=== 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: e3caab925d13535c28a7f61e60f70e91adc8e5da ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/01 job id: 50007491 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/02 job id: 50007492 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/03 job id: 50007493 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/04 job id: 50007494 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/05 job id: 50007495 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/06 job id: 50007496 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/07 job id: 50007497 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/08 job id: 50007498 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/09 job id: 50007499 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/10 job id: 50007500 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/11 job id: 50007501 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/12 job id: 50007502 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/13 job id: 50007503 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/14 job id: 50007504 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/15 job id: 50007505 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/16 job id: 50007506 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/17 job id: 50007507 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/18 job id: 50007508 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/19 job id: 50007509 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/20 job id: 50007510 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/21 job id: 50007511 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/22 job id: 50007512 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/23 job id: 50007513 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/24 job id: 50007514 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/25 job id: 50007515 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/26 job id: 50007516 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/27 job id: 50007517 --- 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/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 133.663 133.664 farming_run 1 2.0 133.045 133.048 133.635 133.638 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.486381E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 114.719 114.719 qs_energies 1 2.0 0.000 0.000 114.521 114.523 mp2_main 1 3.0 0.000 0.000 112.466 112.468 mp2_gpw_main 1 4.0 0.028 0.033 111.490 111.492 mp2_ri_gpw_compute_in 1 5.0 0.174 0.215 92.489 92.873 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.632 55.018 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 41.039 46.157 get_2c_integrals 1 6.0 0.008 0.009 36.846 37.684 integrate_v_rspace 273 8.0 0.436 0.452 24.634 29.601 pw_transfer 6555 10.6 0.377 0.399 26.888 27.656 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.471 26.229 grid_integrate_task_list 273 9.0 20.542 25.977 20.542 25.977 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 23.017 23.769 compute_2c_integrals 1 7.0 0.002 0.002 19.297 19.298 rpa_ri_compute_en 1 5.0 0.019 0.021 18.897 19.090 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.713 19.034 mp2_eri_2c_integrate_gpw 1 9.0 2.368 2.420 18.711 19.031 cp_fm_cholesky_decompose 12 8.2 17.471 18.335 17.471 18.335 cholesky_decomp 1 7.0 0.000 0.000 16.395 17.252 fft3d_s 5443 13.4 16.133 16.574 16.154 16.594 ao_to_mo_and_store_B_mult_1 272 7.0 10.773 15.331 10.773 15.331 calculate_wavefunction 272 8.0 5.407 5.542 12.310 12.997 rpa_num_int 1 6.0 0.000 0.001 10.594 10.604 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.569 10.583 calc_mat_Q 8 8.0 0.000 0.000 9.449 9.528 contract_S_to_Q 8 9.0 0.000 0.000 8.870 8.946 calc_potential_gpw 544 9.5 0.004 0.005 8.248 8.714 parallel_gemm_fm 14 9.1 0.000 0.000 8.461 8.552 parallel_gemm_fm_cosma 14 10.1 8.461 8.552 8.461 8.552 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.141 8.382 potential_pw2rs 545 10.0 0.107 0.109 7.542 8.213 create_integ_mat 1 6.0 0.026 0.039 7.825 7.834 collocate_single_gaussian 272 10.0 0.039 0.041 7.399 7.592 array2fm 1 7.0 0.000 0.000 6.787 7.271 pw_scatter_s 2720 13.7 4.318 4.520 4.318 4.520 pw_gather_s 2722 13.2 3.451 3.878 3.451 3.878 array2fm_buffer_send 1 8.0 2.954 3.161 2.954 3.161 pw_poisson_solve 545 10.5 1.121 1.190 2.244 2.486 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.490422, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2801.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.037 421.386 421.389 farming_run 1 2.0 420.589 420.594 421.346 421.348 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.237025E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.034 216.806 216.807 qs_energies 1 2.0 0.000 0.000 216.576 216.590 scf_env_do_scf 1 3.0 0.000 0.000 114.487 114.488 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.512 113.519 rebuild_ks_matrix 4 6.0 0.000 0.000 113.511 113.518 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.511 113.518 hfx_ks_matrix 4 8.0 0.001 0.001 113.160 113.165 integrate_four_center 4 9.0 0.152 0.471 113.160 113.164 integrate_four_center_main 4 10.0 0.101 0.490 101.940 104.870 integrate_four_center_bin 269 11.0 101.839 104.763 101.839 104.763 mp2_main 1 3.0 0.000 0.000 101.785 101.799 mp2_gpw_main 1 4.0 0.049 0.072 100.874 100.892 init_scf_loop 1 4.0 0.000 0.000 96.606 96.606 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.884 75.015 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.821 54.952 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.422 46.356 integrate_v_rspace 95 8.0 0.393 0.560 27.904 32.647 pw_transfer 2240 10.6 0.145 0.168 29.475 30.056 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.495 29.090 mp2_ri_gpw_compute_en 1 5.0 0.062 0.079 26.840 28.977 grid_integrate_task_list 95 9.0 23.328 28.237 23.328 28.237 ao_to_mo_and_store_B_mult_1 91 7.0 10.728 27.938 10.728 27.938 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.252 26.848 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.836 1.880 25.064 25.075 get_2c_integrals 1 6.0 0.000 0.000 19.975 20.000 fft3d_s 1823 13.4 18.470 18.994 18.483 19.008 compute_2c_integrals 1 7.0 0.002 0.003 18.956 18.959 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.564 18.815 mp2_eri_2c_integrate_gpw 1 9.0 1.741 1.835 18.563 18.814 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.878 17.878 calculate_wavefunction 91 8.0 2.009 2.035 9.614 9.817 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.583 8.818 9.300 potential_pw2rs 186 10.0 0.033 0.034 8.405 8.872 local_gemm 172 8.0 8.260 8.744 8.260 8.744 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.525 8.035 8.541 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.111 8.422 calc_potential_gpw 182 9.5 0.002 0.002 7.874 8.100 collocate_single_gaussian 91 10.0 0.016 0.020 7.735 7.928 integrate_four_center_load 4 10.0 0.000 0.000 6.755 6.760 hfx_load_balance 1 11.0 0.000 0.000 6.755 6.760 mp_sync 37 10.5 3.190 6.627 3.190 6.627 mp_sendrecv_dm3 2068 8.0 6.066 6.567 6.066 6.567 mp2_ri_gpw_compute_en_ener 172 7.0 6.331 6.436 6.331 6.436 pw_gather_s 912 13.2 4.497 4.946 4.497 4.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.880119, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1497.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.984832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 53.244 53.246 qs_mol_dyn_low 1 2.0 0.003 0.003 52.947 52.956 qs_forces 11 3.9 0.002 0.003 52.884 52.885 qs_energies 11 4.9 0.002 0.002 51.413 51.428 scf_env_do_scf 11 5.9 0.001 0.001 45.358 45.358 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.335 43.336 qs_scf_new_mos 108 7.5 0.000 0.001 33.501 33.751 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.501 33.750 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.345 33.729 ot_scf_mini 108 9.5 0.002 0.002 31.843 32.024 multiply_cannon 2286 13.5 0.188 0.202 26.187 27.547 multiply_cannon_loop 2286 14.5 1.816 1.926 25.540 26.914 velocity_verlet 10 3.0 0.001 0.001 25.965 25.966 ot_mini 108 10.5 0.001 0.001 19.359 19.600 qs_ot_get_derivative 108 11.5 0.001 0.001 16.368 16.566 mp_waitall_1 245248 16.5 8.219 14.669 8.219 14.669 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 5.937 12.819 multiply_cannon_multrec 54864 15.5 3.678 5.779 7.800 11.180 qs_ot_get_p 119 10.4 0.001 0.003 7.926 8.186 rebuild_ks_matrix 119 8.3 0.000 0.000 7.737 7.869 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.737 7.868 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.815 6.928 mp_sum_l 7287 12.8 5.252 6.902 5.252 6.902 multiply_cannon_sync_h2d 54864 15.5 5.175 6.359 5.175 6.359 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.839 6.291 dbcsr_mm_accdrv_process 76910 16.1 1.830 2.889 4.036 5.750 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.232 5.360 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.179 5.214 init_scf_run 11 5.9 0.000 0.001 4.799 4.799 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.799 4.799 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.426 4.426 sum_up_and_integrate 119 10.3 0.001 0.002 4.369 4.375 integrate_v_rspace 119 11.3 0.002 0.002 4.358 4.365 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.237 4.237 cp_fm_redistribute_end 50 14.0 2.162 4.214 2.167 4.216 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.014 4.140 calculate_rho_elec 119 8.7 0.011 0.016 4.014 4.139 cp_fm_diag_elpa_base 50 14.0 2.044 4.104 2.047 4.110 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.913 3.129 apply_single 119 13.6 0.000 0.000 2.913 3.129 calculate_dm_sparse 119 9.5 0.000 0.000 2.984 3.126 jit_kernel_multiply 13 15.8 2.142 2.772 2.142 2.772 acc_transpose_blocks 54864 15.5 0.223 0.245 2.229 2.755 ot_diis_step 108 11.5 0.006 0.006 2.675 2.676 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.667 2.627 calculate_first_density_matrix 1 7.0 0.000 0.001 2.598 2.601 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.374 2.434 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.291 2.293 density_rs2pw 119 9.7 0.004 0.004 2.059 2.147 wfi_extrapolate 11 7.9 0.001 0.001 2.133 2.133 grid_integrate_task_list 119 12.3 2.009 2.103 2.009 2.103 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.992 2.056 init_scf_loop 11 6.9 0.000 0.000 2.007 2.007 mp_sum_d 4137 12.0 1.346 1.911 1.346 1.911 potential_pw2rs 119 12.3 0.004 0.004 1.760 1.769 pw_transfer 1439 11.6 0.051 0.057 1.593 1.662 make_m2s 4572 13.5 0.053 0.055 1.590 1.640 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.518 1.588 make_images 4572 14.5 0.133 0.139 1.508 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.494 1.520 mp_alltoall_d11v 2130 13.8 1.254 1.441 1.254 1.441 acc_transpose_blocks_sync 164592 16.5 1.200 1.438 1.200 1.438 grid_collocate_task_list 119 9.7 1.353 1.407 1.353 1.407 mp_waitany 12084 13.8 1.236 1.403 1.236 1.403 transfer_rs2pw 487 10.6 0.005 0.006 1.304 1.390 transfer_pw2rs 487 13.2 0.005 0.006 1.293 1.300 fft3d_ps 1201 14.6 0.369 0.475 1.231 1.297 fft_wrap_pw1pw2_140 487 13.2 0.107 0.114 1.178 1.246 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.748 1.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.246000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 490.401792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 896082. 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 0.090 0.118 38.793 38.796 qs_mol_dyn_low 1 2.0 0.003 0.006 38.336 38.343 qs_forces 11 3.9 0.003 0.006 38.273 38.274 qs_energies 11 4.9 0.006 0.016 36.615 36.620 scf_env_do_scf 11 5.9 0.005 0.013 31.580 31.581 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.154 29.154 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.045 22.397 qs_scf_new_mos 108 7.5 0.001 0.001 20.639 20.863 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.639 20.863 ot_scf_mini 108 9.5 0.002 0.003 19.734 19.895 multiply_cannon 2286 13.5 0.209 0.218 16.922 18.422 velocity_verlet 10 3.0 0.001 0.002 18.180 18.181 multiply_cannon_loop 2286 14.5 1.187 1.251 15.780 17.312 ot_mini 108 10.5 0.001 0.001 12.246 12.474 mp_waitall_1 200699 16.5 5.555 10.781 5.555 10.781 qs_ot_get_derivative 108 11.5 0.001 0.001 9.873 10.035 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.088 9.569 multiply_cannon_multrec 27432 15.5 1.833 4.261 6.140 8.850 rebuild_ks_matrix 119 8.3 0.000 0.000 6.705 6.844 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.705 6.843 dbcsr_mm_accdrv_process 47894 16.0 3.573 5.796 4.227 6.297 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.898 6.023 qs_ot_get_p 119 10.4 0.001 0.002 4.567 4.790 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.779 4.687 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.082 4.133 apply_single 119 13.6 0.000 0.000 3.082 4.133 mp_sum_l 7287 12.8 2.123 4.107 2.123 4.107 init_scf_run 11 5.9 0.000 0.001 3.805 3.806 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.805 3.806 sum_up_and_integrate 119 10.3 0.001 0.004 3.645 3.651 integrate_v_rspace 119 11.3 0.002 0.003 3.631 3.635 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.557 3.591 calculate_rho_elec 119 8.7 0.021 0.025 3.557 3.590 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.946 2.964 make_m2s 4572 13.5 0.052 0.054 2.550 2.877 make_images 4572 14.5 0.206 0.245 2.460 2.790 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.553 2.553 init_scf_loop 11 6.9 0.000 0.000 2.402 2.407 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.291 2.375 ot_diis_step 108 11.5 0.011 0.019 2.326 2.326 calculate_first_density_matrix 1 7.0 0.000 0.000 2.270 2.271 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.244 2.244 cp_fm_redistribute_end 50 14.0 1.137 2.220 1.140 2.223 multiply_cannon_sync_h2d 27432 15.5 1.701 2.189 1.701 2.189 cp_fm_diag_elpa_base 50 14.0 1.050 2.132 1.078 2.171 calculate_dm_sparse 119 9.5 0.000 0.001 2.046 2.125 acc_transpose_blocks 27432 15.5 0.115 0.121 1.625 2.007 grid_integrate_task_list 119 12.3 1.840 1.948 1.840 1.948 density_rs2pw 119 9.7 0.004 0.004 1.858 1.933 pw_transfer 1439 11.6 0.064 0.068 1.819 1.