350 ot_diis_step 108 11.5 0.011 0.011 2.330 2.331 cp_fm_diag_elpa_base 50 14.0 1.111 2.255 1.140 2.292 multiply_cannon_sync_h2d 27432 15.5 1.653 2.202 1.653 2.202 calculate_dm_sparse 119 9.5 0.000 0.001 2.092 2.171 density_rs2pw 119 9.7 0.004 0.004 1.983 2.077 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.034 2.037 acc_transpose_blocks 27432 15.5 0.113 0.119 1.639 1.987 grid_integrate_task_list 119 12.3 1.831 1.954 1.831 1.954 pw_transfer 1439 11.6 0.064 0.068 1.914 1.951 parallel_gemm_fm 81 9.0 0.000 0.000 1.932 1.936 parallel_gemm_fm_cosma 81 10.0 1.932 1.935 1.932 1.935 jit_kernel_multiply 10 16.3 0.713 1.916 0.713 1.916 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.823 1.861 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.814 1.856 make_basis_sm 11 9.8 0.006 0.010 1.819 1.823 prepare_preconditioner 11 7.9 0.000 0.000 1.593 1.618 make_preconditioner 11 8.9 0.000 0.000 1.593 1.618 make_images_data 4572 15.5 0.046 0.053 1.273 1.617 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.496 1.552 wfi_extrapolate 11 7.9 0.001 0.001 1.529 1.529 potential_pw2rs 119 12.3 0.006 0.006 1.486 1.493 fft3d_ps 1201 14.6 0.522 0.579 1.455 1.486 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.128 1.469 mp_alltoall_d11v 2130 13.8 1.278 1.468 1.278 1.468 fft_wrap_pw1pw2_140 487 13.2 0.127 0.130 1.425 1.462 grid_collocate_task_list 119 9.7 1.288 1.357 1.288 1.357 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.240 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.272 transfer_rs2pw 487 10.6 0.005 0.005 1.122 1.219 mp_allgather_i34 2286 14.5 0.636 1.057 0.636 1.057 mp_sum_d 4137 12.0 0.586 1.044 0.586 1.044 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 0.934 0.935 acc_transpose_blocks_sync 82296 16.5 0.812 0.931 0.812 0.931 transfer_pw2rs 487 13.2 0.004 0.005 0.915 0.920 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.689 0.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.890 0.904 make_images_sizes 4572 15.5 0.005 0.005 0.628 0.878 mp_alltoall_i44 4572 16.5 0.623 0.874 0.623 0.874 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.122000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.272727, yerr=1.285649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 520.929280E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 33.752 33.753 qs_mol_dyn_low 1 2.0 0.003 0.003 33.394 33.451 qs_forces 11 3.9 0.002 0.003 33.298 33.299 qs_energies 11 4.9 0.013 0.039 31.740 31.742 scf_env_do_scf 11 5.9 0.001 0.001 26.532 26.532 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.075 24.075 dbcsr_multiply_generic 2286 12.5 0.093 0.096 17.385 17.451 qs_scf_new_mos 108 7.5 0.001 0.001 16.018 16.039 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.017 16.039 velocity_verlet 10 3.0 0.002 0.002 15.577 15.580 ot_scf_mini 108 9.5 0.002 0.002 15.244 15.262 multiply_cannon 2286 13.5 0.195 0.203 13.845 14.519 multiply_cannon_loop 2286 14.5 0.870 0.922 13.022 13.723 ot_mini 108 10.5 0.001 0.001 9.335 9.350 qs_ot_get_derivative 108 11.5 0.001 0.001 7.842 7.860 multiply_cannon_multrec 18288 15.5 1.881 2.894 7.001 7.344 rebuild_ks_matrix 119 8.3 0.000 0.000 6.147 6.161 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.024 6.146 6.161 dbcsr_mm_accdrv_process 38222 16.0 4.968 5.794 5.025 5.857 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.431 5.444 mp_waitall_1 158411 16.6 2.802 3.983 2.802 3.983 init_scf_run 11 5.9 0.000 0.001 3.893 3.893 scf_env_initial_rho_setup 11 6.9 0.001 0.005 3.893 3.893 sum_up_and_integrate 119 10.3 0.001 0.002 3.710 3.716 integrate_v_rspace 119 11.3 0.002 0.003 3.698 3.705 qs_ot_get_p 119 10.4 0.001 0.001 3.629 3.659 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.923 3.549 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.484 3.490 calculate_rho_elec 119 8.7 0.031 0.032 3.483 3.489 calculate_first_density_matrix 1 7.0 0.004 0.031 2.660 2.662 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.439 2.442 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.428 2.434 init_scf_loop 11 6.9 0.000 0.002 2.432 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.038 2.349 apply_single 119 13.6 0.000 0.000 2.038 2.349 density_rs2pw 119 9.7 0.004 0.004 2.033 2.151 make_m2s 4572 13.5 0.044 0.045 1.995 2.143 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.114 2.114 make_images 4572 14.5 0.190 0.202 1.909 2.055 pw_transfer 1439 11.6 0.065 0.068 1.992 2.006 grid_integrate_task_list 119 12.3 1.797 1.927 1.797 1.927 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.899 1.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.912 1.914 calculate_dm_sparse 119 9.5 0.000 0.001 1.860 1.868 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.829 1.837 cp_fm_diag_elpa_base 50 14.0 1.803 1.814 1.827 1.836 acc_transpose_blocks 18288 15.5 0.079 0.080 1.715 1.831 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.821 1.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.685 1.690 prepare_preconditioner 11 7.9 0.000 0.000 1.673 1.676 make_preconditioner 11 8.9 0.000 0.000 1.673 1.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.531 1.614 mp_sum_l 7287 12.8 1.166 1.535 1.166 1.535 potential_pw2rs 119 12.3 0.007 0.008 1.496 1.503 fft_wrap_pw1pw2_140 487 13.2 0.179 0.181 1.481 1.497 ot_diis_step 108 11.5 0.012 0.012 1.478 1.478 fft3d_ps 1201 14.6 0.549 0.574 1.455 1.470 grid_collocate_task_list 119 9.7 1.233 1.334 1.233 1.334 transfer_rs2pw 487 10.6 0.005 0.005 1.126 1.242 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.187 1.191 multiply_cannon_sync_h2d 18288 15.5 1.012 1.184 1.012 1.184 wfi_extrapolate 11 7.9 0.001 0.001 1.177 1.177 make_images_data 4572 15.5 0.046 0.050 0.831 1.011 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.979 0.996 qs_energies_init_hamiltonians 11 5.9 0.019 0.021 0.972 0.973 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.723 0.913 acc_transpose_blocks_kernels 18288 16.5 0.218 0.224 0.866 0.912 transfer_pw2rs 487 13.2 0.004 0.004 0.903 0.909 acc_transpose_blocks_sync 54864 16.5 0.752 0.887 0.752 0.887 mp_alltoall_d11v 2130 13.8 0.745 0.847 0.745 0.847 mp_alltoall_z22v 1201 16.6 0.741 0.811 0.741 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.789 0.790 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.682 0.751 cp_fm_cholesky_invert 11 10.9 0.722 0.726 0.722 0.726 jit_kernel_transpose 5 15.6 0.648 0.692 0.648 0.692 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.753000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 562.708480E+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 0.062 0.126 36.581 36.582 qs_mol_dyn_low 1 2.0 0.003 0.005 35.991 35.999 qs_forces 11 3.9 0.002 0.003 35.933 35.934 qs_energies 11 4.9 0.003 0.008 34.188 34.193 scf_env_do_scf 11 5.9 0.001 0.003 29.142 29.144 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.784 25.785 dbcsr_multiply_generic 2286 12.5 0.099 0.102 19.792 19.909 velocity_verlet 10 3.0 0.001 0.002 18.382 18.384 qs_scf_new_mos 108 7.5 0.001 0.001 17.732 17.791 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.731 17.790 ot_scf_mini 108 9.5 0.002 0.004 16.696 16.750 multiply_cannon 2286 13.5 0.219 0.226 16.046 16.538 multiply_cannon_loop 2286 14.5 1.557 1.633 15.093 15.461 ot_mini 108 10.5 0.001 0.001 10.278 10.346 multiply_cannon_multrec 27432 15.5 2.451 3.139 8.996 9.267 qs_ot_get_derivative 108 11.5 0.001 0.001 8.424 8.482 dbcsr_mm_accdrv_process 47916 15.9 5.852 7.679 6.442 7.751 rebuild_ks_matrix 119 8.3 0.000 0.000 6.344 6.387 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.018 6.343 6.386 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.584 5.621 qs_ot_get_p 119 10.4 0.001 0.002 3.574 3.648 init_scf_run 11 5.9 0.000 0.002 3.643 3.644 scf_env_initial_rho_setup 11 6.9 0.002 0.003 3.643 3.643 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.045 3.537 sum_up_and_integrate 119 10.3 0.001 0.002 3.499 3.514 integrate_v_rspace 119 11.3 0.002 0.003 3.488 3.503 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.364 3.402 calculate_rho_elec 119 8.7 0.040 0.046 3.364 3.402 init_scf_loop 11 6.9 0.001 0.004 3.339 3.343 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.193 2.655 apply_single 119 13.6 0.000 0.000 2.193 2.655 acc_transpose_blocks 27432 15.5 0.119 0.122 2.455 2.593 prepare_preconditioner 11 7.9 0.000 0.000 2.485 2.493 make_preconditioner 11 8.9 0.000 0.001 2.485 2.493 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.089 2.418 make_m2s 4572 13.5 0.054 0.056 2.248 2.370 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.321 2.348 make_images 4572 14.5 0.272 0.337 2.141 2.261 calculate_first_density_matrix 1 7.0 0.001 0.005 2.249 2.251 calculate_dm_sparse 119 9.5 0.000 0.000 2.149 2.200 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.161 2.173 pw_transfer 1439 11.6 0.064 0.068 2.018 2.057 mp_waitall_1 137007 16.6 1.607 2.031 1.607 2.031 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.927 1.969 density_rs2pw 119 9.7 0.004 0.004 1.789 1.920 grid_integrate_task_list 119 12.3 1.816 1.904 1.816 1.904 ot_diis_step 108 11.5 0.012 0.013 1.812 1.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.802 1.803 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.757 1.757 mp_sum_l 7287 12.8 1.109 1.714 1.109 1.714 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.602 1.648 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.646 acc_transpose_blocks_sync 82296 16.5 1.472 1.596 1.472 1.596 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.524 1.533 cp_fm_diag_elpa_base 50 14.0 1.489 1.505 1.520 1.529 jit_kernel_multiply 8 15.8 0.525 1.480 0.525 1.480 fft3d_ps 1201 14.6 0.591 0.650 1.415 1.446 grid_collocate_task_list 119 9.7 1.249 1.352 1.249 1.352 wfi_extrapolate 11 7.9 0.001 0.001 1.340 1.340 potential_pw2rs 119 12.3 0.008 0.009 1.309 1.316 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.291 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.240 1.250 cp_fm_upper_to_full 72 14.2 0.814 1.160 0.814 1.160 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.666 1.145 qs_energies_init_hamiltonians 11 5.9 0.033 0.046 1.132 1.133 transfer_rs2pw 487 10.6 0.004 0.005 0.885 1.019 dbcsr_complete_redistribute 329 12.2 0.116 0.139 0.746 1.017 make_images_data 4572 15.5 0.047 0.051 0.829 0.948 mp_alltoall_d11v 2130 13.8 0.765 0.913 0.765 0.913 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.705 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.855 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.836 0.848 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.561 0.831 mp_alltoall_z22v 1201 16.6 0.707 0.739 0.707 0.