=== 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: 4dc27d601c3e5b0826783d31464cb7b82f8cb647 ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/01 job id: 46923421 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/02 job id: 46923424 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/03 job id: 46923425 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/04 job id: 46923428 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/05 job id: 46923431 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/06 job id: 46923434 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/07 job id: 46923435 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/08 job id: 46923436 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/09 job id: 46923437 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/10 job id: 46923438 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/11 job id: 46923440 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/12 job id: 46923442 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/13 job id: 46923445 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/14 job id: 46923448 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/15 job id: 46923452 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/16 job id: 46923454 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/17 job id: 46923456 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/18 job id: 46923459 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/19 job id: 46923462 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/20 job id: 46923464 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/21 job id: 46923465 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/22 job id: 46923466 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/23 job id: 46923467 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/24 job id: 46923468 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/25 job id: 46923469 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/26 job id: 46923470 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/27 job id: 46923471 --- 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/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 0.025 0.039 137.407 137.408 farming_run 1 2.0 136.872 136.874 137.355 137.358 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464877E+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 1103589. 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 0.047 0.065 116.291 116.291 qs_energies 1 2.0 0.000 0.000 116.032 116.033 mp2_main 1 3.0 0.000 0.000 113.661 113.662 mp2_gpw_main 1 4.0 0.019 0.025 112.578 112.579 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.052 93.477 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.557 55.983 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.875 47.240 get_2c_integrals 1 6.0 0.008 0.009 36.800 37.324 integrate_v_rspace 273 8.0 0.436 0.449 25.239 30.344 pw_transfer 6555 10.6 0.378 0.385 27.453 27.971 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.144 26.664 grid_integrate_task_list 273 9.0 21.018 26.620 21.018 26.620 fft_wrap_pw1pw2_100 2178 12.4 1.159 1.266 23.646 24.167 rpa_ri_compute_en 1 5.0 0.019 0.020 19.415 19.760 compute_2c_integrals 1 7.0 0.002 0.002 19.294 19.295 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.800 18.885 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.432 18.797 18.884 cp_fm_cholesky_decompose 12 8.2 17.434 17.951 17.434 17.951 cholesky_decomp 1 7.0 0.000 0.000 16.343 16.862 fft3d_s 5443 13.4 16.200 16.664 16.221 16.685 ao_to_mo_and_store_B_mult_1 272 7.0 10.852 15.562 10.852 15.562 calculate_wavefunction 272 8.0 5.408 5.565 12.569 13.151 rpa_num_int 1 6.0 0.000 0.000 10.596 10.606 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.582 10.585 calc_mat_Q 8 8.0 0.000 0.000 9.461 9.549 contract_S_to_Q 8 9.0 0.000 0.000 8.881 8.966 parallel_gemm_fm 14 9.1 0.000 0.000 8.467 8.566 parallel_gemm_fm_cosma 14 10.1 8.467 8.566 8.467 8.566 calc_potential_gpw 544 9.5 0.005 0.006 8.252 8.544 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.168 8.387 potential_pw2rs 545 10.0 0.108 0.109 7.680 8.317 create_integ_mat 1 6.0 0.014 0.027 8.300 8.310 collocate_single_gaussian 272 10.0 0.040 0.042 7.424 7.633 array2fm 1 7.0 0.000 0.000 6.834 7.316 pw_scatter_s 2720 13.7 4.456 4.673 4.456 4.673 pw_gather_s 2722 13.2 3.903 4.329 3.903 4.329 array2fm_buffer_send 1 8.0 2.922 3.167 2.922 3.167 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.579108, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2812.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 0.029 0.037 396.992 396.993 farming_run 1 2.0 396.182 396.187 396.949 396.953 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.232040E+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 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 21616078. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 208.867 208.867 qs_energies 1 2.0 0.000 0.001 208.644 208.651 scf_env_do_scf 1 3.0 0.000 0.000 105.955 105.955 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.011 105.018 rebuild_ks_matrix 4 6.0 0.000 0.000 105.009 105.017 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.009 105.017 hfx_ks_matrix 4 8.0 0.001 0.001 104.606 104.610 integrate_four_center 4 9.0 0.143 0.454 104.605 104.609 mp2_main 1 3.0 0.000 0.004 102.394 102.401 mp2_gpw_main 1 4.0 0.032 0.045 101.542 101.550 integrate_four_center_main 4 10.0 0.099 0.538 96.552 98.833 integrate_four_center_bin 265 11.0 96.453 98.745 96.453 98.745 init_scf_loop 1 4.0 0.000 0.000 91.657 91.657 mp2_ri_gpw_compute_in 1 5.0 0.064 0.068 74.915 75.857 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.468 55.410 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 42.214 47.248 integrate_v_rspace 95 8.0 0.398 0.572 28.617 33.464 pw_transfer 2240 10.6 0.143 0.168 29.983 30.515 ao_to_mo_and_store_B_mult_1 91 7.0 10.573 29.566 10.573 29.566 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.984 29.532 grid_integrate_task_list 95 9.0 23.892 28.954 23.892 28.954 mp2_ri_gpw_compute_en 1 5.0 0.056 0.067 26.463 28.090 fft_wrap_pw1pw2_100 730 12.4 1.297 1.468 26.682 27.215 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.892 24.865 24.875 get_2c_integrals 1 6.0 0.001 0.006 20.349 20.382 compute_2c_integrals 1 7.0 0.003 0.004 19.329 19.332 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.938 19.190 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.869 18.937 19.189 fft3d_s 1823 13.4 18.462 18.908 18.475 18.921 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.295 14.295 calculate_wavefunction 91 8.0 2.008 2.045 9.728 9.947 potential_pw2rs 186 10.0 0.034 0.035 8.688 9.206 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.697 9.103 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.283 8.664 local_gemm 172 8.0 8.139 8.543 8.139 8.543 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.513 7.930 8.253 calc_potential_gpw 182 9.5 0.002 0.002 7.917 8.176 collocate_single_gaussian 91 10.0 0.017 0.023 7.929 8.171 mp2_ri_gpw_compute_en_ener 172 7.0 6.353 6.419 6.353 6.419 mp_sendrecv_dm3 2068 8.0 5.967 6.318 5.967 6.318 pw_gather_s 912 13.2 4.907 5.456 4.907 5.456 mp_sync 37 10.5 2.735 4.747 2.735 4.747 pw_scatter_s 910 13.7 3.917 4.475 3.917 4.475 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.531324, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 451.547136E+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 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.040 59.068 59.070 qs_mol_dyn_low 1 2.0 0.003 0.004 58.754 58.763 qs_forces 11 3.9 0.002 0.003 58.675 58.681 qs_energies 11 4.9 0.002 0.007 56.923 56.934 scf_env_do_scf 11 5.9 0.001 0.002 49.893 49.893 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.857 46.859 qs_scf_new_mos 108 7.5 0.000 0.001 33.023 33.285 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.022 33.285 dbcsr_multiply_generic 2286 12.5 0.094 0.099 31.915 32.318 ot_scf_mini 108 9.5 0.002 0.005 31.440 31.603 velocity_verlet 10 3.0 0.001 0.002 27.925 27.926 multiply_cannon 2286 13.5 0.184 0.196 25.069 26.472 multiply_cannon_loop 2286 14.5 1.433 1.504 23.711 25.157 ot_mini 108 10.5 0.001 0.001 17.794 18.038 qs_ot_get_derivative 108 11.5 0.001 0.001 14.938 15.119 mp_waitall_1 245248 16.5 7.795 13.679 7.795 13.679 multiply_cannon_metrocomm3 54864 15.5 0.070 0.075 4.649 11.760 multiply_cannon_multrec 54864 15.5 4.481 6.971 7.868 11.360 rebuild_ks_matrix 119 8.3 0.000 0.000 10.139 10.258 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 10.138 10.258 qs_ot_get_p 119 10.4 0.001 0.001 9.206 9.443 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.949 9.060 multiply_cannon_sync_h2d 54864 15.5 6.122 6.660 6.122 6.660 qs_rho_update_rho_low 119 7.7 0.000 0.001 6.358 6.419 calculate_rho_elec 119 8.7 0.012 0.017 6.358 6.418 sum_up_and_integrate 119 10.3 0.012 0.014 6.347 6.356 integrate_v_rspace 119 11.3 0.002 0.003 6.336 6.346 qs_ot_p2m_diag 50 11.0 0.004 0.007 6.187 6.223 init_scf_run 11 5.9 0.000 0.001 5.541 5.541 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.540 5.541 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.453 5.454 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.916 5.353 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.998 5.090 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.926 4.928 mp_sum_l 7287 12.8 3.187 4.920 3.187 4.920 cp_fm_redistribute_end 50 14.0 2.498 4.847 2.522 4.855 rs_pw_transfer 974 11.9 0.011 0.012 4.625 4.765 cp_fm_diag_elpa_base 50 14.0 2.317 4.664 2.327 4.681 dbcsr_mm_accdrv_process 76910 16.1 1.027 1.650 3.310 4.621 density_rs2pw 119 9.7 0.004 0.005 4.105 4.216 pw_transfer 1439 11.6 0.052 0.059 3.679 3.815 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 3.603 3.743 potential_pw2rs 119 12.3 0.004 0.004 3.471 3.531 make_m2s 4572 13.5 0.054 0.056 3.372 3.502 make_images 4572 14.5 0.134 0.139 3.288 3.419 fft3d_ps 1201 14.6 0.382 0.499 3.257 3.402 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.996 3.193 apply_single 119 13.6 0.000 0.000 2.996 3.193 calculate_dm_sparse 119 9.5 0.000 0.001 2.980 3.117 init_scf_loop 11 6.9 0.001 0.005 2.991 2.998 fft_wrap_pw1pw2_140 487 13.2 0.174 0.196 2.794 2.934 jit_kernel_multiply 13 15.8 2.226 2.900 2.226 2.900 calculate_first_density_matrix 1 7.0 0.000 0.003 2.807 2.809 ot_diis_step 108 11.5 0.006 0.006 2.705 2.706 wfi_extrapolate 11 7.9 0.001 0.001 2.635 2.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.568 2.571 mp_alltoall_d11v 2130 13.8 2.212 2.412 2.212 2.412 mp_alltoall_z22v 1201 16.6 2.143 2.246 2.143 2.246 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.154 2.195 make_images_sizes 4572 15.5 0.004 0.004 1.760 2.170 mp_alltoall_i44 4572 16.5 1.756 2.165 1.756 2.165 acc_transpose_blocks 54864 15.5 0.208 0.227 1.693 2.151 grid_integrate_task_list 119 12.3 2.020 2.149 2.020 2.149 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.110 2.141 multiply_cannon_metrocomm1 54864 15.5 0.054 0.059 1.208 2.088 mp_sum_d 4135 12.0 1.106 1.707 1.106 1.707 prepare_preconditioner 11 7.9 0.000 0.001 1.662 1.689 make_preconditioner 11 8.9 0.000 0.002 1.662 1.689 mp_waitany 12084 13.8 1.491 1.640 1.491 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.612 1.624 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.557 1.593 grid_collocate_task_list 119 9.7 1.378 1.452 1.378 1.452 yz_to_x 249 15.8 0.019 0.028 1.330 1.426 make_images_data 4572 15.5 0.038 0.043 0.948 1.385 mp_allgather_i34 2286 14.5 0.887 1.351 0.887 1.351 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.247 1.259 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=59.070000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 486.486016E+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 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1032414. 