=== 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: eb5c80b9cfa321b86a0f21d0deddafcd7fbed371 ################# 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/01 job id: 50051368 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/02 job id: 50051369 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/03 job id: 50051370 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/04 job id: 50051371 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/05 job id: 50051372 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/06 job id: 50051373 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/07 job id: 50051374 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/08 job id: 50051375 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/09 job id: 50051376 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/10 job id: 50051377 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/11 job id: 50051378 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/12 job id: 50051379 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/13 job id: 50051380 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/14 job id: 50051381 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/15 job id: 50051382 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/16 job id: 50051383 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/17 job id: 50051384 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/18 job id: 50051385 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/19 job id: 50051387 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/20 job id: 50051389 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/21 job id: 50051390 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/22 job id: 50051391 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/23 job id: 50051392 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/24 job id: 50051393 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/25 job id: 50051394 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/26 job id: 50051395 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/27 job id: 50051396 --- 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 133.874 133.875 farming_run 1 2.0 133.210 133.212 133.846 133.849 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489461E+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 0.024 114.989 114.990 qs_energies 1 2.0 0.000 0.000 114.763 114.764 mp2_main 1 3.0 0.000 0.000 112.622 112.624 mp2_gpw_main 1 4.0 0.027 0.033 111.615 111.617 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.631 93.010 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.633 55.013 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.163 41.038 46.136 get_2c_integrals 1 6.0 0.008 0.009 37.003 37.827 integrate_v_rspace 273 8.0 0.440 0.452 24.646 29.548 pw_transfer 6555 10.6 0.379 0.409 26.804 27.243 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.434 25.921 grid_integrate_task_list 273 9.0 20.550 25.907 20.550 25.907 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 22.973 23.424 compute_2c_integrals 1 7.0 0.002 0.002 19.275 19.277 rpa_ri_compute_en 1 5.0 0.019 0.021 18.879 19.055 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.581 18.975 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.430 18.579 18.974 cp_fm_cholesky_decompose 12 8.2 17.667 18.523 17.667 18.523 cholesky_decomp 1 7.0 0.000 0.000 16.565 17.409 fft3d_s 5443 13.4 16.111 16.433 16.132 16.455 ao_to_mo_and_store_B_mult_1 272 7.0 10.768 15.320 10.768 15.320 calculate_wavefunction 272 8.0 5.391 5.564 12.305 12.929 rpa_num_int 1 6.0 0.000 0.001 10.655 10.655 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.635 10.647 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.577 contract_S_to_Q 8 9.0 0.000 0.000 8.907 8.998 calc_potential_gpw 544 9.5 0.004 0.005 8.230 8.703 parallel_gemm_fm 14 9.1 0.000 0.000 8.510 8.585 parallel_gemm_fm_cosma 14 10.1 8.510 8.585 8.510 8.585 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.070 8.367 potential_pw2rs 545 10.0 0.106 0.108 7.478 8.107 create_integ_mat 1 6.0 0.030 0.041 7.752 7.752 collocate_single_gaussian 272 10.0 0.039 0.041 7.323 7.578 array2fm 1 7.0 0.000 0.000 6.596 7.101 pw_scatter_s 2720 13.7 4.335 4.473 4.335 4.473 pw_gather_s 2722 13.2 3.440 3.796 3.440 3.796 array2fm_buffer_send 1 8.0 2.863 3.127 2.863 3.127 pw_poisson_solve 545 10.5 1.121 1.176 2.196 2.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.614656, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 0.026 0.037 422.066 422.068 farming_run 1 2.0 420.896 420.918 422.024 422.026 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235431E+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 26116008. 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 0.031 217.365 217.366 qs_energies 1 2.0 0.000 0.000 217.114 217.124 scf_env_do_scf 1 3.0 0.000 0.000 115.197 115.198 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.266 114.273 rebuild_ks_matrix 4 6.0 0.000 0.000 114.265 114.272 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.265 114.272 hfx_ks_matrix 4 8.0 0.001 0.001 113.918 113.922 integrate_four_center 4 9.0 0.153 0.464 113.917 113.921 integrate_four_center_main 4 10.0 0.133 0.610 101.817 105.815 integrate_four_center_bin 262 11.0 101.684 105.755 101.684 105.755 mp2_main 1 3.0 0.000 0.000 101.627 101.637 mp2_gpw_main 1 4.0 0.051 0.081 100.736 100.749 init_scf_loop 1 4.0 0.000 0.000 97.415 97.417 mp2_ri_gpw_compute_in 1 5.0 0.067 0.078 73.911 74.941 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.772 54.796 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.161 41.595 46.460 integrate_v_rspace 95 8.0 0.397 0.566 28.074 32.730 pw_transfer 2240 10.6 0.147 0.191 29.477 30.053 fft_wrap_pw1pw2 1868 11.4 0.018 0.024 28.490 29.051 ao_to_mo_and_store_B_mult_1 91 7.0 10.504 28.979 10.504 28.979 mp2_ri_gpw_compute_en 1 5.0 0.058 0.069 26.673 28.651 grid_integrate_task_list 95 9.0 23.486 28.371 23.486 28.371 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.239 26.790 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.896 24.990 25.001 get_2c_integrals 1 6.0 0.000 0.000 20.036 20.079 compute_2c_integrals 1 7.0 0.002 0.003 19.018 19.030 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.575 18.893 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.840 18.575 18.891 fft3d_s 1823 13.4 18.482 18.826 18.496 18.840 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.781 17.781 calculate_wavefunction 91 8.0 2.009 2.042 9.615 9.838 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.580 8.767 9.212 potential_pw2rs 186 10.0 0.033 0.034 8.418 8.982 local_gemm 172 8.0 8.213 8.656 8.213 8.656 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.126 8.533 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.517 7.998 8.484 calc_potential_gpw 182 9.5 0.002 0.002 7.885 8.170 mp_sync 37 10.5 4.071 8.101 4.071 8.101 collocate_single_gaussian 91 10.0 0.017 0.024 7.737 7.965 integrate_four_center_load 4 10.0 0.000 0.000 6.768 6.772 hfx_load_balance 1 11.0 0.000 0.000 6.768 6.772 mp_sendrecv_dm3 2068 8.0 6.044 6.507 6.044 6.507 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.396 6.339 6.396 pw_gather_s 912 13.2 4.477 5.049 4.477 5.049 pw_scatter_s 910 13.7 3.867 4.415 3.867 4.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.733582, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 453.308416E+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 1574207. 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 0.061 0.128 56.511 56.513 qs_mol_dyn_low 1 2.0 0.003 0.003 56.052 56.073 qs_forces 11 3.9 0.002 0.003 55.976 55.976 qs_energies 11 4.9 0.005 0.034 54.343 54.401 scf_env_do_scf 11 5.9 0.002 0.007 48.179 48.180 scf_env_do_scf_inner_loop 108 6.5 0.004 0.017 46.085 46.086 qs_scf_new_mos 108 7.5 0.000 0.000 35.847 36.144 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.847 36.143 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.298 35.739 ot_scf_mini 108 9.5 0.002 0.002 34.161 34.350 multiply_cannon 2286 13.5 0.179 0.186 26.951 28.560 multiply_cannon_loop 2286 14.5 1.822 1.927 26.276 27.923 velocity_verlet 10 3.0 0.006 0.014 26.589 26.590 ot_mini 108 10.5 0.001 0.001 20.940 21.194 qs_ot_get_derivative 108 11.5 0.001 0.001 17.895 18.086 mp_waitall_1 245248 16.5 9.174 15.911 9.174 15.911 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.469 13.771 multiply_cannon_multrec 54864 15.5 3.629 5.766 7.662 11.064 qs_ot_get_p 119 10.4 0.001 0.002 8.402 8.704 rebuild_ks_matrix 119 8.3 0.000 0.000 7.967 8.143 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.032 7.967 8.143 mp_sum_l 7287 12.8 6.292 8.108 6.292 8.108 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.998 7.174 multiply_cannon_sync_h2d 54864 15.5 5.128 6.963 5.128 6.963 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.389 6.509 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.881 6.319 dbcsr_mm_accdrv_process 76910 16.1 1.848 2.893 3.947 5.660 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.516 5.581 init_scf_run 11 5.9 0.000 0.001 4.725 4.725 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.725 4.725 cp_dbcsr_syevd 50 12.0 0.003 0.006 4.658 4.658 sum_up_and_integrate 119 10.3 0.001 0.002 4.537 4.548 integrate_v_rspace 119 11.3 0.002 0.002 4.525 4.538 cp_fm_diag_elpa 50 13.0 0.000 0.001 4.420 4.420 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.261 4.403 calculate_rho_elec 119 8.7 0.011 0.016 4.260 4.403 cp_fm_redistribute_end 50 14.0 2.251 4.384 2.259 4.388 cp_fm_diag_elpa_base 50 14.0 2.121 4.263 2.126 4.275 multiply_cannon_metrocomm1 54864 15.5 0.055 0.061 1.985 3.457 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.906 3.106 apply_single 119 13.6 0.000 0.000 2.906 3.106 calculate_dm_sparse 119 9.5 0.000 0.000 2.953 3.101 acc_transpose_blocks 54864 15.5 0.221 0.243 2.296 2.871 ot_diis_step 108 11.5 0.006 0.006 2.749 2.750 jit_kernel_multiply 13 15.8 2.036 2.673 2.036 2.673 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.579 2.660 calculate_first_density_matrix 1 7.0 0.000 0.003 2.520 2.522 density_rs2pw 119 9.7 0.004 0.004 2.250 2.335 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.274 2.276 wfi_extrapolate 11 7.9 0.001 0.001 2.136 2.136 grid_integrate_task_list 119 12.3 2.011 2.108 2.011 2.108 mp_sum_d 4137 12.0 1.472 2.107 1.472 2.107 init_scf_loop 11 6.9 0.000 0.003 2.075 2.076 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.001 2.054 potential_pw2rs 119 12.3 0.004 0.004 1.903 1.922 pw_transfer 1439 11.6 0.052 0.055 1.769 1.851 make_m2s 4572 13.5 0.053 0.055 1.742 1.792 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.694 1.777 make_images 4572 14.5 0.133 0.139 1.660 1.710 transfer_rs2pw 487 10.6 0.005 0.006 1.440 1.559 mp_alltoall_d11v 2130 13.8 1.385 1.531 1.385 1.531 fft3d_ps 1201 14.6 0.370 0.475 1.406 1.486 mp_waitany 12084 13.8 1.301 1.480 1.301 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.437 1.470 acc_transpose_blocks_sync 164592 16.5 1.206 1.443 1.206 1.443 grid_collocate_task_list 119 9.7 1.354 1.426 1.354 1.426 transfer_pw2rs 487 13.2 0.006 0.006 1.385 1.404 fft_wrap_pw1pw2_140 487 13.2 0.107 0.113 1.319 1.400 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.779 1.166 acc_transpose_blocks_kernels 54864 16.5 0.249 0.383 0.824 1.136 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.513000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 490.893312E+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 1546818. 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 0.026 0.064 40.142 40.144 qs_mol_dyn_low 1 2.0 0.003 0.005 39.873 39.881 qs_forces 11 3.9 0.003 0.011 39.794 39.795 qs_energies 11 4.9 0.002 0.007 38.096 38.101 scf_env_do_scf 11 5.9 0.001 0.001 32.902 32.902 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.377 30.379 dbcsr_multiply_generic 2286 12.5 0.109 0.112 22.428 22.811 qs_scf_new_mos 108 7.5 0.001 0.001 21.404 21.628 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.403 21.627 ot_scf_mini 108 9.5 0.003 0.003 20.437 20.602 velocity_verlet 10 3.0 0.004 0.020 18.902 18.903 multiply_cannon 2286 13.5 0.209 0.216 16.995 18.535 multiply_cannon_loop 2286 14.5 1.185 1.253 15.784 17.329 ot_mini 108 10.5 0.001 0.001 12.496 12.728 mp_waitall_1 200699 16.5 5.768 10.897 5.768 10.897 qs_ot_get_derivative 108 11.5 0.001 0.001 10.073 10.238 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.117 9.517 multiply_cannon_multrec 27432 15.5 1.819 3.999 6.143 8.936 rebuild_ks_matrix 119 8.3 0.000 0.000 6.965 7.113 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.965 7.112 dbcsr_mm_accdrv_process 47894 16.0 3.571 5.827 4.244 6.313 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.133 6.267 qs_ot_get_p 119 10.4 0.001 0.002 4.970 5.190 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.800 4.650 mp_sum_l 7287 12.8 2.201 4.270 2.201 4.270 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.022 4.087 apply_single 119 13.6 0.000 0.000 3.022 4.087 init_scf_run 11 5.9 0.000 0.001 3.951 3.951 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.950 3.951 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.839 3.867 calculate_rho_elec 119 8.7 0.021 0.024 3.838 3.867 sum_up_and_integrate 119 10.3 0.001 0.003 3.855 3.861 integrate_v_rspace 119 11.3 0.002 0.004 3.840 3.847 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.232 3.254 make_m2s 4572 13.5 0.052 0.053 2.798 3.125 make_images 4572 14.5 0.206 0.244 2.708 3.038 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.814 2.815 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.457 2.544 init_scf_loop 11 6.9 0.001 0.003 2.501 2.505 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.435 2.436 cp_fm_redistribute_end 50 14.0 1.230 2.398 1.234 2.400 ot_diis_step 108 11.5 0.015 0.043 2.372 2.373 calculate_first_density_matrix 1 7.0 0.000 0.003 2.353 2.355 cp_fm_diag_elpa_base 50 14.0 1.132 2.297 1.160 2.337 density_rs2pw 119 9.7 0.004 0.004 2.131 2.212 multiply_cannon_sync_h2d 27432 15.5 1.703 2.200 1.703 2.200 calculate_dm_sparse 119 9.5 0.000 0.001 2.106 2.187 pw_transfer 1439 11.6 0.065 0.068 1.982 2.030 acc_transpose_blocks 27432 15.5 0.112 0.118 1.611 1.947 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.890 1.