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.582000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.818182, yerr=3.712832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 625.930240E+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 0.011 0.031 30.018 30.019 qs_mol_dyn_low 1 2.0 0.003 0.008 29.704 29.712 qs_forces 11 3.9 0.032 0.050 29.633 29.634 qs_energies 11 4.9 0.002 0.002 27.885 27.904 scf_env_do_scf 11 5.9 0.000 0.001 22.268 22.268 scf_env_do_scf_inner_loop 108 6.5 0.005 0.008 19.735 19.736 velocity_verlet 10 3.0 0.002 0.002 14.882 14.886 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.234 13.311 qs_scf_new_mos 108 7.5 0.001 0.001 12.009 12.034 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.008 12.033 ot_scf_mini 108 9.5 0.002 0.002 11.295 11.317 multiply_cannon 2286 13.5 0.225 0.231 10.542 11.080 multiply_cannon_loop 2286 14.5 0.644 0.665 9.609 9.792 ot_mini 108 10.5 0.001 0.001 6.502 6.530 multiply_cannon_multrec 9144 15.5 1.705 1.984 6.015 6.279 rebuild_ks_matrix 119 8.3 0.000 0.000 5.840 5.862 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.839 5.861 qs_ot_get_derivative 108 11.5 0.001 0.001 5.210 5.232 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.190 5.209 dbcsr_mm_accdrv_process 12550 15.8 3.743 4.235 4.197 4.274 init_scf_run 11 5.9 0.000 0.001 4.106 4.106 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.106 4.106 sum_up_and_integrate 119 10.3 0.001 0.001 3.475 3.480 integrate_v_rspace 119 11.3 0.003 0.003 3.465 3.469 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.392 3.397 calculate_rho_elec 119 8.7 0.060 0.061 3.391 3.397 qs_ot_get_p 119 10.4 0.001 0.001 2.921 2.950 calculate_first_density_matrix 1 7.0 0.000 0.000 2.918 2.919 init_scf_loop 11 6.9 0.000 0.000 2.512 2.513 pw_transfer 1439 11.6 0.065 0.068 2.150 2.158 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.056 2.067 make_m2s 4572 13.5 0.033 0.034 1.873 2.032 mp_waitall_1 115863 16.7 1.461 1.947 1.461 1.947 grid_integrate_task_list 119 12.3 1.862 1.946 1.862 1.946 make_images 4572 14.5 0.267 0.298 1.784 1.941 density_rs2pw 119 9.7 0.003 0.003 1.805 1.888 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.874 1.877 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.776 1.797 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.786 1.787 calculate_dm_sparse 119 9.5 0.000 0.000 1.764 1.783 prepare_preconditioner 11 7.9 0.000 0.000 1.739 1.743 make_preconditioner 11 8.9 0.000 0.000 1.739 1.743 fft_wrap_pw1pw2_140 487 13.2 0.324 0.333 1.694 1.704 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.626 1.654 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.627 1.627 acc_transpose_blocks 9144 15.5 0.042 0.043 1.425 1.455 fft3d_ps 1201 14.6 0.650 0.663 1.409 1.422 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.388 1.396 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.374 1.386 jit_kernel_multiply 8 15.6 0.415 1.376 0.415 1.376 grid_collocate_task_list 119 9.7 1.296 1.361 1.296 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.349 1.358 cp_fm_diag_elpa_base 50 14.0 1.322 1.340 1.347 1.356 qs_energies_init_hamiltonians 11 5.9 0.042 0.044 1.287 1.304 potential_pw2rs 119 12.3 0.010 0.011 1.286 1.288 ot_diis_step 108 11.5 0.013 0.013 1.278 1.279 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.222 1.254 apply_single 119 13.6 0.000 0.000 1.221 1.254 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.148 1.153 wfi_extrapolate 11 7.9 0.001 0.001 1.135 1.135 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.781 1.019 make_images_data 4572 15.5 0.041 0.046 0.797 0.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.925 transfer_rs2pw 487 10.6 0.004 0.004 0.815 0.893 mp_alltoall_d11v 2130 13.8 0.781 0.884 0.781 0.884 cp_fm_cholesky_invert 11 10.9 0.867 0.870 0.867 0.870 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.830 0.837 dbcsr_complete_redistribute 329 12.2 0.304 0.398 0.754 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.770 0.773 acc_transpose_blocks_sync 27432 16.5 0.721 0.753 0.721 0.753 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.358 0.747 qs_env_update_s_mstruct 11 6.9 0.001 0.004 0.693 0.731 parallel_gemm_fm 81 9.0 0.000 0.000 0.726 0.727 parallel_gemm_fm_cosma 81 10.0 0.726 0.727 0.726 0.727 mp_allgather_i34 2286 14.5 0.249 0.686 0.249 0.686 make_basis_sm 11 9.8 0.004 0.005 0.685 0.686 transfer_pw2rs 487 13.2 0.003 0.004 0.672 0.673 mp_alltoall_z22v 1201 16.6 0.634 0.668 0.634 0.668 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.646 0.649 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.019000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.636364, yerr=5.645154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 791.871488E+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 0.013 0.035 43.856 43.857 qs_mol_dyn_low 1 2.0 0.003 0.003 43.543 43.550 qs_forces 11 3.9 0.002 0.003 43.391 43.391 qs_energies 11 4.9 0.001 0.001 41.373 41.377 scf_env_do_scf 11 5.9 0.001 0.001 35.361 35.361 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.425 27.427 velocity_verlet 10 3.0 0.002 0.002 24.595 24.612 dbcsr_multiply_generic 2286 12.5 0.104 0.106 19.339 19.532 qs_scf_new_mos 108 7.5 0.001 0.001 17.584 17.679 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.583 17.678 ot_scf_mini 108 9.5 0.002 0.002 16.415 16.515 multiply_cannon 2286 13.5 0.298 0.308 15.225 16.170 multiply_cannon_loop 2286 14.5 0.871 0.891 13.928 14.865 ot_mini 108 10.5 0.001 0.001 9.929 10.045 multiply_cannon_multrec 9144 15.5 3.452 4.899 8.844 8.942 qs_ot_get_derivative 108 11.5 0.001 0.001 7.811 7.911 init_scf_loop 11 6.9 0.000 0.000 7.908 7.911 rebuild_ks_matrix 119 8.3 0.000 0.000 7.356 7.498 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.355 7.498 prepare_preconditioner 11 7.9 0.000 0.000 6.922 6.937 make_preconditioner 11 8.9 0.000 0.000 6.922 6.937 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.451 6.814 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.627 6.757 dbcsr_mm_accdrv_process 12550 15.8 4.538 6.188 5.254 6.634 cp_fm_upper_to_full 72 14.2 3.169 4.588 3.169 4.588 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.282 4.296 calculate_rho_elec 119 8.7 0.117 0.120 4.282 4.295 init_scf_run 11 5.9 0.000 0.001 3.947 3.947 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.946 3.947 sum_up_and_integrate 119 10.3 0.001 0.001 3.892 3.899 integrate_v_rspace 119 11.3 0.003 0.003 3.881 3.888 qs_ot_get_p 119 10.4 0.001 0.001 3.506 3.639 mp_waitall_1 94719 16.7 2.326 3.436 2.326 3.436 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.783 3.245 pw_transfer 1439 11.6 0.069 0.070 3.063 3.069 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.965 2.970 dbcsr_complete_redistribute 329 12.2 0.281 0.288 2.040 2.875 make_m2s 4572 13.5 0.037 0.038 2.437 2.614 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.316 2.556 apply_single 119 13.6 0.000 0.000 2.316 2.556 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.714 2.538 fft_wrap_pw1pw2_140 487 13.2 0.649 0.651 2.527 2.534 make_images 4572 14.5 0.350 0.383 2.316 2.493 calculate_first_density_matrix 1 7.0 0.000 0.000 2.431 2.434 mp_alltoall_i22 627 13.8 1.503 2.335 1.503 2.335 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.322 2.329 density_rs2pw 119 9.7 0.003 0.003 2.282 2.297 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.466 2.280 calculate_dm_sparse 119 9.5 0.000 0.000 2.256 2.277 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.105 2.156 grid_integrate_task_list 119 12.3 2.082 2.126 2.082 2.126 ot_diis_step 108 11.5 0.014 0.015 2.089 2.090 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.085 2.087 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.059 2.061 acc_transpose_blocks 9144 15.5 0.044 0.046 1.917 1.948 fft3d_ps 1201 14.6 0.868 0.887 1.886 1.891 mp_sum_l 7287 12.8 1.132 1.804 1.132 1.804 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.770 1.771 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.740 1.755 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.716 1.717 grid_collocate_task_list 119 9.7 1.523 1.547 1.523 1.547 cp_fm_cholesky_invert 11 10.9 1.509 1.513 1.509 1.513 wfi_extrapolate 11 7.9 0.001 0.001 1.451 1.451 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.410 1.410 cp_fm_diag_elpa_base 50 14.0 1.263 1.318 1.408 1.408 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.114 1.385 potential_pw2rs 119 12.3 0.014 0.014 1.369 1.371 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.336 1.357 mp_alltoall_d11v 2130 13.8 1.250 1.297 1.250 1.297 make_images_data 4572 15.5 0.046 0.048 1.062 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.259 1.269 acc_transpose_blocks_sync 27432 16.5 1.116 1.155 1.116 1.155 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.103 1.120 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.988 1.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.940 0.952 qs_create_task_list 11 7.9 0.000 0.000 0.929 0.940 generate_qs_task_list 11 8.9 0.368 0.390 0.929 0.940 mp_alltoall_z22v 1201 16.6 0.882 0.905 0.882 0.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.857000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=738.090909, yerr=14.041850 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.484416E+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 943377. 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 0.009 0.031 83.603 83.605 qs_mol_dyn_low 1 2.0 0.003 0.003 83.271 83.281 qs_forces 11 3.9 0.003 0.004 82.687 82.687 qs_energies 11 4.9 0.001 0.002 79.798 79.820 scf_env_do_scf 11 5.9 0.000 0.001 69.885 69.887 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 64.404 64.405 dbcsr_multiply_generic 2055 12.4 0.103 0.106 51.480 51.712 qs_scf_new_mos 99 7.5 0.000 0.001 47.432 47.578 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.431 47.578 ot_scf_mini 99 9.5 0.002 0.002 45.035 45.181 multiply_cannon 2055 13.4 0.184 0.189 43.248 43.996 velocity_verlet 10 3.0 0.001 0.001 43.545 43.551 multiply_cannon_loop 2055 14.4 1.785 1.828 42.252 42.951 ot_mini 99 10.5 0.001 0.001 26.855 26.985 qs_ot_get_derivative 99 11.5 0.001 0.001 20.091 20.204 multiply_cannon_multrec 49320 15.4 11.390 12.087 17.857 18.600 rebuild_ks_matrix 110 8.3 0.000 0.001 14.468 14.624 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.467 14.623 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.666 12.792 mp_waitall_1 220248 16.4 10.796 11.442 10.796 11.442 multiply_cannon_sync_h2d 49320 15.4 9.635 10.191 9.635 10.191 qs_ot_get_p 110 10.4 0.001 0.001 9.776 9.898 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.744 8.309 init_scf_run 11 5.9 0.000 0.001 7.859 7.859 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.859 7.859 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.244 7.801 apply_single 110 13.6 0.000 0.000 7.244 7.801 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.552 7.627 sum_up_and_integrate 110 10.3 0.002 0.003 7.002 7.016 integrate_v_rspace 110 11.3 0.002 0.003 6.977 6.997 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.603 6.640 dbcsr_mm_accdrv_process 87628 16.1 3.051 3.117 6.336 6.624 ot_diis_step 99 11.5 0.006 0.006 6.576 6.576 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.300 6.414 calculate_rho_elec 110 8.6 0.020 0.024 6.300 6.414 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.785 5.785 init_scf_loop 11 6.9 0.000 0.000 5.453 5.454 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.372 5.435 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.214 5.240 cp_fm_diag_elpa_base 48 14.0 5.200 5.228 5.212 5.238 mp_sum_l 6594 12.7 3.904 4.726 3.904 4.726 wfi_extrapolate 11 7.9 0.001 0.001 4.029 4.029 make_m2s 4110 13.4 0.060 0.065 3.872 4.005 calculate_dm_sparse 110 9.5 0.001 0.001 3.909 4.003 make_images 4110 14.4 0.178 0.190 3.776 3.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.824 3.828 calculate_first_density_matrix 1 7.0 0.000 0.000 3.721 3.734 density_rs2pw 110 9.6 0.004 0.004 3.328 3.526 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.460 3.515 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.411 3.509 grid_integrate_task_list 110 12.3 3.275 3.425 3.275 3.425 prepare_preconditioner 11 7.9 0.000 0.000 3.294 3.312 make_preconditioner 11 8.9 0.000 0.000 3.294 3.312 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.124 3.174 pw_transfer 1331 11.6 0.054 0.067 3.075 3.156 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.080 3.129 jit_kernel_multiply 13 15.9 3.006 3.107 3.006 3.107 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.987 3.072 acc_transpose_blocks 49320 15.4 0.210 0.217 2.703 2.753 fft_wrap_pw1pw2_140 451 13.1 0.308 0.331 2.552 2.638 potential_pw2rs 110 12.3 0.005 0.006 2.596 2.615 mp_alltoall_d11v 2046 13.8 2.055 2.511 2.055 2.511 fft3d_ps 1111 14.6 0.798 0.890 2.281 2.358 grid_collocate_task_list 110 9.6 2.160 2.297 2.160 2.297 mp_waitany 14300 13.8 1.810 2.124 1.810 2.124 transfer_rs2pw 451 10.6 0.005 0.006 1.871 2.052 parallel_gemm_fm 81 9.0 0.000 0.000 1.972 1.977 parallel_gemm_fm_cosma 81 10.0 1.971 1.977 1.971 1.977 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.931 1.961 mp_sum_d 3891 11.9 1.336 1.908 1.336 1.908 make_images_data 4110 15.4 0.043 0.047 1.741 1.874 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.808 1.847 cp_fm_cholesky_invert 11 10.9 1.837 1.841 1.837 1.841 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.518 1.804 transfer_pw2rs 451 13.1 0.006 0.007 1.760 1.776 make_basis_sm 11 9.8 0.002 0.011 1.708 1.710 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.605000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.363636, yerr=2.496278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 594.526208E+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 1877833. 