=== 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: 5def639336615614009cc9e019e9532e6735d0e9 ################# 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 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/01 job id: 44634431 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/02 job id: 44634434 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/03 job id: 44634435 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/04 job id: 44634436 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/05 job id: 44634437 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/06 job id: 44634438 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/07 job id: 44634439 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/08 job id: 44634440 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/09 job id: 44634441 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/10 job id: 44634442 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/11 job id: 44634444 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/12 job id: 44634445 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/13 job id: 44634447 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/14 job id: 44634451 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/15 job id: 44634452 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/16 job id: 44634454 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/17 job id: 44634456 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/18 job id: 44634460 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/19 job id: 44634466 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/20 job id: 44634474 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/21 job id: 44634475 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/22 job id: 44634477 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/23 job id: 44634478 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/24 job id: 44634479 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/25 job id: 44634480 --- 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/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/26 job id: 44634481 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 344 9. 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 0.021 0.037 134.597 134.598 farming_run 1 2.0 134.109 134.125 134.569 134.572 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.468211E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 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 0.010 0.026 115.065 115.066 qs_energies 1 2.0 0.000 0.000 114.860 114.862 mp2_main 1 3.0 0.000 0.000 112.929 112.931 mp2_gpw_main 1 4.0 0.020 0.027 112.082 112.084 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.205 93.773 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.437 56.005 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.174 41.747 47.234 get_2c_integrals 1 6.0 0.000 0.000 37.101 37.595 integrate_v_rspace 273 8.0 0.438 0.452 25.103 30.322 pw_transfer 6555 10.6 0.376 0.388 27.393 28.038 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.076 26.726 grid_integrate_task_list 273 9.0 20.910 26.627 20.910 26.627 fft_wrap_pw1pw2_100 2178 12.4 1.188 1.342 23.593 24.259 compute_2c_integrals 1 7.0 0.006 0.008 19.248 19.249 rpa_ri_compute_en 1 5.0 0.001 0.001 18.772 19.003 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.865 18.972 mp2_eri_2c_integrate_gpw 1 9.0 2.386 2.449 18.862 18.968 cp_fm_cholesky_decompose 12 8.2 17.822 18.357 17.822 18.357 cholesky_decomp 1 7.0 0.000 0.000 16.710 17.235 fft3d_s 5443 13.4 16.119 16.673 16.142 16.695 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.559 10.850 15.559 calculate_wavefunction 272 8.0 5.425 5.570 12.566 13.181 rpa_num_int 1 6.0 0.000 0.000 10.519 10.519 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.466 10.483 calc_mat_Q 8 8.0 0.000 0.000 9.302 9.399 contract_S_to_Q 8 9.0 0.000 0.000 8.723 8.818 calc_potential_gpw 544 9.5 0.005 0.006 8.254 8.587 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.209 8.456 potential_pw2rs 545 10.0 0.108 0.111 7.685 8.401 parallel_gemm_fm 14 9.1 0.000 0.000 8.294 8.400 parallel_gemm_fm_cosma 14 10.1 8.294 8.400 8.294 8.400 collocate_single_gaussian 272 10.0 0.040 0.042 7.447 7.724 create_integ_mat 1 6.0 0.022 0.027 7.592 7.592 array2fm 1 7.0 0.000 0.000 6.591 7.118 pw_scatter_s 2720 13.7 4.462 4.692 4.462 4.692 pw_gather_s 2722 13.2 3.902 4.257 3.902 4.257 array2fm_buffer_send 1 8.0 2.869 3.009 2.869 3.