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.934 1.937 grid_integrate_task_list 119 12.3 1.824 1.935 1.824 1.935 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.707 1.747 jit_kernel_multiply 9 16.4 0.615 1.727 0.615 1.727 make_images_data 4572 15.5 0.047 0.053 1.279 1.636 potential_pw2rs 119 12.3 0.006 0.006 1.598 1.607 prepare_preconditioner 11 7.9 0.000 0.000 1.578 1.604 make_preconditioner 11 8.9 0.000 0.001 1.578 1.604 fft3d_ps 1201 14.6 0.522 0.577 1.522 1.566 mp_alltoall_d11v 2130 13.8 1.293 1.566 1.293 1.566 wfi_extrapolate 11 7.9 0.002 0.013 1.535 1.535 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.476 1.532 hybrid_alltoall_any 4725 16.4 0.054 0.118 1.136 1.531 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.450 1.497 grid_collocate_task_list 119 9.7 1.285 1.336 1.285 1.336 transfer_rs2pw 487 10.6 0.005 0.005 1.253 1.327 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.239 1.291 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.264 mp_sum_d 4137 12.0 0.613 1.101 0.613 1.101 mp_allgather_i34 2286 14.5 0.634 1.049 0.634 1.049 transfer_pw2rs 487 13.2 0.004 0.005 1.002 1.011 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.957 0.959 acc_transpose_blocks_sync 82296 16.5 0.812 0.938 0.812 0.938 make_images_sizes 4572 15.5 0.005 0.005 0.653 0.906 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.892 0.905 mp_alltoall_i44 4572 16.5 0.648 0.902 0.648 0.902 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.661 0.871 mp_alltoall_z22v 1201 16.6 0.761 0.835 0.761 0.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.144000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=467.363636, yerr=1.666391 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.269248E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 32.630 32.631 qs_mol_dyn_low 1 2.0 0.003 0.003 32.251 32.259 qs_forces 11 3.9 0.002 0.003 32.172 32.173 qs_energies 11 4.9 0.021 0.045 30.614 30.616 scf_env_do_scf 11 5.9 0.001 0.001 25.517 25.517 scf_env_do_scf_inner_loop 108 6.5 0.008 0.012 23.108 23.108 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.237 17.306 qs_scf_new_mos 108 7.5 0.001 0.001 15.405 15.424 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.405 15.424 velocity_verlet 10 3.0 0.001 0.001 14.993 14.995 multiply_cannon 2286 13.5 0.194 0.198 14.033 14.730 ot_scf_mini 108 9.5 0.002 0.002 14.652 14.669 multiply_cannon_loop 2286 14.5 0.854 0.887 13.261 13.985 ot_mini 108 10.5 0.001 0.001 9.079 9.096 qs_ot_get_derivative 108 11.5 0.001 0.001 7.597 7.612 multiply_cannon_multrec 18288 15.5 1.880 2.875 7.154 7.508 dbcsr_mm_accdrv_process 38222 16.0 5.105 5.988 5.180 6.050 rebuild_ks_matrix 119 8.3 0.000 0.001 5.956 5.970 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.025 5.955 5.969 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.233 5.245 init_scf_run 11 5.9 0.000 0.001 3.871 3.871 scf_env_initial_rho_setup 11 6.9 0.002 0.003 3.871 3.871 sum_up_and_integrate 119 10.3 0.001 0.001 3.504 3.509 integrate_v_rspace 119 11.3 0.002 0.003 3.491 3.498 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.894 3.479 mp_waitall_1 158411 16.6 2.615 3.395 2.615 3.395 qs_ot_get_p 119 10.4 0.001 0.002 3.345 3.364 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.317 3.324 calculate_rho_elec 119 8.7 0.031 0.031 3.317 3.323 calculate_first_density_matrix 1 7.0 0.000 0.000 2.666 2.669 init_scf_loop 11 6.9 0.000 0.000 2.392 2.393 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.987 2.345 apply_single 119 13.6 0.000 0.000 1.987 2.344 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.222 2.229 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.477 2.209 calculate_dm_sparse 119 9.5 0.000 0.001 2.017 2.027 density_rs2pw 119 9.7 0.004 0.004 1.879 1.993 make_m2s 4572 13.5 0.044 0.046 1.797 1.948 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.925 1.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.914 1.915 grid_integrate_task_list 119 12.3 1.797 1.896 1.797 1.896 pw_transfer 1439 11.6 0.065 0.068 1.878 1.888 make_images 4572 14.5 0.192 0.204 1.712 1.862 acc_transpose_blocks 18288 15.5 0.077 0.080 1.793 1.847 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.785 1.797 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.729 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.686 1.691 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.676 1.684 cp_fm_diag_elpa_base 50 14.0 1.651 1.665 1.674 1.683 prepare_preconditioner 11 7.9 0.000 0.000 1.630 1.633 make_preconditioner 11 8.9 0.000 0.000 1.630 1.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.491 1.572 ot_diis_step 108 11.5 0.011 0.012 1.469 1.469 fft_wrap_pw1pw2_140 487 13.2 0.179 0.182 1.430 1.442 mp_sum_l 7287 12.8 1.016 1.410 1.016 1.410 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.359 1.363 fft3d_ps 1201 14.6 0.546 0.568 1.339 1.350 grid_collocate_task_list 119 9.7 1.235 1.332 1.235 1.332 potential_pw2rs 119 12.3 0.007 0.008 1.309 1.314 multiply_cannon_sync_h2d 18288 15.5 1.049 1.181 1.049 1.181 wfi_extrapolate 11 7.9 0.001 0.001 1.152 1.152 transfer_rs2pw 487 10.6 0.005 0.005 0.986 1.102 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.967 0.997 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.961 0.961 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.956 make_images_data 4572 15.5 0.047 0.050 0.770 0.932 hybrid_alltoall_any 4725 16.4 0.059 0.117 0.667 0.862 mp_alltoall_d11v 2130 13.8 0.658 0.811 0.658 0.811 transfer_pw2rs 487 13.2 0.004 0.004 0.801 0.807 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.793 0.795 acc_transpose_blocks_sync 54864 16.5 0.729 0.788 0.729 0.788 jit_kernel_transpose 5 15.6 0.749 0.773 0.749 0.773 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.678 0.745 mp_alltoall_z22v 1201 16.6 0.644 0.717 0.644 0.717 cp_fm_cholesky_invert 11 10.9 0.688 0.692 0.688 0.692 mp_waitany 9880 13.7 0.525 0.655 0.525 0.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.631000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 559.046656E+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 0.012 0.054 36.847 36.869 qs_mol_dyn_low 1 2.0 0.003 0.003 36.669 36.682 qs_forces 11 3.9 0.003 0.004 36.605 36.608 qs_energies 11 4.9 0.001 0.002 34.899 34.905 scf_env_do_scf 11 5.9 0.001 0.007 29.860 29.862 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.457 26.457 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.168 20.316 velocity_verlet 10 3.0 0.001 0.002 18.479 18.481 qs_scf_new_mos 108 7.5 0.001 0.001 18.315 18.378 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.314 18.378 ot_scf_mini 108 9.5 0.002 0.003 17.276 17.335 multiply_cannon 2286 13.5 0.219 0.228 16.324 16.857 multiply_cannon_loop 2286 14.5 1.544 1.605 15.366 15.814 ot_mini 108 10.5 0.001 0.001 10.700 10.764 multiply_cannon_multrec 27432 15.5 2.483 3.136 9.044 9.285 qs_ot_get_derivative 108 11.5 0.001 0.001 8.830 8.889 dbcsr_mm_accdrv_process 47916 15.9 5.915 7.725 6.460 7.804 rebuild_ks_matrix 119 8.3 0.000 0.000 6.330 6.373 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.330 6.372 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.613 5.649 qs_ot_get_p 119 10.4 0.001 0.001 3.721 3.798 init_scf_run 11 5.9 0.000 0.001 3.706 3.706 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.706 3.706 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.046 3.530 sum_up_and_integrate 119 10.3 0.001 0.001 3.503 3.511 integrate_v_rspace 119 11.3 0.002 0.003 3.491 3.500 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.408 3.441 calculate_rho_elec 119 8.7 0.040 0.046 3.407 3.440 init_scf_loop 11 6.9 0.000 0.000 3.382 3.383 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.194 2.649 apply_single 119 13.6 0.000 0.000 2.193 2.649 mp_waitall_1 137007 16.6 1.935 2.562 1.935 2.562 prepare_preconditioner 11 7.9 0.000 0.000 2.554 2.560 make_preconditioner 11 8.9 0.000 0.000 2.554 2.560 acc_transpose_blocks 27432 15.5 0.118 0.120 2.434 2.541 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.159 2.492 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.389 2.419 make_m2s 4572 13.5 0.054 0.056 2.300 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.322 2.325 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.292 2.304 make_images 4572 14.5 0.274 0.334 2.192 2.288 calculate_dm_sparse 119 9.5 0.000 0.000 2.142 2.195 pw_transfer 1439 11.6 0.065 0.069 1.999 2.036 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.908 1.947 grid_integrate_task_list 119 12.3 1.830 1.944 1.830 1.944 density_rs2pw 119 9.7 0.003 0.004 1.837 1.930 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.880 1.881 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.829 1.830 ot_diis_step 108 11.5 0.012 0.013 1.829 1.830 mp_sum_l 7287 12.8 1.114 1.705 1.114 1.705 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.665 1.675 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.575 1.616 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.597 1.606 cp_fm_diag_elpa_base 50 14.0 1.564 1.580 1.595 1.603 acc_transpose_blocks_sync 82296 16.5 1.458 1.565 1.458 1.565 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.857 1.475 fft3d_ps 1201 14.6 0.592 0.651 1.396 1.422 wfi_extrapolate 11 7.9 0.001 0.001 1.330 1.330 grid_collocate_task_list 119 9.7 1.249 1.324 1.249 1.324 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.304 1.319 potential_pw2rs 119 12.3 0.009 0.009 1.310 1.314 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.241 cp_fm_upper_to_full 72 14.2 0.828 1.180 0.828 1.180 jit_kernel_multiply 6 16.1 0.480 1.107 0.480 1.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.071 1.073 dbcsr_complete_redistribute 329 12.2 0.114 0.138 0.761 1.036 transfer_rs2pw 487 10.6 0.004 0.005 0.908 1.006 make_images_data 4572 15.5 0.048 0.051 0.865 0.970 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.778 0.944 mp_alltoall_d11v 2130 13.8 0.804 0.909 0.804 0.909 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.878 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.828 0.839 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.557 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.810 cp_fm_cholesky_invert 11 10.9 0.756 0.760 0.756 0.760 transfer_pw2rs 487 13.2 0.004 0.004 0.742 0.745 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.869000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.727273, yerr=4.024717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 629.878784E+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 0.031 0.052 28.601 28.602 qs_mol_dyn_low 1 2.0 0.003 0.003 28.377 28.384 qs_forces 11 3.9 0.002 0.002 28.317 28.318 qs_energies 11 4.9 0.001 0.002 26.609 26.611 scf_env_do_scf 11 5.9 0.000 0.001 21.935 21.935 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.503 19.503 velocity_verlet 10 3.0 0.002 0.002 14.634 14.637 dbcsr_multiply_generic 2286 12.5 0.095 0.097 12.963 13.031 qs_scf_new_mos 108 7.5 0.001 0.001 11.720 11.744 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.719 11.744 ot_scf_mini 108 9.5 0.002 0.002 11.007 11.034 multiply_cannon 2286 13.5 0.225 0.230 10.347 10.821 multiply_cannon_loop 2286 14.5 0.645 0.669 9.425 9.573 ot_mini 108 10.5 0.001 0.001 6.285 6.315 multiply_cannon_multrec 9144 15.5 1.653 1.841 5.898 6.089 rebuild_ks_matrix 119 8.3 0.000 0.000 5.850 5.878 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.849 5.878 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.203 5.229 qs_ot_get_derivative 108 11.5 0.001 0.001 4.993 5.019 dbcsr_mm_accdrv_process 12550 15.8 3.474 4.129 4.136 4.206 sum_up_and_integrate 119 10.3 0.001 0.001 3.479 3.483 integrate_v_rspace 119 11.3 0.003 0.011 3.469 3.473 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.429 3.435 calculate_rho_elec 119 8.7 0.060 0.061 3.428 3.435 init_scf_run 11 5.9 0.000 0.001 3.246 3.246 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.245 3.246 qs_ot_get_p 119 10.4 0.001 0.002 2.840 2.874 init_scf_loop 11 6.9 0.000 0.000 2.413 2.413 pw_transfer 1439 11.6 0.065 0.067 2.160 2.169 calculate_first_density_matrix 1 7.0 0.000 0.000 2.102 2.103 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.067 2.076 make_m2s 