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.837 1.840 jit_kernel_multiply 9 16.4 0.594 1.782 0.594 1.782 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.728 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.649 1.689 make_images_data 4572 15.5 0.047 0.054 1.213 1.567 prepare_preconditioner 11 7.9 0.000 0.000 1.488 1.515 make_preconditioner 11 8.9 0.000 0.000 1.488 1.515 wfi_extrapolate 11 7.9 0.006 0.044 1.478 1.478 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.391 1.448 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.078 1.431 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.379 1.413 potential_pw2rs 119 12.3 0.006 0.006 1.400 1.407 fft3d_ps 1201 14.6 0.521 0.577 1.359 1.384 grid_collocate_task_list 119 9.7 1.290 1.356 1.290 1.356 mp_alltoall_d11v 2130 13.8 1.168 1.345 1.168 1.345 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.195 1.241 transfer_rs2pw 487 10.6 0.005 0.005 1.053 1.125 mp_sum_d 4137 12.0 0.560 1.038 0.560 1.038 mp_allgather_i34 2286 14.5 0.566 0.969 0.566 0.969 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.957 0.958 acc_transpose_blocks_kernels 27432 16.5 0.189 0.279 0.672 0.942 acc_transpose_blocks_sync 82296 16.5 0.813 0.939 0.813 0.939 transfer_pw2rs 487 13.2 0.004 0.006 0.874 0.880 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.864 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.796000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.363636, yerr=1.822722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 522.768384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 32.171 32.172 qs_mol_dyn_low 1 2.0 0.003 0.003 31.949 31.966 qs_forces 11 3.9 0.002 0.002 31.825 31.826 qs_energies 11 4.9 0.001 0.001 30.232 30.234 scf_env_do_scf 11 5.9 0.002 0.013 25.455 25.455 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.044 23.045 dbcsr_multiply_generic 2286 12.5 0.094 0.097 17.029 17.089 qs_scf_new_mos 108 7.5 0.001 0.001 15.369 15.383 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.368 15.383 velocity_verlet 10 3.0 0.010 0.014 15.099 15.100 ot_scf_mini 108 9.5 0.002 0.002 14.612 14.622 multiply_cannon 2286 13.5 0.195 0.201 13.859 14.525 multiply_cannon_loop 2286 14.5 0.857 0.917 13.094 13.780 ot_mini 108 10.5 0.001 0.001 9.029 9.041 qs_ot_get_derivative 108 11.5 0.001 0.001 7.542 7.552 multiply_cannon_multrec 18288 15.5 1.889 2.895 7.045 7.381 rebuild_ks_matrix 119 8.3 0.000 0.000 6.012 6.026 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.055 6.012 6.026 dbcsr_mm_accdrv_process 38222 16.0 5.001 5.775 5.061 5.840 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.257 5.268 mp_waitall_1 158411 16.6 2.699 3.781 2.699 3.781 init_scf_run 11 5.9 0.000 0.001 3.622 3.622 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.622 3.622 sum_up_and_integrate 119 10.3 0.001 0.001 3.530 3.535 integrate_v_rspace 119 11.3 0.003 0.003 3.517 3.524 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.884 3.419 qs_ot_get_p 119 10.4 0.001 0.001 3.353 3.379 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.249 3.253 calculate_rho_elec 119 8.7 0.031 0.032 3.249 3.252 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.494 2.422 calculate_first_density_matrix 1 7.0 0.000 0.001 2.409 2.410 init_scf_loop 11 6.9 0.000 0.000 2.393 2.395 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.942 2.338 apply_single 119 13.6 0.000 0.000 1.942 2.338 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.245 2.250 calculate_dm_sparse 119 9.5 0.000 0.001 2.003 2.011 make_m2s 4572 13.5 0.044 0.045 1.822 1.961 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.950 1.951 density_rs2pw 119 9.7 0.004 0.004 1.821 1.942 grid_integrate_task_list 119 12.3 1.798 1.898 1.798 1.898 make_images 4572 14.5 0.192 0.204 1.736 1.873 pw_transfer 1439 11.6 0.065 0.068 1.843 1.852 acc_transpose_blocks 18288 15.5 0.079 0.081 1.669 1.764 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.750 1.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.708 1.709 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.703 1.709 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.696 1.706 cp_fm_diag_elpa_base 50 14.0 1.673 1.686 1.694 1.705 prepare_preconditioner 11 7.9 0.000 0.000 1.639 1.641 make_preconditioner 11 8.9 0.000 0.000 1.639 1.641 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.500 1.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.530 1.533 ot_diis_step 108 11.5 0.011 0.011 1.472 1.472 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.412 1.423 mp_sum_l 7287 12.8 0.959 1.377 0.959 1.377 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.342 1.345 potential_pw2rs 119 12.3 0.007 0.008 1.333 1.336 fft3d_ps 1201 14.6 0.545 0.564 1.304 1.315 grid_collocate_task_list 119 9.7 1.233 1.306 1.233 1.306 multiply_cannon_sync_h2d 18288 15.5 1.058 1.239 1.058 1.239 wfi_extrapolate 11 7.9 0.001 0.001 1.161 1.161 transfer_rs2pw 487 10.6 0.005 0.005 0.970 1.080 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.937 0.955 make_images_data 4572 15.5 0.047 0.050 0.794 0.950 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.932 0.933 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.689 0.871 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 0.838 0.852 transfer_pw2rs 487 13.2 0.004 0.004 0.840 0.848 acc_transpose_blocks_sync 54864 16.5 0.732 0.831 0.732 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.827 mp_alltoall_d11v 2130 13.8 0.654 0.814 0.654 0.814 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.671 0.738 cp_fm_cholesky_invert 11 10.9 0.690 0.693 0.690 0.693 mp_alltoall_z22v 1201 16.6 0.611 0.678 0.611 0.678 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.453 0.658 mp_waitany 9880 13.7 0.568 0.653 0.568 0.653 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.172000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.998080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.106 36.537 36.538 qs_mol_dyn_low 1 2.0 0.003 0.003 36.171 36.178 qs_forces 11 3.9 0.002 0.003 36.103 36.104 qs_energies 11 4.9 0.002 0.002 34.405 34.409 scf_env_do_scf 11 5.9 0.001 0.001 29.299 29.301 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.894 25.894 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.041 20.145 velocity_verlet 10 3.0 0.004 0.014 18.289 18.291 qs_scf_new_mos 108 7.5 0.001 0.001 17.861 17.916 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.860 17.915 multiply_cannon 2286 13.5 0.219 0.226 16.402 16.899 ot_scf_mini 108 9.5 0.002 0.003 16.834 16.888 multiply_cannon_loop 2286 14.5 1.548 1.615 15.463 15.777 ot_mini 108 10.5 0.001 0.001 10.431 10.493 multiply_cannon_multrec 27432 15.5 2.471 3.259 9.125 9.441 qs_ot_get_derivative 108 11.5 0.001 0.001 8.581 8.635 dbcsr_mm_accdrv_process 47916 15.9 6.006 7.762 6.551 7.837 rebuild_ks_matrix 119 8.3 0.000 0.000 6.299 6.353 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.298 6.352 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.580 5.629 init_scf_run 11 5.9 0.000 0.001 3.784 3.784 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.784 3.784 qs_ot_get_p 119 10.4 0.009 0.015 3.615 3.689 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.039 3.487 sum_up_and_integrate 119 10.3 0.001 0.001 3.470 3.480 integrate_v_rspace 119 11.3 0.003 0.003 3.459 3.469 init_scf_loop 11 6.9 0.000 0.000 3.386 3.387 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.329 3.364 calculate_rho_elec 119 8.7 0.040 0.046 3.329 3.364 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.140 2.577 apply_single 119 13.6 0.000 0.000 2.140 2.577 acc_transpose_blocks 27432 15.5 0.120 0.123 2.430 2.530 prepare_preconditioner 11 7.9 0.000 0.000 2.521 2.528 make_preconditioner 11 8.9 0.000 0.000 2.521 2.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.131 2.458 calculate_first_density_matrix 1 7.0 0.000 0.001 2.397 2.399 make_m2s 4572 13.5 0.054 0.056 2.239 2.356 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.309 2.337 mp_waitall_1 137007 16.6 1.875 2.286 1.875 2.286 make_images 4572 14.5 0.273 0.335 2.132 2.246 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.197 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.173 2.182 pw_transfer 1439 11.6 0.064 0.069 1.985 2.021 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.970 1.971 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.894 1.933 density_rs2pw 119 9.7 0.004 0.004 1.798 1.904 grid_integrate_task_list 119 12.3 1.828 1.882 1.828 1.882 ot_diis_step 108 11.5 0.012 0.012 1.808 1.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.790 1.803 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.774 1.775 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.572 1.612 mp_sum_l 7287 12.8 1.047 1.602 1.047 1.602 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.534 1.542 cp_fm_diag_elpa_base 50 14.0 1.501 1.516 1.532 1.541 acc_transpose_blocks_sync 82296 16.5 1.449 1.537 1.449 1.537 fft3d_ps 1201 14.6 0.591 0.653 1.382 1.406 multiply_cannon_metrocomm3 27432 15.5 0.039 0.041 0.823 1.381 jit_kernel_multiply 7 15.7 0.479 1.341 0.479 1.341 wfi_extrapolate 11 7.9 0.001 0.001 1.335 1.335 grid_collocate_task_list 119 9.7 1.247 1.333 1.247 1.333 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.283 1.300 potential_pw2rs 119 12.3 0.009 0.010 1.275 1.277 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.240 1.249 cp_fm_upper_to_full 72 14.2 0.815 1.159 0.815 1.159 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.062 1.063 dbcsr_complete_redistribute 329 12.2 0.124 0.151 0.777 1.054 transfer_rs2pw 487 10.6 0.004 0.005 0.876 0.984 make_images_data 4572 15.5 0.048 0.052 0.829 0.946 hybrid_alltoall_any 4725 16.4 0.067 0.157 0.709 0.897 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.874 acc_transpose_blocks_kernels 27432 16.5 0.272 0.282 0.830 0.849 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.272 0.841 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.562 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.811 mp_alltoall_d11v 2130 13.8 0.716 0.803 0.716 0.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.538000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.272727, yerr=2.894852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 617.771008E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.044 28.535 28.546 qs_mol_dyn_low 1 2.0 0.003 0.003 28.241 28.249 qs_forces 11 3.9 0.002 0.003 28.183 28.184 qs_energies 11 4.9 0.010 0.019 26.483 26.485 scf_env_do_scf 11 5.9 0.001 0.004 21.665 21.665 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.224 19.224 velocity_verlet 10 3.0 0.002 0.002 14.575 14.578 dbcsr_multiply_generic 2286 12.5 0.093 0.096 13.036 13.110 qs_scf_new_mos 108 7.5 0.001 0.001 11.585 11.608 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.585 11.607 multiply_cannon 2286 13.5 0.223 0.230 10.501 10.962 ot_scf_mini 108 9.5 0.002 0.002 10.886 10.910 multiply_cannon_loop 2286 14.5 0.644 0.671 9.587 9.765 multiply_cannon_multrec 9144 15.5 1.693 1.877 6.060 6.322 ot_mini 108 10.5 0.001 0.001 6.254 6.284 rebuild_ks_matrix 119 8.3 0.000 0.000 5.778 5.803 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.777 5.803 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.140 5.162 qs_ot_get_derivative 108 11.5 0.001 0.001 4.976 5.000 dbcsr_mm_accdrv_process 12550 15.8 3.619 4.307 4.257 4.363 sum_up_and_integrate 119 10.3 0.001 0.001 3.428 3.431 integrate_v_rspace 119 11.3 0.003 0.003 3.418 3.421 init_scf_run 11 5.9 0.000 0.001 3.358 3.358 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.358 3.358 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.337 3.344 calculate_rho_elec 119 8.7 0.060 0.061 3.336 3.344 qs_ot_get_p 119 10.4 0.004 0.009 2.768 2.803 init_scf_loop 11 6.9 0.000 0.000 2.421 2.423 calculate_first_density_matrix 1 7.0 0.000 0.001 2.228 2.229 pw_transfer 1439 11.6 0.065 0.067 2.117 2.127 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.024 2.037 make_m2s 4572 13.5 0.033 0.035 1.760 1.921 grid_integrate_task_list 119 12.3 1.862 1.920 1.862 1.920 calculate_dm_sparse 119 9.5 0.000 0.000 1.871 1.889 mp_waitall_1 115863 16.7 1.328 1.883 1.328 1.883 make_images 4572 14.5 0.269 0.299 1.670 1.831 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.805 1.808 density_rs2pw 119 9.7 0.003 0.004 1.743 1.808 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.727 1.746 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.679 1.692 prepare_preconditioner 11 7.9 0.000 0.000 1.684 1.688 make_preconditioner 11 8.9 0.000 0.000 1.684 1.688 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.572 1.599 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.564 1.564 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.536 1.537 acc_transpose_blocks 9144 15.5 0.043 0.044 1.423 1.464 jit_kernel_multiply 8 15.7 0.599 1.427 0.599 1.427 grid_collocate_task_list 119 9.7 1.296 1.397 1.296 1.397 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.376 1.386 fft3d_ps 1201 14.6 0.653 0.666 1.375 1.384 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.330 1.344 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.300 1.307 cp_fm_diag_elpa_base 50 14.0 1.274 1.290 1.298 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.268 1.272 ot_diis_step 108 11.5 0.013 0.013 1.265 1.