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 0.042 0.136 69.700 69.737 qs_mol_dyn_low 1 2.0 0.003 0.006 69.368 69.378 qs_forces 11 3.9 0.003 0.003 68.601 68.603 qs_energies 11 4.9 0.002 0.007 65.287 65.291 scf_env_do_scf 11 5.9 0.001 0.002 56.681 56.684 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 49.058 49.058 dbcsr_multiply_generic 2055 12.4 0.115 0.132 37.991 38.156 velocity_verlet 10 3.0 0.001 0.002 36.632 36.648 qs_scf_new_mos 99 7.5 0.001 0.001 33.255 33.372 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.255 33.371 multiply_cannon 2055 13.4 0.224 0.243 31.334 32.312 ot_scf_mini 99 9.5 0.003 0.004 31.580 31.696 multiply_cannon_loop 2055 14.4 1.157 1.191 30.180 31.043 ot_mini 99 10.5 0.001 0.001 18.478 18.598 multiply_cannon_multrec 24660 15.4 6.972 8.186 14.092 15.383 rebuild_ks_matrix 110 8.3 0.000 0.000 13.478 13.628 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.477 13.628 qs_ot_get_derivative 99 11.5 0.001 0.001 12.713 12.833 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.855 11.994 mp_waitall_1 176588 16.5 7.622 10.257 7.622 10.257 multiply_cannon_metrocomm3 24660 15.4 0.074 0.075 5.227 8.169 init_scf_loop 11 6.9 0.001 0.003 7.588 7.589 multiply_cannon_sync_h2d 24660 15.4 6.362 7.533 6.362 7.533 dbcsr_mm_accdrv_process 52282 16.1 5.670 6.684 6.951 7.277 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.458 7.103 apply_single 110 13.6 0.000 0.001 6.457 7.103 qs_ot_get_p 110 10.4 0.001 0.002 6.484 6.633 sum_up_and_integrate 110 10.3 0.002 0.003 6.352 6.364 integrate_v_rspace 110 11.3 0.002 0.003 6.325 6.336 init_scf_run 11 5.9 0.000 0.001 6.179 6.180 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.179 6.180 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.891 5.900 calculate_rho_elec 110 8.6 0.039 0.047 5.890 5.899 ot_diis_step 99 11.5 0.011 0.011 5.716 5.716 prepare_preconditioner 11 7.9 0.000 0.000 5.579 5.596 make_preconditioner 11 8.9 0.000 0.002 5.579 5.596 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.797 5.524 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.169 5.323 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.596 4.616 make_m2s 4110 13.4 0.057 0.060 4.233 4.562 make_images 4110 14.4 0.403 0.466 4.121 4.445 cp_dbcsr_syevd 48 12.0 0.003 0.006 4.131 4.131 pw_transfer 1331 11.6 0.066 0.074 3.567 3.707 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.461 3.605 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.561 3.573 cp_fm_diag_elpa_base 48 14.0 3.511 3.526 3.556 3.567 wfi_extrapolate 11 7.9 0.001 0.001 3.489 3.489 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.300 3.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.303 3.305 grid_integrate_task_list 110 12.3 3.167 3.296 3.167 3.296 density_rs2pw 110 9.6 0.004 0.004 3.142 3.291 fft_wrap_pw1pw2_140 451 13.1 0.357 0.378 2.952 3.095 calculate_dm_sparse 110 9.5 0.001 0.001 3.004 3.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.958 3.032 fft3d_ps 1111 14.6 1.115 1.347 2.554 2.716 make_images_data 4110 15.4 0.048 0.053 2.288 2.631 cp_fm_cholesky_invert 11 10.9 2.598 2.605 2.598 2.605 calculate_first_density_matrix 1 7.0 0.000 0.003 2.592 2.595 hybrid_alltoall_any 4261 16.3 0.104 0.449 1.986 2.576 mp_sum_l 6594 12.7 1.815 2.479 1.815 2.479 grid_collocate_task_list 110 9.6 2.168 2.312 2.168 2.312 potential_pw2rs 110 12.3 0.008 0.009 2.245 2.253 acc_transpose_blocks 24660 15.4 0.116 0.119 1.977 2.043 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.976 1.991 jit_kernel_multiply 10 16.3 0.916 1.938 0.916 1.938 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.910 1.912 mp_alltoall_d11v 2046 13.8 1.737 1.878 1.737 1.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.780 1.794 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.773 1.717 transfer_rs2pw 451 10.6 0.006 0.007 1.472 1.632 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.564 1.574 mp_irecv_dv 57340 16.2 0.649 1.502 0.649 1.502 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.491 mp_allgather_i34 2055 14.4 0.461 1.476 0.461 1.476 dbcsr_complete_redistribute 325 12.2 0.238 0.301 1.130 1.429 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.737000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=562.000000, yerr=6.809085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.951680E+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 0.010 0.028 60.300 60.300 qs_mol_dyn_low 1 2.0 0.003 0.004 60.035 60.045 qs_forces 11 3.9 0.003 0.004 59.969 59.969 qs_energies 11 4.9 0.001 0.002 56.777 56.781 scf_env_do_scf 11 5.9 0.000 0.001 48.907 48.907 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.293 40.293 velocity_verlet 10 3.0 0.001 0.002 32.818 32.820 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.973 29.204 qs_scf_new_mos 99 7.5 0.001 0.001 25.521 25.615 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.520 25.615 ot_scf_mini 99 9.5 0.002 0.003 24.257 24.364 multiply_cannon 2055 13.4 0.212 0.222 22.568 23.679 multiply_cannon_loop 2055 14.4 0.814 0.836 21.379 22.424 ot_mini 99 10.5 0.001 0.001 13.953 14.061 rebuild_ks_matrix 110 8.3 0.000 0.000 12.212 12.344 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.211 12.344 multiply_cannon_multrec 16440 15.4 3.668 4.792 9.896 10.939 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.763 10.881 mp_waitall_1 139946 16.5 6.991 10.174 6.991 10.174 qs_ot_get_derivative 99 11.5 0.001 0.001 9.532 9.639 init_scf_loop 11 6.9 0.000 0.000 8.579 8.579 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.357 7.362 prepare_preconditioner 11 7.9 0.000 0.000 6.821 6.849 make_preconditioner 11 8.9 0.000 0.000 6.821 6.849 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.158 6.518 dbcsr_mm_accdrv_process 34862 16.1 5.368 5.776 6.072 6.382 sum_up_and_integrate 110 10.3 0.001 0.002 6.321 6.335 integrate_v_rspace 110 11.3 0.003 0.003 6.295 6.309 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.822 5.832 calculate_rho_elec 110 8.6 0.059 0.059 5.821 5.831 qs_ot_get_p 110 10.4 0.001 0.001 5.448 5.581 init_scf_run 11 5.9 0.000 0.001 5.464 5.464 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.464 5.464 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.940 5.408 apply_single 110 13.6 0.000 0.000 4.940 5.408 make_m2s 4110 13.4 0.049 0.050 4.273 4.649 make_images 4110 14.4 0.397 0.520 4.156 4.531 ot_diis_step 99 11.5 0.011 0.012 4.393 4.393 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.341 3.986 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.787 3.792 multiply_cannon_sync_h2d 16440 15.4 3.225 3.688 3.225 3.688 pw_transfer 1331 11.6 0.065 0.073 3.676 3.685 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.569 3.581 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.416 3.416 grid_integrate_task_list 110 12.3 3.186 3.377 3.186 3.377 density_rs2pw 110 9.6 0.004 0.004 3.046 3.192 fft_wrap_pw1pw2_140 451 13.1 0.459 0.468 3.069 3.083 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.878 2.887 cp_fm_diag_elpa_base 48 14.0 2.811 2.845 2.876 2.886 make_images_data 4110 15.4 0.046 0.050 2.380 2.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.802 2.803 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.051 2.740 cp_fm_cholesky_invert 11 10.9 2.674 2.681 2.674 2.681 calculate_dm_sparse 110 9.5 0.001 0.001 2.533 2.567 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.486 2.547 fft3d_ps 1111 14.6 1.121 1.128 2.488 2.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.443 2.495 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.888 2.408 grid_collocate_task_list 110 9.6 2.226 2.404 2.226 2.404 calculate_first_density_matrix 1 7.0 0.000 0.000 2.371 2.372 mp_irecv_dv 48980 15.7 0.814 2.277 0.814 2.277 potential_pw2rs 110 12.3 0.010 0.011 2.147 2.152 mp_sum_l 6594 12.7 1.508 2.126 1.508 2.126 mp_alltoall_d11v 2046 13.8 1.793 2.054 1.793 2.054 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.959 1.959 dbcsr_complete_redistribute 325 12.2 0.324 0.345 1.467 1.927 cp_fm_upper_to_full 70 14.2 1.405 1.776 1.405 1.776 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.656 cp_fm_cholesky_decompose 22 10.9 1.626 1.646 1.626 1.646 acc_transpose_blocks 16440 15.4 0.078 0.081 1.575 1.643 mp_allgather_i34 2055 14.4 0.503 1.511 0.503 1.511 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.373 1.499 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.458 1.471 transfer_rs2pw 451 10.6 0.005 0.006 1.298 1.447 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.967 1.417 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.287 1.297 mp_waitany 17072 13.8 1.134 1.286 1.134 1.286 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.300000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.727273, yerr=9.126295 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 734.097408E+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 0.010 0.029 70.071 70.072 qs_mol_dyn_low 1 2.0 0.003 0.003 69.837 69.852 qs_forces 11 3.9 0.003 0.003 69.420 69.421 qs_energies 11 4.9 0.001 0.001 65.945 65.949 scf_env_do_scf 11 5.9 0.000 0.001 57.439 57.442 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.798 44.799 velocity_verlet 10 3.0 0.001 0.002 39.725 39.742 dbcsr_multiply_generic 2055 12.4 0.115 0.120 31.745 31.993 qs_scf_new_mos 99 7.5 0.001 0.001 29.344 29.463 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.343 29.463 ot_scf_mini 99 9.5 0.002 0.003 27.631 27.728 multiply_cannon 2055 13.4 0.239 0.253 24.138 25.269 multiply_cannon_loop 2055 14.4 1.425 1.452 22.500 23.129 ot_mini 99 10.5 0.001 0.001 15.828 15.944 multiply_cannon_multrec 24660 15.4 4.080 6.627 13.060 14.197 init_scf_loop 11 6.9 0.000 0.000 12.597 12.598 rebuild_ks_matrix 110 8.3 0.000 0.000 12.476 12.590 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.476 12.589 qs_ot_get_derivative 99 11.5 0.001 0.001 11.610 11.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.011 11.114 prepare_preconditioner 11 7.9 0.000 0.000 10.789 10.801 make_preconditioner 11 8.9 0.000 0.000 10.789 10.