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.083905, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2739.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 344 10. 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 0.033 0.060 399.157 399.161 farming_run 1 2.0 398.242 398.245 399.063 399.084 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.219371E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. 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 0.015 0.052 210.365 210.365 qs_energies 1 2.0 0.000 0.000 210.105 210.116 scf_env_do_scf 1 3.0 0.000 0.000 106.915 106.915 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.057 106.064 rebuild_ks_matrix 4 6.0 0.000 0.000 106.056 106.063 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 106.056 106.063 hfx_ks_matrix 4 8.0 0.001 0.001 105.679 105.683 integrate_four_center 4 9.0 0.144 0.458 105.679 105.682 mp2_main 1 3.0 0.000 0.000 102.903 102.915 mp2_gpw_main 1 4.0 0.033 0.048 102.082 102.095 integrate_four_center_main 4 10.0 0.149 0.551 96.703 99.873 integrate_four_center_bin 265 11.0 96.554 99.550 96.554 99.550 init_scf_loop 1 4.0 0.000 0.000 92.728 92.728 mp2_ri_gpw_compute_in 1 5.0 0.070 0.089 75.244 76.484 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.762 56.006 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 42.344 47.479 integrate_v_rspace 95 8.0 0.400 0.567 28.678 33.798 pw_transfer 2240 10.6 0.144 0.160 29.882 30.294 ao_to_mo_and_store_B_mult_1 91 7.0 10.732 29.745 10.732 29.745 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.892 29.265 grid_integrate_task_list 95 9.0 23.980 29.252 23.980 29.252 mp2_ri_gpw_compute_en 1 5.0 0.062 0.074 26.688 28.666 fft_wrap_pw1pw2_100 730 12.4 1.270 1.461 26.603 27.008 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.915 24.775 24.786 get_2c_integrals 1 6.0 0.000 0.000 20.379 20.419 compute_2c_integrals 1 7.0 0.007 0.021 19.360 19.371 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.853 19.215 mp2_eri_2c_integrate_gpw 1 9.0 1.748 1.889 18.850 19.214 fft3d_s 1823 13.4 18.434 18.738 18.448 18.751 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.186 14.186 calculate_wavefunction 91 8.0 2.051 2.096 9.775 9.989 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.587 8.909 9.253 potential_pw2rs 186 10.0 0.033 0.035 8.610 9.236 local_gemm 172 8.0 8.353 8.677 8.353 8.677 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.225 8.606 mp2_ri_gpw_compute_en_comm 22 7.0 0.528 0.552 7.635 8.169 calc_potential_gpw 182 9.5 0.002 0.002 7.909 8.147 collocate_single_gaussian 91 10.0 0.017 0.019 7.865 8.141 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.434 6.340 6.434 mp_sync 38 10.4 3.687 6.282 3.687 6.282 mp_sendrecv_dm3 2068 8.0 5.598 6.159 5.598 6.159 pw_gather_s 912 13.2 4.883 5.459 4.883 5.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.083394, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 452.067328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578368. 