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.257 apply_single 119 13.6 0.000 0.000 1.216 1.257 potential_pw2rs 119 12.3 0.010 0.010 1.245 1.246 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.226 1.227 wfi_extrapolate 11 7.9 0.001 0.001 1.079 1.079 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.743 0.978 make_images_data 4572 15.5 0.042 0.046 0.760 0.940 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.923 mp_alltoall_d11v 2130 13.8 0.750 0.882 0.750 0.882 cp_fm_cholesky_invert 11 10.9 0.838 0.840 0.838 0.840 transfer_rs2pw 487 10.6 0.004 0.004 0.750 0.824 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.813 0.820 acc_transpose_blocks_sync 27432 16.5 0.724 0.765 0.724 0.765 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.761 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.676 0.720 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.347 0.676 mp_allgather_i34 2286 14.5 0.232 0.648 0.232 0.648 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.641 0.647 transfer_pw2rs 487 13.2 0.003 0.004 0.643 0.645 mp_alltoall_z22v 1201 16.6 0.596 0.627 0.596 0.627 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.546000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.636364, yerr=4.140108 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 782.553088E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.042 43.295 43.296 qs_mol_dyn_low 1 2.0 0.003 0.003 43.028 43.040 qs_forces 11 3.9 0.002 0.003 42.966 42.968 qs_energies 11 4.9 0.002 0.002 40.944 40.949 scf_env_do_scf 11 5.9 0.001 0.001 35.257 35.258 scf_env_do_scf_inner_loop 108 6.5 0.005 0.015 27.394 27.395 velocity_verlet 10 3.0 0.006 0.009 24.347 24.353 dbcsr_multiply_generic 2286 12.5 0.102 0.104 19.137 19.394 qs_scf_new_mos 108 7.5 0.001 0.001 17.540 17.645 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.540 17.644 ot_scf_mini 108 9.5 0.002 0.002 16.362 16.472 multiply_cannon 2286 13.5 0.300 0.309 15.098 16.064 multiply_cannon_loop 2286 14.5 0.865 0.887 13.824 14.826 ot_mini 108 10.5 0.001 0.001 9.928 10.049 multiply_cannon_multrec 9144 15.5 3.390 4.836 8.801 8.929 qs_ot_get_derivative 108 11.5 0.001 0.001 7.862 7.974 init_scf_loop 11 6.9 0.000 0.000 7.838 7.840 rebuild_ks_matrix 119 8.3 0.000 0.000 7.369 7.517 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.369 7.517 prepare_preconditioner 11 7.9 0.000 0.000 6.845 6.859 make_preconditioner 11 8.9 0.000 0.000 6.845 6.859 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.635 6.767 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.413 6.739 dbcsr_mm_accdrv_process 12550 15.8 4.702 6.482 5.272 6.608 cp_fm_upper_to_full 72 14.2 3.173 4.551 3.173 4.551 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.307 4.312 calculate_rho_elec 119 8.7 0.117 0.120 4.306 4.312 sum_up_and_integrate 119 10.3 0.001 0.001 3.907 3.912 integrate_v_rspace 119 11.3 0.004 0.004 3.897 3.903 init_scf_run 11 5.9 0.000 0.001 3.626 3.626 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.626 3.626 qs_ot_get_p 119 10.4 0.001 0.001 3.479 3.623 mp_waitall_1 94719 16.7 2.328 3.405 2.328 3.405 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.895 3.300 pw_transfer 1439 11.6 0.068 0.069 3.084 3.090 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.986 2.992 dbcsr_complete_redistribute 329 12.2 0.294 0.299 1.989 2.800 make_m2s 4572 13.5 0.037 0.038 2.415 2.576 fft_wrap_pw1pw2_140 487 13.2 0.649 0.655 2.541 2.548 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.206 2.534 apply_single 119 13.6 0.000 0.000 2.206 2.534 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.502 make_images 4572 14.5 0.352 0.384 2.295 2.456 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.364 2.339 density_rs2pw 119 9.7 0.003 0.004 2.315 2.334 calculate_dm_sparse 119 9.5 0.000 0.000 2.285 2.301 mp_alltoall_i22 627 13.8 1.407 2.265 1.407 2.265 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.426 2.236 grid_integrate_task_list 119 12.3 2.083 2.160 2.083 2.160 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.105 2.160 calculate_first_density_matrix 1 7.0 0.000 0.000 2.151 2.153 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.057 2.060 ot_diis_step 108 11.5 0.014 0.015 2.039 2.039 mp_sum_l 7287 12.8 1.077 1.914 1.077 1.914 fft3d_ps 1201 14.6 0.867 0.885 1.907 1.910 acc_transpose_blocks 9144 15.5 0.044 0.045 1.817 1.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.844 1.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.772 1.774 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.694 1.695 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.579 1.625 grid_collocate_task_list 119 9.7 1.524 1.573 1.524 1.573 cp_fm_cholesky_invert 11 10.9 1.466 1.469 1.466 1.469 wfi_extrapolate 11 7.9 0.001 0.001 1.413 1.413 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.396 1.397 cp_fm_diag_elpa_base 50 14.0 1.249 1.300 1.394 1.395 potential_pw2rs 119 12.3 0.014 0.014 1.375 1.376 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.331 1.353 hybrid_alltoall_any 4725 16.4 0.091 0.152 1.088 1.297 mp_alltoall_d11v 2130 13.8 1.240 1.285 1.240 1.285 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.275 1.281 make_images_data 4572 15.5 0.046 0.050 1.025 1.230 acc_transpose_blocks_sync 27432 16.5 1.105 1.145 1.105 1.145 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.106 1.128 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.991 1.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.944 0.960 qs_create_task_list 11 7.9 0.000 0.000 0.928 0.941 generate_qs_task_list 11 8.9 0.370 0.390 0.928 0.941 mp_alltoall_z22v 1201 16.6 0.903 0.928 0.903 0.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.296000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=732.090909, yerr=18.007345 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 506.478592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 81.715 81.716 qs_mol_dyn_low 1 2.0 0.003 0.003 81.412 81.421 qs_forces 11 3.9 0.003 0.004 81.339 81.340 qs_energies 11 4.9 0.001 0.002 78.466 78.479 scf_env_do_scf 11 5.9 0.000 0.001 69.625 69.627 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.146 64.146 dbcsr_multiply_generic 2055 12.4 0.111 0.120 50.883 51.191 qs_scf_new_mos 99 7.5 0.000 0.001 47.084 47.188 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.084 47.188 ot_scf_mini 99 9.5 0.002 0.002 44.730 44.824 multiply_cannon 2055 13.4 0.182 0.187 42.567 43.533 velocity_verlet 10 3.0 0.001 0.003 43.080 43.080 multiply_cannon_loop 2055 14.4 1.786 1.820 41.568 42.408 ot_mini 99 10.5 0.001 0.001 26.665 26.755 qs_ot_get_derivative 99 11.5 0.001 0.001 19.850 19.932 multiply_cannon_multrec 49320 15.4 11.412 12.220 17.455 18.476 rebuild_ks_matrix 110 8.3 0.000 0.001 14.584 14.669 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.583 14.668 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.777 12.860 mp_waitall_1 220248 16.4 10.680 11.540 10.680 11.540 multiply_cannon_sync_h2d 49320 15.4 9.651 10.190 9.651 10.190 qs_ot_get_p 110 10.4 0.001 0.001 9.626 9.769 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.731 8.292 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.261 7.848 apply_single 110 13.6 0.000 0.000 7.261 7.848 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.452 7.566 sum_up_and_integrate 110 10.3 0.001 0.003 7.027 7.039 integrate_v_rspace 110 11.3 0.002 0.003 7.001 7.020 init_scf_run 11 5.9 0.000 0.001 6.807 6.808 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.807 6.808 ot_diis_step 99 11.5 0.006 0.006 6.590 6.590 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.543 6.558 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.338 6.487 calculate_rho_elec 110 8.6 0.020 0.024 6.337 6.487 dbcsr_mm_accdrv_process 87628 16.1 3.058 3.142 5.912 6.194 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.764 5.764 init_scf_loop 11 6.9 0.000 0.000 5.451 5.452 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.297 5.338 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.171 5.197 cp_fm_diag_elpa_base 48 14.0 5.157 5.183 5.169 5.195 mp_sum_l 6594 12.7 3.902 4.523 3.902 4.523 make_m2s 4110 13.4 0.061 0.066 3.951 4.066 wfi_extrapolate 11 7.9 0.001 0.001 4.026 4.026 make_images 4110 14.4 0.178 0.192 3.855 3.972 calculate_dm_sparse 110 9.5 0.001 0.001 3.715 3.819 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.583 3.587 density_rs2pw 110 9.6 0.004 0.004 3.336 3.512 grid_integrate_task_list 110 12.3 3.264 3.429 3.264 3.429 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.344 3.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.242 3.275 prepare_preconditioner 11 7.9 0.000 0.000 3.239 3.256 make_preconditioner 11 8.9 0.000 0.000 3.239 3.256 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.144 3.191 pw_transfer 1331 11.6 0.055 0.072 3.091 3.171 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.047 3.088 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.003 3.087 calculate_first_density_matrix 1 7.0 0.000 0.000 2.671 2.687 jit_kernel_multiply 13 15.9 2.574 2.661 2.574 2.661 fft_wrap_pw1pw2_140 451 13.1 0.309 0.336 2.552 2.652 potential_pw2rs 110 12.3 0.005 0.006 2.617 2.642 acc_transpose_blocks 49320 15.4 0.211 0.217 2.570 2.630 mp_alltoall_d11v 2046 13.8 2.093 2.534 2.093 2.534 fft3d_ps 1111 14.6 0.798 0.893 2.297 2.354 grid_collocate_task_list 110 9.6 2.158 2.272 2.158 2.272 mp_waitany 14300 13.8 1.796 2.068 1.796 2.068 transfer_rs2pw 451 10.6 0.005 0.006 1.875 2.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.951 make_images_data 4110 15.4 0.043 0.048 1.769 1.902 mp_sum_d 3891 11.9 1.355 1.873 1.355 1.873 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.537 1.815 cp_fm_cholesky_invert 11 10.9 1.809 1.812 1.809 1.812 transfer_pw2rs 451 13.1 0.006 0.007 1.776 1.789 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.716000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.818182, yerr=3.995866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 590.979072E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3295319. 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 0.054 0.106 68.037 68.038 qs_mol_dyn_low 1 2.0 0.003 0.003 67.678 67.690 qs_forces 11 3.9 0.003 0.003 67.606 67.607 qs_energies 11 4.9 0.001 0.002 64.296 64.299 scf_env_do_scf 11 5.9 0.001 0.001 55.858 55.862 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.415 48.416 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.673 37.834 velocity_verlet 10 3.0 0.001 0.002 35.630 35.631 qs_scf_new_mos 99 7.5 0.001 0.001 32.759 32.885 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.758 32.884 multiply_cannon 2055 13.4 0.225 0.244 31.217 32.340 ot_scf_mini 99 9.5 0.003 0.003 31.095 31.215 multiply_cannon_loop 2055 14.4 1.158 1.188 29.962 30.726 ot_mini 99 10.5 0.001 0.001 18.204 18.332 multiply_cannon_multrec 24660 15.4 7.000 8.652 14.049 15.570 rebuild_ks_matrix 110 8.3 0.000 0.000 13.400 13.480 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.399 13.479 qs_ot_get_derivative 99 11.5 0.001 0.001 12.450 12.578 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.779 11.845 mp_waitall_1 176588 16.5 7.357 10.070 7.357 10.070 multiply_cannon_metrocomm3 24660 15.4 0.073 0.074 5.067 7.966 multiply_cannon_sync_h2d 24660 15.4 6.353 7.568 6.353 7.568 init_scf_loop 11 6.9 0.000 0.000 7.410 7.411 dbcsr_mm_accdrv_process 52282 16.1 5.662 6.376 6.878 7.213 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.448 7.137 apply_single 110 13.6 0.000 0.001 6.448 7.137 qs_ot_get_p 110 10.4 0.001 0.001 6.275 6.445 sum_up_and_integrate 110 10.3 0.001 0.002 6.279 6.292 integrate_v_rspace 110 11.3 0.002 0.003 6.252 6.264 init_scf_run 11 5.9 0.000 0.001 6.093 6.094 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.093 6.093 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.797 5.805 calculate_rho_elec 110 8.6 0.039 0.047 5.796 5.804 ot_diis_step 99 11.5 0.010 0.010 5.707 5.707 prepare_preconditioner 11 7.9 0.000 0.000 5.425 5.453 make_preconditioner 11 8.9 0.000 0.000 5.425 5.453 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.676 5.372 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.047 5.196 make_m2s 4110 13.4 0.057 0.059 4.125 4.587 make_images 4110 14.4 0.406 0.470 4.013 4.469 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.402 4.423 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.946 3.946 pw_transfer 1331 11.6 0.066 0.076 3.487 3.620 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.380 3.517 wfi_extrapolate 11 7.9 0.001 0.001 3.467 3.467 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.395 3.408 cp_fm_diag_elpa_base 48 14.0 