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 0.053 0.094 54.520 54.529 qs_mol_dyn_low 1 2.0 0.003 0.005 54.097 54.110 qs_forces 11 3.9 0.003 0.005 54.026 54.031 qs_energies 11 4.9 0.003 0.010 52.013 52.025 scf_env_do_scf 11 5.9 0.001 0.006 45.410 45.410 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 40.998 40.999 qs_scf_new_mos 108 7.5 0.001 0.001 27.239 27.477 qs_scf_loop_do_ot 108 8.5 0.001 0.001 27.239 27.477 dbcsr_multiply_generic 2286 12.5 0.105 0.116 26.103 27.012 ot_scf_mini 108 9.5 0.003 0.004 26.127 26.302 velocity_verlet 10 3.0 0.001 0.002 25.995 25.997 multiply_cannon 2286 13.5 0.209 0.217 18.177 20.999 multiply_cannon_loop 2286 14.5 0.898 0.968 16.412 19.190 ot_mini 108 10.5 0.001 0.001 15.350 15.622 mp_waitall_1 200699 16.5 9.004 14.348 9.004 14.348 qs_ot_get_derivative 108 11.5 0.001 0.001 12.396 12.578 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 5.249 10.764 rebuild_ks_matrix 119 8.3 0.000 0.000 9.934 10.213 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 9.934 10.213 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.862 9.120 multiply_cannon_multrec 27432 15.5 1.887 4.100 5.983 8.779 qs_ot_get_p 119 10.4 0.001 0.003 7.291 7.529 sum_up_and_integrate 119 10.3 0.024 0.027 6.550 6.571 integrate_v_rspace 119 11.3 0.002 0.003 6.525 6.544 qs_rho_update_rho_low 119 7.7 0.001 0.001 6.336 6.376 calculate_rho_elec 119 8.7 0.021 0.024 6.335 6.375 rs_pw_transfer 974 11.9 0.010 0.011 5.862 6.309 dbcsr_mm_accdrv_process 47894 16.0 3.167 5.544 4.029 5.903 mp_sum_l 7287 12.8 3.376 5.563 3.376 5.563 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.546 5.407 init_scf_run 11 5.9 0.000 0.001 5.095 5.095 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.095 5.095 density_rs2pw 119 9.7 0.004 0.009 4.456 4.859 qs_ot_p2m_diag 50 11.0 0.009 0.013 4.809 4.833 make_m2s 4572 13.5 0.053 0.056 4.119 4.594 make_images 4572 14.5 0.201 0.235 4.029 4.501 init_scf_loop 11 6.9 0.000 0.003 4.342 4.347 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.274 4.315 apply_single 119 13.6 0.000 0.000 3.274 4.315 cp_dbcsr_syevd 50 12.0 0.003 0.007 4.268 4.269 potential_pw2rs 119 12.3 0.006 0.006 4.059 4.109 cp_fm_diag_elpa 50 13.0 0.000 0.001 3.457 3.461 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 3.318 3.415 pw_transfer 1439 11.6 0.066 0.069 3.293 3.382 cp_fm_redistribute_end 50 14.0 1.728 3.353 1.745 3.362 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 3.200 3.292 cp_fm_diag_elpa_base 50 14.0 1.558 3.154 1.598 3.224 prepare_preconditioner 11 7.9 0.000 0.000 3.055 3.082 make_preconditioner 11 8.9 0.000 0.002 3.055 3.082 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.919 2.981 fft3d_ps 1201 14.6 0.531 0.596 2.754 2.836 ot_diis_step 108 11.5 0.011 0.011 2.802 2.803 multiply_cannon_sync_h2d 27432 15.5 2.008 2.737 2.008 2.737 calculate_first_density_matrix 1 7.0 0.001 0.004 2.729 2.732 make_images_data 4572 15.5 0.045 0.052 1.695 2.689 calculate_dm_sparse 119 9.5 0.000 0.001 2.473 2.581 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.362 2.366 mp_alltoall_d11v 2130 13.8 2.131 2.357 2.131 2.357 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.458 2.257 wfi_extrapolate 11 7.9 0.001 0.001 2.239 2.240 mp_allgather_i34 2286 14.5 1.203 2.203 1.203 2.203 make_images_sizes 4572 15.5 0.005 0.005 1.611 2.125 mp_alltoall_i44 4572 16.5 1.607 2.121 1.607 2.121 grid_integrate_task_list 119 12.3 1.859 1.996 1.859 1.996 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.861 1.944 fft_wrap_pw1pw2_140 487 13.2 0.200 0.213 1.860 1.939 mp_sendrecv_dv 16779 12.7 1.858 1.901 1.858 1.901 jit_kernel_multiply 9 16.1 0.809 1.862 0.809 1.862 mp_alltoall_z22v 1201 16.6 1.595 1.732 1.595 1.732 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.641 1.705 mp_sum_d 4135 12.0 1.048 1.643 1.048 1.643 cp_fm_cholesky_invert 11 10.9 1.598 1.611 1.598 1.611 acc_transpose_blocks 27432 15.5 0.107 0.112 1.242 1.551 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.526 1.544 rs_pw_transfer_PW2RS_50 119 14.3 0.584 0.601 1.410 1.514 dbcsr_complete_redistribute 329 12.2 0.082 0.115 1.368 1.466 grid_collocate_task_list 119 9.7 1.287 1.417 1.287 1.417 rs_pw_transfer_RS2PW_50 119 11.7 0.268 0.286 1.185 1.262 copy_dbcsr_to_fm 153 11.3 0.002 0.003 1.090 1.198 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.140 1.163 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.711 1.145 mp_waitany 5720 13.7 0.729 1.135 0.729 1.135 arnoldi_extremal 119 11.4 0.002 0.003 1.036 1.128 arnoldi_normal_ev 119 12.4 0.003 0.007 1.034 1.126 multiply_cannon_metrocomm1 27432 15.5 0.030 0.034 0.452 1.116 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=54.529000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.636364, yerr=1.553455 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 917470. 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 0.035 0.042 46.712 46.713 qs_mol_dyn_low 1 2.0 0.004 0.016 46.410 46.420 qs_forces 11 3.9 0.007 0.010 46.341 46.343 qs_energies 11 4.9 0.002 0.005 44.373 44.386 scf_env_do_scf 11 5.9 0.000 0.001 38.534 38.534 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 33.864 33.864 velocity_verlet 10 3.0 0.001 0.001 23.144 23.146 qs_scf_new_mos 108 7.5 0.001 0.001 21.734 21.759 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.733 21.758 dbcsr_multiply_generic 2286 12.5 0.101 0.102 20.858 21.197 ot_scf_mini 108 9.5 0.002 0.002 20.764 20.786 multiply_cannon 2286 13.5 0.196 0.204 14.852 15.973 multiply_cannon_loop 2286 14.5 0.639 0.670 13.521 14.583 ot_mini 108 10.5 0.001 0.001 11.886 11.955 qs_ot_get_derivative 108 11.5 0.001 0.001 9.897 9.925 rebuild_ks_matrix 119 8.3 0.000 0.000 8.870 8.989 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.869 8.988 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.861 7.966 mp_waitall_1 158411 16.6 6.130 7.578 6.130 7.578 multiply_cannon_multrec 18288 15.5 1.859 2.729 6.790 6.976 qs_ot_get_p 119 10.4 0.001 0.001 6.044 6.151 sum_up_and_integrate 119 10.3 0.030 0.031 6.110 6.114 integrate_v_rspace 119 11.3 0.003 0.003 6.080 6.085 dbcsr_mm_accdrv_process 38222 16.0 4.166 5.625 4.852 5.684 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.600 5.621 calculate_rho_elec 119 8.7 0.031 0.031 5.600 5.620 rs_pw_transfer 974 11.9 0.009 0.010 5.257 5.480 init_scf_loop 11 6.9 0.000 0.002 4.628 4.629 init_scf_run 11 5.9 0.000 0.001 4.443 4.444 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.443 4.443 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.705 4.333 density_rs2pw 119 9.7 0.004 0.004 3.937 4.181 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 2.527 4.023 qs_ot_p2m_diag 50 11.0 0.012 0.012 3.934 3.941 potential_pw2rs 119 12.3 0.007 0.009 3.633 3.658 make_m2s 4572 13.5 0.045 0.045 3.318 3.581 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.564 3.565 make_images 4572 14.5 0.193 0.206 3.231 3.495 prepare_preconditioner 11 7.9 0.000 0.001 3.456 3.463 make_preconditioner 11 8.9 0.000 0.000 3.456 3.463 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.242 3.344 cp_fm_diag_elpa 50 13.0 0.000 0.000 3.017 3.025 cp_fm_diag_elpa_base 50 14.0 2.979 3.000 3.014 3.022 mp_sum_l 7287 12.8 2.306 2.995 2.306 2.995 pw_transfer 1439 11.6 0.066 0.070 2.916 2.951 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.823 2.858 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.278 2.678 apply_single 119 13.6 0.000 0.000 2.278 2.677 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.485 2.504 calculate_first_density_matrix 1 7.0 0.002 0.013 2.474 2.474 fft3d_ps 1201 14.6 0.537 0.556 2.315 2.352 calculate_dm_sparse 119 9.5 0.000 0.000 2.190 2.223 grid_integrate_task_list 119 12.3 1.839 1.974 1.839 1.974 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.965 1.968 jit_kernel_multiply 10 16.1 0.633 1.967 0.633 1.967 make_images_data 4572 15.5 0.045 0.049 1.541 1.956 wfi_extrapolate 11 7.9 0.001 0.001 1.887 1.887 ot_diis_step 108 11.5 0.011 0.012 1.879 1.880 fft_wrap_pw1pw2_140 487 13.2 0.248 0.254 1.737 1.776 cp_fm_cholesky_invert 11 10.9 1.758 1.766 1.758 1.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.632 1.667 mp_alltoall_d11v 2130 13.8 1.447 1.644 1.447 1.644 mp_sendrecv_dv 11067 12.7 1.586 1.618 1.586 1.618 mp_alltoall_z22v 1201 16.6 1.502 1.584 1.502 1.584 hybrid_alltoall_any 4725 16.4 0.057 0.116 1.306 1.572 make_images_sizes 4572 15.5 0.005 0.005 1.084 1.521 mp_alltoall_i44 4572 16.5 1.079 1.516 1.079 1.516 grid_collocate_task_list 119 9.7 1.263 1.416 1.263 1.416 mp_allgather_i34 2286 14.5 0.788 1.414 0.788 1.414 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.352 1.411 multiply_cannon_sync_h2d 18288 15.5 1.192 1.390 1.192 1.390 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.369 1.387 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.675 1.301 acc_transpose_blocks 18288 15.5 0.075 0.078 1.275 1.296 rs_pw_transfer_PW2RS_50 119 14.3 0.402 0.412 1.131 1.210 dbcsr_complete_redistribute 329 12.2 0.095 0.110 0.999 1.128 rs_pw_transfer_RS2PW_50 119 11.7 0.240 0.246 1.049 1.109 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.089 1.104 mp_waitany 9880 13.7 0.779 1.044 0.779 1.044 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.034 1.041 cp_fm_cholesky_decompose 22 10.9 0.981 0.985 0.981 0.985 arnoldi_extremal 119 11.4 0.001 0.002 0.952 0.977 arnoldi_normal_ev 119 12.4 0.002 0.002 0.951 0.975 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.123 0.694 0.965 mp_sum_d 4135 12.0 0.740 0.949 0.740 0.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=46.713000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.475520E+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 63496. 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 0.042 0.072 46.520 46.520 qs_mol_dyn_low 1 2.0 0.003 0.003 46.159 46.168 qs_forces 11 3.9 0.002 0.002 46.088 46.089 qs_energies 11 4.9 0.002 0.006 44.101 44.109 scf_env_do_scf 11 5.9 0.001 0.004 37.893 37.895 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 32.885 32.886 velocity_verlet 10 3.0 0.001 0.001 23.668 23.670 dbcsr_multiply_generic 2286 12.5 0.101 0.105 21.199 21.422 qs_scf_new_mos 108 7.5 0.001 0.001 21.343 21.404 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.342 21.403 ot_scf_mini 108 9.5 0.003 0.004 20.219 20.283 multiply_cannon 2286 13.5 0.234 0.272 15.361 15.908 multiply_cannon_loop 2286 14.5 0.940 0.968 13.859 14.383 ot_mini 108 10.5 0.001 0.001 11.468 11.551 qs_ot_get_derivative 108 11.5 0.001 0.001 9.332 9.389 multiply_cannon_multrec 27432 15.5 2.322 3.039 8.577 8.930 rebuild_ks_matrix 119 8.3 0.000 0.000 8.476 8.566 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.475 8.565 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.537 7.618 dbcsr_mm_accdrv_process 47916 15.9 5.395 6.966 6.163 7.367 qs_ot_get_p 119 10.4 0.001 0.001 5.735 5.832 sum_up_and_integrate 119 10.3 0.035 0.037 5.402 5.410 integrate_v_rspace 119 11.3 0.003 0.003 5.367 5.376 mp_waitall_1 137007 16.6 4.090 5.340 4.090 5.340 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.254 5.303 calculate_rho_elec 119 8.7 0.040 0.046 5.253 5.303 init_scf_loop 11 6.9 0.001 0.005 4.972 4.973 init_scf_run 11 5.9 0.000 0.001 4.590 4.590 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.589 4.590 rs_pw_transfer 974 11.9 0.009 0.009 4.342 4.540 make_m2s 4572 13.5 0.054 0.056 3.759 4.180 make_images 4572 14.5 0.270 0.331 3.650 4.069 prepare_preconditioner 11 7.9 0.000 0.001 3.895 3.906 make_preconditioner 11 8.9 0.001 0.002 3.895 3.906 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.407 3.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.444 3.811 qs_ot_p2m_diag 50 11.0 0.015 0.022 3.725 3.735 density_rs2pw 119 9.7 0.004 0.004 3.523 3.706 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.294 3.295 potential_pw2rs 119 12.3 0.009 0.010 2.951 2.973 pw_transfer 1439 11.6 0.066 0.071 2.806 2.852 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.323 2.825 apply_single 119 13.6 0.000 0.000 2.323 2.825 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.771 2.780 cp_fm_diag_elpa_base 50 14.0 2.723 2.747 2.768 2.778 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.713 2.763 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.626 2.657 calculate_first_density_matrix 1 7.0 0.001 0.003 2.556 2.558 calculate_dm_sparse 119 9.5 0.000 0.000 2.273 2.332 make_images_data 4572 15.5 0.045 0.048 1.683 2.259 fft3d_ps 1201 14.6 0.562 0.618 2.154 2.192 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.099 2.102 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.964 2.082 ot_diis_step 108 11.5 0.012 0.012 2.056 2.057 fft_wrap_pw1pw2_140 487 13.2 0.286 0.304 1.983 2.036 grid_integrate_task_list 119 12.3 1.858 1.963 1.858 1.963 wfi_extrapolate 11 7.9 0.001 0.001 1.957 1.957 mp_sum_l 7287 12.8 1.662 1.941 1.662 1.941 hybrid_alltoall_any 4725 16.4 0.063 0.152 1.409 1.938 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.756 1.789 dbcsr_complete_redistribute 329 12.2 0.122 0.168 1.251 1.555 acc_transpose_blocks 27432 15.5 0.109 0.111 1.465 1.542 mp_alltoall_z22v 1201 16.6 1.476 1.511 1.476 1.511 mp_alltoall_d11v 2130 13.8 1.385 1.508 1.385 1.508 jit_kernel_multiply 8 15.8 0.707 1.506 0.707 1.506 cp_fm_cholesky_invert 11 10.9 1.491 1.498 1.491 1.498 make_images_sizes 4572 15.5 0.005 0.005 1.053 1.487 mp_alltoall_i44 4572 16.5 1.048 1.481 1.048 1.481 mp_sendrecv_dv 8211 12.7 1.354 1.376 1.354 1.376 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.339 1.376 grid_collocate_task_list 119 9.7 1.260 1.362 1.260 1.362 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.307 1.317 cp_fm_upper_to_full 72 14.2 0.884 1.283 0.884 1.283 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.204 1.204 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.819 1.113 mp_allgather_i34 2286 14.5 0.785 1.104 0.785 1.104 multiply_cannon_sync_h2d 27432 15.5 0.953 1.046 0.953 1.046 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.036 1.046 multiply_cannon_metrocomm1 27432 15.5 0.032 0.034 0.440 0.983 rs_pw_transfer_RS2PW_50 119 11.7 0.228 0.234 0.916 0.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=46.520000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.727273, yerr=3.359949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 598.220800E+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 63495. 