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 0.021 0.073 52.047 52.048 qs_mol_dyn_low 1 2.0 0.003 0.004 51.534 51.554 qs_forces 11 3.9 0.002 0.005 51.413 51.415 qs_energies 11 4.9 0.008 0.012 49.538 49.553 scf_env_do_scf 11 5.9 0.000 0.001 43.097 43.098 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.006 41.007 dbcsr_multiply_generic 2286 12.5 0.094 0.097 32.898 33.272 qs_scf_new_mos 108 7.5 0.000 0.001 31.087 31.403 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.087 31.402 ot_scf_mini 108 9.5 0.002 0.002 29.481 29.676 multiply_cannon 2286 13.5 0.181 0.191 25.916 27.279 multiply_cannon_loop 2286 14.5 1.500 1.570 25.244 26.654 velocity_verlet 10 3.0 0.001 0.001 24.857 24.858 ot_mini 108 10.5 0.001 0.001 18.464 18.717 qs_ot_get_derivative 108 11.5 0.001 0.001 15.539 15.736 mp_waitall_1 245248 16.5 8.073 15.045 8.073 15.045 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 5.841 13.056 multiply_cannon_multrec 54864 15.5 4.234 6.465 7.860 11.435 rebuild_ks_matrix 119 8.3 0.000 0.000 7.811 7.943 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.810 7.942 multiply_cannon_sync_h2d 54864 15.5 5.940 7.455 5.940 7.455 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.888 7.016 qs_ot_get_p 119 10.4 0.001 0.001 6.456 6.789 mp_sum_l 7207 12.9 5.109 6.731 5.109 6.731 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.128 5.572 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.311 5.444 init_scf_run 11 5.9 0.000 0.001 5.206 5.207 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.206 5.207 dbcsr_mm_accdrv_process 76910 16.1 1.183 1.839 3.548 4.978 sum_up_and_integrate 119 10.3 0.012 0.015 4.422 4.428 integrate_v_rspace 119 11.3 0.002 0.003 4.409 4.416 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.088 4.199 calculate_rho_elec 119 8.7 0.011 0.017 4.087 4.198 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.671 3.746 calculate_first_density_matrix 1 7.0 0.000 0.001 3.065 3.083 jit_kernel_multiply 13 15.8 2.302 3.062 2.302 3.062 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.861 3.058 apply_single 119 13.6 0.000 0.000 2.861 3.057 calculate_dm_sparse 119 9.5 0.000 0.001 2.816 2.978 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.841 2.841 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.642 2.787 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.767 2.769 rs_pw_transfer 974 11.9 0.011 0.012 2.623 2.729 ot_diis_step 108 11.5 0.006 0.006 2.650 2.651 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.643 2.643 cp_fm_redistribute_end 50 14.0 2.406 2.622 2.411 2.623 cp_fm_diag_elpa_base 50 14.0 0.210 2.539 0.211 2.548 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.493 2.533 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.410 2.463 mp_sum_d 4127 12.0 1.679 2.337 1.679 2.337 density_rs2pw 119 9.7 0.004 0.004 2.147 2.248 acc_transpose_blocks 54864 15.5 0.234 0.253 1.755 2.173 grid_integrate_task_list 119 12.3 2.036 2.129 2.036 2.129 init_scf_loop 11 6.9 0.000 0.000 2.073 2.073 wfi_extrapolate 11 7.9 0.001 0.001 2.069 2.070 potential_pw2rs 119 12.3 0.004 0.004 1.801 1.813 pw_transfer 1439 11.6 0.053 0.059 1.730 1.804 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.653 1.731 make_m2s 4572 13.5 0.054 0.056 1.569 1.610 make_images 4572 14.5 0.132 0.138 1.487 1.529 fft3d_ps 1201 14.6 0.367 0.471 1.423 1.494 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.368 1.401 grid_collocate_task_list 119 9.7 1.295 1.377 1.295 1.377 mp_alltoall_d11v 2130 13.8 1.168 1.343 1.168 1.343 fft_wrap_pw1pw2_140 487 13.2 0.083 0.095 1.263 1.342 mp_waitany 12084 13.8 1.155 1.287 1.155 1.287 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.711 1.102 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.048000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.316928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1030242. 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 0.102 0.123 40.708 40.713 qs_mol_dyn_low 1 2.0 0.003 0.003 39.939 39.946 qs_forces 11 3.9 0.025 0.042 39.202 39.202 qs_energies 11 4.9 0.002 0.006 37.469 37.491 scf_env_do_scf 11 5.9 0.001 0.001 31.350 31.351 scf_env_do_scf_inner_loop 108 6.5 0.006 0.030 28.883 28.883 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.843 22.198 qs_scf_new_mos 108 7.5 0.001 0.001 19.829 20.075 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.828 20.075 ot_scf_mini 108 9.5 0.002 0.003 18.964 19.137 velocity_verlet 10 3.0 0.001 0.001 18.582 18.597 multiply_cannon 2286 13.5 0.224 0.300 16.387 17.901 multiply_cannon_loop 