3.349 3.366 3.392 3.406 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.244 3.305 grid_integrate_task_list 110 12.3 3.170 3.294 3.170 3.294 density_rs2pw 110 9.6 0.004 0.004 3.052 3.281 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.219 3.221 fft_wrap_pw1pw2_140 451 13.1 0.356 0.378 2.913 3.058 calculate_dm_sparse 110 9.5 0.001 0.001 3.004 3.038 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.887 2.937 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.983 2.800 make_images_data 4110 15.4 0.048 0.053 2.240 2.728 fft3d_ps 1111 14.6 1.115 1.348 2.474 2.627 calculate_first_density_matrix 1 7.0 0.003 0.027 2.530 2.532 cp_fm_cholesky_invert 11 10.9 2.491 2.497 2.491 2.497 mp_sum_l 6594 12.7 1.743 2.379 1.743 2.379 grid_collocate_task_list 110 9.6 2.172 2.328 2.172 2.328 potential_pw2rs 110 12.3 0.008 0.009 2.160 2.167 acc_transpose_blocks 24660 15.4 0.116 0.119 1.974 2.007 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.954 1.974 mp_alltoall_d11v 2046 13.8 1.713 1.904 1.713 1.904 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.842 1.843 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.775 1.792 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.776 1.713 transfer_rs2pw 451 10.6 0.006 0.007 1.377 1.679 mp_irecv_dv 57340 16.2 0.650 1.590 0.650 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.575 1.586 mp_allgather_i34 2055 14.4 0.563 1.574 0.563 1.574 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.492 mp_waitany 10164 13.8 1.146 1.454 1.146 1.454 jit_kernel_multiply 8 16.4 0.851 1.453 0.851 1.453 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.038000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.181818, yerr=5.982067 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.243072E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.053 59.074 59.075 qs_mol_dyn_low 1 2.0 0.003 0.003 58.721 58.731 qs_forces 11 3.9 0.003 0.004 58.651 58.652 qs_energies 11 4.9 0.001 0.002 55.500 55.503 scf_env_do_scf 11 5.9 0.000 0.001 47.655 47.655 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.285 39.285 velocity_verlet 10 3.0 0.001 0.002 32.095 32.096 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.536 28.802 qs_scf_new_mos 99 7.5 0.001 0.001 24.802 24.892 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.802 24.892 multiply_cannon 2055 13.4 0.214 0.223 22.536 23.721 ot_scf_mini 99 9.5 0.002 0.002 23.562 23.672 multiply_cannon_loop 2055 14.4 0.812 0.842 21.384 22.371 ot_mini 99 10.5 0.001 0.001 13.571 13.680 rebuild_ks_matrix 110 8.3 0.000 0.000 12.021 12.162 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.021 12.162 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.574 10.700 multiply_cannon_multrec 16440 15.4 3.621 4.502 9.845 10.680 mp_waitall_1 139946 16.5 6.833 9.997 6.833 9.997 qs_ot_get_derivative 99 11.5 0.001 0.001 9.172 9.286 init_scf_loop 11 6.9 0.000 0.000 8.338 8.338 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.429 7.421 prepare_preconditioner 11 7.9 0.000 0.000 6.590 6.610 make_preconditioner 11 8.9 0.000 0.000 6.590 6.610 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.933 6.279 dbcsr_mm_accdrv_process 34862 16.1 5.413 5.815 6.070 6.230 sum_up_and_integrate 110 10.3 0.001 0.002 6.198 6.212 integrate_v_rspace 110 11.3 0.003 0.003 6.172 6.186 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.681 5.695 calculate_rho_elec 110 8.6 0.059 0.059 5.680 5.694 init_scf_run 11 5.9 0.000 0.001 5.477 5.477 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.477 5.477 qs_ot_get_p 110 10.4 0.002 0.005 5.206 5.364 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.825 5.354 apply_single 110 13.6 0.000 0.000 4.825 5.353 make_m2s 4110 13.4 0.049 0.052 4.023 4.390 ot_diis_step 99 11.5 0.011 0.011 4.370 4.370 make_images 4110 14.4 0.397 0.522 3.906 4.270 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.214 3.786 multiply_cannon_sync_h2d 16440 15.4 3.225 3.742 3.225 3.742 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.599 3.604 pw_transfer 1331 11.6 0.065 0.071 3.560 3.571 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.454 3.468 grid_integrate_task_list 110 12.3 3.209 3.430 3.209 3.430 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.236 3.236 density_rs2pw 110 9.6 0.004 0.004 2.915 3.093 fft_wrap_pw1pw2_140 451 13.1 0.457 0.465 3.006 3.023 wfi_extrapolate 11 7.9 0.001 0.001 2.942 2.942 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.788 2.789 hybrid_alltoall_any 4261 16.3 0.108 0.378 1.992 2.734 make_images_data 4110 15.4 0.046 0.050 2.244 2.731 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.718 2.726 cp_fm_diag_elpa_base 48 14.0 2.653 2.683 2.716 2.725 calculate_dm_sparse 110 9.5 0.001 0.001 2.542 2.573 cp_fm_cholesky_invert 11 10.9 2.539 2.545 2.539 2.545 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.910 2.515 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.458 2.511 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.403 2.464 calculate_first_density_matrix 1 7.0 0.000 0.000 2.445 2.446 grid_collocate_task_list 110 9.6 2.229 2.406 2.229 2.406 fft3d_ps 1111 14.6 1.122 1.149 2.373 2.384 mp_irecv_dv 48980 15.7 0.834 2.382 0.834 2.382 mp_sum_l 6594 12.7 1.357 2.084 1.357 2.084 potential_pw2rs 110 12.3 0.011 0.011 2.043 2.046 mp_alltoall_d11v 2046 13.8 1.714 2.039 1.714 2.039 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.943 1.943 dbcsr_complete_redistribute 325 12.2 0.331 0.380 1.405 1.857 cp_fm_upper_to_full 70 14.2 1.399 1.760 1.399 1.760 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.655 acc_transpose_blocks 16440 15.4 0.079 0.081 1.577 1.650 cp_fm_cholesky_decompose 22 10.9 1.536 1.559 1.536 1.559 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.484 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.453 mp_allgather_i34 2055 14.4 0.465 1.420 0.465 1.420 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.952 1.392 transfer_rs2pw 451 10.6 0.005 0.006 1.219 1.392 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.271 mp_waitany 17072 13.8 1.097 1.254 1.097 1.254 rs_gather_matrices 110 12.3 0.136 0.151 0.870 1.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.075000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.909091, yerr=9.169659 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 742.658048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 65.037 65.038 qs_mol_dyn_low 1 2.0 0.003 0.003 64.818 64.828 qs_forces 11 3.9 0.003 0.003 64.743 64.744 qs_energies 11 4.9 0.003 0.012 61.364 61.368 scf_env_do_scf 11 5.9 0.001 0.001 53.173 53.176 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.605 41.605 velocity_verlet 10 3.0 0.002 0.003 36.996 36.998 dbcsr_multiply_generic 2055 12.4 0.117 0.121 30.411 30.607 qs_scf_new_mos 99 7.5 0.001 0.001 27.104 27.210 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.104 27.210 ot_scf_mini 99 9.5 0.002 0.003 25.422 25.522 multiply_cannon 2055 13.4 0.238 0.252 23.419 24.690 multiply_cannon_loop 2055 14.4 1.411 1.492 22.074 22.683 ot_mini 99 10.5 0.001 0.001 14.648 14.775 multiply_cannon_multrec 24660 15.4 4.066 6.718 12.979 14.113 rebuild_ks_matrix 110 8.3 0.000 0.000 11.901 12.008 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.900 12.007 init_scf_loop 11 6.9 0.000 0.000 11.530 11.531 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.510 10.602 qs_ot_get_derivative 99 11.5 0.001 0.001 10.487 10.592 prepare_preconditioner 11 7.9 0.000 0.000 9.808 9.827 make_preconditioner 11 8.9 0.000 0.000 9.808 9.827 dbcsr_mm_accdrv_process 52304 16.0 7.667 9.131 8.755 9.755 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.001 9.510 sum_up_and_integrate 110 10.3 0.001 0.002 6.179 6.191 integrate_v_rspace 110 11.3 0.003 0.003 6.153 6.164 mp_waitall_1 121746 16.5 4.139 6.159 4.139 6.159 qs_ot_get_p 110 10.4 0.001 0.001 5.675 5.823 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.772 5.781 calculate_rho_elec 110 8.6 0.078 0.081 5.772 5.781 make_m2s 4110 13.4 0.059 0.061 5.401 5.689 make_images 4110 14.4 0.574 0.693 5.260 5.542 init_scf_run 11 5.9 0.000 0.001 5.463 5.464 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.463 5.464 cp_fm_upper_to_full 70 14.2 3.405 4.962 3.405 4.962 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.035 4.125 apply_single 110 13.6 0.000 0.000 4.035 4.124 ot_diis_step 99 11.5 0.011 0.012 4.122 4.122 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.847 3.863 dbcsr_complete_redistribute 325 12.2 0.416 0.466 2.639 3.823 pw_transfer 1331 11.6 0.065 0.073 3.709 3.742 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.602 3.639 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.479 3.526 grid_integrate_task_list 110 12.3 3.261 3.446 3.261 3.446 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.315 3.372 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.145 3.317 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.291 3.291 fft_wrap_pw1pw2_140 451 13.1 0.530 0.544 3.117 3.152 density_rs2pw 110 9.6 0.004 0.004 2.914 3.080 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.336 3.050 calculate_dm_sparse 110 9.5 0.001 0.001 2.990 3.023 make_images_data 4110 15.4 0.048 0.053 2.698 3.011 hybrid_alltoall_any 4261 16.3 0.123 0.459 2.261 2.962 wfi_extrapolate 11 7.9 0.001 0.001 2.959 2.959 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.799 2.957 mp_alltoall_i22 605 13.7 1.701 2.913 1.701 2.913 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.782 2.784 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.772 2.782 cp_fm_diag_elpa_base 48 14.0 2.626 2.686 2.770 2.780 cp_fm_cholesky_invert 11 10.9 2.586 2.593 2.586 2.593 acc_transpose_blocks 24660 15.4 0.113 0.115 2.414 2.593 multiply_cannon_sync_h2d 24660 15.4 2.395 2.564 2.395 2.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.431 2.472 fft3d_ps 1111 14.6 1.151 1.190 2.429 2.446 grid_collocate_task_list 110 9.6 2.261 2.431 2.261 2.431 calculate_first_density_matrix 1 7.0 0.000 0.001 2.410 2.413 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.227 2.227 mp_alltoall_d11v 2046 13.8 1.722 2.088 1.722 2.088 potential_pw2rs 110 12.3 0.013 0.013 1.999 2.003 jit_kernel_multiply 11 15.8 0.744 1.987 0.744 1.987 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.782 1.813 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.708 cp_fm_cholesky_decompose 22 10.9 1.614 1.656 1.614 1.656 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.630 1.643 mp_allgather_i34 2055 14.4 0.434 1.583 0.434 1.583 acc_transpose_blocks_sync 73980 16.4 1.397 1.565 1.397 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.548 1.563 mp_sum_l 6594 12.7 0.901 1.507 0.901 1.507 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.859 1.496 mp_irecv_dv 62702 16.1 0.756 1.414 0.756 1.414 transfer_rs2pw 451 10.6 0.005 0.006 1.140 1.311 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.038000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.545455, yerr=11.364727 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 862.474240E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 54.516 54.517 qs_mol_dyn_low 1 2.0 0.003 0.003 54.264 54.273 qs_forces 11 3.9 0.003 0.003 54.198 54.199 qs_energies 11 4.9 0.001 0.002 50.555 50.561 scf_env_do_scf 11 5.9 0.000 0.001 42.371 42.372 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 34.767 34.768 velocity_verlet 10 3.0 0.002 0.002 30.888 30.891 dbcsr_multiply_generic 2055 12.4 0.106 0.108 22.886 23.035 qs_scf_new_mos 99 7.5 0.001 0.001 20.222 20.274 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.221 20.274 ot_scf_mini 99 9.5 0.002 0.002 18.975 18.994 multiply_cannon 2055 13.4 0.239 0.255 17.433 18.707 multiply_cannon_loop 2055 14.4 0.604 0.638 16.171 16.474 rebuild_ks_matrix 110 8.3 0.000 0.000 11.621 11.632 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.621 11.631 ot_mini 99 10.5 0.001 0.001 10.446 10.458 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.336 10.347 multiply_cannon_multrec 8220 15.4 3.160 4.358 7.552 8.414 init_scf_loop 11 6.9 0.000 0.000 7.560 7.560 mp_waitall_1 103326 16.6 5.632 7.253 5.632 7.253 qs_ot_get_derivative 99 11.5 0.001 0.001 6.775 6.792 sum_up_and_integrate 110 10.3 0.001 0.002 6.181 6.192 integrate_v_rspace 110 11.3 0.003 0.003 6.155 6.165 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.980 5.991 calculate_rho_elec 110 8.6 0.113 0.113 5.979 5.991 prepare_preconditioner 11 7.9 0.000 0.000 5.918 5.923 make_preconditioner 11 8.9 0.000 0.000 5.918 5.923 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.501 5.569 dbcsr_mm_accdrv_process 17442 15.9 3.090 4.114 4.254 5.161 init_scf_run 11 5.9 0.000 0.001 5.049 5.049 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.049 5.049 qs_ot_get_p 110 10.4 0.001 0.002 4.597 4.615 make_m2s 4110 13.4 0.038 0.039 4.152 4.440 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.919 4.348 make_images 4110 14.4 0.645 0.708 4.022 4.311 pw_transfer 1331 11.6 0.065 0.070 4.048 4.062 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.940 3.957 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.708 3.759 apply_single 110 13.6 0.000 0.000 3.708 3.759 ot_diis_step 99 11.5 0.012 0.012 3.649 3.649 grid_integrate_task_list 110 12.3 3.388 3.483 3.388 3.483 fft_wrap_pw1pw2_140 451 13.1 0.719 0.735 3.432 3.449 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.259 