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.994 10.484 dbcsr_mm_accdrv_process 52304 16.0 7.713 9.262 8.822 9.874 mp_waitall_1 121746 16.5 5.024 7.184 5.024 7.184 qs_ot_get_p 110 10.4 0.001 0.001 6.579 6.708 sum_up_and_integrate 110 10.3 0.001 0.002 6.640 6.654 integrate_v_rspace 110 11.3 0.003 0.003 6.615 6.627 make_m2s 4110 13.4 0.059 0.061 5.927 6.430 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.332 6.340 calculate_rho_elec 110 8.6 0.077 0.081 6.332 6.340 make_images 4110 14.4 0.570 0.687 5.785 6.284 init_scf_run 11 5.9 0.000 0.001 5.713 5.714 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.713 5.714 cp_fm_upper_to_full 70 14.2 3.449 4.982 3.449 4.982 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.544 4.558 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.103 4.210 apply_single 110 13.6 0.000 0.000 4.102 4.210 ot_diis_step 99 11.5 0.011 0.013 4.179 4.180 pw_transfer 1331 11.6 0.065 0.074 4.060 4.094 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.953 3.991 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.954 3.954 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.868 3.928 dbcsr_complete_redistribute 325 12.2 0.411 0.465 2.771 3.921 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.559 3.617 density_rs2pw 110 9.6 0.004 0.004 3.416 3.582 make_images_data 4110 15.4 0.048 0.052 2.937 3.508 hybrid_alltoall_any 4261 16.3 0.123 0.455 2.490 3.458 grid_integrate_task_list 110 12.3 3.250 3.407 3.250 3.407 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.336 3.345 cp_fm_diag_elpa_base 48 14.0 3.179 3.241 3.333 3.343 fft_wrap_pw1pw2_140 451 13.1 0.531 0.545 3.302 3.341 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.188 3.324 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.576 3.300 wfi_extrapolate 11 7.9 0.001 0.001 3.222 3.222 cp_fm_cholesky_invert 11 10.9 3.203 3.211 3.203 3.211 calculate_dm_sparse 110 9.5 0.001 0.001 3.014 3.045 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.786 2.900 mp_alltoall_i22 605 13.7 1.692 2.872 1.692 2.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.854 2.856 fft3d_ps 1111 14.6 1.154 1.195 2.782 2.802 multiply_cannon_sync_h2d 24660 15.4 2.327 2.540 2.327 2.540 acc_transpose_blocks 24660 15.4 0.111 0.115 2.436 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.453 2.495 potential_pw2rs 110 12.3 0.012 0.013 2.414 2.427 grid_collocate_task_list 110 9.6 2.266 2.392 2.266 2.392 calculate_first_density_matrix 1 7.0 0.000 0.000 2.388 2.391 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.249 2.250 mp_alltoall_d11v 2046 13.8 1.968 2.189 1.968 2.189 cp_fm_cholesky_decompose 22 10.9 1.947 1.989 1.947 1.989 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.859 1.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.706 1.720 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.713 mp_allgather_i34 2055 14.4 0.716 1.670 0.716 1.670 transfer_rs2pw 451 10.6 0.005 0.005 1.458 1.651 mp_sum_l 6594 12.7 0.994 1.598 0.994 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.557 1.572 acc_transpose_blocks_sync 73980 16.4 1.428 1.516 1.428 1.516 mp_alltoall_z22v 1111 16.6 1.414 1.487 1.414 1.487 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.871 1.464 jit_kernel_multiply 8 15.8 0.766 1.433 0.766 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=70.072000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.727273, yerr=8.001033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 849.858560E+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 0.011 0.029 58.635 58.636 qs_mol_dyn_low 1 2.0 0.003 0.003 58.393 58.402 qs_forces 11 3.9 0.003 0.003 57.642 57.643 qs_energies 11 4.9 0.001 0.001 53.964 53.967 scf_env_do_scf 11 5.9 0.000 0.001 45.554 45.555 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 37.294 37.296 velocity_verlet 10 3.0 0.002 0.002 32.987 33.033 dbcsr_multiply_generic 2055 12.4 0.106 0.110 24.162 24.294 qs_scf_new_mos 99 7.5 0.001 0.001 22.371 22.418 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.371 22.418 ot_scf_mini 99 9.5 0.002 0.002 21.078 21.116 multiply_cannon 2055 13.4 0.237 0.245 18.240 19.548 multiply_cannon_loop 2055 14.4 0.604 0.622 16.898 17.146 rebuild_ks_matrix 110 8.3 0.000 0.000 11.876 11.916 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.875 11.916 ot_mini 99 10.5 0.001 0.001 11.456 11.487 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.565 10.598 multiply_cannon_multrec 8220 15.4 3.209 4.455 7.637 8.616 mp_waitall_1 103326 16.6 6.496 8.342 6.496 8.342 init_scf_loop 11 6.9 0.000 0.000 8.207 8.210 qs_ot_get_derivative 99 11.5 0.001 0.001 7.655 7.692 prepare_preconditioner 11 7.9 0.000 0.000 6.520 6.527 make_preconditioner 11 8.9 0.000 0.000 6.520 6.527 sum_up_and_integrate 110 10.3 0.001 0.002 6.370 6.382 integrate_v_rspace 110 11.3 0.003 0.003 6.344 6.356 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.173 6.187 calculate_rho_elec 110 8.6 0.112 0.113 6.173 6.187 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.090 6.159 qs_ot_get_p 110 10.4 0.001 0.001 5.561 5.595 init_scf_run 11 5.9 0.000 0.001 5.247 5.247 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.247 5.247 dbcsr_mm_accdrv_process 17442 15.9 3.102 4.119 4.287 5.189 make_m2s 4110 13.4 0.038 0.039 4.420 4.663 make_images 4110 14.4 0.641 0.705 4.290 4.530 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.293 4.474 pw_transfer 1331 11.6 0.066 0.070 4.225 4.231 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.118 4.127 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.040 4.044 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.813 3.867 apply_single 110 13.6 0.000 0.000 3.813 3.867 ot_diis_step 99 11.5 0.012 0.012 3.773 3.773 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.666 3.667 fft_wrap_pw1pw2_140 451 13.1 0.718 0.729 3.528 3.539 grid_integrate_task_list 110 12.3 3.385 3.466 3.385 3.466 density_rs2pw 110 9.6 0.004 0.004 3.174 3.266 cp_fm_cholesky_invert 11 10.9 3.236 3.240 3.236 3.240 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.083 3.091 cp_fm_diag_elpa_base 48 14.0 3.026 3.053 3.081 3.089 wfi_extrapolate 11 7.9 0.001 0.001 2.861 2.861 make_images_data 4110 15.4 0.040 0.046 2.354 2.717 hybrid_alltoall_any 4261 16.3 0.199 0.844 2.310 2.706 fft3d_ps 1111 14.6 1.292 1.299 2.662 2.674 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.667 2.668 calculate_dm_sparse 110 9.5 0.001 0.001 2.554 2.594 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.564 2.565 multiply_cannon_sync_h2d 8220 15.4 2.358 2.482 2.358 2.482 grid_collocate_task_list 110 9.6 2.372 2.467 2.372 2.467 calculate_first_density_matrix 1 7.0 0.000 0.000 2.286 2.289 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.162 2.180 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.117 2.150 potential_pw2rs 110 12.3 0.015 0.015 2.130 2.137 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.063 2.082 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 1.993 mp_alltoall_d11v 2046 13.8 1.803 1.936 1.803 1.936 cp_fm_cholesky_decompose 22 10.9 1.804 1.816 1.804 1.816 mp_allgather_i34 2055 14.4 0.512 1.659 0.512 1.659 dbcsr_complete_redistribute 325 12.2 0.541 0.578 1.570 1.651 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.649 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.497 1.608 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.472 1.485 acc_transpose_blocks 8220 15.4 0.041 0.042 1.362 1.386 qs_create_task_list 11 7.9 0.000 0.000 1.213 1.311 generate_qs_task_list 11 8.9 0.374 0.444 1.213 1.311 transfer_rs2pw 451 10.6 0.005 0.005 1.186 1.293 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.265 1.283 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.469 1.209 mp_waitany 9240 13.8 1.067 1.199 1.067 1.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.636000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=799.727273, yerr=11.330497 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 1.403286E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.036 92.005 92.006 qs_mol_dyn_low 1 2.0 0.004 0.009 91.693 91.703 qs_forces 11 3.9 0.022 0.026 91.623 91.626 qs_energies 11 4.9 0.008 0.010 87.435 87.441 scf_env_do_scf 11 5.9 0.000 0.001 75.393 75.394 velocity_verlet 10 3.0 0.002 0.002 57.329 57.335 scf_env_do_scf_inner_loop 99 6.5 0.007 0.012 46.147 46.148 dbcsr_multiply_generic 2055 12.4 0.121 0.125 30.379 30.518 init_scf_loop 11 6.9 0.000 0.000 29.177 29.181 qs_scf_new_mos 99 7.5 0.001 0.001 27.869 27.948 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.869 27.947 prepare_preconditioner 11 7.9 0.000 0.000 27.001 27.011 make_preconditioner 11 8.9 0.000 0.000 27.000 27.011 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.195 26.480 ot_scf_mini 99 9.5 0.002 0.002 26.062 26.131 multiply_cannon 2055 13.4 0.332 0.348 22.894 23.595 multiply_cannon_loop 2055 14.4 0.822 0.844 20.979 21.300 cp_fm_upper_to_full 70 14.2 12.768 18.306 12.768 18.306 ot_mini 99 10.5 0.001 0.001 14.694 14.761 rebuild_ks_matrix 110 8.3 0.000 0.001 14.228 14.305 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.228 14.305 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.908 12.976 dbcsr_complete_redistribute 325 12.2 1.123 1.153 7.529 10.752 multiply_cannon_multrec 8220 15.4 4.092 4.255 9.832 9.953 qs_ot_get_derivative 99 11.5 0.001 0.001 9.866 9.936 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.500 9.715 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.790 8.972 mp_waitall_1 84994 16.7 7.903 8.859 7.903 8.859 mp_alltoall_i22 605 13.7 5.444 8.689 5.444 8.689 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.636 7.667 calculate_rho_elec 110 8.6 0.222 0.223 7.635 7.666 init_scf_run 11 5.9 0.000 0.001 7.560 7.560 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.559 7.560 sum_up_and_integrate 110 10.3 0.002 0.002 7.087 7.105 integrate_v_rspace 110 11.3 0.003 0.004 7.059 7.077 qs_ot_get_p 110 10.4 0.001 0.001 6.099 6.183 make_m2s 4110 13.4 0.043 0.043 5.572 6.092 pw_transfer 1331 11.6 0.074 0.075 5.942 5.948 make_images 4110 14.4 0.875 0.931 5.382 5.900 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.825 5.832 dbcsr_mm_accdrv_process 11614 15.7 3.848 4.099 5.592 5.785 cp_fm_cholesky_invert 11 10.9 5.747 5.752 5.747 5.752 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.968 5.477 apply_single 110 13.6 0.000 0.000 4.968 5.477 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.112 5.415 fft_wrap_pw1pw2_140 451 13.1 1.364 1.369 5.093 5.098 ot_diis_step 99 11.5 0.015 0.016 4.802 4.802 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.321 4.328 density_rs2pw 110 9.6 0.004 0.004 4.069 4.100 calculate_first_density_matrix 1 7.0 0.000 0.000 4.022 4.022 qs_energies_init_hamiltonians 11 5.9 0.075 0.079 3.796 3.801 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.796 3.797 grid_integrate_task_list 110 12.3 3.701 3.765 3.701 3.765 hybrid_alltoall_any 4261 16.3 0.263 0.564 2.904 3.648 make_images_data 4110 15.4 0.044 0.047 2.904 3.583 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.069 3.570 fft3d_ps 1111 14.6 1.881 1.887 3.534 3.538 wfi_extrapolate 11 7.9 0.001 0.001 3.418 3.419 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.262 3.266 calculate_dm_sparse 110 9.5 0.001 0.001 3.205 3.228 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.182 3.182 cp_fm_diag_elpa_base 48 14.0 2.637 2.845 3.179 3.180 multiply_cannon_sync_h2d 8220 15.4 3.131 3.156 3.131 3.156 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.764 2.803 grid_collocate_task_list 110 9.6 2.686 2.711 2.686 2.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.613 2.635 potential_pw2rs 110 12.3 0.021 0.021 2.527 2.533 qs_env_update_s_mstruct 11 6.9 0.002 0.009 2.239 2.283 mp_alltoall_d11v 2046 13.8 2.183 2.270 2.183 2.270 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.095 2.193 cp_fm_cholesky_decompose 22 10.9 2.151 2.171 2.151 2.171 make_basis_sm 11 9.8 0.001 0.001 2.137 2.142 qs_create_task_list 11 7.9 0.000 0.000 1.882 1.928 generate_qs_task_list 11 8.9 0.732 0.785 1.882 1.928 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.918 1.927 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.808 1.856 parallel_gemm_fm 81 9.0 0.000 0.000 1.846 1.849 parallel_gemm_fm_cosma 81 10.0 1.846 1.849 1.846 1.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=92.006000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1259.727273, yerr=62.509768 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 633.372672E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1722306. 