2286 14.5 0.907 0.983 15.305 16.815 ot_mini 108 10.5 0.001 0.001 11.840 12.075 mp_waitall_1 200699 16.5 5.769 11.296 5.769 11.296 qs_ot_get_derivative 108 11.5 0.001 0.001 9.390 9.563 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.171 9.556 multiply_cannon_multrec 27432 15.5 1.977 4.375 5.900 8.626 rebuild_ks_matrix 119 8.3 0.000 0.000 7.150 7.288 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.149 7.287 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.320 6.445 dbcsr_mm_accdrv_process 47894 16.0 2.981 4.980 3.854 5.707 init_scf_run 11 5.9 0.000 0.001 4.771 4.771 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.771 4.771 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.559 4.586 qs_ot_get_p 119 10.4 0.001 0.001 4.270 4.512 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.221 4.316 apply_single 119 13.6 0.000 0.000 3.221 4.315 mp_sum_l 7207 12.9 2.223 4.202 2.223 4.202 sum_up_and_integrate 119 10.3 0.024 0.027 4.088 4.093 integrate_v_rspace 119 11.3 0.002 0.003 4.064 4.071 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.751 3.783 calculate_rho_elec 119 8.7 0.021 0.024 3.751 3.783 calculate_first_density_matrix 1 7.0 0.000 0.002 3.281 3.283 make_m2s 4572 13.5 0.052 0.054 2.830 3.056 make_images 4572 14.5 0.199 0.235 2.742 2.966 multiply_cannon_sync_h2d 27432 15.5 2.212 2.903 2.212 2.903 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.884 2.903 rs_pw_transfer 974 11.9 0.010 0.011 2.476 2.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.530 2.533 density_rs2pw 119 9.7 0.004 0.004 2.068 2.482 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.455 2.455 init_scf_loop 11 6.9 0.000 0.002 2.449 2.449 ot_diis_step 108 11.5 0.011 0.011 2.400 2.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.231 2.274 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.088 2.178 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.141 2.142 cp_fm_redistribute_end 50 14.0 1.775 2.120 1.778 2.121 calculate_dm_sparse 119 9.5 0.000 0.001 2.029 2.107 cp_fm_diag_elpa_base 50 14.0 0.329 2.013 0.341 2.065 jit_kernel_multiply 10 15.9 0.821 1.983 0.821 1.983 pw_transfer 1439 11.6 0.065 0.072 1.914 1.951 grid_integrate_task_list 119 12.3 1.828 1.935 1.828 1.935 make_images_data 4572 15.5 0.044 0.050 1.278 1.904 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.822 1.862 potential_pw2rs 119 12.3 0.006 0.006 1.831 1.840 acc_transpose_blocks 27432 15.5 0.108 0.114 1.206 1.691 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.134 1.618 fft3d_ps 1201 14.6 0.510 0.562 1.524 1.559 prepare_preconditioner 11 7.9 0.000 0.000 1.487 1.513 make_preconditioner 11 8.9 0.000 0.001 1.487 1.513 mp_alltoall_d11v 2130 13.8 1.155 1.488 1.155 1.488 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.445 1.483 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.383 1.440 wfi_extrapolate 11 7.9 0.001 0.001 1.435 1.435 grid_collocate_task_list 119 9.7 1.230 1.370 1.230 1.370 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.270 1.279 mp_allgather_i34 2286 14.5 0.505 1.272 0.505 1.272 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.143 1.187 mp_sum_d 4127 12.0 0.649 1.100 0.649 1.100 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.672 1.061 qs_energies_init_hamiltonians 11 5.9 0.026 0.041 1.010 1.013 mp_waitany 5720 13.7 0.525 0.982 0.525 0.982 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.158 0.554 0.