4572 13.5 0.034 0.034 1.816 1.982 grid_integrate_task_list 119 12.3 1.866 1.945 1.866 1.945 make_images 4572 14.5 0.268 0.300 1.726 1.891 density_rs2pw 119 9.7 0.003 0.003 1.828 1.881 qs_ot_p2m_diag 50 11.0 0.023 0.023 1.842 1.844 mp_waitall_1 115863 16.7 1.334 1.809 1.334 1.809 calculate_dm_sparse 119 9.5 0.000 0.000 1.740 1.760 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.720 1.741 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.709 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.677 1.682 make_preconditioner 11 8.9 0.000 0.000 1.677 1.682 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.595 1.596 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.567 1.595 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.533 1.534 acc_transpose_blocks 9144 15.5 0.042 0.043 1.459 1.499 fft3d_ps 1201 14.6 0.652 0.666 1.418 1.428 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.389 1.403 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.373 1.381 grid_collocate_task_list 119 9.7 1.297 1.363 1.297 1.363 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.326 1.334 cp_fm_diag_elpa_base 50 14.0 1.298 1.318 1.324 1.332 potential_pw2rs 119 12.3 0.010 0.010 1.281 1.284 ot_diis_step 108 11.5 0.013 0.013 1.278 1.278 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.215 1.250 apply_single 119 13.6 0.000 0.000 1.215 1.250 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.219 1.219 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.123 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.092 1.092 jit_kernel_multiply 6 16.2 0.623 1.020 0.623 1.020 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.757 1.003 make_images_data 4572 15.5 0.042 0.045 0.789 0.980 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.927 transfer_rs2pw 487 10.6 0.004 0.004 0.814 0.871 mp_alltoall_d11v 2130 13.8 0.763 0.869 0.763 0.869 cp_fm_cholesky_invert 11 10.9 0.840 0.843 0.840 0.843 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.817 0.824 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.767 0.770 acc_transpose_blocks_sync 27432 16.5 0.727 0.764 0.727 0.764 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.668 0.716 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.675 0.678 mp_alltoall_z22v 1201 16.6 0.640 0.674 0.640 0.674 mp_allgather_i34 2286 14.5 0.238 0.666 0.238 0.666 transfer_pw2rs 487 13.2 0.003 0.004 0.664 0.665 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.312 0.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.602000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.000000, yerr=9.361430 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 764.735488E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 43.559 43.561 qs_mol_dyn_low 1 2.0 0.003 0.003 43.329 43.338 qs_forces 11 3.9 0.002 0.003 43.267 43.268 qs_energies 11 4.9 0.001 0.001 41.227 41.230 scf_env_do_scf 11 5.9 0.001 0.001 35.595 35.595 scf_env_do_scf_inner_loop 108 6.5 0.006 0.008 27.612 27.614 velocity_verlet 10 3.0 0.001 0.002 24.636 24.641 dbcsr_multiply_generic 2286 12.5 0.104 0.107 19.267 19.524 qs_scf_new_mos 108 7.5 0.001 0.001 17.680 17.770 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.679 17.770 ot_scf_mini 108 9.5 0.002 0.002 16.511 16.608 multiply_cannon 2286 13.5 0.301 0.311 15.121 16.120 multiply_cannon_loop 2286 14.5 0.873 0.883 13.815 14.862 ot_mini 108 10.5 0.001 0.001 10.035 10.149 multiply_cannon_multrec 9144 15.5 3.345 4.757 8.740 8.863 qs_ot_get_derivative 108 11.5 0.001 0.001 7.909 8.007 init_scf_loop 11 6.9 0.000 0.000 7.954 7.960 rebuild_ks_matrix 119 8.3 0.000 0.000 7.429 7.570 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.429 7.569 prepare_preconditioner 11 7.9 0.000 0.000 6.893 6.909 make_preconditioner 11 8.9 0.000 0.000 6.893 6.909 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.699 6.828 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.485 6.780 dbcsr_mm_accdrv_process 12550 15.8 4.683 6.468 5.259 6.568 cp_fm_upper_to_full 72 14.2 3.193 4.538 3.193 4.538 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.311 4.320 calculate_rho_elec 119 8.7 0.117 0.120 4.310 4.319 sum_up_and_integrate 119 10.3 0.001 0.001 3.943 3.949 integrate_v_rspace 119 11.3 0.003 0.003 3.933 3.939 qs_ot_get_p 119 10.4 0.044 0.048 3.552 3.693 mp_waitall_1 94719 16.7 2.466 3.598 2.466 3.598 init_scf_run 11 5.9 0.000 0.001 3.571 3.571 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.571 3.571 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.893 3.314 pw_transfer 1439 11.6 0.069 0.069 3.100 3.108 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.001 3.011 dbcsr_complete_redistribute 329 12.2 0.282 0.286 1.973 2.812 make_m2s 4572 13.5 0.037 0.037 2.511 2.695 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.278 2.606 apply_single 119 13.6 0.000 0.000 2.278 2.605 make_images 4572 14.5 0.352 0.386 2.390 2.572 fft_wrap_pw1pw2_140 487 13.2 0.650 0.657 2.541 2.553 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.651 2.475 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.405 2.400 density_rs2pw 119 9.7 0.003 0.004 2.317 2.351 calculate_dm_sparse 119 9.5 0.000 0.000 2.238 2.260 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.403 2.222 mp_alltoall_i22 627 13.8 1.385 2.171 1.385 2.171 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.120 2.170 grid_integrate_task_list 119 12.3 2.099 2.133 2.099 2.133 ot_diis_step 108 11.5 0.014 0.015 2.096 2.096 calculate_first_density_matrix 1 7.0 0.000 0.000 2.095 2.096 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.062 2.063 fft3d_ps 1201 14.6 0.869 0.885 1.921 1.925 mp_sum_l 7287 12.8 1.091 1.882 1.091 1.882 acc_transpose_blocks 9144 15.5 0.044 0.045 1.826 1.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.819 1.820 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.774 1.775 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.695 1.696 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.562 1.599 grid_collocate_task_list 119 9.7 1.536 1.559 1.536 1.559 cp_fm_cholesky_invert 11 10.9 1.520 1.523 1.520 1.523 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.165 1.459 potential_pw2rs 119 12.3 0.014 0.014 1.416 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.413 1.413 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.389 1.389 cp_fm_diag_elpa_base 50 14.0 1.244 1.297 1.387 1.387 make_images_data 4572 15.5 0.046 0.049 1.115 1.362 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.332 1.353 mp_alltoall_d11v 2130 13.8 1.232 1.262 1.232 1.262 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.246 acc_transpose_blocks_sync 27432 16.5 1.115 1.133 1.115 1.133 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.105 1.121 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.997 1.049 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.939 0.951 mp_alltoall_z22v 1201 16.6 0.916 0.944 0.916 0.944 qs_create_task_list 11 7.9 0.000 0.000 0.930 0.940 generate_qs_task_list 11 8.9 0.367 0.386 0.930 0.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.561000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.181818, yerr=15.001928 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 502.484992E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 82.207 82.210 qs_mol_dyn_low 1 2.0 0.003 0.006 81.560 81.570 qs_forces 11 3.9 0.003 0.004 81.485 81.486 qs_energies 11 4.9 0.001 0.002 78.634 78.650 scf_env_do_scf 11 5.9 0.001 0.001 69.713 69.716 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.264 64.265 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.184 51.594 qs_scf_new_mos 99 7.5 0.000 0.001 47.269 47.377 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.269 47.377 ot_scf_mini 99 9.5 0.002 0.002 44.840 44.939 multiply_cannon 2055 13.4 0.176 0.184 42.658 43.906 velocity_verlet 10 3.0 0.002 0.002 43.440 43.441 multiply_cannon_loop 2055 14.4 1.777 1.814 41.718 42.996 ot_mini 99 10.5 0.001 0.001 26.679 26.781 qs_ot_get_derivative 99 11.5 0.001 0.001 19.801 19.898 multiply_cannon_multrec 49320 15.4 11.397 12.110 17.397 18.054 rebuild_ks_matrix 110 8.3 0.000 0.001 14.480 14.733 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.479 14.732 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.691 12.933 mp_waitall_1 220248 16.4 10.893 11.795 10.893 11.795 multiply_cannon_sync_h2d 49320 15.4 9.651 10.351 9.651 10.351 qs_ot_get_p 110 10.4 0.001 0.002 9.588 9.725 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.751 8.313 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.599 7.941 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.245 7.702 apply_single 110 13.6 0.000 0.000 7.245 7.702 sum_up_and_integrate 110 10.3 0.002 0.003 6.976 6.990 integrate_v_rspace 110 11.3 0.002 0.003 6.951 6.971 init_scf_run 11 5.9 0.000 0.001 6.862 6.862 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.862 6.862 ot_diis_step 99 11.5 0.006 0.006 6.621 6.621 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.332 6.473 calculate_rho_elec 110 8.6 0.020 0.024 6.332 6.472 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.386 6.416 dbcsr_mm_accdrv_process 87628 16.1 3.046 3.128 5.870 6.141 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.528 5.528 init_scf_loop 11 6.9 0.000 0.000 5.418 5.419 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.303 5.356 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.964 4.984 cp_fm_diag_elpa_base 48 14.0 4.951 4.969 4.962 4.982 mp_sum_l 6594 12.7 4.167 4.940 4.167 4.940 wfi_extrapolate 11 7.9 0.001 0.001 4.068 4.069 make_m2s 4110 13.4 0.060 0.065 3.922 4.015 make_images 4110 14.4 0.178 0.190 3.826 3.924 calculate_dm_sparse 110 9.5 0.001 0.001 3.802 3.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.589 3.593 density_rs2pw 110 9.6 0.004 0.004 3.368 3.552 grid_integrate_task_list 110 12.3 3.259 3.458 3.259 3.458 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.451 3.407 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.285 3.321 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.216 3.277 prepare_preconditioner 11 7.9 0.000 0.000 3.250 3.273 make_preconditioner 11 8.9 0.000 0.000 3.250 3.273 pw_transfer 1331 11.6 0.055 0.070 3.095 3.175 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.041 3.091 fft_wrap_pw1pw2 1111 12.6 0.007 0.009 3.006 3.088 calculate_first_density_matrix 1 7.0 0.000 0.001 2.689 2.698 fft_wrap_pw1pw2_140 451 13.1 0.309 0.335 2.569 2.657 potential_pw2rs 110 12.3 0.005 0.006 2.576 2.595 acc_transpose_blocks 49320 15.4 0.207 0.216 2.540 2.594 jit_kernel_multiply 13 15.9 2.546 2.585 2.546 2.585 mp_alltoall_d11v 2046 13.8 2.036 2.464 2.036 2.464 fft3d_ps 1111 14.6 0.797 0.886 2.299 2.364 grid_collocate_task_list 110 9.6 2.158 2.287 2.158 2.287 transfer_rs2pw 451 10.6 0.005 0.006 1.903 2.095 mp_sum_d 3891 11.9 1.416 1.985 1.416 1.985 mp_waitany 14300 13.8 1.812 1.950 1.812 1.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.918 1.936 make_images_data 4110 15.4 0.043 0.048 1.774 1.895 cp_fm_cholesky_invert 11 10.9 1.778 1.782 1.778 1.782 hybrid_alltoall_any 4261 16.3 0.084 0.479 1.527 1.772 transfer_pw2rs 451 13.1 0.006 0.007 1.742 1.753 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.664 1.695 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.210000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.603240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.220160E+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 3406133. 