3.263 density_rs2pw 110 9.6 0.004 0.004 2.985 3.106 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.924 2.924 cp_fm_cholesky_invert 11 10.9 2.836 2.840 2.836 2.840 wfi_extrapolate 11 7.9 0.001 0.001 2.692 2.692 hybrid_alltoall_any 4261 16.3 0.200 0.851 2.237 2.691 qs_energies_init_hamiltonians 11 5.9 0.013 0.025 2.686 2.686 make_images_data 4110 15.4 0.040 0.046 2.230 2.595 calculate_dm_sparse 110 9.5 0.001 0.001 2.474 2.526 multiply_cannon_sync_h2d 8220 15.4 2.383 2.509 2.383 2.509 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.500 2.501 fft3d_ps 1111 14.6 1.288 1.297 2.485 2.496 grid_collocate_task_list 110 9.6 2.363 2.441 2.363 2.441 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.386 2.393 cp_fm_diag_elpa_base 48 14.0 2.331 2.354 2.384 2.392 calculate_first_density_matrix 1 7.0 0.000 0.000 2.260 2.262 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.129 2.145 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.980 2.010 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.001 potential_pw2rs 110 12.3 0.015 0.016 1.991 1.994 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.891 1.901 mp_alltoall_d11v 2046 13.8 1.648 1.726 1.648 1.726 cp_fm_cholesky_decompose 22 10.9 1.662 1.674 1.662 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.612 1.615 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.497 1.611 mp_allgather_i34 2055 14.4 0.434 1.511 0.434 1.511 dbcsr_complete_redistribute 325 12.2 0.573 0.595 1.406 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.430 1.446 acc_transpose_blocks 8220 15.4 0.041 0.042 1.366 1.420 jit_kernel_multiply 8 15.6 0.848 1.378 0.848 1.378 qs_create_task_list 11 7.9 0.000 0.000 1.201 1.297 generate_qs_task_list 11 8.9 0.375 0.441 1.201 1.297 transfer_rs2pw 451 10.6 0.005 0.005 1.090 1.230 mp_waitany 9240 13.8 1.043 1.199 1.043 1.199 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.145 1.167 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.517000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=808.454545, yerr=8.478461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 1.379815E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.059 87.055 87.062 qs_mol_dyn_low 1 2.0 0.003 0.003 86.722 86.733 qs_forces 11 3.9 0.003 0.003 86.651 86.652 qs_energies 11 4.9 0.001 0.001 82.525 82.526 scf_env_do_scf 11 5.9 0.000 0.001 72.479 72.479 velocity_verlet 10 3.0 0.002 0.002 55.591 55.598 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.432 44.434 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.360 29.484 init_scf_loop 11 6.9 0.000 0.000 27.978 27.983 qs_scf_new_mos 99 7.5 0.001 0.001 26.588 26.631 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.588 26.630 prepare_preconditioner 11 7.9 0.000 0.000 25.967 25.975 make_preconditioner 11 8.9 0.000 0.000 25.966 25.975 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.361 25.452 ot_scf_mini 99 9.5 0.002 0.002 24.797 24.838 multiply_cannon 2055 13.4 0.331 0.347 22.425 23.029 multiply_cannon_loop 2055 14.4 0.831 0.845 20.679 21.006 cp_fm_upper_to_full 70 14.2 12.738 18.071 12.738 18.071 ot_mini 99 10.5 0.001 0.001 14.059 14.100 rebuild_ks_matrix 110 8.3 0.000 0.000 13.891 13.947 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.891 13.946 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.600 12.650 dbcsr_complete_redistribute 325 12.2 1.032 1.058 7.161 10.340 multiply_cannon_multrec 8220 15.4 4.046 4.196 9.717 9.834 qs_ot_get_derivative 99 11.5 0.001 0.001 9.440 9.481 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.156 9.346 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.590 8.741 mp_alltoall_i22 605 13.7 5.230 8.452 5.230 8.452 mp_waitall_1 84994 16.7 7.410 8.245 7.410 8.245 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.450 7.487 calculate_rho_elec 110 8.6 0.222 0.223 7.450 7.487 sum_up_and_integrate 110 10.3 0.002 0.002 6.861 6.875 integrate_v_rspace 110 11.3 0.004 0.004 6.833 6.847 pw_transfer 1331 11.6 0.074 0.075 5.801 5.805 init_scf_run 11 5.9 0.000 0.001 5.767 5.767 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.767 5.767 dbcsr_mm_accdrv_process 11614 15.7 4.102 4.853 5.523 5.746 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.684 5.688 make_m2s 4110 13.4 0.043 0.043 5.204 5.667 qs_ot_get_p 110 10.4 0.002 0.002 5.462 5.513 make_images 4110 14.4 0.882 0.938 5.014 5.475 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.746 5.194 apply_single 110 13.6 0.000 0.000 4.746 5.194 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.928 5.193 cp_fm_cholesky_invert 11 10.9 5.140 5.144 5.140 5.144 fft_wrap_pw1pw2_140 451 13.1 1.366 1.370 5.004 5.012 ot_diis_step 99 11.5 0.015 0.015 4.599 4.599 density_rs2pw 110 9.6 0.004 0.004 3.911 3.940 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.912 3.919 grid_integrate_task_list 110 12.3 3.680 3.755 3.680 3.755 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.683 3.683 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.005 3.458 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.397 3.397 fft3d_ps 1111 14.6 1.886 1.898 3.390 3.397 hybrid_alltoall_any 4261 16.3 0.264 0.567 2.792 3.381 wfi_extrapolate 11 7.9 0.001 0.001 3.315 3.315 make_images_data 4110 15.4 0.045 0.048 2.674 3.265 calculate_dm_sparse 110 9.5 0.001 0.001 3.159 3.189 multiply_cannon_sync_h2d 8220 15.4 3.112 3.138 3.112 3.138 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.904 2.909 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.820 2.821 cp_fm_diag_elpa_base 48 14.0 2.288 2.481 2.819 2.819 grid_collocate_task_list 110 9.6 2.671 2.695 2.671 2.695 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.631 2.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.384 2.408 potential_pw2rs 110 12.3 0.021 0.022 2.402 2.404 calculate_first_density_matrix 1 7.0 0.000 0.000 2.334 2.335 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.210 2.273 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.171 mp_alltoall_d11v 2046 13.8 2.000 2.059 2.000 2.059 cp_fm_cholesky_decompose 22 10.9 1.946 1.967 1.946 1.967 qs_create_task_list 11 7.9 0.000 0.000 1.880 1.924 generate_qs_task_list 11 8.9 0.736 0.790 1.880 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.889 1.894 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.757 1.802 acc_transpose_blocks 8220 15.4 0.042 0.043 1.732 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.062000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1248.181818, yerr=43.931389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/15/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 1.103326E+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 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 629.964800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 3035365. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.106 209.921 209.923 qs_mol_dyn_low 1 2.0 0.003 0.005 209.210 209.244 qs_forces 11 3.9 0.006 0.007 209.110 209.112 qs_energies 11 4.9 0.004 0.028 203.431 203.454 scf_env_do_scf 11 5.9 0.005 0.030 186.549 186.553 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 166.106 166.108 dbcsr_multiply_generic 2527 12.6 0.178 0.183 127.252 128.017 qs_scf_new_mos 118 7.6 0.001 0.001 126.131 126.460 qs_scf_loop_do_ot 118 8.6 0.001 0.001 126.131 126.459 velocity_verlet 10 3.0 0.003 0.012 125.453 125.454 ot_scf_mini 118 9.6 0.003 0.004 119.299 119.654 multiply_cannon 2527 13.6 0.239 0.246 102.987 105.315 multiply_cannon_loop 2527 14.6 2.407 2.463 100.736 102.979 ot_mini 118 10.6 0.001 0.001 66.596 66.925 multiply_cannon_multrec 60648 15.6 32.246 34.974 42.116 44.215 qs_ot_get_derivative 118 11.6 0.001 0.001 41.436 41.746 rebuild_ks_matrix 129 8.3 0.001 0.001 34.094 34.567 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.022 34.093 34.566 mp_waitall_1 269254 16.5 29.378 32.475 29.378 32.475 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.628 31.090 qs_ot_get_p 129 10.4 0.002 0.012 30.133 30.512 multiply_cannon_sync_h2d 60648 15.6 26.727 29.144 26.727 29.144 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.699 25.489 apply_single 129 13.6 0.001 0.001 24.699 25.489 ot_diis_step 118 11.6 0.008 0.008 24.943 24.944 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.320 23.410 init_scf_loop 11 6.9 0.001 0.003 20.348 20.351 cp_dbcsr_syevd 83 12.4 0.005 0.007 20.300 20.302 multiply_cannon_metrocomm3 60648 15.6 0.117 0.122 16.611 20.096 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.687 19.904 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.308 17.336 cp_fm_diag_elpa_base 83 14.4 17.236 17.275 17.302 17.330 prepare_preconditioner 11 7.9 0.000 0.000 15.643 15.690 make_preconditioner 11 8.9 0.000 0.002 15.643 15.690 make_full_inverse_cholesky 11 9.9 0.000 0.001 14.879 15.071 sum_up_and_integrate 129 10.3 0.002 0.004 14.188 14.208 make_m2s 5054 13.6 0.105 0.116 13.880 14.196 integrate_v_rspace 129 11.3 0.003 0.004 14.130 14.151 make_images 5054 14.6 0.400 0.419 13.693 14.017 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.279 13.460 calculate_rho_elec 129 8.7 0.045 0.064 13.278 13.460 init_scf_run 11 5.9 0.000 0.001 12.645 12.646 scf_env_initial_rho_setup 11 6.9 0.001 0.004 12.645 12.645 mp_sum_l 8010 12.9 9.152 10.523 9.152 10.523 dbcsr_mm_accdrv_process 125468 16.2 4.743 4.899 9.436 9.955 wfi_extrapolate 11 7.9 0.001 0.001 9.193 9.193 cp_fm_cholesky_invert 11 10.9 8.864 8.871 8.864 8.871 calculate_dm_sparse 129 9.5 0.001 0.001 8.637 8.735 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.385 8.554 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.291 8.421 multiply_cannon_metrocomm1 60648 15.6 0.094 0.099 6.184 8.315 pw_transfer 1559 11.6 0.073 0.090 7.707 7.900 make_images_data 5054 15.6 0.066 0.071 6.789 7.757 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.504 7.697 density_rs2pw 129 9.7 0.006 0.007 6.848 7.671 grid_integrate_task_list 129 12.3 7.093 7.525 7.093 7.525 hybrid_alltoall_any 5240 16.5 0.296 2.281 5.929 7.197 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.838 6.855 fft_wrap_pw1pw2_140 527 13.2 0.862 0.896 6.556 6.729 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.892 6.000 fft3d_ps 1301 14.7 2.215 2.870 5.499 5.856 mp_alltoall_d11v 2423 14.1 4.372 5.706 4.372 5.706 grid_collocate_task_list 129 9.7 4.875 5.243 4.875 5.243 potential_pw2rs 129 12.3 0.009 0.010 4.639 4.665 cp_fm_cholesky_decompose 22 10.9 4.613 4.627 4.613 4.627 mp_sum_d 4501 12.1 3.557 4.405 3.557 4.405 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.923000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=5.407372 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 843.366400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5277533. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.064 190.479 190.496 qs_mol_dyn_low 1 2.0 0.003 0.003 190.060 190.082 qs_forces 11 3.9 0.004 0.005 189.976 189.977 qs_energies 11 4.9 0.002 0.002 183.228 183.240 scf_env_do_scf 11 5.9 0.001 0.001 166.861 166.870 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 134.123 134.126 velocity_verlet 10 3.0 0.001 0.002 120.215 120.216 dbcsr_multiply_generic 2527 12.6 0.189 0.196 97.934 99.268 qs_scf_new_mos 118 7.6 0.001 0.001 95.305 95.870 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.304 95.869 ot_scf_mini 118 9.6 0.004 0.004 90.441 91.079 multiply_cannon 2527 13.6 0.507 0.569 77.658 82.829 multiply_cannon_loop 2527 14.6 1.572 1.639 74.286 76.988 ot_mini 118 10.6 0.001 0.001 50.665 51.221 mp_waitall_1 216438 16.6 24.625 39.397 24.625 39.397 multiply_cannon_multrec 30324 15.6 20.802 24.652 31.450 35.799 rebuild_ks_matrix 129 8.3 0.001 0.001 32.418 33.083 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.417 33.082 init_scf_loop 11 6.9 0.000 0.000 32.649 32.650 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.177 29.788 multiply_cannon_metrocomm3 30324 15.6 0.099 0.107 16.139 29.095 qs_ot_get_derivative 118 11.6 0.001 0.002 28.337 28.957 prepare_preconditioner 11 7.9 0.000 0.000 28.371 28.437 make_preconditioner 11 8.9 0.000 0.000 28.371 28.437 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.076 27.603 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.391 23.544 apply_single 129 13.6 0.001 0.001 22.390 23.544 qs_ot_get_p 129 10.4 0.001 0.002 21.721 22.400 ot_diis_step 118 11.6 0.014 0.015 22.152 22.154 multiply_cannon_sync_h2d 30324 15.6 17.875 20.033 17.875 20.033 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.810 16.845 cp_fm_cholesky_invert 11 10.9 16.542 16.555 16.542 16.555 make_m2s 5054 13.6 0.087 0.093 14.308 15.916 make_images 5054 14.6 1.169 1.368 14.089 15.692 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.598 15.599 sum_up_and_integrate 129 10.3 0.002 0.004 14.081 14.111 integrate_v_rspace 129 11.3 0.003 0.004 14.020 14.054 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.479 13.508 calculate_rho_elec 129 8.7 0.087 0.104 13.479 13.507 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.415 12.448 cp_fm_diag_elpa_base 83 14.4 12.161 12.245 12.408 12.436 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.180 11.605 init_scf_run 11 5.9 0.000 0.001 11.567 11.569 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.567 11.569 multiply_cannon_metrocomm4 27797 15.6 0.100 0.115 3.998 11.242 mp_irecv_dv 70031 16.3 3.795 10.842 3.795 10.842 dbcsr_mm_accdrv_process 62734 16.2 5.553 6.171 10.098 10.634 make_images_data 5054 15.6 0.066 0.075 8.443 10.486 hybrid_alltoall_any 5240 16.5 0.352 1.503 7.234 9.617 pw_transfer 1559 11.6 0.085 0.095 8.646 8.704 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.422 8.483 wfi_extrapolate 11 7.9 0.001 0.001 8.360 8.360 grid_integrate_task_list 129 12.3 7.244 7.621 7.244 7.621 fft_wrap_pw1pw2_140 527 13.2 0.940 0.967 7.430 7.497 density_rs2pw 129 9.7 0.006 0.006 7.151 7.481 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.374 7.148 cp_fm_cholesky_decompose 22 10.9 6.953 7.029 6.953 7.029 calculate_dm_sparse 129 9.5 0.001 0.001 6.571 6.722 mp_sum_l 8010 12.9 4.163 6.344 4.163 6.