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.901 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.713000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.895936E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 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 0.540 0.609 35.031 35.040 qs_mol_dyn_low 1 2.0 0.003 0.005 33.900 33.907 qs_forces 11 3.9 0.003 0.006 33.799 33.801 qs_energies 11 4.9 0.007 0.039 32.073 32.135 scf_env_do_scf 11 5.9 0.012 0.055 26.043 26.044 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.320 23.321 dbcsr_multiply_generic 2286 12.5 0.094 0.096 17.350 17.440 qs_scf_new_mos 108 7.5 0.001 0.001 15.260 15.274 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.259 15.273 velocity_verlet 10 3.0 0.001 0.002 15.044 15.047 multiply_cannon 2286 13.5 0.195 0.205 14.090 14.845 ot_scf_mini 108 9.5 0.006 0.017 14.548 14.560 multiply_cannon_loop 2286 14.5 0.637 0.664 13.313 14.070 ot_mini 108 10.5 0.001 0.001 9.236 9.252 multiply_cannon_multrec 18288 15.5 2.023 3.365 7.266 7.939 qs_ot_get_derivative 108 11.5 0.001 0.001 7.746 7.761 rebuild_ks_matrix 119 8.3 0.000 0.000 6.506 6.520 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.505 6.520 dbcsr_mm_accdrv_process 38222 16.0 4.513 5.861 5.158 6.018 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.678 5.694 init_scf_run 11 5.9 0.000 0.002 4.811 4.812 scf_env_initial_rho_setup 11 6.9 0.018 0.139 4.811 4.811 mp_waitall_1 158411 16.6 2.908 4.276 2.908 4.276 sum_up_and_integrate 119 10.3 0.032 0.035 4.008 4.051 integrate_v_rspace 119 11.3 0.003 0.005 3.976 4.025 qs_rho_update_rho_low 119 7.7 0.008 0.064 3.676 3.696 calculate_rho_elec 119 8.7 0.045 0.146 3.668 3.696 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.732 3.437 calculate_first_density_matrix 1 7.0 0.001 0.003 3.293 3.294 qs_ot_get_p 119 10.4 0.001 0.007 3.221 3.248 init_scf_loop 11 6.9 0.003 0.011 2.653 2.655 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.585 2.593 rs_pw_transfer 974 11.9 0.009 0.010 2.255 2.500 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.108 2.476 apply_single 119 13.6 0.000 0.000 2.108 2.476 density_rs2pw 119 9.7 0.004 0.004 1.938 2.199 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.146 2.151 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.090 2.092 pw_transfer 1439 11.6 0.066 0.071 1.954 1.998 calculate_dm_sparse 119 9.5 0.000 0.000 1.982 1.996 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.954 1.959 grid_integrate_task_list 119 12.3 1.792 1.910 1.792 1.910 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.860 1.907 mp_sum_l 7207 12.9 1.186 1.893 1.186 1.893 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.886 1.887 make_m2s 4572 13.5 0.044 0.045 1.702 1.869 potential_pw2rs 119 12.3 0.007 0.008 1.798 1.845 make_images 4572 14.5 0.189 0.200 1.618 1.781 prepare_preconditioner 11 7.9 0.000 0.000 1.697 1.700 make_preconditioner 11 8.9 0.000 0.001 1.697 1.700 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.666 1.667 cp_fm_redistribute_end 50 14.0 1.243 1.646 1.244 1.647 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.554 1.636 cp_fm_diag_elpa_base 50 14.0 0.387 1.568 0.401 1.609 fft3d_ps 1201 14.6 0.521 0.539 1.540 1.587 multiply_cannon_sync_h2d 18288 15.5 1.420 1.563 1.420 1.563 fft_wrap_pw1pw2_140 487 13.2 0.090 0.093 1.505 1.552 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.522 1.528 ot_diis_step 108 11.5 0.011 0.013 1.470 1.470 jit_kernel_multiply 6 16.3 0.594 1.450 0.594 1.450 acc_transpose_blocks 18288 15.5 0.075 0.080 1.384 1.414 grid_collocate_task_list 119 9.7 1.216 1.387 1.216 1.387 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.361 1.367 wfi_extrapolate 11 7.9 0.001 0.001 1.142 1.142 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.625 1.109 mp_alltoall_d11v 2130 13.8 0.864 1.025 0.864 1.025 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.974 0.976 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.944 0.962 make_images_data 4572 15.5 0.044 0.048 0.730 0.