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 0.016 0.042 68.731 68.732 qs_mol_dyn_low 1 2.0 0.003 0.003 68.412 68.421 qs_forces 11 3.9 0.003 0.003 68.300 68.301 qs_energies 11 4.9 0.003 0.010 64.979 64.982 scf_env_do_scf 11 5.9 0.001 0.002 56.340 56.344 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.813 48.814 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.832 37.996 velocity_verlet 10 3.0 0.013 0.020 35.842 35.844 qs_scf_new_mos 99 7.5 0.001 0.001 33.077 33.221 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.077 33.221 multiply_cannon 2055 13.4 0.225 0.242 31.246 32.314 ot_scf_mini 99 9.5 0.003 0.004 31.414 31.542 multiply_cannon_loop 2055 14.4 1.156 1.184 30.000 30.796 ot_mini 99 10.5 0.002 0.010 18.332 18.476 multiply_cannon_multrec 24660 15.4 6.995 8.616 14.034 15.745 rebuild_ks_matrix 110 8.3 0.000 0.000 13.471 13.577 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 13.470 13.577 qs_ot_get_derivative 99 11.5 0.001 0.001 12.550 12.681 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.855 11.950 mp_waitall_1 176588 16.5 7.433 10.030 7.433 10.030 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.101 7.656 init_scf_loop 11 6.9 0.001 0.004 7.493 7.494 multiply_cannon_sync_h2d 24660 15.4 6.381 7.419 6.381 7.419 dbcsr_mm_accdrv_process 52282 16.1 5.558 6.469 6.869 7.216 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.429 7.057 apply_single 110 13.6 0.000 0.001 6.429 7.057 qs_ot_get_p 110 10.4 0.002 0.004 6.437 6.580 sum_up_and_integrate 110 10.3 0.003 0.012 6.327 6.342 integrate_v_rspace 110 11.3 0.002 0.003 6.299 6.315 init_scf_run 11 5.9 0.000 0.001 6.220 6.221 scf_env_initial_rho_setup 11 6.9 0.002 0.002 6.220 6.220 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.840 5.847 calculate_rho_elec 110 8.6 0.039 0.047 5.839 5.847 ot_diis_step 99 11.5 0.010 0.014 5.734 5.735 prepare_preconditioner 11 7.9 0.000 0.000 5.468 5.487 make_preconditioner 11 8.9 0.000 0.002 5.468 5.487 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.733 5.430 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.064 5.225 make_m2s 4110 13.4 0.057 0.060 4.191 4.635 qs_ot_p2m_diag 48 11.0 0.029 0.062 4.532 4.565 make_images 4110 14.4 0.407 0.467 4.079 4.518 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.070 4.071 pw_transfer 1331 11.6 0.066 0.073 3.527 3.675 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.420 3.570 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.508 3.523 cp_fm_diag_elpa_base 48 14.0 3.459 3.480 3.505 3.520 wfi_extrapolate 11 7.9 0.001 0.001 3.497 3.497 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.267 3.327 grid_integrate_task_list 110 12.3 3.157 3.325 3.157 3.325 density_rs2pw 110 9.6 0.004 0.004 3.086 3.280 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.254 3.256 fft_wrap_pw1pw2_140 451 13.1 0.359 0.379 2.938 3.093 calculate_dm_sparse 110 9.5 0.001 0.001 2.964 2.991 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.929 2.953 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.968 2.756 make_images_data 4110 15.4 0.048 0.052 2.259 2.731 fft3d_ps 1111 14.6 1.116 1.357 2.510 2.674 calculate_first_density_matrix 1 7.0 0.001 0.011 2.626 2.629 cp_fm_cholesky_invert 11 10.9 2.551 2.558 2.551 2.558 mp_sum_l 6594 12.7 1.780 2.501 1.780 2.501 grid_collocate_task_list 110 9.6 2.171 2.292 2.171 2.292 potential_pw2rs 110 12.3 0.008 0.009 2.215 2.225 acc_transpose_blocks 24660 15.4 0.114 0.117 1.973 1.999 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.977 1.998 mp_alltoall_d11v 2046 13.8 1.711 1.905 1.711 1.905 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.902 1.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.774 1.787 transfer_rs2pw 451 10.6 0.006 0.007 1.436 1.653 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.771 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.552 jit_kernel_multiply 8 16.3 0.946 1.536 0.946 1.536 mp_allgather_i34 2055 14.4 0.549 1.525 0.549 1.525 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.491 mp_irecv_dv 57340 16.2 0.647 1.454 0.647 1.454 mp_waitany 10164 13.8 1.192 1.437 1.192 1.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.732000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.454545, yerr=6.919848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 668.647424E+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 0.025 0.076 60.269 60.279 qs_mol_dyn_low 1 2.0 0.003 0.003 59.574 59.635 qs_forces 11 3.9 0.003 0.003 59.493 59.493 qs_energies 11 4.9 0.001 0.002 56.259 56.262 scf_env_do_scf 11 5.9 0.001 0.001 48.376 48.377 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.839 39.839 velocity_verlet 10 3.0 0.001 0.002 32.424 32.426 dbcsr_multiply_generic 2055 12.4 0.127 0.133 28.690 28.904 qs_scf_new_mos 99 7.5 0.001 0.001 25.200 25.302 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.200 25.302 ot_scf_mini 99 9.5 0.002 0.003 23.953 24.071 multiply_cannon 2055 13.4 0.212 0.220 22.448 23.556 multiply_cannon_loop 2055 14.4 0.813 0.842 21.285 22.379 ot_mini 99 10.5 0.001 0.001 13.670 13.790 rebuild_ks_matrix 110 8.3 0.000 0.000 12.187 12.315 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.022 12.187 12.315 multiply_cannon_multrec 16440 15.4 3.675 4.721 9.901 10.814 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.678 10.794 mp_waitall_1 139946 16.5 6.798 9.572 6.798 9.572 qs_ot_get_derivative 99 11.5 0.001 0.001 9.271 9.389 init_scf_loop 11 6.9 0.000 0.000 8.505 8.505 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.269 6.901 prepare_preconditioner 11 7.9 0.000 0.000 6.762 6.781 make_preconditioner 11 8.9 0.000 0.000 6.762 6.781 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.063 6.412 sum_up_and_integrate 110 10.3 0.002 0.003 6.313 6.333 integrate_v_rspace 110 11.3 0.003 0.003 6.287 6.307 dbcsr_mm_accdrv_process 34862 16.1 5.386 5.800 6.070 6.229 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.771 5.782 calculate_rho_elec 110 8.6 0.059 0.060 5.770 5.781 qs_ot_get_p 110 10.4 0.001 0.002 5.447 5.568 init_scf_run 11 5.9 0.000 0.001 5.493 5.493 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.492 5.492 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.938 5.359 apply_single 110 13.6 0.000 0.000 4.937 5.359 make_m2s 4110 13.4 0.050 0.052 4.153 4.533 make_images 4110 14.4 0.400 0.522 4.038 4.413 ot_diis_step 99 11.5 0.011 0.011 4.370 4.370 multiply_cannon_sync_h2d 16440 15.4 3.259 3.876 3.259 3.876 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.175 3.830 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.803 3.807 pw_transfer 1331 11.6 0.065 0.072 3.654 3.668 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.547 3.564 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.433 3.434 grid_integrate_task_list 110 12.3 3.221 3.352 3.221 3.352 density_rs2pw 110 9.6 0.004 0.004 2.993 3.196 fft_wrap_pw1pw2_140 451 13.1 0.461 0.471 3.071 3.089 wfi_extrapolate 11 7.9 0.001 0.001 2.958 2.958 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.893 2.903 cp_fm_diag_elpa_base 48 14.0 2.827 2.858 2.891 2.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.812 2.813 make_images_data 4110 15.4 0.046 0.051 2.316 2.811 hybrid_alltoall_any 4261 16.3 0.109 0.379 2.073 2.754 cp_fm_cholesky_invert 11 10.9 2.612 2.618 2.612 2.618 calculate_dm_sparse 110 9.5 0.001 0.001 2.517 2.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.487 2.531 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.421 2.490 fft3d_ps 1111 14.6 1.124 1.139 2.464 2.486 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.429 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.876 2.385 grid_collocate_task_list 110 9.6 2.229 2.361 2.229 2.361 mp_irecv_dv 48980 15.7 0.802 2.250 0.802 2.250 potential_pw2rs 110 12.3 0.011 0.011 2.164 2.174 mp_sum_l 6594 12.7 1.489 2.032 1.489 2.032 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.958 1.959 dbcsr_complete_redistribute 325 12.2 0.333 0.396 1.467 1.934 mp_alltoall_d11v 2046 13.8 1.752 1.912 1.752 1.912 cp_fm_upper_to_full 70 14.2 1.408 1.788 1.408 1.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.700 1.714 acc_transpose_blocks 16440 15.4 0.077 0.080 1.583 1.677 cp_fm_cholesky_decompose 22 10.9 1.592 1.617 1.592 1.617 transfer_rs2pw 451 10.6 0.005 0.006 1.286 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.015 1.374 1.487 mp_allgather_i34 2055 14.4 0.478 1.476 0.478 1.476 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.454 1.463 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.997 1.452 mp_waitany 17072 13.8 1.119 1.349 1.119 1.349 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.269 1.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.279000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.454545, yerr=8.172722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 734.208000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 65.518 65.519 qs_mol_dyn_low 1 2.0 0.003 0.003 65.303 65.313 qs_forces 11 3.9 0.003 0.003 65.223 65.227 qs_energies 11 4.9 0.001 0.002 61.817 61.823 scf_env_do_scf 11 5.9 0.000 0.001 53.574 53.577 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.943 41.944 velocity_verlet 10 3.0 0.001 0.002 37.340 37.343 dbcsr_multiply_generic 2055 12.4 0.117 0.122 30.661 30.839 qs_scf_new_mos 99 7.5 0.001 0.001 27.260 27.372 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.259 27.372 ot_scf_mini 99 9.5 0.002 0.003 25.583 25.679 multiply_cannon 2055 13.4 0.239 0.257 23.560 24.755 multiply_cannon_loop 2055 14.4 1.436 1.491 22.126 22.726 ot_mini 99 10.5 0.001 0.001 14.723 14.844 multiply_cannon_multrec 24660 15.4 4.103 6.706 13.007 14.230 rebuild_ks_matrix 110 8.3 0.000 0.000 12.079 12.183 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 12.078 12.183 init_scf_loop 11 6.9 0.000 0.000 11.592 11.593 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.663 10.757 qs_ot_get_derivative 99 11.5 0.001 0.001 10.550 10.650 prepare_preconditioner 11 7.9 0.000 0.000 9.854 9.874 make_preconditioner 11 8.9 0.000 0.000 9.854 9.874 dbcsr_mm_accdrv_process 52304 16.0 7.770 9.200 8.744 9.685 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.070 9.551 sum_up_and_integrate 110 10.3 0.001 0.002 6.274 6.288 integrate_v_rspace 110 11.3 0.003 0.003 6.248 6.261 mp_waitall_1 121746 16.5 4.253 6.258 4.253 6.258 make_m2s 4110 13.4 0.059 0.062 5.509 5.907 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.826 5.841 calculate_rho_elec 110 8.6 0.078 0.081 5.826 5.840 qs_ot_get_p 110 10.4 0.001 0.002 5.688 5.835 make_images 4110 14.4 0.576 0.692 5.369 5.763 init_scf_run 11 5.9 0.000 0.001 5.511 5.511 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.511 5.511 cp_fm_upper_to_full 70 14.2 3.352 4.874 3.352 4.874 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.054 4.143 apply_single 110 13.6 0.000 0.000 4.054 4.143 ot_diis_step 99 11.5 0.011 0.011 4.136 4.136 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.833 3.849 dbcsr_complete_redistribute 325 12.2 0.416 0.494 2.681 3.830 pw_transfer 1331 11.6 0.065 0.073 3.755 3.790 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.648 3.688 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.486 3.543 grid_integrate_task_list 110 12.3 3.286 3.466 3.286 3.466 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.342 3.398 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.163 3.306 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.271 3.271 hybrid_alltoall_any 4261 16.3 0.123 0.459 2.329 3.211 fft_wrap_pw1pw2_140 451 13.1 0.531 0.544 3.143 3.184 make_images_data 4110 15.4 0.048 0.052 2.798 3.176 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.361 3.156 density_rs2pw 110 9.6 0.004 0.004 2.938 3.128 calculate_dm_sparse 110 9.5 0.001 0.001 2.980 3.017 wfi_extrapolate 11 7.9 0.001 0.001 2.997 2.997 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.775 2.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.827 2.829 mp_alltoall_i22 605 13.7 1.666 2.827 1.666 2.827 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.745 2.757 cp_fm_diag_elpa_base 48 14.0 2.598 2.655 2.743 2.755 cp_fm_cholesky_invert 11 10.9 2.657 2.665 2.657 2.665 multiply_cannon_sync_h2d 24660 15.4 2.407 2.576 2.407 2.576 acc_transpose_blocks 24660 15.4 0.111 0.113 2.455 2.566 fft3d_ps 1111 14.6 1.151 1.190 2.475 2.490 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.443 2.488 grid_collocate_task_list 110 9.6 2.264 2.435 2.264 2.435 calculate_first_density_matrix 1 7.0 0.000 0.000 2.419 2.423 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.239 2.244 mp_alltoall_d11v 2046 13.8 1.775 2.175 1.775 2.175 potential_pw2rs 110 12.3 0.012 0.013 2.077 2.081 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.796 1.830 mp_allgather_i34 2055 14.4 0.516 1.719 0.516 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.713 cp_fm_cholesky_decompose 22 10.9 1.644 1.696 1.644 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.655 1.666 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.562 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.828 1.551 acc_transpose_blocks_sync 73980 16.4 1.437 1.543 1.437 1.543 mp_sum_l 6594 12.7 0.900 1.516 0.900 1.516 mp_irecv_dv 62702 16.1 0.726 1.472 0.726 1.472 transfer_rs2pw 451 10.6 0.005 0.006 1.160 1.