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 0.010 0.030 207.213 207.214 qs_mol_dyn_low 1 2.0 0.003 0.004 206.785 206.799 qs_forces 11 3.9 0.034 0.039 206.701 206.702 qs_energies 11 4.9 0.001 0.003 201.102 201.122 scf_env_do_scf 11 5.9 0.001 0.001 183.464 183.468 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.267 163.269 dbcsr_multiply_generic 2507 12.6 0.177 0.184 125.903 126.410 qs_scf_new_mos 117 7.6 0.001 0.001 124.180 124.482 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.179 124.481 velocity_verlet 10 3.0 0.001 0.002 123.132 123.133 ot_scf_mini 117 9.6 0.003 0.003 117.520 117.791 multiply_cannon 2507 13.6 0.237 0.243 101.792 103.262 multiply_cannon_loop 2507 14.6 2.388 2.438 99.632 101.107 ot_mini 117 10.6 0.001 0.001 66.570 66.887 multiply_cannon_multrec 60168 15.6 31.756 33.767 41.558 43.765 qs_ot_get_derivative 117 11.6 0.001 0.001 41.730 42.034 rebuild_ks_matrix 128 8.3 0.001 0.001 33.389 33.684 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.388 33.683 mp_waitall_1 267128 16.5 28.920 32.328 28.920 32.328 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.990 30.261 qs_ot_get_p 128 10.4 0.001 0.001 28.770 29.143 multiply_cannon_sync_h2d 60168 15.6 26.408 28.462 26.408 28.462 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.355 25.027 apply_single 128 13.6 0.001 0.001 24.355 25.027 ot_diis_step 117 11.6 0.008 0.008 24.534 24.535 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.228 22.309 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.055 20.235 init_scf_loop 11 6.9 0.000 0.000 20.124 20.126 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.432 19.433 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.141 17.611 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.503 16.535 cp_fm_diag_elpa_base 83 14.4 16.435 16.466 16.499 16.532 prepare_preconditioner 11 7.9 0.000 0.000 15.535 15.584 make_preconditioner 11 8.9 0.000 0.000 15.535 15.584 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.768 14.915 make_m2s 5014 13.6 0.105 0.112 13.758 14.209 make_images 5014 14.6 0.398 0.416 13.572 14.033 sum_up_and_integrate 128 10.3 0.002 0.004 13.861 13.875 integrate_v_rspace 128 11.3 0.003 0.004 13.803 13.821 init_scf_run 11 5.9 0.000 0.001 13.389 13.390 scf_env_initial_rho_setup 11 6.9 0.050 0.050 13.389 13.390 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.973 13.135 calculate_rho_elec 128 8.7 0.045 0.064 12.973 13.134 mp_sum_l 7950 12.9 9.093 10.359 9.093 10.359 dbcsr_mm_accdrv_process 124484 16.2 4.709 4.840 9.370 9.901 wfi_extrapolate 11 7.9 0.001 0.001 9.030 9.030 cp_fm_cholesky_invert 11 10.9 8.785 8.793 8.785 8.793 calculate_dm_sparse 128 9.5 0.001 0.001 8.470 8.564 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.153 8.367 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.093 8.202 multiply_cannon_metrocomm1 60168 15.6 0.092 0.095 6.435 7.984 pw_transfer 1547 11.6 0.073 0.092 7.523 7.699 make_images_data 5014 15.6 0.065 0.072 6.706 7.635 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.321 7.493 grid_integrate_task_list 128 12.3 7.099 7.434 7.099 7.434 density_rs2pw 128 9.7 0.006 0.006 6.617 7.212 hybrid_alltoall_any 5200 16.5 0.292 2.246 5.883 7.210 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.995 7.006 fft_wrap_pw1pw2_140 523 13.2 0.855 0.896 6.454 6.612 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.980 6.141 fft3d_ps 1291 14.7 2.198 2.845 5.328 5.662 mp_alltoall_d11v 2415 14.1 4.246 5.310 4.246 5.310 grid_collocate_task_list 128 9.7 4.844 5.277 4.844 5.277 cp_fm_cholesky_decompose 22 10.9 4.572 4.588 4.572 4.588 potential_pw2rs 128 12.3 0.009 0.010 4.371 4.389 mp_sum_d 4467 12.1 3.427 4.239 3.427 4.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.214000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.181818, yerr=6.698958 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 847.613952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58170. MP_Allreduce 11091 960. MP_Sync 86 MP_Alltoall 1955 6587854. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.123 188.300 188.307 qs_mol_dyn_low 1 2.0 0.003 0.004 187.811 187.825 qs_forces 11 3.9 0.004 0.004 187.691 187.692 qs_energies 11 4.9 0.001 0.002 180.946 180.958 scf_env_do_scf 11 5.9 0.001 0.001 162.952 162.962 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.519 130.521 velocity_verlet 10 3.0 0.002 0.002 116.810 116.812 dbcsr_multiply_generic 2485 12.5 0.185 0.191 96.436 97.616 qs_scf_new_mos 116 7.6 0.001 0.001 92.630 93.184 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.630 93.183 ot_scf_mini 116 9.6 0.004 0.004 87.860 88.486 multiply_cannon 2485 13.5 0.501 0.568 76.759 80.997 multiply_cannon_loop 2485 14.5 1.545 1.611 73.419 76.379 ot_mini 116 10.6 0.001 0.001 49.768 50.305 mp_waitall_1 212858 16.6 23.609 38.636 23.609 38.636 multiply_cannon_multrec 29820 15.5 20.863 25.491 31.394 36.255 rebuild_ks_matrix 127 8.3 0.001 0.001 31.928 32.505 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.927 32.504 init_scf_loop 11 6.9 0.000 0.000 32.346 32.347 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.680 29.200 qs_ot_get_derivative 116 11.6 0.001 0.002 27.995 28.614 multiply_cannon_metrocomm3 29820 15.5 0.099 0.107 15.496 28.385 prepare_preconditioner 11 7.9 0.000 0.000 27.978 28.058 make_preconditioner 11 8.9 0.000 0.000 27.978 28.058 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.705 27.217 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.881 23.029 apply_single 127 13.6 0.001 0.001 21.881 23.028 ot_diis_step 116 11.6 0.015 0.015 21.598 21.600 qs_ot_get_p 127 10.4 0.001 0.001 20.307 20.905 multiply_cannon_sync_h2d 29820 15.5 17.966 20.236 17.966 20.236 cp_fm_cholesky_invert 11 10.9 16.388 16.401 16.388 16.401 make_m2s 4970 13.5 0.086 0.093 13.862 15.522 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.487 15.520 make_images 4970 14.5 1.151 1.353 13.646 15.304 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.307 14.308 sum_up_and_integrate 127 10.3 0.002 0.004 13.755 13.783 integrate_v_rspace 127 11.3 0.003 0.004 13.696 13.725 init_scf_run 11 5.9 0.000 0.001 13.208 13.210 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.208 13.209 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.112 13.144 calculate_rho_elec 127 8.7 0.086 0.103 13.112 13.143 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 11.004 11.424 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.220 11.251 cp_fm_diag_elpa_base 82 14.4 10.971 11.076 11.212 11.239 multiply_cannon_metrocomm4 27335 15.5 0.098 0.111 3.832 10.779 dbcsr_mm_accdrv_process 61726 16.2 5.365 6.071 9.985 10.604 mp_irecv_dv 68888 16.3 3.634 10.390 3.634 10.390 make_images_data 4970 15.5 0.065 0.074 8.166 10.317 hybrid_alltoall_any 5155 16.4 0.345 1.545 6.954 9.426 pw_transfer 1535 11.6 0.083 0.098 8.410 8.464 wfi_extrapolate 11 7.9 0.001 0.001 8.292 8.292 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.190 8.249 grid_integrate_task_list 127 12.3 7.143 7.435 7.143 7.435 fft_wrap_pw1pw2_140 519 13.2 0.927 0.955 7.275 7.345 density_rs2pw 127 9.7 0.006 0.006 6.881 7.331 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.321 7.121 cp_fm_cholesky_decompose 22 10.9 6.756 6.832 6.756 6.832 calculate_dm_sparse 127 9.5 0.001 0.001 6.424 6.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.166 6.175 mp_sum_l 7884 12.9 4.056 6.157 4.056 6.157 fft3d_ps 1281 14.7 2.804 2.977 5.765 5.805 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.306 5.390 grid_collocate_task_list 127 9.7 4.972 5.373 4.972 5.373 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.205 5.360 mp_allgather_i34 2485 14.5 1.858 4.777 1.858 4.777 calculate_first_density_matrix 1 7.0 0.000 0.000 4.732 4.735 mp_alltoall_d11v 2401 14.1 4.000 4.436 4.000 4.436 potential_pw2rs 127 12.3 0.015 0.017 4.330 4.347 mp_sum_d 4455 12.1 2.659 4.120 2.659 4.120 parallel_gemm_fm 81 9.0 0.000 0.000 4.073 4.099 parallel_gemm_fm_cosma 81 10.0 4.072 4.099 4.072 4.099 make_basis_sm 11 9.8 0.001 0.001 4.061 4.072 dbcsr_complete_redistribute 393 12.7 0.845 0.948 3.135 3.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.307000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=806.000000, yerr=3.330302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 981.553152E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.082 173.145 173.147 qs_mol_dyn_low 1 2.0 0.003 0.005 172.530 172.619 qs_forces 11 3.9 0.004 0.005 172.061 172.064 qs_energies 11 4.9 0.008 0.029 165.550 165.558 scf_env_do_scf 11 5.9 0.003 0.017 149.688 149.690 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.245 115.246 velocity_verlet 10 3.0 0.002 0.011 110.330 110.344 dbcsr_multiply_generic 2485 12.5 0.188 0.215 81.097 82.122 qs_scf_new_mos 116 7.6 0.001 0.001 78.700 79.028 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.699 79.027 ot_scf_mini 116 9.6 0.003 0.005 74.511 74.857 multiply_cannon 2485 13.5 0.501 0.520 61.604 66.477 multiply_cannon_loop 2485 14.5 1.121 1.202 58.950 61.241 ot_mini 116 10.6 0.001 0.001 42.036 42.384 init_scf_loop 11 6.9 0.001 0.005 34.332 34.333 mp_waitall_1 169034 16.6 24.833 34.191 24.833 34.191 rebuild_ks_matrix 127 8.3 0.001 0.001 29.930 30.399 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.037 29.929 30.399 prepare_preconditioner 11 7.9 0.000 0.000 30.334 30.381 make_preconditioner 11 8.9 0.000 0.002 30.334 30.380 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.012 29.430 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.936 27.364 multiply_cannon_multrec 19880 15.5 12.902 16.198 22.183 25.506 multiply_cannon_metrocomm3 19880 15.5 0.061 0.065 15.512 24.487 qs_ot_get_derivative 116 11.6 0.001 0.002 22.517 22.868 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.653 20.700 apply_single 127 13.6 0.001 0.001 19.653 20.700 ot_diis_step 116 11.6 0.018 0.018 19.413 19.413 qs_ot_get_p 127 10.4 0.001 0.002 17.929 18.367 make_m2s 4970 13.5 0.077 0.082 14.652 15.420 make_images 4970 14.5 1.134 1.218 14.417 15.186 multiply_cannon_sync_h2d 19880 15.5 13.517 15.112 13.517 15.112 cp_fm_cholesky_invert 11 10.9 14.081 14.090 14.081 14.090 sum_up_and_integrate 127 10.3 0.002 0.004 13.786 13.811 qs_ot_p2m_diag 82 11.4 0.262 0.270 13.791 13.799 integrate_v_rspace 127 11.3 0.003 0.004 13.727 13.751 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.325 13.360 calculate_rho_elec 127 8.7 0.128 0.142 13.325 13.359 cp_dbcsr_syevd 82 12.4 0.006 0.008 12.755 12.756 init_scf_run 11 5.9 0.000 0.001 10.485 10.486 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.485 10.486 make_images_data 4970 15.5 0.062 0.073 8.880 10.021 cp_fm_diag_elpa 82 13.4 0.001 0.001 9.658 9.681 cp_fm_diag_elpa_base 82 14.4 9.259 9.415 9.653 9.676 multiply_cannon_metrocomm4 17395 15.5 0.064 0.074 3.493 9.490 mp_irecv_dv 49801 16.2 3.365 9.225 3.365 9.225 hybrid_alltoall_any 5155 16.4 0.445 2.025 7.670 9.184 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.812 9.073 dbcsr_mm_accdrv_process 41158 16.2 5.612 5.942 8.744 8.902 pw_transfer 1535 11.6 0.083 0.100 8.654 8.751 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.433 8.537 grid_integrate_task_list 127 12.3 7.276 7.669 7.276 7.669 fft_wrap_pw1pw2_140 519 13.2 1.051 1.079 7.509 7.618 wfi_extrapolate 11 7.9 0.001 0.001 7.403 7.403 cp_fm_upper_to_full 104 14.8 5.736 7.216 5.736 7.216 density_rs2pw 127 9.7 0.005 0.006 6.778 7.192 cp_fm_cholesky_decompose 22 10.9 7.111 7.130 7.111 7.130 dbcsr_complete_redistribute 393 12.7 1.158 1.192 4.488 6.226 calculate_dm_sparse 127 9.5 0.001 0.001 5.795 5.875 fft3d_ps 1281 14.7 2.797 3.031 5.772 5.864 grid_collocate_task_list 127 9.7 5.127 5.531 5.127 5.531 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.475 5.482 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.570 5.250 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.351 5.085 mp_alltoall_d11v 2401 14.1 4.167 4.893 4.167 4.