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.139 6.147 fft3d_ps 1301 14.7 2.850 3.052 5.963 6.011 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.453 5.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.226 5.405 grid_collocate_task_list 129 9.7 5.039 5.363 5.039 5.363 mp_allgather_i34 2527 14.6 1.885 5.102 1.885 5.102 mp_alltoall_d11v 2423 14.1 4.136 4.633 4.136 4.633 potential_pw2rs 129 12.3 0.016 0.019 4.528 4.547 mp_sum_d 4501 12.1 2.653 4.039 2.653 4.039 dbcsr_complete_redistribute 395 12.7 0.774 0.868 3.163 4.016 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.496000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.181818, yerr=5.749596 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 955.031552E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.156 175.571 175.572 qs_mol_dyn_low 1 2.0 0.005 0.021 174.929 174.970 qs_forces 11 3.9 0.012 0.054 174.807 174.811 qs_energies 11 4.9 0.005 0.025 168.213 168.224 scf_env_do_scf 11 5.9 0.004 0.016 152.438 152.445 scf_env_do_scf_inner_loop 117 6.6 0.006 0.027 117.655 117.656 velocity_verlet 10 3.0 0.002 0.008 112.391 112.395 dbcsr_multiply_generic 2507 12.6 0.185 0.191 82.028 83.065 qs_scf_new_mos 117 7.6 0.001 0.001 80.604 81.012 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.603 81.011 ot_scf_mini 117 9.6 0.003 0.004 76.340 76.758 multiply_cannon 2507 13.6 0.509 0.528 62.413 67.307 multiply_cannon_loop 2507 14.6 1.122 1.189 59.238 61.696 ot_mini 117 10.6 0.001 0.002 42.633 43.061 mp_waitall_1 170520 16.6 24.888 34.822 24.888 34.822 init_scf_loop 11 6.9 0.001 0.004 34.676 34.679 prepare_preconditioner 11 7.9 0.000 0.000 30.655 30.713 make_preconditioner 11 8.9 0.000 0.001 30.655 30.713 rebuild_ks_matrix 128 8.3 0.001 0.001 30.212 30.693 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.027 30.211 30.692 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.312 29.663 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.226 27.675 multiply_cannon_multrec 20056 15.6 12.988 16.093 22.301 25.427 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.577 25.177 qs_ot_get_derivative 117 11.6 0.001 0.002 22.798 23.207 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.881 20.990 apply_single 128 13.6 0.001 0.001 19.881 20.990 ot_diis_step 117 11.6 0.018 0.021 19.721 19.721 qs_ot_get_p 128 10.4 0.001 0.002 18.995 19.402 make_m2s 5014 13.6 0.078 0.084 14.639 16.004 make_images 5014 14.6 1.142 1.221 14.403 15.767 multiply_cannon_sync_h2d 20056 15.6 13.541 15.086 13.541 15.086 qs_ot_p2m_diag 83 11.4 0.265 0.273 14.776 14.785 cp_fm_cholesky_invert 11 10.9 14.261 14.269 14.261 14.269 sum_up_and_integrate 128 10.3 0.002 0.003 13.904 13.928 integrate_v_rspace 128 11.3 0.003 0.005 13.845 13.870 cp_dbcsr_syevd 83 12.4 0.005 0.007 13.717 13.717 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.544 13.581 calculate_rho_elec 128 8.7 0.129 0.143 13.543 13.581 make_images_data 5014 15.6 0.063 0.074 8.808 10.553 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.525 10.548 cp_fm_diag_elpa_base 83 14.4 10.117 10.269 10.521 10.543 init_scf_run 11 5.9 0.000 0.001 10.536 10.536 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.535 10.536 hybrid_alltoall_any 5200 16.5 0.449 2.048 7.695 9.853 multiply_cannon_metrocomm4 17549 15.6 0.065 0.075 3.531 9.443 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.997 9.297 mp_irecv_dv 50230 16.2 3.403 9.185 3.403 9.185 dbcsr_mm_accdrv_process 41502 16.2 5.634 5.963 8.771 8.888 pw_transfer 1547 11.6 0.084 0.103 8.755 8.855 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.532 8.639 grid_integrate_task_list 128 12.3 7.307 7.690 7.307 7.690 fft_wrap_pw1pw2_140 523 13.2 1.063 1.098 7.581 7.689 wfi_extrapolate 11 7.9 0.004 0.027 7.465 7.465 density_rs2pw 128 9.7 0.005 0.006 6.895 7.330 cp_fm_cholesky_decompose 22 10.9 7.212 7.245 7.212 7.245 cp_fm_upper_to_full 105 14.8 5.742 7.218 5.742 7.218 dbcsr_complete_redistribute 395 12.7 1.192 1.219 4.647 6.402 calculate_dm_sparse 128 9.5 0.001 0.001 5.876 5.985 fft3d_ps 1291 14.7 2.823 3.057 5.846 5.943 grid_collocate_task_list 128 9.7 5.165 5.510 5.165 5.510 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.495 5.506 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.478 5.232 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.604 5.231 mp_alltoall_d11v 2415 14.1 4.325 5.022 4.325 5.022 mp_allgather_i34 2507 14.6 1.671 4.811 1.671 4.811 mp_sum_l 7950 12.9 3.161 4.772 3.161 4.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.541 4.682 potential_pw2rs 128 12.3 0.021 0.023 4.357 4.363 transfer_fm_to_dbcsr 11 9.9 0.019 0.026 2.324 4.012 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.963 3.995 mp_alltoall_i22 716 14.1 1.975 3.839 1.975 3.839 qs_energies_init_hamiltonians 11 5.9 0.012 0.017 3.795 3.800 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.486 3.532 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.572000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.272727, yerr=11.855884 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.153851E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3994 58319. MP_Allreduce 11062 1083. MP_Sync 87 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.108 190.808 190.835 qs_mol_dyn_low 1 2.0 0.003 0.004 190.271 190.373 qs_forces 11 3.9 0.004 0.004 190.143 190.149 qs_energies 11 4.9 0.002 0.008 183.040 183.049 scf_env_do_scf 11 5.9 0.001 0.002 166.158 166.169 velocity_verlet 10 3.0 0.002 0.002 126.189 126.193 scf_env_do_scf_inner_loop 117 6.6 0.004 0.011 118.962 118.964 qs_scf_new_mos 117 7.6 0.001 0.001 82.412 82.730 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.411 82.730 dbcsr_multiply_generic 2507 12.6 0.193 0.198 80.668 81.437 ot_scf_mini 117 9.6 0.003 0.004 77.841 78.153 multiply_cannon 2507 13.6 0.550 0.594 55.767 58.279 multiply_cannon_loop 2507 14.6 1.848 1.896 52.070 53.836 init_scf_loop 11 6.9 0.001 0.005 47.072 47.073 ot_mini 117 10.6 0.001 0.002 43.135 43.420 prepare_preconditioner 11 7.9 0.000 0.000 42.967 42.994 make_preconditioner 11 8.9 0.000 0.002 42.967 42.994 make_full_inverse_cholesky 11 9.9 0.011 0.021 36.476 41.573 multiply_cannon_multrec 30084 15.6 13.675 18.387 26.537 30.992 rebuild_ks_matrix 128 8.3 0.001 0.001 29.432 29.703 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 29.432 29.702 mp_waitall_1 147882 16.7 17.202 26.936 17.202 26.936 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.556 26.801 qs_ot_get_derivative 117 11.6 0.001 0.002 23.525 23.835 make_m2s 5014 13.6 0.092 0.097 20.383 21.277 qs_ot_get_p 128 10.4 0.011 0.014 20.784 21.108 make_images 5014 14.6 1.935 2.253 20.072 20.967 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.004 19.588 apply_single 128 13.6 0.001 0.001 19.004 19.588 ot_diis_step 117 11.6 0.018 0.019 19.477 19.479 cp_fm_upper_to_full 105 14.8 11.553 17.042 11.553 17.042 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.567 16.619 cp_fm_cholesky_invert 11 10.9 16.533 16.542 16.533 16.542 cp_dbcsr_syevd 83 12.4 0.005 0.007 15.138 15.139 multiply_cannon_metrocomm3 30084 15.6 0.048 0.052 6.333 14.501 sum_up_and_integrate 128 10.3 0.002 0.003 14.104 14.132 integrate_v_rspace 128 11.3 0.003 0.004 14.044 14.073 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.923 13.980 calculate_rho_elec 128 8.7 0.170 0.186 13.923 13.980 dbcsr_complete_redistribute 395 12.7 1.484 1.597 9.257 13.072 dbcsr_mm_accdrv_process 62264 16.2 8.499 9.189 12.430 12.891 make_images_data 5014 15.6 0.065 0.071 10.930 12.478 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.843 11.860 cp_fm_diag_elpa_base 83 14.4 10.864 11.188 11.835 11.854 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.864 11.689 multiply_cannon_sync_h2d 30084 15.6 10.467 11.306 10.467 11.306 hybrid_alltoall_any 5200 16.5 0.527 2.213 9.799 11.257 init_scf_run 11 5.9 0.000 0.001 10.912 10.913 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.911 10.913 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.469 10.210 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.886 10.106 mp_alltoall_i22 716 14.1 5.694 9.597 5.694 9.597 pw_transfer 1547 11.6 0.085 0.101 9.300 9.402 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.074 9.182 fft_wrap_pw1pw2_140 523 13.2 1.247 1.301 8.063 8.182 grid_integrate_task_list 128 12.3 7.566 7.792 7.566 7.792 cp_fm_cholesky_decompose 22 10.9 7.672 7.764 7.672 7.764 wfi_extrapolate 11 7.9 0.001 0.001 7.707 7.707 density_rs2pw 128 9.7 0.005 0.006 6.965 7.316 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.761 7.183 mp_irecv_dv 76098 16.2 2.611 6.911 2.611 6.911 calculate_dm_sparse 128 9.5 0.001 0.001 6.250 6.335 fft3d_ps 1291 14.7 2.996 3.108 6.128 6.210 grid_collocate_task_list 128 9.7 5.286 5.662 5.286 5.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.472 5.510 mp_alltoall_d11v 2415 14.1 4.988 5.336 4.988 5.336 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.529 4.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.407 4.469 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.449 4.449 potential_pw2rs 128 12.3 0.023 0.023 4.377 4.391 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.298 4.359 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.835000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.636364, yerr=19.583262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.560822E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58273. MP_Allreduce 11070 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.046 170.664 170.672 qs_mol_dyn_low 1 2.0 0.003 0.003 170.215 170.239 qs_forces 11 3.9 0.004 0.004 170.123 170.125 qs_energies 11 4.9 0.002 0.002 162.656 162.663 scf_env_do_scf 11 5.9 0.001 0.001 145.510 145.519 velocity_verlet 10 3.0 0.002 0.002 111.381 111.385 scf_env_do_scf_inner_loop 116 6.6 0.007 0.018 109.351 109.353 qs_scf_new_mos 116 7.6 0.001 0.001 72.440 72.562 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.439 72.562 dbcsr_multiply_generic 2485 12.5 0.182 0.191 71.755 72.187 ot_scf_mini 116 9.6 0.003 0.004 68.052 68.165 multiply_cannon 2485 13.5 0.554 0.571 53.323 56.400 multiply_cannon_loop 2485 14.5 0.800 0.831 50.193 51.197 ot_mini 116 10.6 0.001 0.001 37.564 37.673 init_scf_loop 11 6.9 0.000 0.000 36.006 36.010 prepare_preconditioner 11 7.9 0.000 0.000 32.061 32.093 make_preconditioner 11 8.9 0.000 0.000 32.061 32.093 mp_waitall_1 124680 16.7 24.845 31.098 24.845 31.098 make_full_inverse_cholesky 11 9.9 0.010 0.023 29.950 30.212 rebuild_ks_matrix 127 8.3 0.001 0.001 29.059 29.180 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.058 29.179 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.466 26.575 multiply_cannon_multrec 9940 15.5 10.228 15.930 17.753 22.106 qs_ot_get_derivative 116 11.6 0.001 0.002 20.494 20.602 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.843 20.101 cp_fm_cholesky_invert 11 10.9 18.502 18.508 18.502 18.508 qs_ot_get_p 127 10.4 0.067 0.071 17.938 18.088 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.150 17.383 apply_single 127 13.6 0.001 0.001 17.149 17.383 ot_diis_step 116 11.6 0.019 0.020 16.999 16.999 make_m2s 4970 13.5 0.063 0.067 14.564 15.493 make_images 4970 14.5 2.140 2.574 14.258 15.178 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.226 14.249 calculate_rho_elec 127 8.7 0.251 0.262 14.225 14.248 sum_up_and_integrate 127 10.3 0.002 0.002 14.177 14.226 integrate_v_rspace 127 11.3 0.004 0.004 14.117 14.169 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.056 14.070 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.893 12.894 multiply_cannon_sync_h2d 9940 15.5 10.611 11.222 10.611 11.222 init_scf_run 11 5.9 0.000 0.001 10.383 10.383 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.383 10.383 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.984 9.993 cp_fm_diag_elpa_base 82 14.4 9.751 9.817 9.980 9.988 pw_transfer 1535 11.6 0.084 0.093 9.889 9.934 make_images_data 4970 15.5 0.054 0.064 8.348 9.752 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.669 9.716 hybrid_alltoall_any 5155 16.4 0.834 3.744 8.230 9.316 fft_wrap_pw1pw2_140 519 13.2 1.653 1.687 8.542 8.592 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.017 8.095 cp_fm_cholesky_decompose 22 10.9 7.926 8.011 7.926 8.011 grid_integrate_task_list 127 12.3 7.740 7.985 7.740 7.985 dbcsr_mm_accdrv_process 20590 16.1 3.237 4.429 7.177 7.902 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.383 7.448 wfi_extrapolate 11 7.9 0.001 0.001 7.327 7.327 density_rs2pw 127 9.7 0.005 0.005 6.980 7.191 fft3d_ps 1281 14.7 3.112 3.208 6.163 6.191 calculate_dm_sparse 127 9.5 0.001 0.001 6.042 6.120 grid_collocate_task_list 127 9.7 5.534 5.757 5.534 5.757 dbcsr_complete_redistribute 393 12.7 2.086 2.144 5.239 5.594 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.848 5.580 mp_irecv_dv 28618 15.9 1.808 5.491 1.808 5.491 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.210 5.211 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.142 5.151 mp_alltoall_d11v 2401 14.1 4.663 5.093 4.663 5.093 mp_allgather_i34 2485 14.5 1.315 4.604 1.315 4.604 potential_pw2rs 127 12.3 0.026 0.028 4.395 4.