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.519000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.727273, yerr=9.478117 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 853.065728E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.039 55.438 55.439 qs_mol_dyn_low 1 2.0 0.003 0.003 55.122 55.131 qs_forces 11 3.9 0.003 0.003 55.053 55.054 qs_energies 11 4.9 0.001 0.002 51.384 51.389 scf_env_do_scf 11 5.9 0.000 0.001 43.176 43.176 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 35.501 35.502 velocity_verlet 10 3.0 0.002 0.003 31.183 31.186 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.305 23.418 qs_scf_new_mos 99 7.5 0.001 0.001 20.841 20.897 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.841 20.896 ot_scf_mini 99 9.5 0.002 0.003 19.584 19.603 multiply_cannon 2055 13.4 0.237 0.252 17.674 19.076 multiply_cannon_loop 2055 14.4 0.604 0.632 16.399 16.732 rebuild_ks_matrix 110 8.3 0.000 0.000 11.686 11.711 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.686 11.710 ot_mini 99 10.5 0.001 0.001 10.812 10.823 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.396 10.422 multiply_cannon_multrec 8220 15.4 3.199 4.407 7.596 8.476 init_scf_loop 11 6.9 0.000 0.000 7.627 7.628 mp_waitall_1 103326 16.6 5.884 7.334 5.884 7.334 qs_ot_get_derivative 99 11.5 0.001 0.001 7.127 7.146 sum_up_and_integrate 110 10.3 0.001 0.002 6.234 6.246 integrate_v_rspace 110 11.3 0.003 0.003 6.207 6.220 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.042 6.058 calculate_rho_elec 110 8.6 0.113 0.113 6.042 6.057 prepare_preconditioner 11 7.9 0.000 0.000 5.983 5.986 make_preconditioner 11 8.9 0.000 0.000 5.982 5.986 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.553 5.644 dbcsr_mm_accdrv_process 17442 15.9 3.105 4.092 4.258 5.170 init_scf_run 11 5.9 0.000 0.001 5.078 5.078 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.078 5.078 qs_ot_get_p 110 10.4 0.002 0.002 4.821 4.834 make_m2s 4110 13.4 0.037 0.039 4.273 4.560 make_images 4110 14.4 0.648 0.708 4.143 4.429 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.996 4.382 pw_transfer 1331 11.6 0.065 0.070 4.104 4.113 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.996 4.009 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.719 3.767 apply_single 110 13.6 0.000 0.000 3.719 3.766 ot_diis_step 99 11.5 0.012 0.012 3.662 3.662 grid_integrate_task_list 110 12.3 3.387 3.500 3.387 3.500 fft_wrap_pw1pw2_140 451 13.1 0.719 0.730 3.460 3.474 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.437 3.442 density_rs2pw 110 9.6 0.004 0.004 3.069 3.170 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.095 3.096 cp_fm_cholesky_invert 11 10.9 2.855 2.858 2.855 2.858 hybrid_alltoall_any 4261 16.3 0.201 0.866 2.238 2.763 wfi_extrapolate 11 7.9 0.001 0.001 2.726 2.726 make_images_data 4110 15.4 0.040 0.046 2.282 2.686 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.670 2.671 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.566 2.573 cp_fm_diag_elpa_base 48 14.0 2.510 2.538 2.564 2.572 fft3d_ps 1111 14.6 1.292 1.298 2.540 2.554 calculate_dm_sparse 110 9.5 0.001 0.001 2.486 2.530 grid_collocate_task_list 110 9.6 2.367 2.510 2.367 2.510 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.499 2.500 multiply_cannon_sync_h2d 8220 15.4 2.386 2.459 2.386 2.459 calculate_first_density_matrix 1 7.0 0.000 0.000 2.256 2.258 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.137 2.151 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.045 2.078 potential_pw2rs 110 12.3 0.015 0.015 2.035 2.039 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.021 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.928 1.941 mp_alltoall_d11v 2046 13.8 1.606 1.793 1.606 1.793 cp_fm_cholesky_decompose 22 10.9 1.691 1.704 1.691 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.618 1.621 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.501 1.612 dbcsr_complete_redistribute 325 12.2 0.554 0.578 1.430 1.533 mp_allgather_i34 2055 14.4 0.443 1.530 0.443 1.530 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.447 acc_transpose_blocks 8220 15.4 0.040 0.041 1.368 1.406 qs_create_task_list 11 7.9 0.000 0.000 1.203 1.301 generate_qs_task_list 11 8.9 0.373 0.440 1.203 1.301 transfer_rs2pw 451 10.6 0.005 0.005 1.145 1.265 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.770 1.247 jit_kernel_multiply 7 15.4 0.837 1.196 0.837 1.196 mp_waitany 9240 13.8 1.053 1.187 1.053 1.187 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.126 1.143 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.439000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.272727, yerr=10.480206 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 1.432207E+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 0.027 0.067 88.436 88.450 qs_mol_dyn_low 1 2.0 0.003 0.003 88.068 88.091 qs_forces 11 3.9 0.007 0.012 87.997 87.998 qs_energies 11 4.9 0.002 0.002 83.858 83.863 scf_env_do_scf 11 5.9 0.000 0.001 73.732 73.733 velocity_verlet 10 3.0 0.004 0.006 56.291 56.298 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.414 45.416 dbcsr_multiply_generic 2055 12.4 0.128 0.130 29.747 29.922 init_scf_loop 11 6.9 0.000 0.000 28.247 28.251 qs_scf_new_mos 99 7.5 0.001 0.001 27.232 27.313 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.232 27.312 prepare_preconditioner 11 7.9 0.000 0.000 26.207 26.221 make_preconditioner 11 8.9 0.000 0.000 26.207 26.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.593 25.684 ot_scf_mini 99 9.5 0.002 0.002 25.432 25.510 multiply_cannon 2055 13.4 0.329 0.344 22.504 23.154 multiply_cannon_loop 2055 14.4 0.822 0.837 20.673 21.086 cp_fm_upper_to_full 70 14.2 12.730 18.068 12.730 18.068 ot_mini 99 10.5 0.001 0.001 14.352 14.428 rebuild_ks_matrix 110 8.3 0.000 0.000 14.129 14.228 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.129 14.227 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.829 12.917 dbcsr_complete_redistribute 325 12.2 1.005 1.056 7.168 10.365 multiply_cannon_multrec 8220 15.4 4.054 4.217 9.713 9.845 qs_ot_get_derivative 99 11.5 0.001 0.001 9.576 9.656 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.161 9.369 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.598 8.779 mp_waitall_1 84994 16.7 7.632 8.703 7.632 8.703 mp_alltoall_i22 605 13.7 5.250 8.423 5.250 8.423 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.573 7.611 calculate_rho_elec 110 8.6 0.222 0.223 7.573 7.611 sum_up_and_integrate 110 10.3 0.002 0.002 7.049 7.065 integrate_v_rspace 110 11.3 0.004 0.004 7.020 7.038 make_m2s 4110 13.4 0.043 0.043 5.411 5.940 qs_ot_get_p 110 10.4 0.003 0.003 5.781 5.867 pw_transfer 1331 11.6 0.075 0.075 5.854 5.864 init_scf_run 11 5.9 0.000 0.001 5.804 5.804 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.804 5.804 make_images 4110 14.4 0.883 0.935 5.223 5.751 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.736 5.745 dbcsr_mm_accdrv_process 11614 15.7 3.813 4.131 5.511 5.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.872 5.391 apply_single 110 13.6 0.000 0.000 4.872 5.391 cp_fm_cholesky_invert 11 10.9 5.342 5.347 5.342 5.347 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.931 5.273 fft_wrap_pw1pw2_140 451 13.1 1.367 1.372 5.033 5.045 ot_diis_step 99 11.5 0.015 0.016 4.761 4.761 qs_ot_p2m_diag 48 11.0 0.151 0.157 4.134 4.142 density_rs2pw 110 9.6 0.004 0.004 4.017 4.053 grid_integrate_task_list 110 12.3 3.704 3.796 3.704 3.796 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.719 3.723 hybrid_alltoall_any 4261 16.3 0.265 0.577 2.877 3.617 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.612 3.613 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.067 3.532 make_images_data 4110 15.4 0.045 0.047 2.858 3.497 fft3d_ps 1111 14.6 1.884 1.894 3.441 3.451 wfi_extrapolate 11 7.9 0.001 0.001 3.366 3.366 calculate_dm_sparse 110 9.5 0.001 0.001 3.166 3.196 multiply_cannon_sync_h2d 8220 15.4 3.131 3.150 3.131 3.150 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.013 3.014 cp_fm_diag_elpa_base 48 14.0 2.472 2.656 3.011 3.011 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.919 2.924 grid_collocate_task_list 110 9.6 2.690 2.759 2.690 2.759 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.680 2.725 potential_pw2rs 110 12.3 0.021 0.021 2.464 2.470 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.399 2.435 calculate_first_density_matrix 1 7.0 0.000 0.000 2.317 2.318 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.245 2.295 mp_alltoall_d11v 2046 13.8 2.143 2.223 2.143 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.112 2.185 cp_fm_cholesky_decompose 22 10.9 1.996 2.014 1.996 2.014 qs_create_task_list 11 7.9 0.000 0.000 1.919 1.964 generate_qs_task_list 11 8.9 0.729 0.783 1.919 1.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.897 1.907 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.779 1.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.450000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1266.363636, yerr=60.165337 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 633.536512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529097. 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 0.029 0.050 208.178 208.179 qs_mol_dyn_low 1 2.0 0.003 0.004 207.600 207.632 qs_forces 11 3.9 0.005 0.012 207.460 207.461 qs_energies 11 4.9 0.014 0.100 201.784 201.801 scf_env_do_scf 11 5.9 0.016 0.051 184.590 184.600 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.778 163.780 dbcsr_multiply_generic 2507 12.6 0.177 0.183 125.800 126.752 qs_scf_new_mos 117 7.6 0.001 0.001 124.512 124.809 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.511 124.808 velocity_verlet 10 3.0 0.016 0.044 123.416 123.417 ot_scf_mini 117 9.6 0.003 0.004 117.799 118.057 multiply_cannon 2507 13.6 0.236 0.244 101.983 103.057 multiply_cannon_loop 2507 14.6 2.385 2.426 99.828 101.055 ot_mini 117 10.6 0.001 0.001 66.469 66.757 multiply_cannon_multrec 60168 15.6 31.957 33.576 41.783 43.607 qs_ot_get_derivative 117 11.6 0.001 0.001 41.530 41.815 rebuild_ks_matrix 128 8.3 0.001 0.001 33.684 33.994 qs_ks_build_kohn_sham_matrix 128 9.3 0.032 0.087 33.683 33.993 mp_waitall_1 267128 16.5 29.017 31.621 29.017 31.621 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.232 30.527 qs_ot_get_p 128 10.4 0.042 0.277 29.143 29.502 multiply_cannon_sync_h2d 60168 15.6 26.411 28.171 26.411 28.171 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.426 25.617 apply_single 128 13.6 0.001 0.001 24.425 25.616 ot_diis_step 117 11.6 0.008 0.008 24.712 24.714 qs_ot_p2m_diag 83 11.4 0.081 0.125 22.361 22.411 init_scf_loop 11 6.9 0.002 0.013 20.715 20.737 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 20.009 20.246 cp_dbcsr_syevd 83 12.4 0.005 0.009 19.577 19.580 multiply_cannon_metrocomm3 60168 15.6 0.114 0.117 16.170 17.839 cp_fm_diag_elpa 83 13.4 0.001 0.003 16.638 16.669 cp_fm_diag_elpa_base 83 14.4 16.568 16.611 16.630 16.663 prepare_preconditioner 11 7.9 0.000 0.000 15.936 15.979 make_preconditioner 11 8.9 0.000 0.001 15.936 15.979 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.182 15.386 make_m2s 5014 13.6 0.104 0.112 13.718 14.120 sum_up_and_integrate 128 10.3 0.002 0.006 13.981 13.998 make_images 5014 14.6 0.398 0.418 13.533 13.947 integrate_v_rspace 128 11.3 0.005 0.030 13.922 13.942 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.026 13.152 calculate_rho_elec 128 8.7 0.045 0.064 13.026 13.151 init_scf_run 11 5.9 0.000 0.001 12.507 12.509 scf_env_initial_rho_setup 11 6.9 0.008 0.037 12.506 12.509 mp_sum_l 7950 12.9 8.846 10.017 8.846 10.017 dbcsr_mm_accdrv_process 124484 16.2 4.730 4.830 9.390 9.875 cp_fm_cholesky_invert 11 10.9 9.202 9.210 9.202 9.210 wfi_extrapolate 11 7.9 0.001 0.001 8.980 8.980 calculate_dm_sparse 128 9.5 0.001 0.001 8.549 8.660 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.166 8.314 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.064 8.173 multiply_cannon_metrocomm1 60168 15.6 0.093 0.099 6.411 8.101 pw_transfer 1547 11.6 0.073 0.094 7.548 7.727 make_images_data 5014 15.6 0.066 0.072 6.714 7.657 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.347 7.527 grid_integrate_task_list 128 12.3 7.058 7.424 7.058 7.424 density_rs2pw 128 9.7 0.006 0.007 6.689 7.330 hybrid_alltoall_any 5200 16.5 0.293 2.264 5.884 7.308 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.655 6.681 fft_wrap_pw1pw2_140 523 13.2 0.857 0.894 6.456 6.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.781 5.873 fft3d_ps 1291 14.7 2.195 2.852 5.353 5.686 mp_alltoall_d11v 2415 14.1 4.272 5.533 4.272 5.533 grid_collocate_task_list 128 9.7 4.842 5.249 4.842 5.249 mp_sum_d 4472 12.1 3.818 4.861 3.818 4.861 cp_fm_cholesky_decompose 22 10.9 4.582 4.594 4.582 4.594 potential_pw2rs 128 12.3 0.009 0.010 4.473 4.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.179000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.909091, yerr=6.653154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 841.474048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5000758. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.117 190.681 190.695 qs_mol_dyn_low 1 2.0 0.003 0.004 189.964 190.016 qs_forces 11 3.9 0.004 0.005 189.878 189.879 qs_energies 11 4.9 0.002 0.006 183.149 183.159 scf_env_do_scf 11 5.9 0.001 0.003 166.294 166.304 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 133.424 133.427 velocity_verlet 10 3.0 0.002 0.003 119.938 119.939 dbcsr_multiply_generic 2527 12.6 0.195 0.240 98.073 99.323 qs_scf_new_mos 118 7.6 0.001 0.001 94.742 95.324 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.741 95.323 ot_scf_mini 118 9.6 0.004 0.006 89.887 90.528 multiply_cannon 2527 13.6 0.508 0.570 77.593 82.686 multiply_cannon_loop 2527 14.6 1.573 1.651 74.307 77.253 ot_mini 118 10.6 0.001 0.001 50.793 51.367 mp_waitall_1 216438 16.6 24.444 38.732 24.444 38.732 multiply_cannon_multrec 30324 15.6 20.989 26.734 31.602 37.598 rebuild_ks_matrix 129 8.3 0.001 0.001 32.323 33.048 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.322 33.048 init_scf_loop 11 6.9 0.001 0.003 32.782 32.784 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.091 29.750 multiply_cannon_metrocomm3 30324 15.6 0.099 0.106 16.013 29.540 qs_ot_get_derivative 118 11.6 0.001 0.002 28.485 29.134 prepare_preconditioner 11 7.9 0.000 0.000 28.457 28.532 make_preconditioner 11 8.9 0.000 0.001 28.457 28.532 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.141 27.687 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.413 23.691 apply_single 129 13.6 0.001 0.001 22.412 23.691 ot_diis_step 118 11.6 0.015 0.015 22.119 22.121 qs_ot_get_p 129 10.4 0.001 0.002 20.973 21.729 multiply_cannon_sync_h2d 30324 15.6 17.971 20.923 17.971 20.923 cp_fm_cholesky_invert 11 10.9 16.614 16.626 16.614 16.626 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.079 16.122 make_m2s 5054 13.6 0.088 0.094 14.377 15.937 make_images 5054 14.6 1.184 1.387 14.157 15.713 cp_dbcsr_syevd 83 12.4 0.007 0.022 14.858 14.858 sum_up_and_integrate 129 10.3 0.002 0.004 13.973 13.999 integrate_v_rspace 129 11.3 0.003 0.003 13.912 13.942 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.541 13.582 calculate_rho_elec 129 8.7 0.088 0.104 13.540 13.581 cp_fm_diag_elpa 83 13.4 0.001 0.001 11.723 11.758 cp_fm_diag_elpa_base 83 14.4 11.459 11.555 11.703 11.731 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.233 11.693 init_scf_run 11 5.9 0.000 0.001 11.615 11.616 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.614 11.616 multiply_cannon_metrocomm4 27797 15.6 0.098 0.113 3.921 11.165 mp_irecv_dv 70031 16.3 3.722 10.769 3.722 10.769 dbcsr_mm_accdrv_process 62734 16.2 5.517 6.092 10.064 10.642 make_images_data 5054 15.6 0.067 0.077 8.448 10.375 hybrid_alltoall_any 5240 16.5 0.349 1.518 7.224 9.822 pw_transfer 1559 11.6 0.085 0.097 8.625 8.681 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.400 8.460 wfi_extrapolate 11 7.9 0.001 0.001 8.304 8.304 density_rs2pw 129 9.7 0.006 0.006 7.123 7.962 grid_integrate_task_list 129 12.3 7.188 7.568 7.188 7.568 fft_wrap_pw1pw2_140 527 13.2 0.943 0.967 7.450 7.521 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.420 7.253 cp_fm_cholesky_decompose 22 10.9 6.904 6.988 6.904 6.988 calculate_dm_sparse 129 9.5 0.001 0.001 6.508 6.667 mp_sum_l 8010 12.9 4.329 6.420 4.329 6.420 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.148 6.157 fft3d_ps 1301 14.7 2.850 3.042 5.937 5.997 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.479 5.572 grid_collocate_task_list 129 9.7 5.041 5.453 5.041 5.