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.583 4.707 mp_sum_l 7884 12.9 3.058 4.375 3.058 4.375 mp_allgather_i34 2485 14.5 1.162 4.305 1.162 4.305 potential_pw2rs 127 12.3 0.019 0.021 4.290 4.301 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.302 4.007 qs_energies_init_hamiltonians 11 5.9 0.032 0.040 4.000 4.001 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.920 3.947 mp_alltoall_i22 712 14.1 1.929 3.788 1.929 3.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.490 3.531 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.147000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=926.454545, yerr=22.386537 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.146200E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 188.763 188.764 qs_mol_dyn_low 1 2.0 0.003 0.004 188.303 188.316 qs_forces 11 3.9 0.004 0.005 187.728 187.736 qs_energies 11 4.9 0.001 0.002 180.637 180.645 scf_env_do_scf 11 5.9 0.001 0.001 162.345 162.356 velocity_verlet 10 3.0 0.002 0.002 124.339 124.361 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.153 116.154 dbcsr_multiply_generic 2507 12.6 0.191 0.196 80.245 80.954 qs_scf_new_mos 117 7.6 0.001 0.001 80.059 80.376 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.058 80.375 ot_scf_mini 117 9.6 0.003 0.003 75.516 75.828 multiply_cannon 2507 13.6 0.552 0.583 55.721 58.019 multiply_cannon_loop 2507 14.6 1.850 1.908 52.092 53.834 init_scf_loop 11 6.9 0.000 0.000 46.072 46.072 ot_mini 117 10.6 0.001 0.001 42.755 43.073 prepare_preconditioner 11 7.9 0.000 0.000 41.940 41.967 make_preconditioner 11 8.9 0.000 0.000 41.940 41.967 make_full_inverse_cholesky 11 9.9 0.013 0.024 35.548 40.609 multiply_cannon_multrec 30084 15.6 13.563 18.444 26.470 31.056 rebuild_ks_matrix 128 8.3 0.001 0.001 29.265 29.527 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.264 29.527 mp_waitall_1 147882 16.7 16.993 27.767 16.993 27.767 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.388 26.623 qs_ot_get_derivative 117 11.6 0.001 0.002 23.056 23.374 make_m2s 5014 13.6 0.092 0.097 20.229 21.256 make_images 5014 14.6 1.948 2.255 19.923 20.953 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.078 19.594 apply_single 128 13.6 0.001 0.001 19.078 19.594 ot_diis_step 117 11.6 0.018 0.019 19.575 19.576 qs_ot_get_p 128 10.4 0.001 0.001 18.916 19.247 cp_fm_upper_to_full 105 14.8 11.536 16.972 11.536 16.972 cp_fm_cholesky_invert 11 10.9 15.871 15.880 15.871 15.880 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.420 15.288 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.829 14.877 sum_up_and_integrate 128 10.3 0.002 0.003 13.968 13.995 integrate_v_rspace 128 11.3 0.003 0.004 13.908 13.936 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.654 13.701 calculate_rho_elec 128 8.7 0.170 0.186 13.654 13.701 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.430 13.431 dbcsr_mm_accdrv_process 62264 16.2 8.404 9.248 12.475 12.957 dbcsr_complete_redistribute 395 12.7 1.492 1.604 9.077 12.846 make_images_data 5014 15.6 0.064 0.070 10.788 12.476 init_scf_run 11 5.9 0.000 0.001 12.381 12.382 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.381 12.382 multiply_cannon_sync_h2d 30084 15.6 10.420 11.551 10.420 11.551 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.735 11.488 hybrid_alltoall_any 5200 16.5 0.529 2.195 9.711 11.371 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.249 10.264 cp_fm_diag_elpa_base 83 14.4 9.280 9.591 10.241 10.258 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.371 10.027 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.697 9.931 mp_alltoall_i22 716 14.1 5.622 9.445 5.622 9.445 pw_transfer 1547 11.6 0.086 0.102 9.152 9.225 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.926 9.005 fft_wrap_pw1pw2_140 523 13.2 1.245 1.285 7.964 8.061 grid_integrate_task_list 128 12.3 7.488 7.956 7.488 7.956 wfi_extrapolate 11 7.9 0.001 0.001 7.602 7.602 cp_fm_cholesky_decompose 22 10.9 7.411 7.518 7.411 7.518 density_rs2pw 128 9.7 0.005 0.006 6.764 7.108 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.809 6.983 mp_irecv_dv 76098 16.2 2.659 6.712 2.659 6.712 calculate_dm_sparse 128 9.5 0.001 0.001 6.233 6.320 fft3d_ps 1291 14.7 2.989 3.063 5.977 6.031 grid_collocate_task_list 128 9.7 5.274 5.676 5.274 5.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.536 5.584 mp_alltoall_d11v 2415 14.1 4.868 5.497 4.868 5.497 calculate_first_density_matrix 1 7.0 0.000 0.000 4.581 4.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.484 4.575 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.421 4.519 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 4.403 4.404 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.237 4.307 potential_pw2rs 128 12.3 0.023 0.024 4.273 4.280 parallel_gemm_fm 81 9.0 0.000 0.000 3.774 3.798 parallel_gemm_fm_cosma 81 10.0 3.774 3.798 3.774 3.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.764000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.636364, yerr=19.973950 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.552675E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58302. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.111 168.938 168.945 qs_mol_dyn_low 1 2.0 0.003 0.003 168.350 168.393 qs_forces 11 3.9 0.004 0.004 167.125 167.130 qs_energies 11 4.9 0.001 0.002 159.719 159.726 scf_env_do_scf 11 5.9 0.001 0.001 142.346 142.360 velocity_verlet 10 3.0 0.002 0.002 110.973 111.047 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.340 107.341 dbcsr_multiply_generic 2485 12.5 0.181 0.189 71.309 71.830 qs_scf_new_mos 116 7.6 0.001 0.001 70.324 70.428 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.324 70.427 ot_scf_mini 116 9.6 0.003 0.003 65.975 66.072 multiply_cannon 2485 13.5 0.556 0.578 53.144 55.959 multiply_cannon_loop 2485 14.5 0.800 0.826 50.062 51.174 ot_mini 116 10.6 0.001 0.001 37.260 37.352 init_scf_loop 11 6.9 0.000 0.000 34.859 34.860 mp_waitall_1 124680 16.7 24.436 31.136 24.436 31.136 prepare_preconditioner 11 7.9 0.000 0.000 31.031 31.062 make_preconditioner 11 8.9 0.000 0.000 31.031 31.062 make_full_inverse_cholesky 11 9.9 0.014 0.029 28.922 29.182 rebuild_ks_matrix 127 8.3 0.001 0.001 28.857 29.004 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.857 29.004 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.265 26.398 multiply_cannon_multrec 9940 15.5 10.227 15.968 17.741 21.860 qs_ot_get_derivative 116 11.6 0.002 0.002 20.165 20.250 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.888 20.216 cp_fm_cholesky_invert 11 10.9 17.630 17.636 17.630 17.636 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.174 17.358 apply_single 127 13.6 0.001 0.001 17.173 17.357 ot_diis_step 116 11.6 0.020 0.022 17.024 17.025 qs_ot_get_p 127 10.4 0.001 0.001 16.155 16.287 make_m2s 4970 13.5 0.063 0.067 14.355 15.289 make_images 4970 14.5 2.136 2.559 14.050 14.978 sum_up_and_integrate 127 10.3 0.002 0.002 14.030 14.077 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.029 14.075 calculate_rho_elec 127 8.7 0.251 0.262 14.029 14.074 integrate_v_rspace 127 11.3 0.004 0.005 13.970 14.019 qs_ot_p2m_diag 82 11.4 0.489 0.495 12.502 12.516 cp_dbcsr_syevd 82 12.4 0.005 0.006 11.365 11.366 multiply_cannon_sync_h2d 9940 15.5 10.642 11.055 10.642 11.055 init_scf_run 11 5.9 0.000 0.001 10.573 10.573 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.572 10.573 pw_transfer 1535 11.6 0.084 0.096 9.746 9.796 make_images_data 4970 15.5 0.054 0.064 8.215 9.626 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.526 9.582 hybrid_alltoall_any 5155 16.4 0.835 3.747 8.103 9.269 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.573 8.585 cp_fm_diag_elpa_base 82 14.4 8.343 8.419 8.570 8.581 fft_wrap_pw1pw2_140 519 13.2 1.657 1.691 8.456 8.511 grid_integrate_task_list 127 12.3 7.730 7.944 7.730 7.944 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.851 7.908 dbcsr_mm_accdrv_process 20590 16.1 3.361 4.380 7.166 7.893 cp_fm_cholesky_decompose 22 10.9 7.739 7.833 7.739 7.833 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.124 7.433 wfi_extrapolate 11 7.9 0.001 0.001 7.384 7.384 density_rs2pw 127 9.7 0.005 0.005 6.832 7.263 calculate_dm_sparse 127 9.5 0.001 0.001 6.009 6.083 fft3d_ps 1281 14.7 3.111 3.211 6.009 6.057 grid_collocate_task_list 127 9.7 5.516 5.843 5.516 5.843 multiply_cannon_metrocomm4 7455 15.5 0.026 0.030 1.854 5.823 mp_irecv_dv 28618 15.9 1.813 5.729 1.813 5.729 dbcsr_complete_redistribute 393 12.7 2.090 2.155 5.213 5.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.206 5.261 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 5.257 5.259 mp_alltoall_d11v 2401 14.1 4.561 4.835 4.561 4.835 mp_allgather_i34 2485 14.5 1.287 4.502 1.287 4.502 potential_pw2rs 127 12.3 0.026 0.026 4.272 4.278 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.097 4.164 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.874 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.378 3.697 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.599 3.652 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.563 3.651 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.421 3.537 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.485 3.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.945000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1444.181818, yerr=37.272506 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1399806074880 0.0% 0.0% 100.0% flops 32 x 32 x 32 1906965479424 0.0% 0.0% 100.0% flops 22 x 9 x 32 1943685697536 0.0% 0.0% 100.0% flops 9 x 22 x 32 1949309632512 0.0% 0.0% 100.0% flops 22 x 22 x 32 2694948675584 0.0% 0.0% 100.0% flops 32 x 32 x 9 4338990710784 0.0% 0.0% 100.0% flops 32 x 32 x 22 5303210868736 0.0% 0.0% 100.0% flops 9 x 32 x 32 5347334946816 0.0% 0.0% 100.0% flops 22 x 32 x 32 6535631601664 0.0% 0.0% 100.0% flops 9 x 32 x 9 11360521838592 0.0% 0.0% 100.0% flops 22 x 32 x 9 14909195169792 0.0% 0.0% 100.0% flops 9 x 32 x 22 14909195169792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19481731219456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.080527E+12 0.0% 0.0% 100.0% flops max/rank 11.516854E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6655092288 0.0% 0.0% 100.0% number of processed stacks 1931936 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.8 marketing flops 142.438295E+12 ------------------------------------------------------------------------------- # multiplications 2464 max memory usage/rank 3.129205E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 98560 MPI messages size (bytes): total size 1.118682E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.350266E+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 44368 34435235840 4194304 < size <= 16777216 43616 373376942080 16777216 < size 9960 710866914816 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4018 58907. MP_Allreduce 11122 1503. MP_Sync 87 MP_Alltoall 1688 36934192. MP_SendRecv 1764 218624. MP_ISendRecv 1764 218624. MP_Wait 9654 MP_ISend 6312 1080635. MP_IRecv 6312 1080635. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.094 287.814 287.830 qs_mol_dyn_low 1 2.0 0.003 0.004 287.090 287.113 qs_forces 11 3.9 0.009 0.010 287.000 287.005 qs_energies 11 4.9 0.012 0.013 278.062 278.068 scf_env_do_scf 11 5.9 0.001 0.001 255.753 255.760 velocity_verlet 10 3.0 0.002 0.002 207.683 207.690 scf_env_do_scf_inner_loop 115 6.6 0.004 0.008 131.620 131.623 init_scf_loop 11 6.9 0.000 0.000 123.862 123.868 prepare_preconditioner 11 7.9 0.000 0.000 119.060 119.081 make_preconditioner 11 8.9 0.000 0.000 119.060 119.081 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.253 116.212 qs_scf_new_mos 115 7.6 0.001 0.001 87.816 87.875 qs_scf_loop_do_ot 115 8.6 0.001 0.001 87.815 87.874 ot_scf_mini 115 9.6 0.004 0.004 83.086 83.114 dbcsr_multiply_generic 2464 12.5 0.212 0.222 80.195 80.683 cp_fm_upper_to_full 105 14.8 53.199 75.953 53.199 75.953 multiply_cannon 2464 13.5 0.670 0.697 57.604 58.799 multiply_cannon_loop 2464 14.5 1.035 1.045 53.726 55.090 ot_mini 115 10.6 0.001 0.001 43.464 43.493 dbcsr_complete_redistribute 395 12.7 3.973 4.019 29.678 42.275 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.201 38.745 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.765 36.099 mp_alltoall_i22 712 14.1 21.514 34.180 21.514 34.180 rebuild_ks_matrix 126 8.3 0.001 0.001 34.091 34.124 qs_ks_build_kohn_sham_matrix 126 9.3 0.017 0.017 34.090 34.124 cp_fm_cholesky_invert 11 10.9 33.098 33.104 33.098 33.104 qs_ks_update_qs_env 126 7.6 0.001 0.001 31.542 31.566 mp_waitall_1 101902 16.8 27.265 31.358 27.265 31.358 qs_ot_get_p 126 10.4 0.001 0.001 25.074 25.142 qs_ot_get_derivative 115 11.6 0.002 0.002 24.131 24.160 qs_ot_p2m_diag 83 11.4 0.879 0.884 21.115 21.145 multiply_cannon_metrocomm3 9856 15.5 0.025 0.025 18.286 19.600 make_m2s 4928 13.5 0.073 0.075 18.152 19.532 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.322 19.324 ot_diis_step 115 11.6 0.021 0.022 19.300 19.301 make_images 4928 14.5 3.002 3.187 17.676 19.055 apply_preconditioner_dbcsr 126 12.6 0.000 0.000 18.699 18.803 apply_single 126 13.6 0.001 0.001 18.699 18.803 multiply_cannon_multrec 9856 15.5 10.108 11.911 17.809 17.889 qs_rho_update_rho_low 126 7.7 0.001 0.001 16.943 16.963 calculate_rho_elec 126 8.7 0.470 0.471 16.943 16.963 sum_up_and_integrate 126 10.3 0.002 0.002 16.123 16.209 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.180 16.181 cp_fm_diag_elpa_base 83 14.4 11.767 13.325 16.176 16.176 integrate_v_rspace 126 11.3 0.004 0.004 16.061 16.147 multiply_cannon_sync_h2d 9856 15.5 14.074 14.116 14.074 14.116 pw_transfer 1523 11.6 0.094 0.094 12.889 12.896 fft_wrap_pw1pw2 1271 12.7 0.011 0.012 12.654 12.662 hybrid_alltoall_any 5114 16.4 1.294 3.013 10.469 12.497 make_images_data 4928 15.5 0.061 0.067 10.263 12.327 init_scf_run 11 5.9 0.000 0.001 11.960 11.960 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.960 11.960 fft_wrap_pw1pw2_140 515 13.2 3.164 3.204 11.272 11.278 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.665 9.683 dbcsr_mm_accdrv_process 20422 16.0 4.063 5.795 7.455 9.351 cp_fm_cholesky_decompose 22 10.9 8.793 8.816 8.793 8.816 wfi_extrapolate 11 7.9 0.001 0.001 8.810 8.810 grid_integrate_task_list 126 12.3 8.435 8.624 8.435 8.624 density_rs2pw 126 9.7 0.005 0.005 8.145 8.192 qs_energies_init_hamiltonians 11 5.9 0.007 0.017 8.187 8.189 fft3d_ps 1271 14.7 3.912 3.932 7.208 7.244 mp_alltoall_d11v 2399 14.1 6.633 6.777 6.633 6.777 calculate_dm_sparse 126 9.5 0.001 0.001 6.579 6.606 grid_collocate_task_list 126 9.7 6.318 6.387 6.318 6.387 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.220 6.268 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.101 6.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.830000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2816.818182, yerr=152.039360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 1.260487E+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 3057086. 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 0.034 0.070 85.176 85.176 qs_energies 1 2.0 0.000 0.000 84.570 84.578 ls_scf 1 3.0 0.000 0.000 83.644 83.650 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.679 72.895 multiply_cannon 111 7.7 0.017 0.020 55.951 56.983 multiply_cannon_loop 111 8.7 0.227 0.241 52.498 53.809 ls_scf_main 1 4.0 0.000 0.000 52.432 52.434 density_matrix_trs4 2 5.0 0.002 0.003 46.874 46.973 ls_scf_init_scf 1 4.0 0.000 0.000 28.165 28.167 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.036 27.085 mp_waitall_1 11031 10.9 22.619 25.855 22.619 25.855 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.952 24.971 multiply_cannon_multrec 2664 9.7 8.139 8.980 15.513 17.323 multiply_cannon_sync_h2d 2664 9.7 13.524 15.799 13.524 15.799 make_m2s 222 7.7 0.009 0.011 13.113 13.678 make_images 222 8.7 0.099 0.108 13.091 13.659 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.848 13.070 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.412 8.536 make_images_data 222 9.7 0.004 0.005 7.691 8.175 dbcsr_mm_accdrv_process 4760 10.4 0.582 0.693 6.995 7.984 hybrid_alltoall_any 227 10.6 0.214 1.819 6.588 7.907 dbcsr_mm_accdrv_process_sort 4760 11.4 6.214 7.097 6.214 7.097 calculate_norms 4752 9.8 5.497 6.099 5.497 6.099 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.141 mp_sum_l 887 5.1 3.089 4.398 3.089 4.398 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.042 4.025 mp_irecv_dv 6231 10.9 2.025 4.016 2.025 4.016 make_images_sizes 222 9.7 0.000 0.000 0.691 3.540 mp_alltoall_i44 222 10.7 0.691 3.540 0.691 3.540 arnoldi_extremal 4 6.8 0.000 0.002 3.210 3.243 arnoldi_normal_ev 4 7.8 0.001 0.003 3.210 3.243 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.286 3.239 build_subspace 16 8.4 0.009 0.012 3.118 3.120 ls_scf_post 1 4.0 0.000 0.000 3.046 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.855 2.909 dbcsr_special_finalize 555 9.7 0.005 0.006 2.378 2.773 dbcsr_merge_single_wm 555 10.7 0.462 0.596 2.370 2.765 make_images_pack 222 9.7 2.206 2.631 2.208 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.316 2.558 dbcsr_sort_data 658 11.4 2.166 2.496 2.166 2.496 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.452 2.063 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.300 2.376 buffer_matrices_ensure_size 222 8.7 1.748 2.083 1.748 2.083 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.771 1.772 rebuild_ks_matrix 3 7.3 0.000 0.000 1.762 1.763 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.762 1.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.176000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 2.131087E+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 0.030 0.078 93.898 93.900 qs_energies 1 2.0 0.001 0.006 93.382 93.416 ls_scf 1 3.0 0.002 0.012 92.047 92.080 dbcsr_multiply_generic 111 6.7 0.015 0.016 78.074 78.384 multiply_cannon 111 7.7 0.027 0.041 56.650 60.931 multiply_cannon_loop 111 8.7 0.135 0.147 53.827 56.636 ls_scf_main 1 4.0 0.000 0.003 54.312 54.317 density_matrix_trs4 2 5.0 0.002 0.004 48.664 48.850 ls_scf_init_scf 1 4.0 0.000 0.002 34.025 34.028 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.712 32.817 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.353 30.363 mp_waitall_1 9105 10.9 21.156 29.518 21.156 29.518 multiply_cannon_multrec 1332 9.7 13.259 17.236 22.543 27.640 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.679 20.092 make_m2s 222 7.7 0.006 0.008 15.073 15.667 make_images 222 8.7 1.366 1.702 15.043 15.638 dbcsr_mm_accdrv_process 4041 10.4 0.312 0.515 8.880 10.423 dbcsr_mm_accdrv_process_sort 4041 11.4 8.400 9.899 8.400 9.899 make_images_data 222 9.7 0.004 0.005 8.725 9.586 hybrid_alltoall_any 227 10.6 0.543 2.550 8.094 9.419 mp_sum_l 887 5.1 4.899 7.718 4.899 7.718 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.205 7.595 mp_irecv_dv 3311 11.0 3.184 7.541 3.184 7.541 calculate_norms 2376 9.8 6.056 6.755 6.056 6.755 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.399 6.607 multiply_cannon_sync_h2d 1332 9.7 4.901 6.405 4.901 6.405 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.683 5.995 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.872 5.068 compute_matrix_preconditioner 1 6.0 0.000 0.001 4.791 4.797 arnoldi_extremal 4 6.8 0.000 0.001 4.603 4.614 arnoldi_normal_ev 4 7.8 0.001 0.005 4.602 4.614 build_subspace 16 8.4 0.014 0.021 4.349 4.353 acc_transpose_blocks 1332 9.7 0.008 0.009 3.779 4.299 acc_transpose_blocks_kernels 1332 10.7 0.020 0.022 3.703 4.219 jit_kernel_transpose 1 13.0 3.683 4.198 3.683 4.198 ls_scf_post 1 4.0 0.000 0.001 3.708 3.731 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.530 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.115 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.211 2.738 3.213 mp_allgather_i34 111 8.7 0.818 3.022 0.818 3.022 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.599 2.691 dbcsr_data_new 4174 10.1 2.109 2.398 2.109 2.398 make_images_pack 222 9.7 1.819 2.123 1.822 2.126 dbcsr_sort_data 436 11.2 1.827 2.122 1.827 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.966 1.968 rebuild_ks_matrix 3 7.3 0.000 0.000 1.953 1.955 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.952 1.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.900000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 2.919600E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.041 93.251 93.252 qs_energies 1 2.0 0.000 0.000 92.567 92.571 ls_scf 1 3.0 0.000 0.000 91.146 91.150 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.642 75.999 ls_scf_main 1 4.0 0.000 0.000 56.997 57.002 multiply_cannon 111 7.7 0.034 0.099 52.052 55.910 multiply_cannon_loop 111 8.7 0.117 0.130 49.290 53.541 density_matrix_trs4 2 5.0 0.002 0.003 51.037 51.241 mp_waitall_1 7281 11.0 23.907 34.300 23.907 34.300 ls_scf_init_scf 1 4.0 0.000 0.000 30.486 30.489 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.296 29.362 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.958 26.971 multiply_cannon_multrec 888 9.7 12.559 15.474 21.121 24.603 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.354 23.578 make_m2s 222 7.7 0.006 0.007 16.629 17.344 make_images 222 8.7 1.584 1.854 16.591 17.306 make_images_data 222 9.7 0.004 0.004 9.777 10.751 hybrid_alltoall_any 227 10.6 0.637 2.954 9.222 10.424 dbcsr_mm_accdrv_process 3754 10.4 0.291 0.495 8.091 9.369 dbcsr_mm_accdrv_process_sort 3754 11.4 7.658 8.875 7.658 8.875 mp_sum_l 887 5.1 5.261 8.293 5.261 8.293 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.504 7.535 mp_irecv_dv 2335 11.1 2.488 7.477 2.488 7.477 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.413 7.328 multiply_cannon_sync_h2d 888 9.7 6.079 7.089 6.079 7.089 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.063 6.588 arnoldi_extremal 4 6.8 0.000 0.001 5.284 5.304 arnoldi_normal_ev 4 7.8 0.001 0.005 5.284 5.304 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.795 5.013 build_subspace 16 8.4 0.014 0.020 4.981 4.986 calculate_norms 1584 9.8 4.348 4.769 4.348 4.769 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.525 3.885 mp_allgather_i34 111 8.7 0.874 3.831 0.874 3.831 ls_scf_post 1 4.0 0.000 0.000 3.662 3.665 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.592 3.017 3.594 ls_scf_store_result 1 5.0 0.000 0.000 3.416 3.489 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.864 2.977 dbcsr_data_new 4116 9.9 2.108 2.450 2.108 2.450 make_images_sizes 222 9.7 0.000 0.000 1.032 2.446 mp_alltoall_i44 222 10.7 1.031 2.445 1.031 2.445 dbcsr_sort_data 325 11.1 1.866 2.173 1.866 2.173 dbcsr_finalize 304 7.8 0.026 0.032 1.605 1.912 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.898 1.900 rebuild_ks_matrix 3 7.3 0.000 0.000 1.879 1.882 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.879 1.882 make_images_pack 222 9.7 1.625 1.876 1.628 1.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.252000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2233.