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.054 4.096 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.587 3.933 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.628 3.717 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.367 3.691 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.634 3.679 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.467 3.564 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.487 3.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.672000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1449.090909, yerr=36.594720 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410026932224 0.0% 0.0% 100.0% flops 32 x 32 x 32 1930587799552 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957877121024 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 11444723269632 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019199483904 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019199483904 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 92.803070E+12 0.0% 0.0% 100.0% flops max/rank 11.607223E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705605184 0.0% 0.0% 100.0% number of processed stacks 1948912 0.0% 0.0% 100.0% average stack size 0.0 0.0 3440.7 marketing flops 143.514185E+12 ------------------------------------------------------------------------------- # multiplications 2488 max memory usage/rank 3.168358E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99520 MPI messages size (bytes): total size 1.127523E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.329611E+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 44888 34846277632 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716109087456 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4046 58514. MP_Allreduce 11192 1501. MP_Sync 87 MP_Alltoall 1700 36954293. 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 0.033 0.065 291.652 291.656 qs_mol_dyn_low 1 2.0 0.003 0.003 291.089 291.111 qs_forces 11 3.9 0.004 0.004 290.991 290.995 qs_energies 11 4.9 0.002 0.002 282.118 282.123 scf_env_do_scf 11 5.9 0.001 0.001 259.615 259.621 velocity_verlet 10 3.0 0.002 0.002 209.825 209.834 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 135.346 135.349 init_scf_loop 11 6.9 0.000 0.000 123.995 123.998 prepare_preconditioner 11 7.9 0.000 0.000 119.104 119.125 make_preconditioner 11 8.9 0.000 0.000 119.104 119.125 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.406 116.249 qs_scf_new_mos 116 7.6 0.001 0.001 90.681 90.741 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.680 90.741 ot_scf_mini 116 9.6 0.004 0.004 85.855 85.878 dbcsr_multiply_generic 2488 12.5 0.214 0.225 82.072 82.642 cp_fm_upper_to_full 105 14.8 52.534 75.221 52.534 75.221 multiply_cannon 2488 13.5 0.682 0.727 58.544 59.804 multiply_cannon_loop 2488 14.5 1.040 1.062 54.471 55.937 ot_mini 116 10.6 0.001 0.001 44.651 44.681 dbcsr_complete_redistribute 395 12.7 3.990 4.073 29.711 42.592 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.192 39.051 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.656 36.395 rebuild_ks_matrix 127 8.3 0.001 0.001 34.666 34.711 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.665 34.710 mp_alltoall_i22 714 14.1 21.411 34.403 21.411 34.403 cp_fm_cholesky_invert 11 10.9 33.729 33.736 33.729 33.736 mp_waitall_1 102888 16.8 28.267 32.674 28.267 32.674 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.125 32.174 qs_ot_get_p 127 10.4 0.002 0.002 26.402 26.462 qs_ot_get_derivative 116 11.6 0.002 0.002 24.724 24.745 qs_ot_p2m_diag 83 11.4 0.879 0.885 22.243 22.274 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.431 20.432 make_m2s 4976 13.5 0.074 0.077 18.841 20.387 multiply_cannon_metrocomm3 9952 15.5 0.024 0.025 18.776 20.256 make_images 4976 14.5 3.029 3.222 18.360 19.908 ot_diis_step 116 11.6 0.022 0.022 19.887 19.888 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.286 19.413 apply_single 127 13.6 0.001 0.001 19.286 19.412 multiply_cannon_multrec 9952 15.5 10.173 11.974 17.920 17.968 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.310 17.324 calculate_rho_elec 127 8.7 0.474 0.475 17.309 17.323 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.208 17.208 cp_fm_diag_elpa_base 83 14.4 12.815 14.430 17.204 17.204 sum_up_and_integrate 127 10.3 0.002 0.002 16.371 16.463 integrate_v_rspace 127 11.3 0.004 0.005 16.308 16.400 multiply_cannon_sync_h2d 9952 15.5 14.192 14.226 14.192 14.226 pw_transfer 1535 11.6 0.095 0.096 13.190 13.200 hybrid_alltoall_any 5162 16.4 1.314 3.091 10.940 13.160 make_images_data 4976 15.5 0.062 0.068 10.770 13.040 fft_wrap_pw1pw2 1281 12.7 0.012 0.012 12.954 12.964 init_scf_run 11 5.9 0.000 0.001 12.206 12.207 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.206 12.207 fft_wrap_pw1pw2_140 519 13.2 3.192 3.237 11.503 11.518 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.883 9.907 dbcsr_mm_accdrv_process 20602 16.0 4.211 6.090 7.498 9.354 wfi_extrapolate 11 7.9 0.001 0.001 9.018 9.018 cp_fm_cholesky_decompose 22 10.9 8.925 8.945 8.925 8.945 grid_integrate_task_list 127 12.3 8.491 8.694 8.491 8.694 density_rs2pw 127 9.7 0.005 0.005 8.395 8.439 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.102 8.104 fft3d_ps 1281 14.7 3.943 3.954 7.463 7.501 mp_alltoall_d11v 2407 14.1 6.962 7.105 6.962 7.105 calculate_dm_sparse 127 9.5 0.001 0.001 6.703 6.740 grid_collocate_task_list 127 9.7 6.364 6.442 6.364 6.442 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.385 6.418 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.175 6.245 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.656000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2886.909091, yerr=159.706808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 1.260954E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 85.038 85.039 qs_energies 1 2.0 0.000 0.000 84.563 84.570 ls_scf 1 3.0 0.000 0.000 83.675 83.683 dbcsr_multiply_generic 111 6.7 0.014 0.019 72.548 72.712 multiply_cannon 111 7.7 0.017 0.020 55.891 56.933 multiply_cannon_loop 111 8.7 0.229 0.244 52.477 53.765 ls_scf_main 1 4.0 0.000 0.000 52.341 52.353 density_matrix_trs4 2 5.0 0.002 0.003 46.815 46.924 ls_scf_init_scf 1 4.0 0.000 0.000 28.307 28.309 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.013 27.060 mp_waitall_1 11031 10.9 22.356 25.560 22.356 25.560 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.006 24.898 24.922 multiply_cannon_multrec 2664 9.7 8.160 8.993 15.559 17.371 multiply_cannon_sync_h2d 2664 9.7 13.624 15.423 13.624 15.423 make_m2s 222 7.7 0.009 0.011 13.101 13.598 make_images 222 8.7 0.098 0.107 13.079 13.578 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.572 12.322 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.471 8.737 make_images_data 222 9.7 0.004 0.005 7.649 8.221 hybrid_alltoall_any 227 10.6 0.216 1.857 6.522 8.096 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.735 7.015 8.036 dbcsr_mm_accdrv_process_sort 4760 11.4 6.221 7.120 6.221 7.120 calculate_norms 4752 9.8 5.530 6.232 5.530 6.232 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.201 mp_sum_l 887 5.1 3.017 4.336 3.017 4.336 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.065 3.722 mp_irecv_dv 6231 10.9 2.048 3.689 2.048 3.689 make_images_sizes 222 9.7 0.000 0.000 0.792 3.439 mp_alltoall_i44 222 10.7 0.792 3.439 0.792 3.439 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.240 arnoldi_normal_ev 4 7.8 0.001 0.004 3.209 3.240 build_subspace 16 8.4 0.009 0.012 3.111 3.113 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.238 3.041 ls_scf_post 1 4.0 0.000 0.000 3.027 3.034 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.889 dbcsr_special_finalize 555 9.7 0.005 0.006 2.306 2.715 dbcsr_merge_single_wm 555 10.7 0.454 0.587 2.298 2.705 make_images_pack 222 9.7 2.207 2.625 2.209 2.626 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.326 2.567 dbcsr_sort_data 658 11.4 2.102 2.477 2.102 2.477 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.455 2.070 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.302 2.385 buffer_matrices_ensure_size 222 8.7 1.749 2.072 1.749 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.786 1.787 rebuild_ks_matrix 3 7.3 0.000 0.000 1.777 1.778 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.777 1.778 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.039000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 2.173821E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.040 89.450 89.450 qs_energies 1 2.0 0.000 0.000 89.010 89.014 ls_scf 1 3.0 0.000 0.000 87.591 87.595 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.728 74.096 multiply_cannon 111 7.7 0.027 0.036 52.444 56.504 ls_scf_main 1 4.0 0.000 0.000 54.214 54.219 multiply_cannon_loop 111 8.7 0.135 0.145 49.723 52.987 density_matrix_trs4 2 5.0 0.002 0.003 48.501 48.702 ls_scf_init_scf 1 4.0 0.000 0.000 29.689 29.691 mp_waitall_1 9105 10.9 20.463 29.631 20.463 29.631 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.509 28.591 multiply_cannon_multrec 1332 9.7 13.392 17.571 22.697 28.019 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.165 26.179 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.184 20.243 make_m2s 222 7.7 0.006 0.007 14.982 15.530 make_images 222 8.7 1.368 1.693 14.951 15.500 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.534 8.903 10.485 dbcsr_mm_accdrv_process_sort 4041 11.4 8.419 9.951 8.419 9.951 make_images_data 222 9.7 0.004 0.005 8.716 9.540 hybrid_alltoall_any 227 10.6 0.542 2.536 8.077 9.392 mp_sum_l 887 5.1 4.823 7.765 4.823 7.765 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.174 7.714 mp_irecv_dv 3311 11.0 3.154 7.659 3.154 7.659 calculate_norms 2376 9.8 6.078 6.779 6.078 6.779 multiply_cannon_sync_h2d 1332 9.7 4.889 6.362 4.889 6.362 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.614 6.118 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.924 5.189 arnoldi_extremal 4 6.8 0.000 0.000 4.663 4.685 arnoldi_normal_ev 4 7.8 0.001 0.005 4.663 4.684 build_subspace 16 8.4 0.014 0.021 4.413 4.415 ls_scf_post 1 4.0 0.000 0.000 3.688 3.692 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.513 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.177 3.410 dbcsr_matrix_vector_mult_local 304 10.0 2.789 3.260 2.791 3.261 mp_allgather_i34 111 8.7 0.712 2.892 0.712 2.892 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.633 2.756 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.208 2.557 dbcsr_data_new 4174 10.1 2.113 2.401 2.113 2.401 make_images_pack 222 9.7 1.815 2.131 1.817 2.133 dbcsr_sort_data 436 11.2 1.823 2.041 1.823 2.041 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.854 1.856 rebuild_ks_matrix 3 7.3 0.000 0.000 1.842 1.844 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.842 1.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.450000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 2.914419E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 92.213 92.214 qs_energies 1 2.0 0.000 0.000 91.668 91.672 ls_scf 1 3.0 0.000 0.000 90.251 90.255 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.845 75.153 ls_scf_main 1 4.0 0.000 0.000 56.332 56.336 multiply_cannon 111 7.7 0.032 0.072 51.848 55.928 multiply_cannon_loop 111 8.7 0.116 0.129 49.085 53.222 density_matrix_trs4 2 5.0 0.002 0.003 50.382 50.569 mp_waitall_1 7281 11.0 23.151 33.649 23.151 33.649 ls_scf_init_scf 1 4.0 0.000 0.000 30.195 30.198 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.973 29.067 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.616 26.632 multiply_cannon_multrec 888 9.7 12.680 15.472 21.266 24.671 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.720 23.074 make_m2s 222 7.7 0.006 0.007 16.006 16.805 make_images 222 8.7 1.578 1.844 15.968 16.765 make_images_data 222 9.7 0.004 0.004 9.343 10.515 hybrid_alltoall_any 227 10.6 0.640 2.939 8.976 10.090 dbcsr_mm_accdrv_process 3754 10.4 0.304 0.492 8.110 9.340 dbcsr_mm_accdrv_process_sort 3754 11.4 7.675 8.849 7.675 8.849 mp_sum_l 887 5.1 5.290 8.761 5.290 8.761 multiply_cannon_sync_h2d 888 9.7 6.006 7.440 6.006 7.440 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.499 7.226 mp_irecv_dv 2335 11.1 2.483 7.150 2.483 7.150 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.111 6.819 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.730 6.622 arnoldi_extremal 4 6.8 0.000 0.000 5.104 5.125 arnoldi_normal_ev 4 7.8 0.001 0.005 5.104 5.125 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.801 5.098 calculate_norms 1584 9.8 4.401 4.843 4.401 4.843 build_subspace 16 8.4 0.014 0.020 4.804 4.811 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.502 3.808 ls_scf_post 1 4.0 0.000 0.000 3.724 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.079 3.647 3.082 3.649 mp_allgather_i34 111 8.7 0.882 3.609 0.882 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.454 3.547 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.986 dbcsr_data_new 4116 9.9 2.114 2.457 2.114 2.457 dbcsr_sort_data 325 11.1 1.880 2.186 1.880 2.186 make_images_sizes 222 9.7 0.000 0.000 0.831 2.083 mp_alltoall_i44 222 10.7 0.830 2.082 0.830 2.082 dbcsr_finalize 304 7.8 0.026 0.032 1.605 1.927 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.902 1.904 rebuild_ks_matrix 3 7.3 0.000 0.000 1.884 1.886 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.884 1.886 make_images_pack 222 9.7 1.623 1.866 1.626 1.