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.236 5.422 mp_alltoall_d11v 2423 14.1 4.273 5.291 4.273 5.291 mp_allgather_i34 2527 14.6 1.787 4.983 1.787 4.983 potential_pw2rs 129 12.3 0.016 0.018 4.452 4.466 dbcsr_complete_redistribute 395 12.7 0.770 0.864 3.177 4.011 mp_sum_d 4501 12.1 2.659 4.003 2.659 4.003 transfer_rs2pw 527 10.6 0.007 0.007 3.288 3.973 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.873 3.874 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.695000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.727273, yerr=4.024717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 961.236992E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.031 175.216 175.217 qs_mol_dyn_low 1 2.0 0.003 0.003 174.754 174.768 qs_forces 11 3.9 0.004 0.005 174.634 174.635 qs_energies 11 4.9 0.001 0.002 168.107 168.117 scf_env_do_scf 11 5.9 0.001 0.001 152.547 152.548 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.645 117.645 velocity_verlet 10 3.0 0.001 0.002 112.662 112.664 dbcsr_multiply_generic 2527 12.6 0.191 0.203 82.540 83.538 qs_scf_new_mos 118 7.6 0.001 0.001 80.316 80.622 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.315 80.621 ot_scf_mini 118 9.6 0.003 0.004 76.025 76.382 multiply_cannon 2527 13.6 0.513 0.536 62.896 67.200 multiply_cannon_loop 2527 14.6 1.131 1.198 59.676 61.993 ot_mini 118 10.6 0.001 0.001 42.804 43.149 init_scf_loop 11 6.9 0.000 0.000 34.803 34.804 mp_waitall_1 171878 16.6 24.864 33.956 24.864 33.956 prepare_preconditioner 11 7.9 0.000 0.000 30.830 30.881 make_preconditioner 11 8.9 0.000 0.000 30.830 30.881 rebuild_ks_matrix 129 8.3 0.001 0.001 30.425 30.869 qs_ks_build_kohn_sham_matrix 129 9.3 0.022 0.035 30.425 30.868 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.407 29.805 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.439 27.841 multiply_cannon_multrec 20216 15.6 13.141 16.377 22.521 25.802 multiply_cannon_metrocomm3 20216 15.6 0.061 0.065 15.632 24.613 qs_ot_get_derivative 118 11.6 0.001 0.002 22.845 23.198 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.041 21.110 apply_single 129 13.6 0.001 0.001 20.041 21.110 ot_diis_step 118 11.6 0.019 0.035 19.844 19.844 qs_ot_get_p 129 10.4 0.001 0.001 18.424 18.851 make_m2s 5054 13.6 0.079 0.085 14.685 16.158 make_images 5054 14.6 1.164 1.257 14.448 15.917 multiply_cannon_sync_h2d 20216 15.6 13.677 15.469 13.677 15.469 cp_fm_cholesky_invert 11 10.9 14.293 14.303 14.293 14.303 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.251 14.263 sum_up_and_integrate 129 10.3 0.002 0.003 14.001 14.029 integrate_v_rspace 129 11.3 0.003 0.003 13.941 13.971 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.590 13.626 calculate_rho_elec 129 8.7 0.130 0.144 13.590 13.625 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.188 13.188 make_images_data 5054 15.6 0.064 0.074 8.743 10.582 init_scf_run 11 5.9 0.000 0.001 10.570 10.570 scf_env_initial_rho_setup 11 6.9 0.002 0.002 10.570 10.570 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.026 10.042 cp_fm_diag_elpa_base 83 14.4 9.622 9.771 10.023 10.039 hybrid_alltoall_any 5240 16.5 0.452 2.052 7.629 9.840 multiply_cannon_metrocomm4 17689 15.6 0.064 0.074 3.543 9.631 mp_irecv_dv 50625 16.2 3.415 9.375 3.415 9.375 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.957 9.210 dbcsr_mm_accdrv_process 41830 16.2 5.724 6.028 8.835 8.996 pw_transfer 1559 11.6 0.085 0.101 8.810 8.909 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.586 8.691 fft_wrap_pw1pw2_140 527 13.2 1.064 1.107 7.634 7.753 grid_integrate_task_list 129 12.3 7.350 7.742 7.350 7.742 wfi_extrapolate 11 7.9 0.001 0.001 7.395 7.395 cp_fm_upper_to_full 105 14.8 5.774 7.297 5.774 7.297 cp_fm_cholesky_decompose 22 10.9 7.258 7.289 7.258 7.289 density_rs2pw 129 9.7 0.006 0.006 6.930 7.281 dbcsr_complete_redistribute 395 12.7 1.162 1.195 4.643 6.403 calculate_dm_sparse 129 9.5 0.001 0.001 5.883 5.986 fft3d_ps 1301 14.7 2.840 3.085 5.885 5.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.622 5.630 grid_collocate_task_list 129 9.7 5.207 5.593 5.207 5.593 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.629 5.315 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.489 5.253 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.714 4.836 mp_allgather_i34 2527 14.6 1.667 4.748 1.667 4.748 mp_alltoall_d11v 2423 14.1 4.325 4.668 4.325 4.668 mp_sum_l 8010 12.9 3.110 4.424 3.110 4.424 potential_pw2rs 129 12.3 0.020 0.022 4.391 4.403 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.404 4.155 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.998 4.024 mp_alltoall_i22 718 14.1 1.966 3.880 1.966 3.880 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 3.707 3.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.486 3.527 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.217000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=913.727273, yerr=10.349272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/18/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 4.353788E+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 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 1.167258E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 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 0.030 0.137 189.305 189.307 qs_mol_dyn_low 1 2.0 0.003 0.005 188.672 188.731 qs_forces 11 3.9 0.004 0.004 188.547 188.555 qs_energies 11 4.9 0.009 0.037 181.351 181.366 scf_env_do_scf 11 5.9 0.005 0.032 164.235 164.247 velocity_verlet 10 3.0 0.002 0.002 124.492 124.495 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.936 117.949 qs_scf_new_mos 117 7.6 0.001 0.001 81.561 81.861 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.560 81.860 dbcsr_multiply_generic 2507 12.6 0.194 0.200 80.547 81.341 ot_scf_mini 117 9.6 0.003 0.004 77.003 77.292 multiply_cannon 2507 13.6 0.554 0.595 55.894 58.982 multiply_cannon_loop 2507 14.6 1.866 1.919 52.090 53.949 init_scf_loop 11 6.9 0.001 0.004 46.147 46.159 ot_mini 117 10.6 0.001 0.001 43.108 43.380 prepare_preconditioner 11 7.9 0.000 0.000 42.023 42.055 make_preconditioner 11 8.9 0.000 0.003 42.023 42.055 make_full_inverse_cholesky 11 9.9 0.012 0.026 35.633 40.680 multiply_cannon_multrec 30084 15.6 13.626 18.787 26.520 31.253 rebuild_ks_matrix 128 8.3 0.001 0.001 29.424 29.703 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.037 29.423 29.703 mp_waitall_1 147882 16.7 16.999 26.823 16.999 26.823 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.483 26.734 qs_ot_get_derivative 117 11.6 0.001 0.002 23.380 23.665 make_m2s 5014 13.6 0.092 0.096 20.270 21.479 make_images 5014 14.6 1.947 2.281 19.967 21.178 qs_ot_get_p 128 10.4 0.006 0.041 20.016 20.349 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.177 19.818 apply_single 128 13.6 0.001 0.001 19.177 19.818 ot_diis_step 117 11.6 0.018 0.019 19.606 19.607 cp_fm_upper_to_full 105 14.8 11.519 16.963 11.519 16.963 cp_fm_cholesky_invert 11 10.9 15.975 15.985 15.975 15.985 qs_ot_p2m_diag 83 11.4 0.347 0.389 15.661 15.716 multiply_cannon_metrocomm3 30084 15.6 0.048 0.052 6.266 14.518 cp_dbcsr_syevd 83 12.4 0.006 0.012 14.243 14.255 sum_up_and_integrate 128 10.3 0.002 0.003 14.084 14.118 integrate_v_rspace 128 11.3 0.003 0.004 14.024 14.059 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.839 13.898 calculate_rho_elec 128 8.7 0.170 0.186 13.838 13.897 dbcsr_mm_accdrv_process 62264 16.2 8.583 9.146 12.459 12.962 dbcsr_complete_redistribute 395 12.7 1.489 1.611 9.123 12.862 make_images_data 5014 15.6 0.065 0.072 10.838 12.662 hybrid_alltoall_any 5200 16.5 0.532 2.218 9.732 11.943 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.755 11.513 multiply_cannon_sync_h2d 30084 15.6 10.509 11.268 10.509 11.268 init_scf_run 11 5.9 0.000 0.001 10.979 10.981 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.979 10.981 cp_fm_diag_elpa 83 13.4 0.001 0.004 10.929 10.960 cp_fm_diag_elpa_base 83 14.4 9.953 10.269 10.914 10.948 transfer_fm_to_dbcsr 11 9.9 0.002 0.009 6.367 10.061 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.776 9.993 pw_transfer 1547 11.6 0.086 0.102 9.306 9.374 mp_alltoall_i22 716 14.1 5.612 9.341 5.612 9.341 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.080 9.153 fft_wrap_pw1pw2_140 523 13.2 1.247 1.292 8.056 8.140 grid_integrate_task_list 128 12.3 7.580 7.901 7.580 7.901 wfi_extrapolate 11 7.9 0.001 0.001 7.698 7.698 cp_fm_cholesky_decompose 22 10.9 7.416 7.502 7.416 7.502 density_rs2pw 128 9.7 0.005 0.006 6.912 7.297 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.789 7.135 mp_irecv_dv 76098 16.2 2.639 6.860 2.639 6.860 calculate_dm_sparse 128 9.5 0.001 0.001 6.267 6.344 fft3d_ps 1291 14.7 2.999 3.045 6.128 6.201 grid_collocate_task_list 128 9.7 5.290 5.715 5.290 5.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.479 5.533 mp_alltoall_d11v 2415 14.1 4.890 5.317 4.890 5.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.476 4.558 qs_energies_init_hamiltonians 11 5.9 0.039 0.047 4.546 4.547 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.380 4.449 potential_pw2rs 128 12.3 0.023 0.023 4.364 4.381 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.270 4.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.307000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1098.818182, yerr=20.148621 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 1.537647E+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 3989 58388. MP_Allreduce 11050 1168. 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 0.030 0.073 168.940 168.956 qs_mol_dyn_low 1 2.0 0.003 0.003 168.467 168.493 qs_forces 11 3.9 0.006 0.006 168.366 168.371 qs_energies 11 4.9 0.002 0.002 160.948 160.955 scf_env_do_scf 11 5.9 0.001 0.001 143.634 143.647 velocity_verlet 10 3.0 0.002 0.002 110.611 110.615 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.834 107.835 dbcsr_multiply_generic 2485 12.5 0.182 0.189 71.707 72.285 qs_scf_new_mos 116 7.6 0.001 0.001 70.985 71.137 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.985 71.137 ot_scf_mini 116 9.6 0.003 0.003 66.620 66.816 multiply_cannon 2485 13.5 0.556 0.575 53.022 56.178 multiply_cannon_loop 2485 14.5 0.798 0.833 50.077 51.334 ot_mini 116 10.6 0.001 0.001 37.447 37.631 init_scf_loop 11 6.9 0.000 0.000 35.653 35.654 mp_waitall_1 124680 16.7 25.058 31.795 25.058 31.795 prepare_preconditioner 11 7.9 0.000 0.000 31.683 31.707 make_preconditioner 11 8.9 0.000 0.000 31.683 31.707 make_full_inverse_cholesky 11 9.9 0.014 0.027 29.564 29.823 rebuild_ks_matrix 127 8.3 0.001 0.001 29.147 29.335 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.146 29.334 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.577 26.748 multiply_cannon_multrec 9940 15.5 10.168 14.705 17.687 22.445 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.975 20.784 qs_ot_get_derivative 116 11.6 0.001 0.002 20.389 20.572 cp_fm_cholesky_invert 11 10.9 18.205 18.211 18.205 18.211 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.089 17.345 apply_single 127 13.6 0.001 0.001 17.088 17.344 ot_diis_step 116 11.6 0.020 0.020 16.987 16.987 qs_ot_get_p 127 10.4 0.002 0.002 16.510 16.674 make_m2s 4970 13.5 0.063 0.067 14.848 15.680 make_images 4970 14.5 2.145 2.585 14.543 15.380 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.171 14.221 calculate_rho_elec 127 8.7 0.251 0.262 14.171 14.220 sum_up_and_integrate 127 10.3 0.002 0.002 14.076 14.125 integrate_v_rspace 127 11.3 0.004 0.004 14.016 14.067 qs_ot_p2m_diag 82 11.4 0.489 0.495 12.785 12.803 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.635 11.635 multiply_cannon_sync_h2d 