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.907 0.909 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.862 0.886 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.633 0.819 mp_waitany 9880 13.7 0.538 0.804 0.538 0.804 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.520 0.769 mp_alltoall_z22v 1201 16.6 0.626 0.756 0.626 0.756 parallel_gemm_fm 81 9.0 0.000 0.000 0.749 0.753 parallel_gemm_fm_cosma 81 10.0 0.748 0.753 0.748 0.753 jit_kernel_transpose 5 15.6 0.734 0.747 0.734 0.747 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.663 0.730 cp_fm_cholesky_invert 11 10.9 0.709 0.713 0.709 0.713 mp_sum_d 4127 12.0 0.488 0.710 0.488 0.710 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.040000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 552.841216E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.060 0.088 35.696 35.702 qs_mol_dyn_low 1 2.0 0.003 0.003 35.210 35.338 qs_forces 11 3.9 0.002 0.002 35.147 35.148 qs_energies 11 4.9 0.001 0.001 33.433 33.438 scf_env_do_scf 11 5.9 0.000 0.001 27.607 27.609 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.207 24.207 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.630 18.735 velocity_verlet 10 3.0 0.001 0.001 17.634 17.636 qs_scf_new_mos 108 7.5 0.001 0.001 15.807 15.861 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.807 15.860 multiply_cannon 2286 13.5 0.232 0.273 15.138 15.566 ot_scf_mini 108 9.5 0.002 0.003 14.880 14.932 multiply_cannon_loop 2286 14.5 0.939 0.973 14.212 14.592 multiply_cannon_multrec 27432 15.5 2.345 3.056 9.224 9.656 ot_mini 108 10.5 0.001 0.001 9.052 9.117 dbcsr_mm_accdrv_process 47916 15.9 5.793 7.829 6.786 8.038 qs_ot_get_derivative 108 11.5 0.001 0.001 7.248 7.300 rebuild_ks_matrix 119 8.3 0.000 0.000 6.539 6.593 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.538 6.593 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.804 5.853 init_scf_run 11 5.9 0.000 0.001 4.489 4.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.488 4.489 sum_up_and_integrate 119 10.3 0.036 0.039 3.780 3.787 integrate_v_rspace 119 11.3 0.002 0.002 3.744 3.753 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.516 3.549 calculate_rho_elec 119 8.7 0.040 0.046 3.515 3.549 init_scf_loop 11 6.9 0.000 0.000 3.381 3.381 qs_ot_get_p 119 10.4 0.001 0.001 3.252 3.325 calculate_first_density_matrix 1 7.0 0.000 0.000 3.126 3.127 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.657 3.110 prepare_preconditioner 11 7.9 0.000 0.000 2.536 2.545 make_preconditioner 11 8.9 0.000 0.000 2.536 2.544 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.084 2.526 apply_single 119 13.6 0.000 0.000 2.083 2.526 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.523 2.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.144 2.471 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.357 2.371 make_m2s 4572 13.5 0.054 0.055 2.097 2.215 calculate_dm_sparse 119 9.5 0.000 0.000 2.150 2.209 jit_kernel_multiply 8 16.1 0.933 2.166 0.933 2.166 density_rs2pw 119 9.7 0.004 0.004 1.945 2.128 rs_pw_transfer 974 11.9 0.009 0.009 1.952 2.127 mp_waitall_1 137007 16.6 1.632 2.124 1.632 2.124 make_images 4572 14.5 0.271 0.332 1.990 2.108 pw_transfer 1439 11.6 0.066 0.072 2.064 2.103 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.016 2.024 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.970 2.014 grid_integrate_task_list 119 12.3 1.831 1.933 1.831 1.933 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.874 1.900 ot_diis_step 108 11.5 0.012 0.012 1.764 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.705 1.706 fft3d_ps 1201 14.6 0.557 0.610 1.645 1.684 fft_wrap_pw1pw2_140 487 13.2 0.089 0.095 1.634 1.680 acc_transpose_blocks 27432 15.5 0.111 0.113 1.635 1.670 potential_pw2rs 119 12.3 0.008 0.010 1.550 1.555 mp_sum_l 7207 12.9 0.969 1.541 0.969 1.541 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.477 1.478 cp_fm_redistribute_end 50 14.0 0.978 1.454 0.979 1.455 cp_fm_diag_elpa_base 50 14.0 0.454 1.391 0.474 1.429 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.342 1.354 grid_collocate_task_list 119 9.7 1.220 1.325 1.220 1.325 wfi_extrapolate 11 7.9 0.001 0.001 1.315 1.315 multiply_cannon_metrocomm3 27432 15.5 0.037 0.039 0.688 1.174 cp_fm_upper_to_full 72 13.5 0.803 1.114 0.803 1.114 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.089 1.111 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.090 1.091 multiply_cannon_sync_h2d 27432 15.5 1.003 1.063 1.003 1.063 dbcsr_complete_redistribute 329 12.2 0.120 0.147 0.758 1.034 acc_transpose_blocks_kernels 27432 16.5 0.267 0.276 1.011 1.030 make_images_data 4572 15.5 0.045 0.049 0.787 0.913 multiply_cannon_metrocomm1 27432 15.5 0.032 0.033 0.208 0.894 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.873 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.684 0.861 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.562 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.824 0.829 mp_alltoall_d11v 2130 13.8 0.678 0.825 0.678 0.825 jit_kernel_transpose 5 15.6 0.743 0.756 0.743 0.756 cp_fm_cholesky_invert 11 10.9 0.743 0.747 0.743 0.