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.214000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2230.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 3.374961E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.051 96.946 96.946 qs_energies 1 2.0 0.000 0.000 96.371 96.376 ls_scf 1 3.0 0.000 0.000 94.707 94.712 dbcsr_multiply_generic 111 6.7 0.016 0.019 78.288 78.560 ls_scf_main 1 4.0 0.000 0.000 58.799 58.800 multiply_cannon 111 7.7 0.042 0.093 51.728 56.042 density_matrix_trs4 2 5.0 0.002 0.003 52.653 52.781 multiply_cannon_loop 111 8.7 0.154 0.168 46.647 49.044 ls_scf_init_scf 1 4.0 0.000 0.000 32.684 32.685 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.458 31.521 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.982 28.997 mp_waitall_1 6369 11.0 22.729 28.773 22.729 28.773 multiply_cannon_multrec 1332 9.7 14.117 17.767 22.046 24.931 make_m2s 222 7.7 0.006 0.008 21.054 22.474 make_images 222 8.7 3.132 3.584 21.004 22.426 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.337 17.088 make_images_data 222 9.7 0.004 0.004 11.758 13.332 hybrid_alltoall_any 227 10.6 0.795 3.760 11.060 12.974 dbcsr_mm_accdrv_process 3641 10.4 0.312 0.482 7.562 9.065 dbcsr_mm_accdrv_process_sort 3641 11.4 7.120 8.567 7.120 8.567 mp_sum_l 887 5.1 3.929 7.537 3.929 7.537 multiply_cannon_sync_h2d 1332 9.7 5.435 6.325 5.435 6.325 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.070 6.044 mp_irecv_dv 3229 10.9 2.046 5.958 2.046 5.958 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.039 5.924 arnoldi_extremal 4 6.8 0.000 0.000 5.371 5.384 arnoldi_normal_ev 4 7.8 0.001 0.004 5.371 5.384 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.462 5.180 build_subspace 16 8.4 0.014 0.021 5.038 5.046 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.557 4.744 mp_allgather_i34 111 8.7 2.191 4.564 2.191 4.564 calculate_norms 2376 9.8 4.190 4.481 4.190 4.481 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.680 4.001 dbcsr_matrix_vector_mult_local 304 10.0 3.260 3.757 3.262 3.759 dbcsr_sort_data 658 11.4 3.098 3.476 3.098 3.476 dbcsr_special_finalize 555 9.7 0.006 0.007 2.852 3.243 dbcsr_merge_single_wm 555 10.7 0.540 0.663 2.844 3.235 ls_scf_post 1 4.0 0.000 0.000 3.224 3.229 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.022 3.092 ls_scf_store_result 1 5.0 0.000 0.000 2.971 3.024 dbcsr_data_release 10477 10.7 1.584 2.437 1.584 2.437 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.946000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 4.761285E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.055 91.657 91.657 qs_energies 1 2.0 0.000 0.000 90.922 90.927 ls_scf 1 3.0 0.000 0.000 88.948 88.953 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.474 70.659 ls_scf_main 1 4.0 0.000 0.000 56.206 56.207 multiply_cannon 111 7.7 0.080 0.131 52.571 55.623 multiply_cannon_loop 111 8.7 0.088 0.093 50.001 50.892 density_matrix_trs4 2 5.0 0.002 0.003 49.196 49.280 ls_scf_init_scf 1 4.0 0.000 0.000 29.326 29.327 mp_waitall_1 5436 11.0 24.228 28.826 24.228 28.826 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.071 28.101 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.971 25.983 multiply_cannon_multrec 444 9.7 13.734 16.385 20.828 22.095 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.702 15.407 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.241 14.293 make_m2s 222 7.7 0.005 0.005 13.293 14.221 make_images 222 8.7 2.036 2.470 13.225 14.151 make_images_data 222 9.7 0.003 0.004 8.100 9.575 hybrid_alltoall_any 227 10.6 0.800 3.818 8.048 9.353 multiply_cannon_sync_h2d 444 9.7 6.747 8.042 6.747 8.042 dbcsr_mm_accdrv_process 3003 10.4 0.364 0.402 6.778 7.892 dbcsr_mm_accdrv_process_sort 3003 11.4 6.414 7.494 6.414 7.494 arnoldi_extremal 4 6.8 0.000 0.000 5.859 5.869 arnoldi_normal_ev 4 7.8 0.002 0.005 5.859 5.869 build_subspace 16 8.4 0.015 0.020 5.469 5.482 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.426 4.679 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.234 4.409 mp_sum_l 887 5.1 2.613 4.230 2.613 4.230 dbcsr_matrix_vector_mult_local 304 10.0 3.752 4.209 3.754 4.211 calculate_norms 792 9.8 3.619 3.727 3.619 3.727 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.515 3.706 mp_irecv_dv 1241 11.2 1.503 3.677 1.503 3.677 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.560 3.644 mp_allgather_i34 111 8.7 1.150 3.511 1.150 3.511 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.808 3.454 ls_scf_post 1 4.0 0.000 0.000 3.416 3.421 ls_scf_store_result 1 5.0 0.000 0.000 3.206 3.242 make_images_sizes 222 9.7 0.000 0.000 0.813 3.188 mp_alltoall_i44 222 10.7 0.813 3.187 0.813 3.187 dbcsr_finalize 304 7.8 0.062 0.077 2.202 2.262 dbcsr_data_new 4608 9.7 1.781 2.253 1.781 2.253 dbcsr_merge_all 275 8.9 0.481 0.522 2.061 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.038 2.038 rebuild_ks_matrix 3 7.3 0.000 0.000 2.005 2.006 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.005 2.006 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.957 1.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.657000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 8.883413E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.108 109.340 109.341 qs_energies 1 2.0 0.000 0.000 107.847 107.850 ls_scf 1 3.0 0.000 0.000 104.867 104.869 dbcsr_multiply_generic 111 6.7 0.023 0.026 78.412 78.595 ls_scf_main 1 4.0 0.000 0.000 63.697 63.697 density_matrix_trs4 2 5.0 0.002 0.003 54.647 54.711 multiply_cannon 111 7.7 0.109 0.160 49.069 51.241 multiply_cannon_loop 111 8.7 0.098 0.101 46.036 49.349 ls_scf_init_scf 1 4.0 0.000 0.000 37.397 37.398 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.840 35.868 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 33.128 33.135 mp_waitall_1 4527 11.1 21.657 25.145 21.657 25.145 make_m2s 222 7.7 0.005 0.005 22.298 23.423 make_images 222 8.7 3.571 3.881 22.191 23.313 multiply_cannon_multrec 444 9.7 17.806 18.414 22.465 23.149 hybrid_alltoall_any 227 10.6 1.657 3.619 12.554 15.284 make_images_data 222 9.7 0.003 0.004 12.727 14.851 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.345 10.811 multiply_cannon_sync_h2d 444 9.7 8.796 8.847 8.796 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.440 7.450 arnoldi_normal_ev 4 7.8 0.003 0.009 7.440 7.450 build_subspace 16 8.4 0.026 0.036 6.895 6.906 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.549 5.694 mp_sum_l 887 5.1 4.644 5.490 4.644 5.490 dbcsr_matrix_vector_mult_local 304 10.0 5.152 5.438 5.155 5.440 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.177 5.272 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.864 5.087 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.054 4.817 dbcsr_mm_accdrv_process 1814 10.4 0.275 0.373 4.476 4.589 compute_matrix_preconditioner 1 6.0 0.002 0.002 4.558 4.564 dbcsr_mm_accdrv_process_sort 1814 11.4 4.131 4.258 4.131 4.258 acc_transpose_blocks 444 9.7 0.003 0.003 0.725 4.010 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 0.642 3.927 jit_kernel_transpose 1 13.0 0.637 3.921 0.637 3.921 ls_scf_post 1 4.0 0.000 0.000 3.773 3.776 mp_allgather_i34 111 8.7 1.134 3.564 1.134 3.564 make_images_sizes 222 9.7 0.000 0.000 1.452 3.564 mp_alltoall_i44 222 10.7 1.452 3.563 1.452 3.563 ls_scf_store_result 1 5.0 0.000 0.000 3.489 3.529 calculate_norms 792 9.8 3.228 3.269 3.228 3.269 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.185 dbcsr_merge_all 275 8.9 0.885 0.917 2.866 2.964 qs_energies_init_hamiltonians 1 3.0 0.011 0.018 2.950 2.950 dbcsr_complete_redistribute 5 7.6 1.436 1.479 2.740 2.863 matrix_ls_to_qs 2 6.0 0.000 0.000 2.398 2.538 dbcsr_sort_data 325 11.1 2.442 2.515 2.442 2.515 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.455 2.457 rebuild_ks_matrix 3 7.3 0.000 0.000 2.389 2.391 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.389 2.391 dbcsr_data_new 6591 9.6 1.884 2.357 1.884 2.357 dbcsr_new_transposed 4 7.5 0.242 0.253 2.284 2.299 dbcsr_frobenius_norm 74 6.6 2.055 2.135 2.197 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.204 dbcsr_add_anytype 103 7.2 0.859 0.893 2.123 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.341000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6936.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3caab925d13535c28a7f61e60f70e91adc8e5da_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 594.042880E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4218258. 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 0.036 0.100 226.975 226.991 qs_mol_dyn_low 1 2.0 0.003 0.012 225.824 225.869 qs_forces 5 3.8 0.006 0.024 225.707 225.710 qs_energies 5 4.8 0.005 0.036 222.393 222.412 scf_env_do_scf 5 5.8 0.000 0.002 207.082 207.086 scf_env_do_scf_inner_loop 105 6.6 0.003 0.011 180.533 180.539 qs_scf_new_mos 105 7.6 0.000 0.001 140.687 140.908 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.686 140.907 dbcsr_multiply_generic 1445 12.2 0.128 0.137 133.727 134.223 ot_scf_mini 105 9.6 0.003 0.004 130.777 130.902 multiply_cannon 1445 13.2 0.275 0.285 114.794 116.955 multiply_cannon_loop 1445 14.2 2.826 2.983 113.019 114.455 velocity_verlet 4 3.0 0.006 0.016 107.279 107.280 ot_mini 105 10.6 0.001 0.001 60.157 60.273 multiply_cannon_multrec 69360 15.2 29.686 35.551 39.723 45.219 mp_waitall_1 488190 16.1 34.811 41.685 34.811 41.685 qs_ot_get_p 112 10.4 0.001 0.017 40.672 40.949 qs_ot_get_derivative 55 11.6 0.001 0.001 38.329 38.483 multiply_cannon_sync_h2d 69360 15.2 29.082 33.307 29.082 33.307 multiply_cannon_metrocomm3 69360 15.2 0.199 0.212 25.694 32.783 qs_ot_p2m_diag 40 11.0 0.022 0.053 29.781 29.908 rebuild_ks_matrix 110 8.4 0.000 0.000 29.150 29.306 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.054 29.150 29.306 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.477 26.622 cp_dbcsr_syevd 40 12.0 0.002 0.008 26.572 26.575 init_scf_loop 7 6.6 0.000 0.009 26.511 26.514 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.069 23.322 apply_single 62 13.6 0.000 0.000 23.069 23.322 cp_fm_syevd 40 13.0 0.000 0.003 21.562 21.712 prepare_preconditioner 7 7.6 0.000 0.000 21.630 21.658 make_preconditioner 7 8.6 0.001 0.014 21.630 21.658 ot_new_cg_direction 55 11.6 0.001 0.001 21.066 21.067 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.756 17.111 calculate_rho_elec 110 8.6 0.030 0.032 16.755 17.111 cp_fm_redistribute_end 40 14.0 8.480 16.918 8.485 16.919 cp_fm_syevd_base 40 14.0 8.426 16.865 8.426 16.865 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.780 14.989 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.486 14.552 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.958 14.074 mp_sum_l 4764 12.2 12.088 13.188 12.088 13.188 init_scf_run 5 5.8 0.000 0.001 12.543 12.544 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.543 12.544 pw_transfer 1645 12.4 0.079 0.104 11.950 12.193 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.812 12.058 calculate_dm_sparse 110 9.5 0.000 0.000 11.566 11.822 density_rs2pw 110 9.6 0.005 0.006 11.034 11.477 dbcsr_mm_accdrv_process 154766 15.8 6.208 6.422 9.905 10.765 fft_wrap_pw1pw2_240 915 15.0 0.828 0.930 10.398 10.676 qs_vxc_create 110 10.4 0.003 0.013 10.545 10.602 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.271 10.378 cp_fm_cholesky_invert 7 10.6 10.185 10.193 10.185 10.193 check_diag 80 13.5 8.594 8.886 9.297 9.450 fft3d_pb 915 16.0 2.396 2.686 8.600 8.851 sum_up_and_integrate 60 10.3 0.001 0.004 8.579 8.596 integrate_v_rspace 60 11.3 0.001 0.008 8.562 8.579 acc_transpose_blocks 69360 15.2 0.358 0.374 7.995 8.425 transfer_rs2pw 445 10.6 0.007 0.008 7.666 8.166 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.883 7.900 xc_rho_set_and_dset_create 110 12.4 0.078 0.105 7.386 7.659 calculate_first_density_matrix 1 7.0 0.001 0.014 7.610 7.643 multiply_cannon_metrocomm1 69360 15.2 0.098 0.105 4.575 7.568 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.412 7.464 xc_vxc_pw_create 60 11.3 0.039 0.051 7.090 7.148 make_m2s 2890 13.2 0.078 0.087 6.368 7.044 make_images 2890 14.2 0.237 0.257 6.260 6.932 make_full_single_inverse 7 9.6 0.002 0.023 6.873 6.902 xc_pw_derive 510 13.4 0.005 0.007 6.312 6.409 mp_alltoall_z22v 2340 17.7 5.630 5.899 5.630 5.899 acc_transpose_blocks_kernels 69360 16.2 0.846 0.894 5.245 5.540 mp_waitany 7680 13.5 4.407 5.040 4.407 5.040 potential_pw2rs 60 12.3 0.002 0.003 4.781 4.817 multiply_cannon_metrocomm4 67915 15.2 0.189 0.205 2.029 4.743 jit_kernel_transpose 5 15.0 4.400 4.651 4.400 4.651 wfi_extrapolate 5 7.8 0.000 0.001 4.599 4.599 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=226.991000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=564.400000, yerr=2.154066 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e3caab925d13535c28a7f61e60f70e91adc8e5da Summary: empty Status: OK