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 0.017 0.032 40.780 40.780 qs_mol_dyn_low 1 2.0 0.003 0.003 40.528 40.563 qs_forces 11 3.9 0.002 0.002 40.373 40.375 qs_energies 11 4.9 0.001 0.001 38.365 38.369 scf_env_do_scf 11 5.9 0.000 0.001 32.672 32.673 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.119 28.120 velocity_verlet 10 3.0 0.001 0.001 21.210 21.219 qs_scf_new_mos 108 7.5 0.001 0.001 16.995 17.042 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.994 17.041 dbcsr_multiply_generic 2286 12.5 0.095 0.099 16.150 16.264 ot_scf_mini 108 9.5 0.002 0.002 16.043 16.079 multiply_cannon 2286 13.5 0.234 0.247 11.384 12.328 multiply_cannon_loop 2286 14.5 0.333 0.346 9.954 10.137 ot_mini 108 10.5 0.001 0.001 8.254 8.311 rebuild_ks_matrix 119 8.3 0.000 0.000 7.967 7.999 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.967 7.998 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.108 7.138 qs_ot_get_derivative 108 11.5 0.001 0.001 6.618 6.654 multiply_cannon_multrec 9144 15.5 1.603 1.927 5.775 6.073 mp_waitall_1 115863 16.7 4.533 5.457 4.533 5.457 qs_ot_get_p 119 10.4 0.001 0.001 5.352 5.427 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.231 5.248 calculate_rho_elec 119 8.7 0.060 0.061 5.230 5.247 sum_up_and_integrate 119 10.3 0.037 0.041 5.200 5.205 integrate_v_rspace 119 11.3 0.003 0.003 5.163 5.169 init_scf_loop 11 6.9 0.000 0.000 4.511 4.512 dbcsr_mm_accdrv_process 12550 15.8 3.129 4.073 4.072 4.148 init_scf_run 11 5.9 0.000 0.001 4.031 4.032 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.031 4.031 rs_pw_transfer 974 11.9 0.008 0.008 3.624 3.751 make_m2s 4572 13.5 0.035 0.036 3.286 3.705 make_images 4572 14.5 0.268 0.302 3.195 3.613 qs_ot_p2m_diag 50 11.0 0.022 0.023 3.574 3.578 prepare_preconditioner 11 7.9 0.000 0.000 3.473 3.479 make_preconditioner 11 8.9 0.000 0.000 3.473 3.479 density_rs2pw 119 9.7 0.004 0.004 3.325 3.475 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.305 3.359 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.254 3.255 pw_transfer 1439 11.6 0.067 0.070 2.981 2.998 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.887 2.904 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.703 2.711 cp_fm_diag_elpa_base 50 14.0 2.659 2.689 2.699 2.708 potential_pw2rs 119 12.3 0.010 0.011 2.622 2.632 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.328 2.343 make_images_data 4572 15.5 0.039 0.042 1.713 2.282 calculate_first_density_matrix 1 7.0 0.000 0.000 2.257 2.257 fft3d_ps 1201 14.6 0.570 0.584 2.222 2.234 hybrid_alltoall_any 4725 16.4 0.063 0.175 1.556 2.179 fft_wrap_pw1pw2_140 487 13.2 0.363 0.374 2.107 2.126 cp_fm_cholesky_invert 11 10.9 2.040 2.049 2.040 2.049 grid_integrate_task_list 119 12.3 1.986 2.043 1.986 2.043 calculate_dm_sparse 119 9.5 0.000 0.000 2.000 2.031 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.853 1.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.742 1.745 wfi_extrapolate 11 7.9 0.001 0.001 1.698 1.698 mp_allgather_i34 2286 14.5 0.757 1.653 0.757 1.653 ot_diis_step 108 11.5 0.013 0.013 1.583 1.584 jit_kernel_multiply 9 15.5 0.904 1.565 0.904 1.565 mp_alltoall_z22v 1201 16.6 1.524 1.560 1.524 1.560 grid_collocate_task_list 119 9.7 1.366 1.486 1.366 1.486 mp_alltoall_d11v 2130 13.8 1.409 1.485 1.409 1.485 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.460 1.473 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.423 1.453 apply_single 119 13.6 0.000 0.000 1.423 1.453 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.799 1.435 make_images_sizes 4572 15.5 0.005 0.005 0.906 1.431 mp_alltoall_i44 4572 16.5 0.901 1.426 0.901 1.426 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 1.186 1.413 qs_energies_init_hamiltonians 11 5.9 0.011 0.015 1.303 1.303 mp_sum_l 7287 12.8 1.054 1.294 1.054 1.294 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.184 1.195 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.162 1.176 mp_sendrecv_dv 5355 12.7 1.084 1.098 1.084 1.098 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.995 0.997 dbcsr_complete_redistribute 329 12.2 0.157 0.165 0.945 0.995 cp_fm_cholesky_decompose 22 10.9 0.917 0.922 0.917 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.915 arnoldi_extremal 119 11.4 0.002 0.002 0.889 0.909 arnoldi_normal_ev 119 12.4 0.002 0.002 0.887 0.907 acc_transpose_blocks 9144 15.5 0.038 0.039 0.891 0.901 yz_to_x 606 15.1 0.053 0.056 0.849 0.871 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.807 0.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=40.780000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.727273, yerr=3.277774 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 761.634816E+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 63729. 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 0.018 0.034 54.362 54.363 qs_mol_dyn_low 1 2.0 0.003 0.003 54.093 54.100 qs_forces 11 3.9 0.002 0.003 54.018 54.021 qs_energies 11 4.9 0.001 0.001 51.702 51.713 scf_env_do_scf 11 5.9 0.001 0.001 45.246 45.246 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 34.683 34.685 velocity_verlet 10 3.0 0.001 0.001 30.601 30.608 qs_scf_new_mos 108 7.5 0.001 0.001 22.103 22.182 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.102 22.182 dbcsr_multiply_generic 2286 12.5 0.103 0.105 21.745 21.991 ot_scf_mini 108 9.5 0.002 0.002 20.799 20.889 multiply_cannon 2286 13.5 0.307 0.325 15.449 16.982 multiply_cannon_loop 2286 14.5 0.349 0.353 13.784 14.804 ot_mini 108 10.5 0.001 0.001 11.735 11.873 init_scf_loop 11 6.9 0.000 0.000 10.508 10.513 qs_ot_get_derivative 108 11.5 0.001 0.001 9.319 9.408 prepare_preconditioner 11 7.9 0.000 0.000 9.307 9.322 make_preconditioner 11 8.9 0.000 0.000 9.307 9.322 rebuild_ks_matrix 119 8.3 0.000 0.000 9.072 9.232 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 9.071 9.231 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.747 9.157 multiply_cannon_multrec 9144 15.5 3.430 4.706 8.617 8.727 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.118 8.261 mp_waitall_1 94719 16.7 5.500 6.788 5.500 6.788 dbcsr_mm_accdrv_process 12550 15.8 4.164 5.700 5.063 6.421 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.762 5.776 calculate_rho_elec 119 8.7 0.118 0.121 5.761 5.776 qs_ot_get_p 119 10.4 0.001 0.001 5.548 5.710 sum_up_and_integrate 119 10.3 0.064 0.065 5.312 5.318 integrate_v_rspace 119 11.3 0.003 0.003 5.248 5.254 cp_fm_upper_to_full 72 14.2 3.271 4.733 3.271 4.733 make_m2s 4572 13.5 0.038 0.038 4.011 4.350 init_scf_run 11 5.9 0.000 0.001 4.257 4.257 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.257 4.257 make_images 4572 14.5 0.355 0.387 3.889 4.229 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.195 3.658 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 2.641 3.617 pw_transfer 1439 11.6 0.069 0.069 3.600 3.607 qs_ot_p2m_diag 50 11.0 0.042 0.043 3.587 3.590 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.501 3.508 density_rs2pw 119 9.7 0.004 0.004 3.412 3.428 dbcsr_complete_redistribute 329 12.2 0.299 0.303 2.473 3.400 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.195 3.196 cp_fm_cholesky_invert 11 10.9 3.074 3.079 3.074 3.079 rs_pw_transfer 974 11.9 0.009 0.009 2.950 2.981 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.945 2.880 make_images_data 4572 15.5 0.043 0.045 2.263 2.833 fft_wrap_pw1pw2_140 487 13.2 0.616 0.622 2.790 2.798 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.522 2.720 apply_single 119 13.6 0.000 0.000 2.522 2.720 mp_sum_l 7287 12.8 1.745 2.681 1.745 2.681 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.669 2.670 cp_fm_diag_elpa_base 50 14.0 2.505 2.566 2.666 2.667 hybrid_alltoall_any 4725 16.4 0.090 0.150 2.132 2.613 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.502 2.541 mp_alltoall_i22 627 13.8 1.559 2.539 1.559 2.539 fft3d_ps 1201 14.6 0.602 0.610 2.519 2.526 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.554 2.464 calculate_dm_sparse 119 9.5 0.000 0.000 2.418 2.455 potential_pw2rs 119 12.3 0.014 0.014 2.380 2.384 ot_diis_step 108 11.5 0.014 0.015 2.346 2.347 calculate_first_density_matrix 1 7.0 0.000 0.000 2.261 2.263 grid_integrate_task_list 119 12.3 2.238 2.249 2.238 2.249 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.079 2.080 wfi_extrapolate 11 7.9 0.001 0.001 1.907 1.907 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.810 1.813 mp_alltoall_z22v 1201 16.6 1.778 1.794 1.778 1.794 mp_alltoall_d11v 2130 13.8 1.724 1.758 1.724 1.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.689 1.733 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.637 1.683 grid_collocate_task_list 119 9.7 1.598 1.609 1.598 1.609 mp_allgather_i34 2286 14.5 0.642 1.349 0.642 1.349 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.301 1.321 make_images_sizes 4572 15.5 0.005 0.005 0.788 1.305 mp_alltoall_i44 4572 16.5 0.783 1.301 0.783 1.301 cp_fm_cholesky_decompose 22 10.9 1.235 1.249 1.235 1.249 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.172 1.189 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.149 1.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=54.363000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.272727, yerr=9.780230 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 502.112256E+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 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2298049. 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 0.023 0.041 91.814 91.816 qs_mol_dyn_low 1 2.0 0.003 0.004 91.473 91.488 qs_forces 11 3.9 0.003 0.003 91.401 91.402 qs_energies 11 4.9 0.002 0.007 88.218 88.232 scf_env_do_scf 11 5.9 0.001 0.002 78.529 78.532 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 71.097 71.098 dbcsr_multiply_generic 2055 12.4 0.109 0.112 51.310 51.635 qs_scf_new_mos 99 7.5 0.001 0.001 49.820 49.946 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.820 49.945 velocity_verlet 10 3.0 0.001 0.002 47.678 47.678 ot_scf_mini 99 9.5 0.002 0.003 47.387 47.491 multiply_cannon 2055 13.4 0.179 0.185 41.812 42.783 multiply_cannon_loop 2055 14.4 1.495 1.527 40.435 41.373 ot_mini 99 10.5 0.001 0.001 26.764 26.890 qs_ot_get_derivative 99 11.5 0.001 0.001 19.977 20.085 multiply_cannon_multrec 49320 15.4 12.566 13.168 17.345 18.127 rebuild_ks_matrix 110 8.3 0.000 0.000 17.017 17.152 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 17.017 17.151 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.927 15.046 mp_waitall_1 220248 16.4 11.653 12.684 11.653 12.684 qs_ot_get_p 110 10.4 0.001 0.001 12.108 12.239 multiply_cannon_sync_h2d 49320 15.4 10.357 10.893 10.357 10.893 sum_up_and_integrate 110 10.3 0.036 0.042 9.067 9.080 integrate_v_rspace 110 11.3 0.003 0.004 9.031 9.051 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.788 8.908 calculate_rho_elec 110 8.6 0.021 0.026 8.787 8.907 qs_ot_p2m_diag 48 11.0 0.012 0.018 8.354 8.381 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.200 7.773 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.288 7.757 apply_single 110 13.6 0.000 0.001 7.288 7.757 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.498 7.498 init_scf_run 11 5.9 0.000 0.001 7.423 7.423 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.422 7.423 init_scf_loop 11 6.9 0.001 0.003 7.393 7.394 rs_pw_transfer 902 11.9 0.011 0.013 5.977 6.866 multiply_cannon_metrocomm3 49320 15.4 0.078 0.080 5.726 6.757 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.664 6.687 cp_fm_diag_elpa_base 48 14.0 6.638 6.659 6.661 6.684 ot_diis_step 99 11.5 0.006 0.006 6.646 6.647 density_rs2pw 110 9.6 0.004 0.005 5.487 6.270 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.846 5.921 make_m2s 4110 13.4 0.061 0.064 5.543 5.701 make_images 4110 14.4 0.179 0.192 5.446 5.605 pw_transfer 1331 11.6 0.056 0.067 5.020 5.204 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.931 5.119 prepare_preconditioner 11 7.9 0.000 0.000 4.977 5.004 make_preconditioner 11 8.9 0.000 0.003 4.976 5.004 dbcsr_mm_accdrv_process 87628 16.1 1.847 1.943 4.655 4.961 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.715 4.771 wfi_extrapolate 11 7.9 0.001 0.001 4.704 4.704 mp_sum_l 6594 12.7 3.517 4.569 3.517 4.569 potential_pw2rs 110 12.3 0.006 0.007 4.322 4.409 fft3d_ps 1111 14.6 0.797 0.902 4.132 4.295 fft_wrap_pw1pw2_140 451 13.1 0.446 0.493 3.904 4.097 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.383 3.997 calculate_dm_sparse 110 9.5 0.001 0.001 3.831 3.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.675 3.681 grid_integrate_task_list 110 12.3 3.324 3.585 3.324 3.585 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.233 3.273 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.113 3.156 mp_alltoall_z22v 1111 16.6 2.839 3.102 2.839 3.102 mp_alltoall_d11v 2046 13.8 2.606 3.041 2.606 3.041 mp_waitany 14300 13.8 2.045 2.800 2.045 2.800 cp_fm_cholesky_invert 11 10.9 2.772 2.777 2.772 2.777 make_images_data 4110 15.4 0.043 0.046 2.411 2.738 calculate_first_density_matrix 1 7.0 0.000 0.003 2.610 2.621 jit_kernel_multiply 13 15.9 2.538 2.555 2.538 2.555 grid_collocate_task_list 110 9.6 2.173 2.392 2.173 2.392 hybrid_alltoall_any 4261 16.3 0.082 0.472 2.011 2.347 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.223 2.257 make_images_sizes 4110 15.4 0.004 0.004 1.601 2.162 mp_alltoall_i44 4110 16.4 1.597 2.158 1.597 2.158 acc_transpose_blocks 49320 15.4 0.201 0.210 2.019 2.083 mp_sum_d 3889 11.9 1.336 1.867 1.336 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=91.816000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=2.742382 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 585.723904E+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 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1269949. 