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 3.391644E+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 0.027 0.047 99.760 99.760 qs_energies 1 2.0 0.000 0.000 99.067 99.072 ls_scf 1 3.0 0.000 0.000 97.193 97.199 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.668 80.021 ls_scf_main 1 4.0 0.000 0.000 59.586 59.587 multiply_cannon 111 7.7 0.050 0.117 52.210 56.813 density_matrix_trs4 2 5.0 0.002 0.003 53.400 53.546 multiply_cannon_loop 111 8.7 0.153 0.170 47.057 50.958 ls_scf_init_scf 1 4.0 0.000 0.000 34.339 34.343 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.070 33.148 mp_waitall_1 6369 11.0 23.465 30.523 23.465 30.523 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.004 30.017 multiply_cannon_multrec 1332 9.7 14.082 17.344 21.980 24.927 make_m2s 222 7.7 0.006 0.008 21.488 22.902 make_images 222 8.7 3.143 3.616 21.437 22.855 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.543 17.685 make_images_data 222 9.7 0.004 0.004 12.059 13.763 hybrid_alltoall_any 227 10.6 0.795 3.741 11.548 13.349 dbcsr_mm_accdrv_process 3641 10.4 0.295 0.483 7.543 9.051 dbcsr_mm_accdrv_process_sort 3641 11.4 7.102 8.557 7.102 8.557 mp_sum_l 887 5.1 4.408 7.598 4.408 7.598 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.100 6.139 mp_irecv_dv 3229 10.9 2.076 6.087 2.076 6.087 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.424 6.078 multiply_cannon_sync_h2d 1332 9.7 5.436 6.026 5.436 6.026 arnoldi_extremal 4 6.8 0.458 0.533 5.859 5.874 arnoldi_normal_ev 4 7.8 0.058 0.073 5.401 5.870 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.690 5.857 build_subspace 16 8.4 0.014 0.021 4.946 4.954 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.733 4.916 mp_allgather_i34 111 8.7 2.216 4.606 2.216 4.606 calculate_norms 2376 9.8 4.155 4.486 4.155 4.486 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.645 3.907 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.685 3.185 3.687 dbcsr_sort_data 658 11.4 3.117 3.424 3.117 3.424 ls_scf_post 1 4.0 0.000 0.000 3.268 3.274 dbcsr_special_finalize 555 9.7 0.006 0.008 2.869 3.193 dbcsr_merge_single_wm 555 10.7 0.539 0.669 2.861 3.185 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.045 3.109 ls_scf_store_result 1 5.0 0.000 0.000 3.005 3.071 dbcsr_data_release 10477 10.7 1.565 2.417 1.565 2.417 make_images_pack 222 9.7 1.673 2.053 1.675 2.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.760000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 4.736053E+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 0.052 0.056 92.917 92.917 qs_energies 1 2.0 0.000 0.000 91.916 91.921 ls_scf 1 3.0 0.000 0.000 89.947 89.951 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.476 71.660 ls_scf_main 1 4.0 0.000 0.000 57.142 57.142 multiply_cannon 111 7.7 0.082 0.152 52.953 55.851 multiply_cannon_loop 111 8.7 0.087 0.092 50.371 52.087 density_matrix_trs4 2 5.0 0.002 0.003 50.172 50.263 ls_scf_init_scf 1 4.0 0.000 0.000 29.417 29.419 mp_waitall_1 5436 11.0 24.947 28.990 24.947 28.990 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.173 28.203 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.007 26.018 multiply_cannon_multrec 444 9.7 13.709 16.208 20.791 22.026 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.145 15.866 make_m2s 222 7.7 0.005 0.005 13.702 14.664 make_images 222 8.7 2.033 2.469 13.635 14.595 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.152 14.433 hybrid_alltoall_any 227 10.6 0.803 3.828 8.279 9.973 make_images_data 222 9.7 0.003 0.004 8.467 9.760 multiply_cannon_sync_h2d 444 9.7 6.756 8.059 6.756 8.059 dbcsr_mm_accdrv_process 3003 10.4 0.356 0.424 6.776 7.877 dbcsr_mm_accdrv_process_sort 3003 11.4 6.407 7.481 6.407 7.481 arnoldi_extremal 4 6.8 0.000 0.001 5.822 5.836 arnoldi_normal_ev 4 7.8 0.002 0.005 5.822 5.836 build_subspace 16 8.4 0.015 0.020 5.421 5.431 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.380 4.580 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.174 4.375 mp_sum_l 887 5.1 2.822 4.293 2.822 4.293 dbcsr_matrix_vector_mult_local 304 10.0 3.699 4.170 3.701 4.172 mp_allgather_i34 111 8.7 1.164 3.766 1.164 3.766 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.529 3.750 calculate_norms 792 9.8 3.605 3.714 3.605 3.714 mp_irecv_dv 1241 11.2 1.515 3.699 1.515 3.699 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.536 3.599 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.016 3.433 ls_scf_post 1 4.0 0.000 0.000 3.388 3.393 make_images_sizes 222 9.7 0.000 0.000 0.855 3.353 mp_alltoall_i44 222 10.7 0.855 3.352 0.855 3.352 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.218 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.302 dbcsr_data_new 4608 9.7 1.787 2.256 1.787 2.256 dbcsr_merge_all 275 8.9 0.480 0.530 2.053 2.142 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.028 2.029 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.996 1.997 qs_energies_init_hamiltonians 1 3.0 0.008 0.031 1.953 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.917000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3797.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_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 8.926175E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.090 105.527 105.528 qs_energies 1 2.0 0.000 0.000 103.807 103.812 ls_scf 1 3.0 0.000 0.000 100.787 100.791 dbcsr_multiply_generic 111 6.7 0.023 0.027 74.444 74.579 ls_scf_main 1 4.0 0.000 0.000 63.520 63.521 density_matrix_trs4 2 5.0 0.002 0.003 54.466 54.526 multiply_cannon 111 7.7 0.144 0.230 48.501 50.518 multiply_cannon_loop 111 8.7 0.098 0.100 45.482 46.009 ls_scf_init_scf 1 4.0 0.000 0.000 33.499 33.499 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.932 31.957 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.237 29.248 mp_waitall_1 4527 11.1 21.690 25.399 21.690 25.399 make_m2s 222 7.7 0.005 0.005 22.324 23.422 make_images 222 8.7 3.567 3.870 22.216 23.312 multiply_cannon_multrec 444 9.7 17.763 18.411 22.391 22.933 hybrid_alltoall_any 227 10.6 1.650 3.611 12.541 15.292 make_images_data 222 9.7 0.003 0.004 12.737 14.875 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.356 10.603 multiply_cannon_sync_h2d 444 9.7 8.781 8.839 8.781 8.839 arnoldi_extremal 4 6.8 0.001 0.001 7.390 7.404 arnoldi_normal_ev 4 7.8 0.003 0.009 7.389 7.404 build_subspace 16 8.4 0.026 0.036 6.848 6.859 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.495 5.658 dbcsr_matrix_vector_mult_local 304 10.0 5.052 5.379 5.054 5.382 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.172 5.262 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.872 5.125 dbcsr_mm_accdrv_process 1814 10.4 0.243 0.357 4.453 4.573 dbcsr_mm_accdrv_process_sort 1814 11.4 4.119 4.249 4.119 4.249 ls_scf_post 1 4.0 0.000 0.000 3.768 3.773 make_images_sizes 222 9.7 0.000 0.000 1.464 3.569 mp_alltoall_i44 222 10.7 1.464 3.569 1.464 3.569 mp_allgather_i34 111 8.7 1.125 3.538 1.125 3.538 ls_scf_store_result 1 5.0 0.000 0.000 3.488 3.526 calculate_norms 792 9.8 3.239 3.276 3.239 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.065 3.111 qs_energies_init_hamiltonians 1 3.0 0.030 0.047 2.989 2.989 dbcsr_merge_all 275 8.9 0.884 0.917 2.852 2.891 dbcsr_complete_redistribute 5 7.6 1.434 1.487 2.736 2.850 matrix_ls_to_qs 2 6.0 0.000 0.000 2.382 2.505 dbcsr_sort_data 325 11.1 2.431 2.491 2.431 2.491 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.468 2.470 rebuild_ks_matrix 3 7.3 0.000 0.000 2.402 2.404 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.402 2.404 dbcsr_data_new 6591 9.6 1.869 2.307 1.869 2.307 dbcsr_new_transposed 4 7.5 0.241 0.252 2.253 2.264 dbcsr_frobenius_norm 74 6.6 2.056 2.133 2.187 2.214 dbcsr_add_d 103 6.2 0.000 0.000 2.128 2.210 dbcsr_add_anytype 103 7.2 0.859 0.893 2.128 2.209 dbcsr_data_release 12724 10.6 1.966 2.163 1.966 2.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.528000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6940.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/443e0b30de9442b60cbf758811016979827d8a6a_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 589.836288E+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 4223043. 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 0.019 0.046 230.151 230.153 qs_mol_dyn_low 1 2.0 0.004 0.029 229.085 229.165 qs_forces 5 3.8 0.006 0.047 228.931 228.932 qs_energies 5 4.8 0.002 0.045 225.958 225.991 scf_env_do_scf 5 5.8 0.001 0.012 210.800 210.803 scf_env_do_scf_inner_loop 105 6.6 0.005 0.036 183.838 183.857 qs_scf_new_mos 105 7.6 0.000 0.001 144.150 144.349 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.150 144.349 ot_scf_mini 105 9.6 0.003 0.004 134.299 134.447 dbcsr_multiply_generic 1445 12.2 0.127 0.135 133.842 134.252 multiply_cannon 1445 13.2 0.275 0.285 114.878 116.549 multiply_cannon_loop 1445 14.2 2.810 2.979 113.211 114.418 velocity_verlet 4 3.0 0.005 0.038 105.994 105.997 ot_mini 105 10.6 0.001 0.014 60.376 60.490 multiply_cannon_multrec 69360 15.2 29.762 34.574 39.587 44.926 qs_ot_get_p 112 10.4 0.001 0.002 44.079 44.437 mp_waitall_1 488190 16.1 34.497 42.013 34.497 42.013 qs_ot_get_derivative 55 11.6 0.001 0.001 38.686 38.817 multiply_cannon_sync_h2d 69360 15.2 29.278 33.651 29.278 33.651 qs_ot_p2m_diag 40 11.0 0.021 0.037 33.111 33.210 multiply_cannon_metrocomm3 69360 15.2 0.201 0.215 25.622 33.152 cp_dbcsr_syevd 40 12.0 0.002 0.002 29.961 29.962 rebuild_ks_matrix 110 8.4 0.000 0.000 28.778 28.939 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.050 28.778 28.939 init_scf_loop 7 6.6 0.000 0.005 26.928 26.931 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.440 26.576 cp_fm_syevd 40 13.0 0.000 0.001 24.790 24.943 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.990 23.282 apply_single 62 13.6 0.000 0.000 22.990 23.282 prepare_preconditioner 7 7.6 0.000 0.000 22.038 22.065 make_preconditioner 7 8.6 0.000 0.004 22.038 22.065 ot_new_cg_direction 55 11.6 0.001 0.003 21.006 21.006 cp_fm_redistribute_end 40 14.0 10.052 20.058 10.057 20.059 cp_fm_syevd_base 40 14.0 9.995 20.002 9.995 20.002 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.672 17.125 calculate_rho_elec 110 8.6 0.030 0.032 16.671 17.124 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.907 14.989 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.672 14.871 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.868 13.964 mp_sum_l 4764 12.2 12.229 13.263 12.229 13.263 init_scf_run 5 5.8 0.000 0.001 12.502 12.503 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.502 12.503 pw_transfer 1645 12.4 0.079 0.098 11.670 11.869 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.532 11.735 calculate_dm_sparse 110 9.5 0.000 0.001 11.430 11.582 density_rs2pw 110 9.6 0.005 0.006 10.949 11.481 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.567 10.639 dbcsr_mm_accdrv_process 154766 15.8 6.188 6.387 9.693 10.599 cp_fm_cholesky_invert 7 10.6 10.476 10.483 10.476 10.483 qs_vxc_create 110 10.4 0.005 0.032 10.439 10.473 fft_wrap_pw1pw2_240 915 15.0 0.832 0.907 10.154 10.343 check_diag 80 13.5 8.596 8.887 9.506 9.658 acc_transpose_blocks 69360 15.2 0.358 0.377 8.212 8.813 fft3d_pb 915 16.0 2.402 2.581 8.339 8.523 sum_up_and_integrate 60 10.3 0.001 0.004 8.332 8.352 integrate_v_rspace 60 11.3 0.002 0.019 8.315 8.336 transfer_rs2pw 445 10.6 0.007 0.008 7.691 8.232 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.136 8.152 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.724 7.772 calculate_first_density_matrix 1 7.0 0.000 0.013 7.693 7.711 xc_rho_set_and_dset_create 110 12.4 0.077 0.117 7.347 7.593 multiply_cannon_metrocomm1 69360 15.2 0.095 0.101 4.618 7.246 xc_vxc_pw_create 60 11.3 0.039 0.050 7.003 7.040 make_full_single_inverse 7 9.6 0.001 0.004 6.843 6.870 make_m2s 2890 13.2 0.078 0.088 6.242 6.832 make_images 2890 14.2 0.237 0.259 6.133 6.724 xc_pw_derive 510 13.4 0.005 0.007 6.207 6.272 acc_transpose_blocks_kernels 69360 16.2 0.839 0.885 5.489 5.971 mp_alltoall_z22v 2340 17.7 5.330 5.656 5.330 5.656 mp_waitany 7680 13.5 4.375 5.106 4.375 5.106 jit_kernel_transpose 5 15.0 4.650 5.092 4.650 5.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.153000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.400000, yerr=2.059126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 443e0b30de9442b60cbf758811016979827d8a6a Summary: empty Status: OK