9940 15.5 10.575 11.156 10.575 11.156 init_scf_run 11 5.9 0.000 0.001 10.482 10.482 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.482 10.482 pw_transfer 1535 11.6 0.084 0.094 9.792 9.830 make_images_data 4970 15.5 0.053 0.063 8.613 9.819 hybrid_alltoall_any 5155 16.4 0.840 3.795 8.435 9.786 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.571 9.616 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.825 8.838 cp_fm_diag_elpa_base 82 14.4 8.583 8.658 8.819 8.832 fft_wrap_pw1pw2_140 519 13.2 1.661 1.700 8.488 8.532 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.959 8.092 grid_integrate_task_list 127 12.3 7.703 7.975 7.703 7.975 cp_fm_cholesky_decompose 22 10.9 7.804 7.921 7.804 7.921 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.247 7.876 dbcsr_mm_accdrv_process 20590 16.1 3.394 4.480 7.173 7.870 wfi_extrapolate 11 7.9 0.001 0.001 7.331 7.331 density_rs2pw 127 9.7 0.005 0.005 6.917 7.302 calculate_dm_sparse 127 9.5 0.001 0.001 6.019 6.115 fft3d_ps 1281 14.7 3.113 3.211 6.052 6.073 grid_collocate_task_list 127 9.7 5.526 5.830 5.526 5.830 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.849 5.702 dbcsr_complete_redistribute 393 12.7 2.147 2.256 5.302 5.661 mp_irecv_dv 28618 15.9 1.809 5.610 1.809 5.610 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.226 5.235 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.234 5.234 mp_alltoall_d11v 2401 14.1 4.632 5.177 4.632 5.177 mp_allgather_i34 2485 14.5 1.148 4.601 1.148 4.601 potential_pw2rs 127 12.3 0.026 0.026 4.322 4.327 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.048 4.114 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.586 3.900 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.459 3.765 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.615 3.683 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.592 3.629 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.433 3.573 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.463 3.480 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.956000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1442.818182, yerr=22.163927 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 11.696233E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.229889E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59279. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.055 293.257 293.258 qs_mol_dyn_low 1 2.0 0.003 0.003 292.672 292.702 qs_forces 11 3.9 0.004 0.005 292.060 292.063 qs_energies 11 4.9 0.002 0.002 283.123 283.127 scf_env_do_scf 11 5.9 0.001 0.001 260.442 260.449 velocity_verlet 10 3.0 0.002 0.002 212.284 212.357 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 135.605 135.607 init_scf_loop 11 6.9 0.000 0.000 124.558 124.561 prepare_preconditioner 11 7.9 0.000 0.000 119.601 119.634 make_preconditioner 11 8.9 0.000 0.000 119.601 119.634 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.338 116.747 qs_scf_new_mos 117 7.6 0.001 0.001 90.543 90.664 qs_scf_loop_do_ot 117 8.6 0.001 0.001 90.542 90.663 ot_scf_mini 117 9.6 0.004 0.004 85.652 85.741 dbcsr_multiply_generic 2507 12.6 0.217 0.228 83.379 84.087 cp_fm_upper_to_full 105 14.8 52.771 75.983 52.771 75.983 multiply_cannon 2507 13.6 0.680 0.738 59.925 61.442 multiply_cannon_loop 2507 14.6 1.043 1.066 55.665 57.476 ot_mini 117 10.6 0.001 0.001 45.255 45.368 dbcsr_complete_redistribute 395 12.7 3.981 4.082 30.279 43.465 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.771 39.969 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.221 37.298 rebuild_ks_matrix 128 8.3 0.001 0.001 35.079 35.187 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 35.078 35.186 mp_alltoall_i22 716 14.1 21.993 35.074 21.993 35.074 mp_waitall_1 103674 16.8 29.073 34.040 29.073 34.040 cp_fm_cholesky_invert 11 10.9 33.479 33.486 33.479 33.486 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.554 32.648 qs_ot_get_p 128 10.4 0.004 0.006 25.143 25.248 qs_ot_get_derivative 117 11.6 0.002 0.002 24.943 25.031 multiply_cannon_metrocomm3 10028 15.6 0.025 0.025 19.638 21.353 qs_ot_p2m_diag 83 11.4 0.880 0.886 21.081 21.109 make_m2s 5014 13.6 0.074 0.076 18.755 20.427 ot_diis_step 117 11.6 0.022 0.023 20.271 20.271 make_images 5014 14.6 3.066 3.266 18.273 19.944 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.517 19.746 apply_single 128 13.6 0.001 0.001 19.517 19.746 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.294 19.295 multiply_cannon_multrec 10028 15.6 10.270 12.118 18.090 18.220 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.316 17.322 calculate_rho_elec 128 8.7 0.478 0.478 17.315 17.321 sum_up_and_integrate 128 10.3 0.002 0.002 16.522 16.619 integrate_v_rspace 128 11.3 0.004 0.005 16.459 16.556 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.153 16.154 cp_fm_diag_elpa_base 83 14.4 11.777 13.363 16.150 16.150 multiply_cannon_sync_h2d 10028 15.6 14.316 14.365 14.316 14.365 hybrid_alltoall_any 5200 16.5 1.312 3.041 10.979 13.461 pw_transfer 1547 11.6 0.095 0.096 13.221 13.226 make_images_data 5014 15.6 0.063 0.068 10.700 13.079 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 12.983 12.990 init_scf_run 11 5.9 0.000 0.001 12.332 12.332 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.331 12.332 fft_wrap_pw1pw2_140 523 13.2 3.222 3.276 11.574 11.580 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.782 9.847 dbcsr_mm_accdrv_process 20762 16.1 4.148 5.830 7.568 9.398 wfi_extrapolate 11 7.9 0.001 0.001 9.158 9.158 cp_fm_cholesky_decompose 22 10.9 8.890 8.908 8.890 8.908 grid_integrate_task_list 128 12.3 8.553 8.755 8.553 8.755 density_rs2pw 128 9.7 0.005 0.005 8.346 8.391 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.157 8.159 fft3d_ps 1291 14.7 3.969 3.982 7.440 7.477 mp_alltoall_d11v 2415 14.1 6.987 7.133 6.987 7.133 calculate_dm_sparse 128 9.5 0.001 0.001 6.789 6.839 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.453 6.510 grid_collocate_task_list 128 9.7 6.415 6.492 6.415 6.492 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.175 6.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.258000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2916.090909, yerr=148.818654 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 1.260990E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 85.189 85.190 qs_energies 1 2.0 0.000 0.000 84.673 84.684 ls_scf 1 3.0 0.000 0.000 83.772 83.784 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.530 72.722 multiply_cannon 111 7.7 0.017 0.020 55.882 57.335 multiply_cannon_loop 111 8.7 0.228 0.245 52.485 54.252 ls_scf_main 1 4.0 0.000 0.000 52.364 52.375 density_matrix_trs4 2 5.0 0.002 0.003 46.851 46.923 ls_scf_init_scf 1 4.0 0.000 0.000 28.359 28.360 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.233 27.286 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.131 25.151 mp_waitall_1 11031 10.9 22.241 25.093 22.241 25.093 multiply_cannon_multrec 2664 9.7 8.182 8.977 15.555 17.435 multiply_cannon_sync_h2d 2664 9.7 13.749 15.295 13.749 15.295 make_m2s 222 7.7 0.009 0.011 13.077 13.521 make_images 222 8.7 0.098 0.108 13.055 13.502 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.579 12.090 make_images_data 222 9.7 0.004 0.005 7.604 8.166 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.725 6.986 7.996 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.392 7.983 hybrid_alltoall_any 227 10.6 0.215 1.854 6.532 7.979 dbcsr_mm_accdrv_process_sort 4760 11.4 6.195 7.103 6.195 7.103 calculate_norms 4752 9.8 5.512 6.242 5.512 6.242 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.011 5.161 mp_sum_l 887 5.1 3.031 4.795 3.031 4.795 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.043 3.711 mp_irecv_dv 6231 10.9 2.027 3.703 2.027 3.703 make_images_sizes 222 9.7 0.000 0.000 0.750 3.687 mp_alltoall_i44 222 10.7 0.750 3.687 0.750 3.687 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.277 3.569 arnoldi_extremal 4 6.8 0.000 0.000 3.192 3.225 arnoldi_normal_ev 4 7.8 0.001 0.003 3.192 3.225 build_subspace 16 8.4 0.009 0.012 3.095 3.097 ls_scf_post 1 4.0 0.000 0.000 3.049 3.060 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.907 dbcsr_special_finalize 555 9.7 0.005 0.006 2.369 2.852 dbcsr_merge_single_wm 555 10.7 0.458 0.579 2.361 2.843 make_images_pack 222 9.7 2.206 2.614 2.208 2.616 dbcsr_sort_data 658 11.4 2.162 2.583 2.162 2.583 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.303 2.543 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.456 2.067 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.298 2.417 buffer_matrices_ensure_size 222 8.7 1.747 2.040 1.747 2.040 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.760 1.761 rebuild_ks_matrix 3 7.3 0.000 0.000 1.751 1.752 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.751 1.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.190000, 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/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 2.136093E+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 0.040 0.082 90.942 90.943 qs_energies 1 2.0 0.000 0.000 90.324 90.331 ls_scf 1 3.0 0.001 0.011 88.239 88.247 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.123 74.416 multiply_cannon 111 7.7 0.027 0.041 52.334 57.599 ls_scf_main 1 4.0 0.000 0.001 54.611 54.617 multiply_cannon_loop 111 8.7 0.136 0.146 49.530 53.364 density_matrix_trs4 2 5.0 0.002 0.003 48.885 49.117 ls_scf_init_scf 1 4.0 0.002 0.021 30.000 30.002 mp_waitall_1 9105 10.9 20.163 29.134 20.163 29.134 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.779 28.870 multiply_cannon_multrec 1332 9.7 13.478 17.214 22.750 27.603 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.233 26.245 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 10.939 20.529 make_m2s 222 7.7 0.006 0.008 15.015 15.555 make_images 222 8.7 1.372 1.699 14.985 15.525 dbcsr_mm_accdrv_process 4041 10.4 0.360 0.535 8.869 10.399 dbcsr_mm_accdrv_process_sort 4041 11.4 8.386 9.864 8.386 9.864 hybrid_alltoall_any 227 10.6 0.543 2.597 8.091 9.595 make_images_data 222 9.7 0.004 0.005 8.687 9.570 mp_sum_l 887 5.1 5.320 8.418 5.320 8.418 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.177 7.797 mp_irecv_dv 3311 11.0 3.157 7.740 3.157 7.740 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.061 6.859 calculate_norms 2376 9.8 6.075 6.624 6.075 6.624 multiply_cannon_sync_h2d 1332 9.7 4.921 6.397 4.921 6.397 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.895 5.110 arnoldi_extremal 4 6.8 0.000 0.000 4.662 4.677 arnoldi_normal_ev 4 7.8 0.001 0.005 4.662 4.677 build_subspace 16 8.4 0.014 0.021 4.412 4.415 ls_scf_post 1 4.0 0.000 0.001 3.626 3.635 ls_scf_store_result 1 5.0 0.000 0.000 3.348 3.436 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.195 3.414 dbcsr_matrix_vector_mult_local 304 10.0 2.784 3.261 2.786 3.263 mp_allgather_i34 111 8.7 0.794 3.059 0.794 3.059 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.665 2.743 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.176 2.660 dbcsr_data_new 4174 10.1 2.115 2.409 2.115 2.409 make_images_pack 222 9.7 1.820 2.127 1.823 2.129 qs_energies_init_hamiltonians 1 3.0 0.023 0.070 2.080 2.080 dbcsr_sort_data 436 11.2 1.832 2.080 1.832 2.080 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.909 1.911 rebuild_ks_matrix 3 7.3 0.000 0.000 1.897 1.899 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.897 1.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.943000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 2.860540E+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 0.031 0.060 92.848 92.849 qs_energies 1 2.0 0.001 0.011 92.171 92.194 ls_scf 1 3.0 0.000 0.002 90.709 90.725 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.250 75.576 ls_scf_main 1 4.0 0.001 0.017 56.629 56.633 multiply_cannon 111 7.7 0.029 0.061 52.153 56.072 multiply_cannon_loop 111 8.7 0.116 0.129 49.387 53.650 density_matrix_trs4 2 5.0 0.004 0.040 50.766 50.971 mp_waitall_1 7281 11.0 23.521 33.692 23.521 33.692 ls_scf_init_scf 1 4.0 0.000 0.002 30.442 30.446 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.176 29.269 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.803 26.813 multiply_cannon_multrec 888 9.7 12.699 15.224 21.236 24.338 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.035 23.383 make_m2s 222 7.7 0.006 0.007 16.259 16.996 make_images 222 8.7 1.581 1.850 16.221 16.955 make_images_data 222 9.7 0.004 0.005 9.380 10.511 hybrid_alltoall_any 227 10.6 0.641 2.947 8.908 10.146 dbcsr_mm_accdrv_process 3754 10.4 0.309 0.491 8.074 9.324 dbcsr_mm_accdrv_process_sort 3754 11.4 7.641 8.834 7.641 8.834 mp_sum_l 887 5.1 5.161 8.716 5.161 8.716 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.439 7.244 multiply_cannon_sync_h2d 888 9.7 6.066 7.203 6.066 7.203 mp_irecv_dv 2335 11.1 2.423 7.186 2.423 7.186 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.752 7.154 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.044 7.108 arnoldi_extremal 4 6.8 0.000 0.000 5.074 5.089 arnoldi_normal_ev 4 7.8 0.001 0.005 5.074 5.089 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.691 4.970 calculate_norms 1584 9.8 4.373 4.820 4.373 4.820 build_subspace 16 8.4 0.014 0.020 4.772 4.778 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.473 3.787 mp_allgather_i34 111 8.7 0.873 3.736 0.873 3.736 ls_scf_post 1 4.0 0.000 0.002 3.638 3.650 dbcsr_matrix_vector_mult_local 304 10.0 3.076 3.629 3.078 3.631 ls_scf_store_result 1 5.0 0.000 0.000 3.396 3.