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.702000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.090909, yerr=2.466302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 600.358912E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 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 0.035 0.054 29.591 29.592 qs_mol_dyn_low 1 2.0 0.003 0.003 29.244 29.253 qs_forces 11 3.9 0.002 0.002 28.766 28.767 qs_energies 11 4.9 0.001 0.001 27.057 27.060 scf_env_do_scf 11 5.9 0.000 0.001 21.249 21.250 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.798 18.798 velocity_verlet 10 3.0 0.001 0.001 14.619 14.648 dbcsr_multiply_generic 2286 12.5 0.091 0.093 12.972 13.048 multiply_cannon 2286 13.5 0.231 0.237 10.569 10.977 qs_scf_new_mos 108 7.5 0.001 0.001 10.798 10.820 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.797 10.819 ot_scf_mini 108 9.5 0.002 0.002 10.138 10.163 multiply_cannon_loop 2286 14.5 0.330 0.340 9.679 9.873 multiply_cannon_multrec 9144 15.5 1.696 1.979 6.519 6.963 rebuild_ks_matrix 119 8.3 0.000 0.000 5.994 6.019 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.993 6.018 ot_mini 108 10.5 0.001 0.001 5.643 5.672 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.345 5.368 dbcsr_mm_accdrv_process 12550 15.8 3.486 4.603 4.722 4.982 qs_ot_get_derivative 108 11.5 0.001 0.001 4.359 4.383 init_scf_run 11 5.9 0.000 0.001 4.288 4.288 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.288 4.288 sum_up_and_integrate 119 10.3 0.038 0.041 3.661 3.666 integrate_v_rspace 119 11.3 0.003 0.003 3.623 3.629 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.533 3.543 calculate_rho_elec 119 8.7 0.060 0.061 3.533 3.543 calculate_first_density_matrix 1 7.0 0.000 0.000 3.170 3.172 qs_ot_get_p 119 10.4 0.001 0.001 2.743 2.786 jit_kernel_multiply 10 15.8 1.199 2.514 1.199 2.514 init_scf_loop 11 6.9 0.000 0.000 2.431 2.431 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.297 2.299 pw_transfer 1439 11.6 0.066 0.069 2.142 2.151 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.094 2.103 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.048 2.059 density_rs2pw 119 9.7 0.004 0.004 1.880 2.021 calculate_dm_sparse 119 9.5 0.000 0.000 1.971 1.991 mp_waitall_1 115863 16.7 1.464 1.988 1.464 1.988 grid_integrate_task_list 119 12.3 1.841 1.932 1.841 1.932 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.845 1.848 make_m2s 4572 13.5 0.034 0.036 1.679 1.843 rs_pw_transfer 974 11.9 0.008 0.008 1.653 1.776 make_images 4572 14.5 0.268 0.304 1.590 1.753 fft3d_ps 1201 14.6 0.560 0.571 1.707 1.716 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.698 1.713 prepare_preconditioner 11 7.9 0.000 0.000 1.677 1.683 make_preconditioner 11 8.9 0.000 0.000 1.677 1.683 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.655 1.655 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.566 1.590 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.523 1.536 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.405 1.410 potential_pw2rs 119 12.3 0.010 0.011 1.399 1.403 grid_collocate_task_list 119 9.7 1.270 1.390 1.270 1.390 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.381 1.381 cp_fm_redistribute_end 50 14.0 0.689 1.359 0.690 1.360 cp_fm_diag_elpa_base 50 14.0 0.626 1.289 0.669 1.346 ot_diis_step 108 11.5 0.012 0.012 1.272 1.273 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 1.248 1.249 acc_transpose_blocks 9144 15.5 0.037 0.038 1.131 1.205 