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 0.021 0.034 88.418 88.420 qs_mol_dyn_low 1 2.0 0.003 0.005 88.082 88.094 qs_forces 11 3.9 0.003 0.003 87.999 88.002 qs_energies 11 4.9 0.002 0.011 84.253 84.274 scf_env_do_scf 11 5.9 0.001 0.002 74.636 74.639 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 63.016 63.017 velocity_verlet 10 3.0 0.001 0.002 45.891 45.892 dbcsr_multiply_generic 2055 12.4 0.121 0.127 43.012 43.764 qs_scf_new_mos 99 7.5 0.001 0.001 43.132 43.285 qs_scf_loop_do_ot 99 8.5 0.001 0.001 43.131 43.284 ot_scf_mini 99 9.5 0.003 0.004 41.161 41.282 multiply_cannon 2055 13.4 0.221 0.247 33.094 36.000 multiply_cannon_loop 2055 14.4 0.919 0.943 31.049 33.229 ot_mini 99 10.5 0.001 0.001 22.470 22.559 qs_ot_get_derivative 99 11.5 0.001 0.001 16.206 16.332 rebuild_ks_matrix 110 8.3 0.000 0.000 15.930 16.034 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 15.929 16.033 mp_waitall_1 176588 16.5 11.434 15.808 11.434 15.808 multiply_cannon_multrec 24660 15.4 7.186 9.181 13.392 15.171 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.012 14.103 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 7.110 11.819 init_scf_loop 11 6.9 0.002 0.012 11.553 11.555 qs_ot_get_p 110 10.4 0.001 0.001 11.271 11.359 prepare_preconditioner 11 7.9 0.000 0.001 9.222 9.242 make_preconditioner 11 8.9 0.000 0.002 9.221 9.242 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.731 8.921 sum_up_and_integrate 110 10.3 0.052 0.059 8.449 8.465 integrate_v_rspace 110 11.3 0.002 0.003 8.396 8.409 qs_ot_p2m_diag 48 11.0 0.028 0.043 8.335 8.371 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.105 8.131 calculate_rho_elec 110 8.6 0.040 0.047 8.105 8.131 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.662 7.663 multiply_cannon_sync_h2d 24660 15.4 6.677 7.618 6.677 7.618 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.707 7.261 apply_single 110 13.6 0.000 0.001 6.707 7.261 init_scf_run 11 5.9 0.000 0.001 6.992 6.993 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.992 6.993 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.597 6.611 cp_fm_diag_elpa_base 48 14.0 6.530 6.551 6.593 6.606 make_m2s 4110 13.4 0.057 0.060 5.980 6.523 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.750 6.471 make_images 4110 14.4 0.396 0.437 5.868 6.406 dbcsr_mm_accdrv_process 52282 16.1 4.660 5.809 6.061 6.343 ot_diis_step 99 11.5 0.010 0.010 6.085 6.086 rs_pw_transfer 902 11.9 0.012 0.013 5.126 5.997 density_rs2pw 110 9.6 0.004 0.005 5.063 5.864 pw_transfer 1331 11.6 0.067 0.074 4.948 5.111 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.840 5.006 cp_fm_cholesky_invert 11 10.9 4.811 4.827 4.811 4.827 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.613 4.689 mp_sum_l 6594 12.7 3.333 4.458 3.333 4.458 wfi_extrapolate 11 7.9 0.001 0.001 4.275 4.275 potential_pw2rs 110 12.3 0.008 0.009 4.005 4.043 fft3d_ps 1111 14.6 1.114 1.341 3.822 3.990 make_images_data 4110 15.4 0.047 0.051 3.206 3.798 fft_wrap_pw1pw2_140 451 13.1 0.513 0.525 3.331 3.503 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.497 3.501 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.707 3.500 grid_integrate_task_list 110 12.3 3.214 3.486 3.214 3.486 calculate_dm_sparse 110 9.5 0.001 0.001 3.308 3.344 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.976 3.008 mp_alltoall_d11v 2046 13.8 2.458 2.980 2.458 2.980 mp_allgather_i34 2055 14.4 1.371 2.739 1.371 2.739 cp_fm_cholesky_decompose 22 10.9 2.704 2.713 2.704 2.713 qs_ot_get_orbitals 99 10.5 0.000 0.001 2.582 2.648 calculate_first_density_matrix 1 7.0 0.001 0.004 2.592 2.597 grid_collocate_task_list 110 9.6 2.114 2.553 2.114 2.553 mp_alltoall_z22v 1111 16.6 2.325 2.416 2.325 2.416 mp_waitany 10164 13.8 1.601 2.326 1.601 2.326 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.091 2.105 dbcsr_complete_redistribute 325 12.2 0.233 0.286 1.718 2.036 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.845 2.015 jit_kernel_multiply 10 16.2 1.049 1.958 1.049 1.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.953 1.955 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 1.118 1.933 make_images_sizes 4110 15.4 0.005 0.005 1.416 1.928 mp_alltoall_i44 4110 16.4 1.412 1.924 1.412 1.924 mp_sum_d 3889 11.9 1.439 1.859 1.439 1.859 mp_irecv_dv 57340 16.2 0.720 1.796 0.720 1.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=88.420000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.272727, yerr=5.925153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 660.213760E+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 66428. 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 0.026 0.046 75.385 75.386 qs_mol_dyn_low 1 2.0 0.003 0.003 74.975 74.987 qs_forces 11 3.9 0.006 0.017 74.892 74.893 qs_energies 11 4.9 0.004 0.009 71.405 71.417 scf_env_do_scf 11 5.9 0.001 0.002 62.612 62.613 scf_env_do_scf_inner_loop 99 6.5 0.004 0.011 50.701 50.702 velocity_verlet 10 3.0 0.001 0.002 40.611 40.613 dbcsr_multiply_generic 2055 12.4 0.112 0.117 32.772 33.760 qs_scf_new_mos 99 7.5 0.001 0.001 33.034 33.179 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.033 33.178 ot_scf_mini 99 9.5 0.003 0.005 31.527 31.657 multiply_cannon 2055 13.4 0.213 0.227 23.642 25.766 multiply_cannon_loop 2055 14.4 0.618 0.631 21.835 24.571 ot_mini 99 10.5 0.001 0.001 16.802 16.966 mp_waitall_1 139946 16.5 9.535 15.024 9.535 15.024 rebuild_ks_matrix 110 8.3 0.000 0.000 14.011 14.378 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.011 14.377 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.369 12.704 qs_ot_get_derivative 99 11.5 0.001 0.001 11.816 11.958 init_scf_loop 11 6.9 0.001 0.003 11.859 11.861 multiply_cannon_multrec 16440 15.4 3.823 5.185 9.556 10.841 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 5.086 10.285 prepare_preconditioner 11 7.9 0.000 0.000 9.841 9.866 make_preconditioner 11 8.9 0.000 0.002 9.841 9.866 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.109 9.503 qs_ot_get_p 110 10.4 0.001 0.001 9.266 9.463 sum_up_and_integrate 110 10.3 0.061 0.063 7.820 7.835 integrate_v_rspace 110 11.3 0.003 0.005 7.759 7.774 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.410 7.462 calculate_rho_elec 110 8.6 0.059 0.059 7.409 7.461 qs_ot_p2m_diag 48 11.0 0.042 0.047 6.646 6.661 make_m2s 4110 13.4 0.050 0.051 5.592 6.330 make_images 4110 14.4 0.392 0.510 5.475 6.215 init_scf_run 11 5.9 0.000 0.001 6.199 6.200 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.199 6.200 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.167 6.169 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.235 5.943 apply_single 110 13.6 0.000 0.000 5.235 5.943 dbcsr_mm_accdrv_process 34862 16.1 4.599 5.329 5.593 5.746 density_rs2pw 110 9.6 0.005 0.005 4.272 5.531 rs_pw_transfer 902 11.9 0.010 0.011 4.280 5.485 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.286 5.294 cp_fm_diag_elpa_base 48 14.0 5.200 5.241 5.283 5.292 ot_diis_step 99 11.5 0.011 0.011 4.845 4.846 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.968 4.635 cp_fm_cholesky_invert 11 10.9 4.528 4.541 4.528 4.541 pw_transfer 1331 11.6 0.067 0.073 4.319 4.351 multiply_cannon_sync_h2d 16440 15.4 3.633 4.316 3.633 4.316 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.212 4.246 make_images_data 4110 15.4 0.043 0.047 3.229 4.094 mp_sum_l 6594 12.7 2.922 3.955 2.922 3.955 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.839 3.849 wfi_extrapolate 11 7.9 0.001 0.001 3.665 3.665 grid_integrate_task_list 110 12.3 3.270 3.558 3.270 3.558 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.375 3.449 fft_wrap_pw1pw2_140 451 13.1 0.632 0.641 3.338 3.371 potential_pw2rs 110 12.3 0.011 0.011 3.287 3.314 fft3d_ps 1111 14.6 1.092 1.104 3.039 3.071 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.025 3.027 mp_waitany 17072 13.8 1.519 2.861 1.519 2.861 calculate_dm_sparse 110 9.5 0.001 0.001 2.803 2.850 mp_alltoall_d11v 2046 13.8 2.453 2.781 2.453 2.781 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.895 2.670 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.515 2.637 cp_fm_cholesky_decompose 22 10.9 2.591 2.614 2.591 2.614 mp_irecv_dv 48980 15.7 0.826 2.544 0.826 2.544 mp_allgather_i34 2055 14.4 1.129 2.543 1.129 2.543 grid_collocate_task_list 110 9.6 2.145 2.536 2.145 2.536 calculate_first_density_matrix 1 7.0 0.001 0.003 2.422 2.424 dbcsr_complete_redistribute 325 12.2 0.312 0.345 1.875 2.399 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.178 1.005 2.252 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.030 2.032 cp_fm_upper_to_full 70 14.2 1.474 1.904 1.474 1.904 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.842 1.874 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.710 1.817 mp_alltoall_z22v 1111 16.6 1.737 1.771 1.737 1.771 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.211 1.726 make_images_sizes 4110 15.4 0.005 0.005 1.010 1.655 mp_alltoall_i44 4110 16.4 1.005 1.650 1.005 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.536 jit_kernel_multiply 8 16.4 0.605 1.517 0.605 1.517 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=75.386000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.000000, yerr=8.268231 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 738.455552E+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 66426. 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 0.023 0.035 83.165 83.166 qs_mol_dyn_low 1 2.0 0.003 0.004 82.824 82.836 qs_forces 11 3.9 0.003 0.006 82.741 82.743 qs_energies 11 4.9 0.004 0.013 78.963 78.970 scf_env_do_scf 11 5.9 0.002 0.006 69.678 69.681 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 52.989 52.990 velocity_verlet 10 3.0 0.001 0.002 46.568 46.571 qs_scf_new_mos 99 7.5 0.001 0.001 34.967 35.080 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.967 35.080 dbcsr_multiply_generic 2055 12.4 0.117 0.123 33.505 33.815 ot_scf_mini 99 9.5 0.003 0.004 33.106 33.194 multiply_cannon 2055 13.4 0.243 0.260 23.845 24.957 multiply_cannon_loop 2055 14.4 0.888 0.916 21.749 22.545 ot_mini 99 10.5 0.001 0.001 17.145 17.280 init_scf_loop 11 6.9 0.001 0.003 16.629 16.632 prepare_preconditioner 11 7.9 0.000 0.000 14.596 14.612 make_preconditioner 11 8.9 0.001 0.005 14.596 14.612 make_full_inverse_cholesky 11 9.9 0.000 0.000 12.668 14.239 rebuild_ks_matrix 110 8.3 0.000 0.000 14.094 14.204 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.094 14.203 multiply_cannon_multrec 24660 15.4 4.088 6.507 12.478 13.719 qs_ot_get_derivative 99 11.5 0.001 0.001 12.568 12.665 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.434 12.537 qs_ot_get_p 110 10.4 0.001 0.001 10.288 10.425 mp_waitall_1 121746 16.5 7.108 9.992 7.108 9.992 dbcsr_mm_accdrv_process 52304 16.0 6.870 8.286 8.248 9.181 sum_up_and_integrate 110 10.3 0.067 0.070 8.003 8.016 integrate_v_rspace 110 11.3 0.003 0.003 7.936 7.951 make_m2s 4110 13.4 0.060 0.062 7.221 7.868 make_images 4110 14.4 0.572 0.695 7.077 7.719 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.690 7.698 calculate_rho_elec 110 8.6 0.078 0.081 7.690 7.698 qs_ot_p2m_diag 48 11.0 0.054 0.063 7.612 7.630 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.955 6.956 init_scf_run 11 5.9 0.000 0.001 6.274 6.275 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.274 6.275 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.946 5.955 cp_fm_diag_elpa_base 48 14.0 5.753 5.826 5.942 5.952 cp_fm_cholesky_invert 11 10.9 5.445 5.462 5.445 5.462 cp_fm_upper_to_full 70 14.2 3.607 5.234 3.607 5.234 density_rs2pw 110 9.6 0.004 0.004 4.423 4.947 rs_pw_transfer 902 11.9 0.010 0.010 4.245 4.742 pw_transfer 1331 11.6 0.066 0.076 4.684 4.723 make_images_data 4110 15.4 0.046 0.049 3.848 4.646 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.577 4.620 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 2.224 4.575 dbcsr_complete_redistribute 325 12.2 0.404 0.449 3.289 4.542 ot_diis_step 99 11.5 0.011 0.012 4.484 4.484 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.303 4.445 apply_single 110 13.6 0.000 0.000 4.303 4.445 hybrid_alltoall_any 4261 16.3 0.121 0.457 3.236 4.386 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.047 4.094 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.987 4.038 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.478 3.711 wfi_extrapolate 11 7.9 0.001 0.001 3.662 3.662 fft_wrap_pw1pw2_140 451 13.1 0.665 0.685 3.537 3.581 grid_integrate_task_list 110 12.3 3.318 3.567 3.318 3.567 potential_pw2rs 110 12.3 0.013 0.013 3.386 3.402 fft3d_ps 1111 14.6 1.090 1.122 3.343 3.371 multiply_cannon_sync_h2d 24660 15.4 3.112 3.366 3.112 3.366 cp_fm_cholesky_decompose 22 10.9 3.232 3.283 3.232 3.283 calculate_dm_sparse 110 9.5 0.001 0.001 3.179 3.214 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.919 3.146 mp_alltoall_i22 605 13.7 1.884 3.126 1.884 3.126 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.019 3.021 mp_alltoall_d11v 2046 13.8 2.567 2.960 2.567 2.960 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.503 2.540 grid_collocate_task_list 110 9.6 2.241 2.479 2.241 2.479 calculate_first_density_matrix 1 7.0 0.002 0.007 2.474 2.476 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.346 2.348 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.196 2.249 mp_sum_l 6594 12.7 1.765 2.230 1.765 2.230 jit_kernel_multiply 12 15.7 1.047 2.193 1.047 2.193 mp_alltoall_z22v 1111 16.6 2.037 2.118 2.037 2.118 mp_allgather_i34 2055 14.4 1.195 2.048 1.195 2.048 mp_waitany 13376 13.8 1.488 2.043 1.488 2.043 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.910 1.925 make_images_sizes 4110 15.4 0.005 0.005 1.145 1.828 mp_alltoall_i44 4110 16.4 1.141 1.823 1.141 1.823 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=83.166000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.545455, yerr=12.794627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 845.754368E+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 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 74.669 74.670 qs_mol_dyn_low 1 2.0 0.003 0.003 74.381 74.391 qs_forces 11 3.9 0.003 0.004 74.295 74.298 qs_energies 11 4.9 0.001 0.001 70.172 70.180 scf_env_do_scf 11 5.9 0.000 0.001 60.708 60.710 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 48.796 48.797 velocity_verlet 10 3.0 0.001 0.001 40.668 40.671 qs_scf_new_mos 99 7.5 0.001 0.001 30.110 30.173 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.109 30.172 ot_scf_mini 99 9.5 0.002 0.002 28.639 28.676 dbcsr_multiply_generic 2055 12.4 0.108 0.113 27.418 27.578 multiply_cannon 2055 13.4 0.254 0.270 19.334 21.329 multiply_cannon_loop 2055 14.4 0.323 0.335 17.344 17.702 rebuild_ks_matrix 110 8.3 0.000 0.000 14.321 14.381 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.321 14.380 ot_mini 99 10.5 0.001 0.001 13.602 13.648 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.727 