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.844 2.951 dbcsr_data_new 4116 9.9 2.112 2.467 2.112 2.467 make_images_sizes 222 9.7 0.000 0.000 1.069 2.247 mp_alltoall_i44 222 10.7 1.068 2.246 1.068 2.246 dbcsr_sort_data 325 11.1 1.862 2.118 1.862 2.118 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.942 1.944 rebuild_ks_matrix 3 7.3 0.000 0.000 1.923 1.925 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.043 1.923 1.925 make_images_pack 222 9.7 1.621 1.886 1.624 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.849000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2220.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 3.340788E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.044 97.351 97.351 qs_energies 1 2.0 0.000 0.000 96.780 96.785 ls_scf 1 3.0 0.000 0.000 94.462 94.469 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.984 78.273 ls_scf_main 1 4.0 0.000 0.000 58.657 58.657 multiply_cannon 111 7.7 0.045 0.078 51.510 56.010 density_matrix_trs4 2 5.0 0.002 0.003 52.517 52.651 multiply_cannon_loop 111 8.7 0.154 0.167 46.445 48.934 ls_scf_init_scf 1 4.0 0.000 0.000 32.558 32.560 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.353 31.416 matrix_sqrt_Newton_Schulz 2 6.5 0.018 0.020 28.846 28.857 mp_waitall_1 6369 11.0 22.535 28.838 22.535 28.838 multiply_cannon_multrec 1332 9.7 14.152 17.747 22.035 24.551 make_m2s 222 7.7 0.007 0.008 21.052 22.511 make_images 222 8.7 3.142 3.611 21.001 22.463 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.108 16.747 make_images_data 222 9.7 0.004 0.004 11.752 13.407 hybrid_alltoall_any 227 10.6 0.796 3.792 11.113 12.881 dbcsr_mm_accdrv_process 3641 10.4 0.286 0.474 7.528 9.071 dbcsr_mm_accdrv_process_sort 3641 11.4 7.088 8.575 7.088 8.575 mp_sum_l 887 5.1 3.843 7.169 3.843 7.169 multiply_cannon_sync_h2d 1332 9.7 5.427 6.117 5.427 6.117 multiply_cannon_metrocomm4 1110 9.7 0.005 0.006 2.086 6.049 mp_irecv_dv 3229 10.9 2.060 5.968 2.060 5.968 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.960 5.664 arnoldi_extremal 4 6.8 0.000 0.000 5.244 5.254 arnoldi_normal_ev 4 7.8 0.003 0.006 5.243 5.254 build_subspace 16 8.4 0.014 0.021 4.914 4.919 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.505 4.866 mp_allgather_i34 111 8.7 2.191 4.703 2.191 4.703 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.536 4.698 calculate_norms 2376 9.8 4.183 4.477 4.183 4.477 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.632 3.934 dbcsr_matrix_vector_mult_local 304 10.0 3.242 3.734 3.244 3.736 dbcsr_sort_data 658 11.4 3.092 3.430 3.092 3.430 ls_scf_post 1 4.0 0.000 0.000 3.247 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.849 3.202 dbcsr_merge_single_wm 555 10.7 0.544 0.665 2.840 3.194 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.043 3.107 ls_scf_store_result 1 5.0 0.000 0.000 2.976 3.052 dbcsr_data_release 10477 10.7 1.578 2.413 1.578 2.413 qs_energies_init_hamiltonians 1 3.0 0.018 0.022 2.308 2.308 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.351000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 4.762554E+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 0.037 0.054 91.609 91.610 qs_energies 1 2.0 0.000 0.000 90.881 90.886 ls_scf 1 3.0 0.000 0.000 88.918 88.922 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.480 70.639 ls_scf_main 1 4.0 0.000 0.000 56.205 56.206 multiply_cannon 111 7.7 0.060 0.108 52.142 55.508 multiply_cannon_loop 111 8.7 0.088 0.097 49.591 51.140 density_matrix_trs4 2 5.0 0.002 0.003 49.193 49.279 ls_scf_init_scf 1 4.0 0.000 0.000 29.335 29.335 mp_waitall_1 5436 11.0 24.204 28.617 24.204 28.617 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.080 28.115 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 25.892 25.904 multiply_cannon_multrec 444 9.7 13.600 16.232 20.678 22.947 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.483 15.131 make_m2s 222 7.7 0.005 0.005 13.641 14.572 make_images 222 8.7 2.041 2.474 13.573 14.502 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.202 14.035 hybrid_alltoall_any 227 10.6 0.803 3.860 8.127 9.828 make_images_data 222 9.7 0.003 0.004 8.336 9.576 multiply_cannon_sync_h2d 444 9.7 6.755 8.267 6.755 8.267 dbcsr_mm_accdrv_process 3003 10.4 0.336 0.429 6.762 7.959 dbcsr_mm_accdrv_process_sort 3003 11.4 6.392 7.530 6.392 7.530 arnoldi_extremal 4 6.8 0.000 0.000 5.795 5.809 arnoldi_normal_ev 4 7.8 0.002 0.005 5.795 5.809 build_subspace 16 8.4 0.015 0.020 5.402 5.412 mp_sum_l 887 5.1 2.692 4.744 2.692 4.744 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.468 4.659 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.177 4.346 dbcsr_matrix_vector_mult_local 304 10.0 3.727 4.147 3.730 4.150 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.521 3.910 mp_irecv_dv 1241 11.2 1.508 3.887 1.508 3.887 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.922 3.816 calculate_norms 792 9.8 3.607 3.732 3.607 3.732 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.572 3.647 mp_allgather_i34 111 8.7 1.127 3.624 1.127 3.624 ls_scf_post 1 4.0 0.000 0.000 3.378 3.383 make_images_sizes 222 9.7 0.000 0.000 0.893 3.369 mp_alltoall_i44 222 10.7 0.893 3.369 0.893 3.369 ls_scf_store_result 1 5.0 0.000 0.000 3.177 3.206 dbcsr_data_new 4608 9.7 1.811 2.325 1.811 2.325 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.281 dbcsr_merge_all 275 8.9 0.479 0.525 2.055 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.031 2.032 rebuild_ks_matrix 3 7.3 0.000 0.000 1.999 2.000 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.999 2.000 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 1.947 1.947 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.610000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3803.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 8.872571E+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 0.089 106.353 106.354 qs_energies 1 2.0 0.000 0.000 104.251 104.255 ls_scf 1 3.0 0.000 0.000 101.213 101.217 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.680 74.808 ls_scf_main 1 4.0 0.000 0.000 63.712 63.712 density_matrix_trs4 2 5.0 0.002 0.003 54.677 54.734 multiply_cannon 111 7.7 0.139 0.359 48.488 50.907 multiply_cannon_loop 111 8.7 0.099 0.101 45.380 45.909 ls_scf_init_scf 1 4.0 0.000 0.000 33.741 33.742 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.188 32.215 matrix_sqrt_Newton_Schulz 2 6.5 0.013 0.037 29.425 29.432 mp_waitall_1 4527 11.1 21.797 25.607 21.797 25.607 make_m2s 222 7.7 0.005 0.005 22.578 23.779 make_images 222 8.7 3.573 3.892 22.471 23.670 multiply_cannon_multrec 444 9.7 17.821 18.472 22.454 23.112 hybrid_alltoall_any 227 10.6 1.654 3.629 12.804 15.632 make_images_data 222 9.7 0.003 0.004 13.000 15.244 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.206 10.506 multiply_cannon_sync_h2d 444 9.7 8.787 8.849 8.787 8.849 arnoldi_extremal 4 6.8 0.000 0.000 7.410 7.442 arnoldi_normal_ev 4 7.8 0.006 0.014 7.410 7.441 build_subspace 16 8.4 0.026 0.036 6.836 6.843 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.489 5.631 dbcsr_matrix_vector_mult_local 304 10.0 5.083 5.370 5.085 5.373 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.180 5.268 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.886 5.142 dbcsr_mm_accdrv_process 1814 10.4 0.296 0.369 4.454 4.566 dbcsr_mm_accdrv_process_sort 1814 11.4 4.112 4.244 4.112 4.244 mp_allgather_i34 111 8.7 1.223 3.927 1.223 3.927 ls_scf_post 1 4.0 0.000 0.000 3.760 3.766 make_images_sizes 222 9.7 0.000 0.000 1.456 3.538 mp_alltoall_i44 222 10.7 1.455 3.538 1.455 3.538 ls_scf_store_result 1 5.0 0.000 0.000 3.487 3.520 calculate_norms 792 9.8 3.232 3.275 3.232 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.183 qs_energies_init_hamiltonians 1 3.0 0.016 0.041 3.007 3.007 dbcsr_merge_all 275 8.9 0.884 0.916 2.865 2.962 dbcsr_complete_redistribute 5 7.6 1.437 1.481 2.748 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.521 dbcsr_sort_data 325 11.1 2.438 2.510 2.438 2.510 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.458 2.460 rebuild_ks_matrix 3 7.3 0.000 0.000 2.392 2.394 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.392 2.394 dbcsr_new_transposed 4 7.5 0.241 0.251 2.278 2.294 dbcsr_data_new 6591 9.6 1.835 2.269 1.835 2.269 dbcsr_frobenius_norm 74 6.6 2.054 2.132 2.199 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.200 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.199 dbcsr_data_release 12724 10.6 1.975 2.166 1.975 2.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.354000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7067.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb5c80b9cfa321b86a0f21d0deddafcd7fbed371_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 593.985536E+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 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 0.029 0.107 230.143 230.145 qs_mol_dyn_low 1 2.0 0.004 0.036 228.940 229.002 qs_forces 5 3.8 0.011 0.042 228.798 228.802 qs_energies 5 4.8 0.008 0.056 225.678 225.729 scf_env_do_scf 5 5.8 0.001 0.010 210.185 210.190 scf_env_do_scf_inner_loop 105 6.6 0.003 0.021 182.948 182.963 qs_scf_new_mos 105 7.6 0.000 0.001 142.545 142.775 qs_scf_loop_do_ot 105 8.6 0.001 0.001 142.544 142.774 dbcsr_multiply_generic 1445 12.2 0.129 0.137 134.490 134.968 ot_scf_mini 105 9.6 0.003 0.018 132.704 132.908 multiply_cannon 1445 13.2 0.273 0.308 115.069 117.002 multiply_cannon_loop 1445 14.2 2.822 2.980 113.256 114.265 velocity_verlet 4 3.0 0.002 0.006 107.469 107.470 ot_mini 105 10.6 0.001 0.036 60.729 60.866 multiply_cannon_multrec 69360 15.2 29.661 34.565 39.588 44.757 qs_ot_get_p 112 10.4 0.001 0.002 41.867 42.147 mp_waitall_1 488190 16.1 35.221 41.085 35.221 41.085 qs_ot_get_derivative 55 11.6 0.001 0.002 38.869 39.014 multiply_cannon_sync_h2d 69360 15.2 29.145 33.152 29.145 33.152 multiply_cannon_metrocomm3 69360 15.2 0.198 0.210 25.947 32.731 qs_ot_p2m_diag 40 11.0 0.022 0.061 30.798 30.898 rebuild_ks_matrix 110 8.4 0.000 0.000 29.375 29.551 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.042 29.374 29.551 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.481 27.483 init_scf_loop 7 6.6 0.003 0.037 27.201 27.206 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.953 27.109 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.115 23.411 apply_single 62 13.6 0.000 0.000 23.115 23.411 cp_fm_syevd 40 13.0 0.000 0.001 22.379 22.529 prepare_preconditioner 7 7.6 0.000 0.000 22.049 22.087 make_preconditioner 7 8.6 0.005 0.034 22.049 22.087 ot_new_cg_direction 55 11.6 0.001 0.025 21.167 21.167 cp_fm_redistribute_end 40 14.0 8.860 17.677 8.865 17.678 cp_fm_syevd_base 40 14.0 8.804 17.621 8.804 17.621 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.177 17.554 calculate_rho_elec 110 8.6 0.030 0.032 17.176 17.554 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.848 15.041 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.631 14.692 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.203 14.303 mp_sum_l 4764 12.2 12.401 13.174 12.401 13.174 init_scf_run 5 5.8 0.000 0.001 12.824 12.825 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.823 12.825 pw_transfer 1645 12.4 0.079 0.101 12.118 12.432 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.979 12.300 density_rs2pw 110 9.6 0.005 0.006 11.271 11.746 calculate_dm_sparse 110 9.5 0.000 0.000 11.449 11.634 fft_wrap_pw1pw2_240 915 15.0 0.830 0.920 10.536 10.830 qs_vxc_create 110 10.4 0.003 0.026 10.680 10.726 dbcsr_mm_accdrv_process 154766 15.8 6.221 6.425 9.794 10.618 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.294 10.366 cp_fm_cholesky_invert 7 10.6 10.218 10.226 10.218 10.226 check_diag 80 13.5 8.594 8.885 9.436 9.586 fft3d_pb 915 16.0 2.398 2.709 8.727 9.093 sum_up_and_integrate 60 10.3 0.001 0.009 8.617 8.641 integrate_v_rspace 60 11.3 0.002 0.016 8.600 8.623 acc_transpose_blocks 69360 15.2 0.354 0.370 8.026 8.412 transfer_rs2pw 445 10.6 0.007 0.008 7.891 8.352 calculate_first_density_matrix 1 7.0 0.006 0.048 7.843 7.874 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.837 7.854 xc_rho_set_and_dset_create 110 12.4 0.083 0.153 7.473 7.745 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.397 7.458 xc_vxc_pw_create 60 11.3 0.040 0.070 7.198 7.245 make_m2s 2890 13.2 0.078 0.086 6.536 7.217 make_full_single_inverse 7 9.6 0.006 0.036 7.144 7.182 make_images 2890 14.2 0.237 0.258 6.428 7.109 multiply_cannon_metrocomm1 69360 15.2 0.097 0.104 4.628 6.716 xc_pw_derive 510 13.4 0.005 0.007 6.394 6.494 mp_alltoall_z22v 2340 17.7 5.754 6.070 5.754 6.070 acc_transpose_blocks_kernels 69360 16.2 0.850 0.899 5.289 5.571 mp_waitany 7680 13.5 4.530 5.179 4.530 5.179 potential_pw2rs 60 12.3 0.002 0.003 4.811 4.857 jit_kernel_transpose 5 15.0 4.439 4.681 4.439 4.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.145000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: eb5c80b9cfa321b86a0f21d0deddafcd7fbed371 Summary: empty Status: OK