12.784 init_scf_loop 11 6.9 0.000 0.000 11.843 11.845 mp_waitall_1 103326 16.6 9.251 11.505 9.251 11.505 qs_ot_get_p 110 10.4 0.001 0.001 10.399 10.456 prepare_preconditioner 11 7.9 0.000 0.000 9.851 9.857 make_preconditioner 11 8.9 0.000 0.000 9.850 9.857 qs_ot_get_derivative 99 11.5 0.001 0.001 9.393 9.432 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.299 9.427 multiply_cannon_multrec 8220 15.4 3.213 4.536 7.452 8.349 sum_up_and_integrate 110 10.3 0.080 0.081 8.305 8.322 integrate_v_rspace 110 11.3 0.003 0.003 8.225 8.240 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.122 8.139 calculate_rho_elec 110 8.6 0.114 0.114 8.121 8.139 qs_ot_p2m_diag 48 11.0 0.081 0.084 7.986 7.992 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.482 7.483 make_m2s 4110 13.4 0.039 0.040 6.003 6.589 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.455 6.463 cp_fm_diag_elpa_base 48 14.0 6.371 6.413 6.451 6.459 make_images 4110 14.4 0.636 0.695 5.870 6.457 init_scf_run 11 5.9 0.000 0.001 6.059 6.060 scf_env_initial_rho_setup 11 6.9 0.003 0.003 6.059 6.060 cp_fm_cholesky_invert 11 10.9 5.407 5.413 5.407 5.413 pw_transfer 1331 11.6 0.066 0.071 5.323 5.340 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.214 5.237 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.398 5.102 dbcsr_mm_accdrv_process 17442 15.9 2.795 3.779 4.106 5.032 density_rs2pw 110 9.6 0.004 0.004 4.527 4.913 make_images_data 4110 15.4 0.038 0.043 3.482 4.378 rs_pw_transfer 902 11.9 0.010 0.010 3.892 4.298 fft_wrap_pw1pw2_140 451 13.1 0.831 0.842 4.166 4.195 ot_diis_step 99 11.5 0.012 0.012 4.146 4.147 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.992 4.039 apply_single 110 13.6 0.000 0.000 3.992 4.039 hybrid_alltoall_any 4261 16.3 0.200 0.861 3.194 3.859 fft3d_ps 1111 14.6 1.145 1.171 3.762 3.776 grid_integrate_task_list 110 12.3 3.475 3.662 3.475 3.662 wfi_extrapolate 11 7.9 0.001 0.001 3.642 3.643 potential_pw2rs 110 12.3 0.015 0.016 3.385 3.392 mp_alltoall_d11v 2046 13.8 2.770 3.019 2.770 3.019 multiply_cannon_sync_h2d 8220 15.4 2.840 2.991 2.840 2.991 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.954 2.975 mp_allgather_i34 2055 14.4 1.171 2.819 1.171 2.819 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.803 2.805 cp_fm_cholesky_decompose 22 10.9 2.760 2.776 2.760 2.776 calculate_dm_sparse 110 9.5 0.001 0.001 2.736 2.772 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.731 2.735 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.697 2.714 grid_collocate_task_list 110 9.6 2.356 2.608 2.356 2.608 mp_alltoall_z22v 1111 16.6 2.378 2.398 2.378 2.398 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.904 2.374 calculate_first_density_matrix 1 7.0 0.000 0.000 2.296 2.297 dbcsr_complete_redistribute 325 12.2 0.539 0.569 2.126 2.266 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.231 2.255 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.991 make_images_sizes 4110 15.4 0.005 0.005 1.101 1.967 mp_alltoall_i44 4110 16.4 1.096 1.963 1.096 1.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.942 mp_waitany 9240 13.8 1.513 1.914 1.513 1.914 mp_sum_l 6594 12.7 1.397 1.717 1.397 1.717 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.571 1.695 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.620 1.655 jit_kernel_multiply 9 15.8 1.002 1.608 1.002 1.608 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.534 rs_gather_matrices 110 12.3 0.326 0.372 1.257 1.497 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=74.670000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=793.090909, yerr=14.145057 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 1.356165E+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 67104. 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 0.022 0.039 105.656 105.657 qs_mol_dyn_low 1 2.0 0.003 0.003 105.280 105.292 qs_forces 11 3.9 0.003 0.003 105.205 105.206 qs_energies 11 4.9 0.001 0.002 100.679 100.681 scf_env_do_scf 11 5.9 0.000 0.001 89.669 89.669 velocity_verlet 10 3.0 0.001 0.001 65.743 65.750 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 55.926 55.928 qs_scf_new_mos 99 7.5 0.001 0.001 34.551 34.588 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.551 34.587 init_scf_loop 11 6.9 0.000 0.000 33.646 33.650 dbcsr_multiply_generic 2055 12.4 0.121 0.125 33.274 33.431 ot_scf_mini 99 9.5 0.002 0.002 32.580 32.621 prepare_preconditioner 11 7.9 0.000 0.000 31.400 31.407 make_preconditioner 11 8.9 0.000 0.000 31.400 31.407 make_full_inverse_cholesky 11 9.9 0.000 0.000 25.414 30.824 multiply_cannon 2055 13.4 0.349 0.380 23.770 25.148 multiply_cannon_loop 2055 14.4 0.344 0.346 21.566 21.914 cp_fm_upper_to_full 70 14.2 12.735 18.413 12.735 18.413 ot_mini 99 10.5 0.001 0.001 16.837 16.873 rebuild_ks_matrix 110 8.3 0.000 0.001 16.093 16.161 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 16.093 16.160 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.506 14.564 mp_waitall_1 84994 16.7 11.127 12.304 11.127 12.304 qs_ot_get_derivative 99 11.5 0.001 0.001 11.857 11.899 dbcsr_complete_redistribute 325 12.2 1.011 1.048 7.994 11.304 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.708 10.019 qs_ot_get_p 110 10.4 0.001 0.001 9.770 9.814 multiply_cannon_multrec 8220 15.4 4.412 4.725 9.683 9.804 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.360 9.388 calculate_rho_elec 110 8.6 0.225 0.225 9.359 9.388 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.971 9.243 mp_alltoall_i22 605 13.7 5.621 8.931 5.621 8.931 sum_up_and_integrate 110 10.3 0.150 0.151 8.725 8.740 cp_fm_cholesky_invert 11 10.9 8.717 8.724 8.717 8.724 integrate_v_rspace 110 11.3 0.003 0.004 8.575 8.589 make_m2s 4110 13.4 0.043 0.044 7.085 7.792 make_images 4110 14.4 0.879 0.913 6.893 7.600 qs_ot_p2m_diag 48 11.0 0.151 0.155 7.436 7.440 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.835 6.837 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 6.514 6.813 init_scf_run 11 5.9 0.000 0.001 6.547 6.547 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.547 6.547 pw_transfer 1331 11.6 0.076 0.076 6.337 6.345 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.219 6.226 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.911 5.911 cp_fm_diag_elpa_base 48 14.0 5.323 5.534 5.907 5.907 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.052 5.500 apply_single 110 13.6 0.000 0.000 5.052 5.500 dbcsr_mm_accdrv_process 11614 15.7 3.244 3.686 5.126 5.362 make_images_data 4110 15.4 0.041 0.044 4.166 5.242 fft_wrap_pw1pw2_140 451 13.1 1.335 1.338 5.218 5.225 density_rs2pw 110 9.6 0.004 0.004 5.071 5.086 ot_diis_step 99 11.5 0.015 0.016 4.930 4.930 hybrid_alltoall_any 4261 16.3 0.257 0.552 3.947 4.919 fft3d_ps 1111 14.6 1.307 1.314 4.151 4.160 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.705 4.155 wfi_extrapolate 11 7.9 0.001 0.001 4.058 4.058 multiply_cannon_sync_h2d 8220 15.4 3.954 3.959 3.954 3.959 grid_integrate_task_list 110 12.3 3.840 3.894 3.840 3.894 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.721 3.723 rs_pw_transfer 902 11.9 0.010 0.011 3.613 3.652 cp_fm_cholesky_decompose 22 10.9 3.407 3.428 3.407 3.428 potential_pw2rs 110 12.3 0.022 0.022 3.421 3.423 calculate_dm_sparse 110 9.5 0.001 0.001 3.380 3.409 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.335 3.361 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.163 3.165 mp_alltoall_d11v 2046 13.8 2.900 2.973 2.900 2.973 grid_collocate_task_list 110 9.6 2.727 2.759 2.727 2.759 mp_alltoall_z22v 1111 16.6 2.591 2.602 2.591 2.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.498 2.524 calculate_first_density_matrix 1 7.0 0.000 0.000 2.353 2.354 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.257 2.312 mp_allgather_i34 2055 14.4 0.981 2.310 0.981 2.310 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.158 2.207 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.193 2.202 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.081 2.186 mp_sum_l 6594 12.7 1.542 2.161 1.542 2.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=105.657000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1217.454545, yerr=60.459398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.018048E+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 57767. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2635444. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 223.239 223.240 qs_mol_dyn_low 1 2.0 0.003 0.004 222.749 222.763 qs_forces 11 3.9 0.005 0.005 222.655 222.655 qs_energies 11 4.9 0.002 0.003 216.621 216.638 scf_env_do_scf 11 5.9 0.001 0.002 199.073 199.077 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 174.636 174.638 velocity_verlet 10 3.0 0.001 0.002 134.412 134.413 qs_scf_new_mos 117 7.6 0.001 0.001 129.703 129.937 qs_scf_loop_do_ot 117 8.6 0.001 0.001 129.702 129.936 dbcsr_multiply_generic 2507 12.6 0.181 0.185 124.481 125.342 ot_scf_mini 117 9.6 0.003 0.003 123.053 123.289 multiply_cannon 2507 13.6 0.242 0.252 100.760 102.759 multiply_cannon_loop 2507 14.6 2.011 2.081 98.210 100.333 ot_mini 117 10.6 0.001 0.001 65.618 65.860 multiply_cannon_multrec 60168 15.6 33.822 35.665 41.748 43.651 qs_ot_get_derivative 117 11.6 0.001 0.001 40.681 40.922 rebuild_ks_matrix 128 8.3 0.001 0.001 36.518 36.865 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 36.518 36.865 qs_ot_get_p 128 10.4 0.001 0.001 35.319 35.734 mp_waitall_1 267128 16.5 29.736 33.286 29.736 33.286 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.702 33.001 multiply_cannon_sync_h2d 60168 15.6 28.012 30.175 28.012 30.175 qs_ot_p2m_diag 83 11.4 0.079 0.091 28.160 28.290 cp_dbcsr_syevd 83 12.4 0.005 0.005 25.606 25.609 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.524 25.251 apply_single 128 13.6 0.001 0.001 24.523 25.251 ot_diis_step 117 11.6 0.008 0.008 24.729 24.730 init_scf_loop 11 6.9 0.000 0.001 24.337 24.339 cp_fm_diag_elpa 83 13.4 0.000 0.000 21.973 22.007 cp_fm_diag_elpa_base 83 14.4 21.856 21.906 21.967 22.001 prepare_preconditioner 11 7.9 0.000 0.000 19.453 19.501 make_preconditioner 11 8.9 0.000 0.002 19.453 19.501 make_full_inverse_cholesky 11 9.9 0.000 0.000 18.631 18.816 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.515 18.677 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.698 16.832 calculate_rho_elec 128 8.7 0.046 0.065 16.697 16.831 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 14.758 16.816 sum_up_and_integrate 128 10.3 0.089 0.106 16.780 16.797 integrate_v_rspace 128 11.3 0.004 0.005 16.691 16.709 make_m2s 5014 13.6 0.103 0.110 15.692 16.066 make_images 5014 14.6 0.404 0.424 15.510 15.890 density_rs2pw 128 9.7 0.006 0.007 9.863 13.266 init_scf_run 11 5.9 0.000 0.001 13.154 13.154 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.153 13.154 rs_pw_transfer 1046 11.9 0.016 0.018 9.062 12.622 cp_fm_cholesky_invert 11 10.9 11.301 11.309 11.301 11.309 wfi_extrapolate 11 7.9 0.001 0.001 9.893 9.893 pw_transfer 1547 11.6 0.075 0.097 9.463 9.775 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 9.259 9.575 make_images_data 5014 15.6 0.066 0.071 7.703 8.674 calculate_dm_sparse 128 9.5 0.001 0.001 8.486 8.553 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.340 8.513 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.043 8.418 mp_sum_l 7950 12.9 6.758 8.142 6.758 8.142 grid_integrate_task_list 128 12.3 7.175 8.064 7.175 8.064 dbcsr_mm_accdrv_process 124484 16.2 2.896 3.079 7.466 8.001 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.885 7.987 hybrid_alltoall_any 5200 16.5 0.290 2.264 6.611 7.700 fft3d_ps 1291 14.7 2.149 2.775 7.054 7.507 fft_wrap_pw1pw2_140 523 13.2 1.247 1.308 7.184 7.497 potential_pw2rs 128 12.3 0.009 0.010 6.684 6.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.810 6.822 mp_alltoall_d11v 2415 14.1 5.006 6.337 5.006 6.337 mp_waitany 16020 13.9 2.781 6.140 2.781 6.140 grid_collocate_task_list 128 9.7 4.722 5.986 4.722 5.986 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.817 5.954 cp_fm_cholesky_decompose 22 10.9 5.841 5.855 5.841 5.855 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.292 2.206 5.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=223.240000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=6.132153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.909888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58220. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5044925. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.046 219.635 219.636 qs_mol_dyn_low 1 2.0 0.003 0.005 219.132 219.146 qs_forces 11 3.9 0.005 0.011 219.040 219.043 qs_energies 11 4.9 0.002 0.009 211.814 211.835 scf_env_do_scf 11 5.9 0.001 0.002 194.276 194.286 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 154.479 154.483 velocity_verlet 10 3.0 0.001 0.001 137.399 137.401 qs_scf_new_mos 117 7.6 0.001 0.001 110.640 111.020 qs_scf_loop_do_ot 117 8.6 0.001 0.001 110.639 111.019 ot_scf_mini 117 9.6 0.004 0.005 105.583 106.011 dbcsr_multiply_generic 2507 12.6 0.193 0.200 102.982 105.324 multiply_cannon 2507 13.6 0.475 0.530 79.111 83.025 multiply_cannon_loop 2507 14.6 1.239 1.283 74.757 80.410 ot_mini 117 10.6 0.001 0.001 54.270 54.690 mp_waitall_1 214728 16.6 28.828 48.691 28.828 48.691 init_scf_loop 11 6.9 0.001 0.008 39.679 39.682 multiply_cannon_multrec 30084 15.6 21.306 27.014 30.884 36.847 rebuild_ks_matrix 128 8.3 0.001 0.001 35.794 36.302 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 35.793 36.302 multiply_cannon_metrocomm3 30084 15.6 0.091 0.099 17.577 35.785 prepare_preconditioner 11 7.9 0.000 0.001 34.935 35.006 make_preconditioner 11 8.9 0.000 0.003 34.935 35.006 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.530 34.165 qs_ot_get_p 128 10.4 0.001 0.002 32.358 32.950 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.160 32.611 qs_ot_get_derivative 117 11.6 0.001 0.002 31.729 32.141 qs_ot_p2m_diag 83 11.4 0.187 0.215 26.464 26.576 cp_dbcsr_syevd 83 12.4 0.006 0.006 24.929 24.932 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.302 23.911 apply_single 128 13.6 0.001 0.001 22.302 23.910 ot_diis_step 117 11.6 0.015 0.015 22.293 22.295 multiply_cannon_sync_h2d 30084 15.6 18.582 21.101 18.582 21.101 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.928 20.952 cp_fm_diag_elpa_base 83 14.4 20.609 20.773 20.920 20.943 cp_fm_cholesky_invert 11 10.9 20.689 20.704 20.689 20.704 make_m2s 5014 13.6 0.089 0.093 16.232 18.402 make_images 5014 14.6 1.146 1.339 16.022 18.198 sum_up_and_integrate 128 10.3 0.116 0.133 16.778 16.809 integrate_v_rspace 128 11.3 0.004 0.006 16.661 16.698 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.038 16.139 calculate_rho_elec 128 8.7 0.088 0.104 16.037 16.139 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 13.060 13.402 init_scf_run 11 5.9 0.000 0.001 12.483 12.484 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.482 12.484 make_images_data 5014 15.6 0.064 0.073 9.797 12.282 density_rs2pw 128 9.7 0.006 0.007 9.156 11.577 hybrid_alltoall_any 5200 16.5 0.343 1.517 8.273 11.534 multiply_cannon_metrocomm4 27577 15.6 0.103 0.120 3.881 11.189 mp_irecv_dv 69486 16.3 3.681 10.792 3.681 10.792 rs_pw_transfer 1046 11.9 0.014 0.016 8.251 10.703 pw_transfer 1547 11.6 0.087 0.100 9.824 9.916 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.597 9.696 dbcsr_mm_accdrv_process 62242 16.2 4.520 5.238 9.067 9.637 cp_fm_cholesky_decompose 22 10.9 9.269 9.356 9.269 9.356 wfi_extrapolate 11 7.9 0.001 0.001 9.240 9.240 mp_sum_l 7950 12.9 5.886 8.111 5.886 8.111 grid_integrate_task_list 128 12.3 7.339 8.024 7.339 8.024 fft_wrap_pw1pw2_140 523 13.2 1.331 1.351 7.913 8.014 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.836 7.607 fft3d_ps 1291 14.7 2.811 2.987 6.912 6.985 calculate_dm_sparse 128 9.5 0.001 0.001 6.756 6.912 potential_pw2rs 128 12.3 0.016 0.019 6.702 6.748 mp_allgather_i34 2507 14.6 2.924 6.480 2.924 6.480 mp_alltoall_d11v 2415 14.1 4.872 6.396 4.872 6.396 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.377 6.387 grid_collocate_task_list 128 9.7 4.827 6.015 4.827 6.015 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.850 5.934 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.345 5.521 mp_waitany 11748 13.9 2.706 5.168 2.706 5.168 dbcsr_complete_redistribute 395 12.7 0.775 0.858 3.731 4.697 mp_sum_d 4470 12.1 3.157 4.685 3.157 4.685 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.381 2.159 4.592 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=219.636000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.454545, yerr=3.201239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 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 957.804544E+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 937889582784 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57818. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394160. 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 0.026 0.045 214.516 214.518 qs_mol_dyn_low 1 2.0 0.003 0.007 213.874 213.944 qs_forces 11 3.9 0.004 0.006 213.748 213.751 qs_energies 11 4.9 0.002 0.009 206.669 206.681 scf_env_do_scf 11 5.9 0.001 0.002 190.088 190.089 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 145.825 145.826 velocity_verlet 10 3.0 0.003 0.004 136.692 136.695 qs_scf_new_mos 118 7.6 0.001 0.001 102.745 103.162 qs_scf_loop_do_ot 118 8.6 0.001 0.001 102.745 103.162 ot_scf_mini 118 9.6 0.003 0.004 98.295 98.739 dbcsr_multiply_generic 2527 12.6 0.184 0.190 88.774 90.326 multiply_cannon 2527 13.6 0.500 0.524 64.834 71.275 multiply_cannon_loop 2527 14.6 0.865 0.891 60.834 64.876 ot_mini 118 10.6 0.001 0.001 48.038 48.534 init_scf_loop 11 6.9 0.001 0.006 44.145 44.146 mp_waitall_1 171878 16.6 30.213 41.850 30.213 41.850 prepare_preconditioner 11 7.9 0.000 0.001 39.797 39.850 make_preconditioner 11 8.9 0.000 0.003 39.797 39.850 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.205 38.770 qs_ot_get_p 129 10.4 0.001 0.001 34.508 35.079 rebuild_ks_matrix 129 8.3 0.001 0.001 33.720 34.276 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 33.720 34.275 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.407 30.915 qs_ot_p2m_diag 83 11.4 0.265 0.272 28.976 29.022 multiply_cannon_metrocomm3 20216 15.6 0.058 0.063 17.558 28.845 qs_ot_get_derivative 118 11.6 0.001 0.002 27.370 27.817 cp_dbcsr_syevd 83 12.4 0.005 0.006 27.592 27.596 multiply_cannon_multrec 20216 15.6 13.244 16.316 21.914 24.884 cp_fm_diag_elpa 83 13.4 0.000 0.000 23.085 23.102 cp_fm_diag_elpa_base 83 14.4 22.553 22.753 23.078 23.096 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.590 21.602 apply_single 129 13.6 0.001 0.001 20.590 21.602 ot_diis_step 118 11.6 0.018 0.018 20.480 20.482 cp_fm_cholesky_invert 11 10.9 20.108 20.123 20.108 20.123 make_m2s 5054 13.6 0.080 0.083 17.163 18.392 make_images 5054 14.6 1.184 1.279 16.931 18.159 sum_up_and_integrate 129 10.3 0.134 0.146 16.892 16.918 integrate_v_rspace 129 11.3 0.004 0.005 16.758 16.785 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.726 16.760 calculate_rho_elec 129 8.7 0.131 0.146 16.726 16.759 multiply_cannon_sync_h2d 20216 15.6 14.003 15.604 14.003 15.604 make_images_data 5054 15.6 0.059 0.067 10.572 12.279 init_scf_run 11 5.9 0.000 0.001 11.350 11.350 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.350 11.350 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.935 11.248 density_rs2pw 129 9.7 0.006 0.007 9.232 11.223 hybrid_alltoall_any 5240 16.5 0.441 2.018 9.022 10.963 pw_transfer 1559 11.6 0.088 0.107 10.089 10.217 cp_fm_cholesky_decompose 22 10.9 10.144 10.188 10.144 10.188 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.861 9.996 rs_pw_transfer 1054 12.0 0.013 0.014 7.886 9.877 multiply_cannon_metrocomm4 17689 15.6 0.062 0.073 3.531 9.672 mp_irecv_dv 50625 16.2 3.411 9.419 3.411 9.419 dbcsr_mm_accdrv_process 41830 16.2 4.573 5.345 8.174 8.350 wfi_extrapolate 11 7.9 0.001 0.001 8.236 8.236 fft_wrap_pw1pw2_140 527 13.2 1.419 1.444 8.059 8.201 grid_integrate_task_list 129 12.3 7.527 8.091 7.527 8.091 cp_fm_upper_to_full 105 14.8 6.033 7.758 6.033 7.758 dbcsr_complete_redistribute 395 12.7 1.166 1.197 5.447 7.402 mp_sum_l 8010 12.9 4.982 7.219 4.982 7.219 fft3d_ps 1301 14.7 2.761 2.989 6.966 7.073 potential_pw2rs 129 12.3 0.021 0.024 6.485 6.519 mp_allgather_i34 2527 14.6 2.514 6.238 2.514 6.238 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 5.482 6.194 calculate_dm_sparse 129 9.5 0.001 0.001 6.101 6.188 mp_alltoall_d11v 2423 14.1 5.312 6.042 5.312 6.042 grid_collocate_task_list 129 9.7 5.078 5.952 5.078 5.952 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.915 5.865 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 5.615 5.620 mp_waitany 11836 13.9 2.716 4.742 2.716 4.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.571 4.724 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.623 4.677 transfer_fm_to_dbcsr 11 9.9 0.018 0.031 2.572 4.482 mp_alltoall_i22 718 14.1 2.309 4.432 2.309 4.432 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=214.518000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=908.727273, yerr=8.147402 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.152934E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57954. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.047 223.958 223.959 qs_mol_dyn_low 1 2.0 0.003 0.004 223.459 223.474 qs_forces 11 3.9 0.004 0.005 223.331 223.341 qs_energies 11 4.9 0.002 0.009 215.718 215.730 scf_env_do_scf 11 5.9 0.001 0.002 197.529 197.538 velocity_verlet 10 3.0 0.001 0.001 147.279 147.283 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 140.486 140.487 qs_scf_new_mos 117 7.6 0.001 0.001 99.524 99.878 qs_scf_loop_do_ot 117 8.6 0.001 0.001 99.523 99.877 ot_scf_mini 117 9.6 0.003 0.005 94.845 95.182 dbcsr_multiply_generic 2507 12.6 0.188 0.194 83.864 84.746 multiply_cannon 2507 13.6 0.547 0.578 56.288 60.274 init_scf_loop 11 6.9 0.001 0.007 56.883 56.885 multiply_cannon_loop 2507 14.6 1.182 1.211 51.801 54.084 prepare_preconditioner 11 7.9 0.000 0.000 52.483 52.508 make_preconditioner 11 8.9 0.000 0.003 52.483 52.508 make_full_inverse_cholesky 11 9.9 0.012 0.023 45.805 51.132 ot_mini 117 10.6 0.001 0.001 46.057 46.397 qs_ot_get_p 128 10.4 0.001 0.001 34.401 34.702 rebuild_ks_matrix 128 8.3 0.001 0.001 32.218 32.575 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.217 32.575 mp_waitall_1 147882 16.7 20.818 31.857 20.818 31.857 multiply_cannon_multrec 30084 15.6 13.880 19.057 25.764 30.692 qs_ot_p2m_diag 83 11.4 0.342 0.389 29.379 29.435 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.971 29.295 cp_dbcsr_syevd 83 12.4 0.005 0.006 27.808 27.811 qs_ot_get_derivative 117 11.6 0.001 0.002 25.973 26.292 make_m2s 5014 13.6 0.095 0.100 22.419 23.880 cp_fm_diag_elpa 83 13.4 0.000 0.000 23.611 23.622 cp_fm_diag_elpa_base 83 14.4 22.489 22.858 23.601 23.611 make_images 5014 14.6 1.930 2.268 22.110 23.572 cp_fm_cholesky_invert 11 10.9 22.234 22.246 22.234 22.246 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.398 20.029 apply_single 128 13.6 0.001 0.001 19.398 20.029 ot_diis_step 117 11.6 0.018 0.019 19.926 19.928 cp_fm_upper_to_full 105 14.8 12.001 17.822 12.001 17.822 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 7.238 16.765 sum_up_and_integrate 128 10.3 0.140 0.151 16.612 16.636 integrate_v_rspace 128 11.3 0.004 0.004 16.472 16.502 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.409 16.436 calculate_rho_elec 128 8.7 0.173 0.189 16.408 16.435 make_images_data 5014 15.6 0.061 0.067 12.364 14.857 dbcsr_complete_redistribute 395 12.7 1.501 1.631 9.931 13.956 hybrid_alltoall_any 5200 16.5 0.528 2.213 10.880 12.922 multiply_cannon_sync_h2d 30084 15.6 11.650 12.829 11.650 12.829 copy_fm_to_dbcsr 209 11.7 0.001 0.002 8.145 12.190 init_scf_run 11 5.9 0.000 0.001 12.044 12.046 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.044 12.045 dbcsr_mm_accdrv_process 62264 16.2 7.370 8.337 11.460 11.997 cp_fm_cholesky_decompose 22 10.9 11.070 11.165 11.070 11.165 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.940 11.164 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.656 10.623 pw_transfer 1547 11.6 0.088 0.103 10.061 10.148 mp_alltoall_i22 716 14.1 5.982 10.019 5.982 10.019 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.832 9.926 density_rs2pw 128 9.7 0.006 0.006 8.552 9.879 wfi_extrapolate 11 7.9 0.001 0.001 8.899 8.899 fft_wrap_pw1pw2_140 523 13.2 1.553 1.587 8.335 8.437 rs_pw_transfer 1046 11.9 0.013 0.014 6.844 8.280 grid_integrate_task_list 128 12.3 7.707 8.169 7.707 8.169 multiply_cannon_metrocomm4 25070 15.6 0.080 0.092 2.813 7.768 mp_irecv_dv 76098 16.2 2.668 7.496 2.668 7.496 mp_alltoall_d11v 2415 14.1 5.909 7.129 5.909 7.129 fft3d_ps 1291 14.7 2.840 2.903 6.759 6.827 calculate_dm_sparse 128 9.5 0.001 0.001 6.408 6.504 potential_pw2rs 128 12.3 0.023 0.023 5.908 5.940 grid_collocate_task_list 128 9.7 5.197 5.891 5.197 5.891 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.702 5.746 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.936 5.039 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.572 4.622 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.532 4.533 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.435 4.527 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=223.959000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1088.454545, yerr=17.814007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.495392E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57813. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.036 210.666 210.667 qs_mol_dyn_low 1 2.0 0.003 0.003 210.218 210.229 qs_forces 11 3.9 0.004 0.005 210.098 210.103 qs_energies 11 4.9 0.001 0.002 201.963 201.971 scf_env_do_scf 11 5.9 0.001 0.001 182.899 182.902 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 137.535 137.536 velocity_verlet 10 3.0 0.001 0.001 137.274 137.279 qs_scf_new_mos 118 7.6 0.001 0.001 94.776 94.938 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.775 94.937 ot_scf_mini 118 9.6 0.003 0.004 90.051 90.132 dbcsr_multiply_generic 2527 12.6 0.185 0.198 80.080 80.910 multiply_cannon 2527 13.6 0.586 0.616 57.065 61.579 multiply_cannon_loop 2527 14.6 0.447 0.462 51.515 53.323 init_scf_loop 11 6.9 0.000 0.000 45.178 45.181 ot_mini 118 10.6 0.001 0.001 44.220 44.318 prepare_preconditioner 11 7.9 0.000 0.000 40.991 41.017 make_preconditioner 11 8.9 0.000 0.000 40.991 41.017 make_full_inverse_cholesky 11 9.9 0.016 0.027 38.664 39.005 mp_waitall_1 126780 16.7 29.593 36.809 29.593 36.809 rebuild_ks_matrix 129 8.3 0.001 0.001 32.622 32.783 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 32.621 32.783 qs_ot_get_p 129 10.4 0.001 0.001 32.123 32.225 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.461 29.611 qs_ot_p2m_diag 83 11.4 0.495 0.500 27.145 27.163 cp_dbcsr_syevd 83 12.4 0.005 0.006 25.607 25.610 cp_fm_cholesky_invert 11 10.9 24.168 24.177 24.168 24.177 qs_ot_get_derivative 118 11.6 0.001 0.002 23.571 23.646 cp_fm_diag_elpa 83 13.4 0.000 0.000 21.503 21.514 cp_fm_diag_elpa_base 83 14.4 21.178 21.320 21.490 21.501 multiply_cannon_multrec 10108 15.6 10.497 14.451 17.804 21.391 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 12.602 21.265 make_m2s 5054 13.6 0.067 0.073 18.329 20.641 ot_diis_step 118 11.6 0.023 0.035 20.503 20.504 make_images 5054 14.6 2.309 2.790 18.019 20.332 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.921 20.265 apply_single 129 13.6 0.001 0.001 19.921 20.264 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.507 17.554 calculate_rho_elec 129 8.7 0.258 0.269 17.507 17.554 sum_up_and_integrate 129 10.3 0.183 0.193 17.223 17.275 integrate_v_rspace 129 11.3 0.004 0.004 17.040 17.102 make_images_data 5054 15.6 0.050 0.061 11.147 13.711 hybrid_alltoall_any 5240 16.5 0.823 3.839 10.582 13.007 multiply_cannon_sync_h2d 10108 15.6 11.574 12.321 11.574 12.321 init_scf_run 11 5.9 0.000 0.001 11.935 11.936 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.934 11.935 pw_transfer 1559 11.6 0.088 0.096 11.064 11.098 cp_fm_cholesky_decompose 22 10.9 10.912 11.065 10.912 11.065 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 10.837 10.874 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.577 9.637 density_rs2pw 129 9.7 0.006 0.006 8.561 9.482 fft_wrap_pw1pw2_140 527 13.2 1.924 1.961 9.158 9.198 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 5.885 8.873 wfi_extrapolate 11 7.9 0.001 0.001 8.866 8.867 mp_allgather_i34 2527 14.6 3.741 8.500 3.741 8.500 grid_integrate_task_list 129 12.3 8.025 8.446 8.025 8.446 dbcsr_mm_accdrv_process 20926 16.1 2.682 3.497 6.942 7.555 mp_alltoall_d11v 2423 14.1 6.599 7.472 6.599 7.472 fft3d_ps 1301 14.7 2.794 2.870 7.289 7.337 rs_pw_transfer 1054 12.0 0.012 0.013 6.008 6.931 dbcsr_complete_redistribute 395 12.7 2.107 2.158 6.095 6.580 calculate_dm_sparse 129 9.5 0.001 0.001 6.447 6.548 grid_collocate_task_list 129 9.7 5.558 6.232 5.558 6.232 potential_pw2rs 129 12.3 0.027 0.028 5.891 5.913 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.518 5.539 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.340 5.343 multiply_cannon_metrocomm4 7581 15.6 0.023 0.027 1.888 4.325 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.254 4.300 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.888 4.293 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.266 4.286 mp_irecv_dv 29086 15.9 1.856 4.257 1.856 4.257 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.192 4.252 copy_dbcsr_to_fm 186 11.8 0.004 0.004 4.110 4.237 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=210.667000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1388.272727, yerr=43.851176 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_performance_tests/20/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 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.174257E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59908. MP_Allreduce 10950 1507. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.043 330.138 330.139 qs_mol_dyn_low 1 2.0 0.003 0.003 329.561 329.575 qs_forces 11 3.9 0.004 0.004 329.459 329.464 qs_energies 11 4.9 0.002 0.002 319.738 319.755 scf_env_do_scf 11 5.9 0.001 0.001 296.089 296.104 velocity_verlet 10 3.0 0.001 0.001 234.825 234.834 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 160.609 160.611 init_scf_loop 11 6.9 0.000 0.000 135.193 135.197 prepare_preconditioner 11 7.9 0.000 0.000 129.790 129.818 make_preconditioner 11 8.9 0.000 0.000 129.790 129.818 make_full_inverse_cholesky 11 9.9 0.036 0.039 105.474 126.746 qs_scf_new_mos 116 7.6 0.001 0.001 105.822 105.973 qs_scf_loop_do_ot 116 8.6 0.001 0.001 105.821 105.972 ot_scf_mini 116 9.6 0.003 0.004 100.826 100.883 dbcsr_multiply_generic 2485 12.5 0.212 0.224 88.702 89.250 cp_fm_upper_to_full 104 14.8 53.590 76.804 53.590 76.804 multiply_cannon 2485 13.5 0.705 0.752 60.766 61.207 multiply_cannon_loop 2485 14.5 0.466 0.477 56.500 58.199 ot_mini 116 10.6 0.001 0.001 48.798 48.879 dbcsr_complete_redistribute 393 12.7 3.971 4.017 30.613 43.393 rebuild_ks_matrix 127 8.3 0.001 0.001 40.898 40.948 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 40.897 40.947 cp_fm_cholesky_invert 11 10.9 39.728 39.735 39.728 39.735 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.910 39.706 qs_ks_update_qs_env 127 7.6 0.001 0.001 37.630 37.667 mp_waitall_1 102768 16.8 32.126 37.113 32.126 37.113 transfer_fm_to_dbcsr 11 9.9 0.029 0.032 24.269 36.962 qs_ot_get_p 127 10.4 0.001 0.001 35.807 35.864 mp_alltoall_i22 712 14.1 22.011 34.900 22.011 34.900 qs_ot_p2m_diag 82 11.4 0.867 0.874 30.855 30.906 cp_dbcsr_syevd 82 12.4 0.006 0.006 28.811 28.814 qs_ot_get_derivative 116 11.6 0.002 0.002 27.924 27.991 cp_fm_diag_elpa 82 13.4 0.000 0.000 24.879 24.881 cp_fm_diag_elpa_base 82 14.4 20.325 22.067 24.873 24.874 make_m2s 4970 13.5 0.076 0.077 22.148 23.455 make_images 4970 14.5 3.691 3.848 21.668 22.975 qs_rho_update_rho_low 127 7.7 0.001 0.001 22.665 22.682 calculate_rho_elec 127 8.7 0.477 0.478 22.664 22.681 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 20.449 22.015 sum_up_and_integrate 127 10.3 0.319 0.321 21.623 21.704 integrate_v_rspace 127 11.3 0.004 0.004 21.304 21.384 ot_diis_step 116 11.6 0.022 0.022 20.790 20.790 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.921 20.158 apply_single 127 13.6 0.001 0.001 19.921 20.158 multiply_cannon_multrec 9940 15.5 10.484 12.254 17.914 17.996 multiply_cannon_sync_h2d 9940 15.5 15.542 15.551 15.542 15.551 make_images_data 4970 15.5 0.056 0.062 12.360 14.731 hybrid_alltoall_any 5155 16.4 1.289 3.008 12.031 14.378 pw_transfer 1535 11.6 0.093 0.094 13.709 13.721 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 13.474 13.486 init_scf_run 11 5.9 0.000 0.001 13.370 13.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.370 13.371 cp_fm_cholesky_decompose 22 10.9 12.000 12.037 12.000 12.037 fft_wrap_pw1pw2_140 519 13.2 3.047 3.067 11.694 11.711 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.179 11.223 mp_alltoall_d11v 2401 14.1 10.175 10.829 10.175 10.829 wfi_extrapolate 11 7.9 0.001 0.001 10.114 10.114 density_rs2pw 127 9.7 0.005 0.006 9.730 9.867 dbcsr_mm_accdrv_process 20590 16.0 3.762 5.614 7.192 9.046 grid_integrate_task_list 127 12.3 8.792 9.005 8.792 9.005 fft3d_ps 1281 14.7 2.828 2.837 8.589 8.619 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.044 8.048 calculate_dm_sparse 127 9.5 0.001 0.001 6.883 6.996 rs_scatter_matrices 138 9.7 3.615 4.453 6.492 6.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.623 6.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=330.139000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2892.909091, yerr=156.751369 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 1.259262E+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 255669. 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 0.023 0.041 85.481 85.482 qs_energies 1 2.0 0.000 0.000 84.963 84.970 ls_scf 1 3.0 0.000 0.000 84.044 84.051 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.409 72.551 multiply_cannon 111 7.7 0.018 0.020 55.773 56.978 multiply_cannon_loop 111 8.7 0.206 0.227 52.269 53.674 ls_scf_main 1 4.0 0.000 0.000 52.578 52.579 density_matrix_trs4 2 5.0 0.002 0.003 47.065 47.130 ls_scf_init_scf 1 4.0 0.000 0.000 28.459 28.460 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.289 27.337 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.189 25.204 mp_waitall_1 11031 10.9 21.833 24.177 21.833 24.177 multiply_cannon_multrec 2664 9.7 8.204 8.910 15.531 17.266 multiply_cannon_sync_h2d 2664 9.7 13.984 16.426 13.984 16.426 make_m2s 222 7.7 0.008 0.010 13.104 13.652 make_images 222 8.7 0.099 0.108 13.082 13.632 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.172 11.497 make_images_data 222 9.7 0.004 0.005 7.631 8.134 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.628 6.943 7.950 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.330 7.948 hybrid_alltoall_any 227 10.6 0.217 1.839 6.513 7.551 dbcsr_mm_accdrv_process_sort 4760 11.4 6.236 7.148 6.236 7.148 calculate_norms 4752 9.8 5.625 6.244 5.625 6.244 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.841 4.943 mp_sum_l 887 5.1 3.017 4.281 3.017 4.281 arnoldi_extremal 4 6.8 0.000 0.000 3.708 3.729 arnoldi_normal_ev 4 7.8 0.001 0.002 3.708 3.729 build_subspace 16 8.4 0.009 0.012 3.610 3.613 make_images_sizes 222 9.7 0.000 0.000 0.788 3.371 mp_alltoall_i44 222 10.7 0.788 3.371 0.788 3.371 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.201 3.206 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.033 3.104 mp_irecv_dv 6231 10.9 2.017 3.076 2.017 3.076 ls_scf_post 1 4.0 0.000 0.000 3.007 3.014 ls_scf_store_result 1 5.0 0.000 0.000 2.819 2.864 dbcsr_special_finalize 555 9.7 0.005 0.006 2.331 2.712 dbcsr_merge_single_wm 555 10.7 0.474 0.602 2.323 2.704 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.477 2.689 make_images_pack 222 9.7 2.207 2.628 2.209 2.630 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.401 2.475 dbcsr_matrix_vector_mult_local 304 10.0 2.047 2.435 2.049 2.437 dbcsr_sort_data 658 11.4 2.109 2.431 2.109 2.431 buffer_matrices_ensure_size 222 8.7 1.799 2.072 1.799 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.866 1.868 rebuild_ks_matrix 3 7.3 0.000 0.000 1.857 1.858 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.857 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.482000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1137.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 2.101535E+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 266696. 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 0.043 0.061 91.105 91.118 qs_energies 1 2.0 0.000 0.000 90.552 90.556 ls_scf 1 3.0 0.000 0.001 89.232 89.235 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.120 75.447 multiply_cannon 111 7.7 0.029 0.038 53.194 56.630 ls_scf_main 1 4.0 0.000 0.004 54.875 54.880 multiply_cannon_loop 111 8.7 0.116 0.122 49.931 52.794 density_matrix_trs4 2 5.0 0.002 0.006 49.079 49.313 ls_scf_init_scf 1 4.0 0.000 0.002 30.810 30.811 mp_waitall_1 9105 10.9 21.229 30.742 21.229 30.742 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.581 29.660 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.167 27.182 multiply_cannon_multrec 1332 9.7 13.104 16.613 22.316 26.951 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.797 21.462 make_m2s 222 7.7 0.006 0.007 15.399 16.103 make_images 222 8.7 1.573 1.964 15.369 16.072 dbcsr_mm_accdrv_process 4041 10.4 0.275 0.448 8.813 10.376 dbcsr_mm_accdrv_process_sort 4041 11.4 8.406 9.928 8.406 9.928 make_images_data 222 9.7 0.004 0.004 8.877 9.822 hybrid_alltoall_any 227 10.6 0.521 2.485 8.297 9.132 mp_sum_l 887 5.1 5.233 8.096 5.233 8.096 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.234 7.853 mp_irecv_dv 3311 11.0 3.214 7.800 3.214 7.800 calculate_norms 2376 9.8 6.002 6.820 6.002 6.820 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.968 6.590 multiply_cannon_sync_h2d 1332 9.7 4.827 6.211 4.827 6.211 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.074 5.276 arnoldi_extremal 4 6.8 0.000 0.000 4.689 4.713 arnoldi_normal_ev 4 7.8 0.001 0.005 4.689 4.712 build_subspace 16 8.4 0.014 0.021 4.428 4.431 ls_scf_post 1 4.0 0.000 0.001 3.548 3.552 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.145 3.382 ls_scf_store_result 1 5.0 0.000 0.000 3.241 3.372 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.229 2.742 3.231 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.181 2.927 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.642 2.718 mp_allgather_i34 111 8.7 0.968 2.409 0.968 2.409 make_images_pack 222 9.7 2.017 2.381 2.020 2.383 dbcsr_sort_data 436 11.2 1.836 2.078 1.836 2.078 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.929 1.931 rebuild_ks_matrix 3 7.3 0.000 0.000 1.917 1.919 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.916 1.919 dbcsr_data_new 4174 10.1 1.608 1.839 1.608 1.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.118000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1707.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 2.736669E+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 265470. 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 0.042 0.072 93.980 93.982 qs_energies 1 2.0 0.000 0.000 93.347 93.351 ls_scf 1 3.0 0.000 0.001 91.935 91.941 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.848 76.134 ls_scf_main 1 4.0 0.000 0.001 57.401 57.405 multiply_cannon 111 7.7 0.041 0.079 52.501 56.412 multiply_cannon_loop 111 8.7 0.100 0.105 48.918 52.636 density_matrix_trs4 2 5.0 0.003 0.007 51.279 51.475 mp_waitall_1 7281 11.0 23.160 32.690 23.160 32.690 ls_scf_init_scf 1 4.0 0.000 0.002 30.965 30.967 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.700 29.800 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.312 27.325 multiply_cannon_multrec 888 9.7 12.658 15.465 21.199 24.568 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.794 22.385 make_m2s 222 7.7 0.006 0.007 16.808 18.045 make_images 222 8.7 1.978 2.300 16.770 18.008 hybrid_alltoall_any 227 10.6 0.621 2.869 9.099 10.516 make_images_data 222 9.7 0.003 0.004 9.461 10.498 dbcsr_mm_accdrv_process 3754 10.4 0.235 0.416 8.061 9.275 dbcsr_mm_accdrv_process_sort 3754 11.4 7.689 8.859 7.689 8.859 mp_sum_l 887 5.1 5.006 8.170 5.006 8.170 multiply_cannon_sync_h2d 888 9.7 6.026 7.398 6.026 7.398 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.447 7.149 mp_irecv_dv 2335 11.1 2.432 7.107 2.432 7.107 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.513 6.628 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.727 6.278 arnoldi_extremal 4 6.8 0.000 0.000 5.642 5.666 arnoldi_normal_ev 4 7.8 0.001 0.005 5.642 5.666 build_subspace 16 8.4 0.014 0.020 5.337 5.343 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.966 5.208 calculate_norms 1584 9.8 4.318 4.700 4.318 4.700 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.690 4.015 mp_allgather_i34 111 8.7 1.413 3.703 1.413 3.703 dbcsr_matrix_vector_mult_local 304 10.0 3.011 3.589 3.013 3.591 ls_scf_post 1 4.0 0.000 0.001 3.569 3.574 ls_scf_store_result 1 5.0 0.000 0.000 3.308 3.394 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.963 3.072 make_images_sizes 222 9.7 0.000 0.000 1.043 2.229 mp_alltoall_i44 222 10.7 1.043 2.229 1.043 2.229 dbcsr_sort_data 325 11.1 1.860 2.141 1.860 2.141 make_images_pack 222 9.7 1.819 2.086 1.822 2.089 dbcsr_data_release 9322 10.9 1.315 1.985 1.315 1.985 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.976 1.979 rebuild_ks_matrix 3 7.3 0.000 0.000 1.957 1.959 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.957 1.959 dbcsr_finalize 304 7.8 0.026 0.032 1.615 1.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.982000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2154.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 3.329081E+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 266696. 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 0.036 0.051 97.202 97.204 qs_energies 1 2.0 0.000 0.000 96.555 96.560 ls_scf 1 3.0 0.000 0.000 94.907 94.914 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.787 78.043 ls_scf_main 1 4.0 0.000 0.000 58.793 58.794 multiply_cannon 111 7.7 0.064 0.137 51.157 55.908 density_matrix_trs4 2 5.0 0.002 0.003 52.519 52.622 multiply_cannon_loop 111 8.7 0.114 0.126 46.039 49.437 ls_scf_init_scf 1 4.0 0.000 0.000 32.896 32.897 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.646 31.727 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.183 29.196 mp_waitall_1 6369 11.0 22.040 28.066 22.040 28.066 multiply_cannon_multrec 1332 9.7 14.230 16.983 22.090 25.099 make_m2s 222 7.7 0.007 0.008 20.979 22.450 make_images 222 8.7 3.140 3.605 20.929 22.402 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.702 16.509 make_images_data 222 9.7 0.004 0.004 11.635 13.312 hybrid_alltoall_any 227 10.6 0.797 3.729 10.862 12.668 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.406 7.487 8.981 dbcsr_mm_accdrv_process_sort 3641 11.4 7.113 8.573 7.113 8.573 mp_sum_l 887 5.1 4.095 7.730 4.095 7.730 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.133 6.136 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.075 6.057 mp_irecv_dv 3229 10.9 2.052 5.975 2.052 5.975 multiply_cannon_sync_h2d 1332 9.7 5.518 5.974 5.518 5.974 arnoldi_extremal 4 6.8 0.000 0.000 5.914 5.932 arnoldi_normal_ev 4 7.8 0.001 0.005 5.914 5.932 build_subspace 16 8.4 0.014 0.020 5.565 5.572 mp_allgather_i34 111 8.7 2.242 4.787 2.242 4.787 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.736 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.525 4.682 calculate_norms 2376 9.8 4.187 4.532 4.187 4.532 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.973 4.181 dbcsr_matrix_vector_mult_local 304 10.0 3.178 3.678 3.180 3.680 dbcsr_sort_data 658 11.4 3.030 3.370 3.030 3.370 ls_scf_post 1 4.0 0.000 0.000 3.218 3.223 dbcsr_special_finalize 555 9.7 0.006 0.007 2.779 3.173 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.112 3.165 dbcsr_merge_single_wm 555 10.7 0.534 0.658 2.771 3.164 ls_scf_store_result 1 5.0 0.000 0.000 2.957 3.013 dbcsr_data_release 10477 10.7 1.589 2.408 1.589 2.408 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.986 1.988 dbcsr_finalize 304 7.8 0.049 0.061 1.791 1.966 rebuild_ks_matrix 3 7.3 0.000 0.000 1.963 1.965 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.963 1.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.204000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2740.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 4.703191E+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 265558. 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 0.070 0.097 100.176 100.178 qs_energies 1 2.0 0.000 0.000 99.236 99.237 ls_scf 1 3.0 0.000 0.000 97.296 97.302 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.308 78.491 ls_scf_main 1 4.0 0.000 0.000 62.620 62.621 multiply_cannon 111 7.7 0.102 0.180 55.385 61.064 density_matrix_trs4 2 5.0 0.002 0.003 55.313 55.397 multiply_cannon_loop 111 8.7 0.069 0.075 50.992 52.704 mp_waitall_1 5436 11.0 27.026 31.905 27.026 31.905 ls_scf_init_scf 1 4.0 0.000 0.000 31.032 31.038 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.699 29.745 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.516 27.527 multiply_cannon_multrec 444 9.7 13.927 16.224 20.953 22.952 make_m2s 222 7.7 0.004 0.005 18.126 20.499 make_images 222 8.7 3.718 4.421 18.065 20.439 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.411 16.813 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.196 13.445 make_images_data 222 9.7 0.003 0.004 10.210 12.660 hybrid_alltoall_any 227 10.6 0.787 3.742 10.032 12.565 dbcsr_mm_accdrv_process 3003 10.4 0.199 0.341 6.733 7.858 dbcsr_mm_accdrv_process_sort 3003 11.4 6.418 7.507 6.418 7.507 multiply_cannon_sync_h2d 444 9.7 6.538 7.404 6.538 7.404 mp_allgather_i34 111 8.7 2.618 7.022 2.618 7.022 arnoldi_extremal 4 6.8 0.000 0.000 6.028 6.039 arnoldi_normal_ev 4 7.8 0.002 0.005 6.028 6.039 build_subspace 16 8.4 0.015 0.019 5.628 5.638 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.610 4.763 mp_sum_l 887 5.1 2.857 4.632 2.857 4.632 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.235 4.432 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.691 4.246 mp_irecv_dv 1241 11.2 1.672 4.213 1.672 4.213 dbcsr_matrix_vector_mult_local 304 10.0 3.673 4.144 3.675 4.147 make_images_sizes 222 9.7 0.000 0.000 1.275 3.803 mp_alltoall_i44 222 10.7 1.275 3.802 1.275 3.802 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.673 3.767 calculate_norms 792 9.8 3.557 3.698 3.557 3.698 ls_scf_post 1 4.0 0.000 0.000 3.643 3.645 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.435 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.921 3.381 dbcsr_finalize 304 7.8 0.062 0.078 2.196 2.290 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.141 2.143 dbcsr_merge_all 275 8.9 0.473 0.528 2.046 2.130 rebuild_ks_matrix 3 7.3 0.000 0.000 2.108 2.110 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.108 2.110 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.178000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3678.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 8.690840E+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 284111. 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 0.084 0.108 109.669 109.671 qs_energies 1 2.0 0.000 0.000 108.135 108.148 ls_scf 1 3.0 0.000 0.000 105.201 105.212 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.612 77.748 ls_scf_main 1 4.0 0.000 0.000 66.665 66.666 density_matrix_trs4 2 5.0 0.002 0.003 57.152 57.216 multiply_cannon 111 7.7 0.132 0.213 49.785 51.245 multiply_cannon_loop 111 8.7 0.068 0.070 46.448 47.779 ls_scf_init_scf 1 4.0 0.000 0.000 34.834 34.834 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.201 33.226 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.361 30.372 mp_waitall_1 4527 11.1 22.226 25.819 22.226 25.819 make_m2s 222 7.7 0.005 0.005 23.981 24.833 make_images 222 8.7 4.596 4.992 23.875 24.725 multiply_cannon_multrec 444 9.7 17.901 18.787 22.491 23.179 hybrid_alltoall_any 227 10.6 1.664 3.648 12.953 15.704 make_images_data 222 9.7 0.003 0.003 13.123 15.473 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.498 11.443 multiply_cannon_sync_h2d 444 9.7 8.867 8.990 8.867 8.990 arnoldi_extremal 4 6.8 0.000 0.000 8.015 8.031 arnoldi_normal_ev 4 7.8 0.003 0.009 8.015 8.031 build_subspace 16 8.4 0.025 0.035 7.452 7.466 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.735 5.919 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.401 5.499 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.077 5.338 dbcsr_matrix_vector_mult_local 304 10.0 5.000 5.318 5.002 5.321 dbcsr_mm_accdrv_process 1814 10.4 0.208 0.316 4.422 4.552 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.258 4.123 4.258 ls_scf_post 1 4.0 0.000 0.000 3.702 3.714 make_images_sizes 222 9.7 0.000 0.000 1.520 3.697 mp_alltoall_i44 222 10.7 1.520 3.697 1.520 3.697 ls_scf_store_result 1 5.0 0.000 0.000 3.440 3.459 calculate_norms 792 9.8 3.222 3.282 3.222 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.187 mp_allgather_i34 111 8.7 0.982 3.159 0.982 3.159 dbcsr_merge_all 275 8.9 0.890 0.919 2.879 2.965 dbcsr_complete_redistribute 5 7.6 1.431 1.483 2.840 2.939 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.904 2.904 dbcsr_data_release 12724 10.6 2.324 2.845 2.324 2.845 matrix_ls_to_qs 2 6.0 0.000 0.000 2.480 2.592 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.570 2.572 dbcsr_sort_data 325 11.1 2.447 2.515 2.447 2.515 rebuild_ks_matrix 3 7.3 0.000 0.000 2.503 2.505 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 2.503 2.505 dbcsr_new_transposed 4 7.5 0.243 0.251 2.443 2.468 mp_sum_l 887 5.1 1.535 2.420 1.535 2.420 mp_alltoall_d11v 48 9.2 2.229 2.284 2.229 2.284 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.199 2.236 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.209 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.671000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6835.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4dc27d601c3e5b0826783d31464cb7b82f8cb647_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 585.650176E+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 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 863268730. 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 0.093 0.207 336.223 336.296 qs_mol_dyn_low 1 2.0 0.003 0.017 334.755 334.827 qs_forces 5 3.8 0.008 0.081 334.615 334.687 qs_energies 5 4.8 0.025 0.072 330.767 330.840 scf_env_do_scf 5 5.8 0.003 0.029 311.779 311.846 scf_env_do_scf_inner_loop 105 6.6 0.003 0.013 270.641 270.702 qs_scf_new_mos 105 7.6 0.000 0.001 212.976 213.326 qs_scf_loop_do_ot 105 8.6 0.001 0.001 212.975 213.326 ot_scf_mini 105 9.6 0.003 0.005 202.525 202.811 dbcsr_multiply_generic 1445 12.2 0.127 0.140 142.905 144.122 velocity_verlet 4 3.0 0.001 0.034 129.791 129.819 multiply_cannon 1445 13.2 0.280 0.291 119.246 123.103 multiply_cannon_loop 1445 14.2 2.341 2.482 116.077 119.261 qs_ot_get_p 112 10.4 0.001 0.002 106.569 106.903 qs_ot_p2m_diag 40 11.0 0.020 0.031 93.189 93.286 cp_dbcsr_syevd 40 12.0 0.002 0.004 89.673 89.694 cp_fm_syevd 40 13.0 0.000 0.003 82.627 82.787 cp_fm_redistribute_end 40 14.0 38.273 76.377 38.284 76.384 cp_fm_syevd_base 40 14.0 38.072 76.174 38.072 76.174 ot_mini 105 10.6 0.001 0.003 64.582 64.766 mp_waitall_1 488190 16.1 42.545 54.755 42.545 54.755 multiply_cannon_multrec 69360 15.2 31.170 37.416 38.821 44.928 qs_ot_get_derivative 55 11.6 0.001 0.001 42.638 42.874 multiply_cannon_metrocomm3 69360 15.2 0.196 0.208 27.822 42.729 init_scf_loop 7 6.6 0.001 0.007 41.068 41.078 rebuild_ks_matrix 110 8.4 0.000 0.000 36.460 36.662 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.018 36.460 36.662 prepare_preconditioner 7 7.6 0.000 0.002 35.017 35.062 make_preconditioner 7 8.6 0.000 0.012 35.017 35.062 qs_ks_update_qs_env 112 7.6 0.001 0.001 33.497 33.668 multiply_cannon_sync_h2d 69360 15.2 29.702 33.417 29.702 33.417 qs_rho_update_rho_low 110 7.6 0.001 0.001 28.788 29.179 calculate_rho_elec 110 8.6 0.030 0.033 28.788 29.179 rs_pw_transfer 690 11.5 0.010 0.012 25.023 26.478 make_full_inverse_cholesky 7 9.6 0.000 0.000 25.417 25.484 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.225 23.546 apply_single 62 13.6 0.000 0.000 23.225 23.546 density_rs2pw 110 9.6 0.006 0.008 21.621 22.943 ot_new_cg_direction 55 11.6 0.001 0.003 21.188 21.193 cp_fm_cholesky_invert 7 10.6 16.596 16.616 16.596 16.616 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.842 16.134 init_scf_run 5 5.8 0.000 0.002 15.865 15.868 scf_env_initial_rho_setup 5 6.8 0.001 0.003 15.865 15.868 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.427 15.599 pw_transfer 1645 12.4 0.085 0.112 14.990 15.313 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 14.845 15.174 sum_up_and_integrate 60 10.3 0.028 0.034 14.503 14.518 mp_sum_l 4764 12.2 13.087 14.492 13.087 14.492 integrate_v_rspace 60 11.3 0.002 0.030 14.475 14.490 multiply_cannon_metrocomm1 69360 15.2 0.088 0.095 5.296 12.627 calculate_dm_sparse 110 9.5 0.000 0.001 12.104 12.311 check_diag 80 13.5 8.589 8.883 12.100 12.283 fft_wrap_pw1pw2_240 915 15.0 1.206 1.300 11.976 12.244 qs_vxc_create 110 10.4 0.002 0.005 11.657 11.719 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 11.208 11.339 make_m2s 2890 13.2 0.078 0.085 10.190 10.795 make_images 2890 14.2 0.242 0.262 10.084 10.690 mp_sendrecv_dv 168740 12.6 10.408 10.659 10.408 10.659 rs_pw_transfer_RS2PW_30 110 11.6 1.539 1.627 9.963 10.587 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 10.449 10.466 fft3d_pb 915 16.0 2.386 2.655 10.018 10.398 calculate_first_density_matrix 1 7.0 0.000 0.005 10.000 10.022 potential_pw2rs 60 12.3 0.003 0.003 9.907 10.009 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.742 9.841 acc_transpose_blocks 69360 15.2 0.341 0.377 9.218 9.833 make_full_single_inverse 7 9.6 0.001 0.006 9.141 9.187 dbcsr_mm_accdrv_process 154766 15.8 3.872 4.215 7.527 8.421 cp_fm_cholesky_decompose 14 10.2 8.251 8.258 8.251 8.258 xc_rho_set_and_dset_create 110 12.4 0.076 0.099 7.946 8.217 xc_vxc_pw_create 60 11.3 0.039 0.049 7.913 7.976 mp_alltoall_z22v 2340 17.7 7.302 7.973 7.302 7.973 acc_transpose_blocks_kernels 69360 16.2 0.818 0.884 7.433 7.858 xc_pw_derive 510 13.4 0.006 0.007 7.204 7.279 jit_kernel_transpose 5 15.0 6.614 6.993 6.614 6.993 mp_alltoall_d11v 1300 13.8 6.013 6.807 6.013 6.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=336.296000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 4dc27d601c3e5b0826783d31464cb7b82f8cb647 Summary: empty Status: OK