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.150 1.169 apply_single 119 13.6 0.000 0.000 1.150 1.169 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.094 1.107 wfi_extrapolate 11 7.9 0.001 0.001 1.069 1.069 hybrid_alltoall_any 4725 16.4 0.062 0.173 0.748 0.995 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.892 0.960 make_images_data 4572 15.5 0.039 0.041 0.772 0.960 mp_alltoall_d11v 2130 13.8 0.741 0.926 0.741 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.920 jit_kernel_transpose 5 15.6 0.774 0.840 0.774 0.840 cp_fm_cholesky_invert 11 10.9 0.836 0.839 0.836 0.839 multiply_cannon_sync_h2d 9144 15.5 0.716 0.796 0.716 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.773 qs_env_update_s_mstruct 11 6.9 0.004 0.014 0.691 0.753 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.703 0.711 mp_alltoall_z22v 1201 16.6 0.594 0.650 0.594 0.650 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.320 0.646 mp_allgather_i34 2286 14.5 0.224 0.635 0.224 0.635 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.390 0.631 yz_to_x 606 15.1 0.266 0.277 0.576 0.596 dbcsr_complete_redistribute 329 12.2 0.181 0.221 0.564 0.592 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.592000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.818182, yerr=5.023910 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 756.039680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. 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 0.038 0.060 44.792 44.793 qs_mol_dyn_low 1 2.0 0.003 0.003 44.422 44.429 qs_forces 11 3.9 0.003 0.003 44.163 44.164 qs_energies 11 4.9 0.001 0.001 42.113 42.117 scf_env_do_scf 11 5.9 0.001 0.001 34.651 34.651 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.448 26.449 velocity_verlet 10 3.0 0.001 0.001 24.406 24.436 dbcsr_multiply_generic 2286 12.5 0.099 0.100 18.218 18.421 qs_scf_new_mos 108 7.5 0.001 0.001 16.316 16.406 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.316 16.406 ot_scf_mini 108 9.5 0.002 0.002 15.233 15.329 multiply_cannon 2286 13.5 0.365 0.373 14.004 15.018 multiply_cannon_loop 2286 14.5 0.345 0.350 12.569 13.595 ot_mini 108 10.5 0.001 0.001 9.134 9.244 multiply_cannon_multrec 9144 15.5 3.414 4.741 8.598 8.762 init_scf_loop 11 6.9 0.000 0.000 8.171 8.173 rebuild_ks_matrix 119 8.3 0.000 0.000 7.722 7.870 qs_ks_build_kohn_sham_matrix 119 9.3 0.078 0.095 7.721 7.869 qs_ot_get_derivative 108 11.5 0.001 0.001 7.072 7.167 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.981 7.118 prepare_preconditioner 11 7.9 0.000 0.000 7.004 7.038 make_preconditioner 11 8.9 0.000 0.000 7.004 7.038 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.604 6.905 dbcsr_mm_accdrv_process 12550 15.8 4.130 5.756 5.062 6.491 init_scf_run 11 5.9 0.000 0.001 5.372 5.372 scf_env_initial_rho_setup 11 6.9 0.005 0.005 5.372 5.372 cp_fm_upper_to_full 72 14.2 3.173 4.506 3.173 4.506 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.462 4.492 calculate_rho_elec 119 8.7 0.118 0.121 4.462 4.491 sum_up_and_integrate 119 10.3 0.065 0.067 4.188 4.200 integrate_v_rspace 119 11.3 0.003 0.003 4.122 4.135 calculate_first_density_matrix 1 7.0 0.000 0.000 3.802 3.834 mp_waitall_1 94719 16.7 2.593 3.463 2.593 3.463 qs_ot_get_p 119 10.4 0.001 0.001 3.327 3.455 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.589 3.105 dbcsr_complete_redistribute 329 12.2 0.403 0.447 2.148 2.981 pw_transfer 1439 11.6 0.069 0.070 2.906 2.925 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.807 2.826 make_m2s 4572 13.5 0.038 0.038 2.560 2.812 make_images 4572 14.5 0.353 0.386 2.439 2.691 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.324 2.645 apply_single 119 13.6 0.000 0.000 2.324 2.645 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.806 2.622 fft3d_ps 1201 14.6 0.594 0.603 2.431 2.447 density_rs2pw 119 9.7 0.004 0.004 2.352 2.378 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.348 2.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.348 2.349 calculate_dm_sparse 119 9.5 0.000 0.000 2.254 2.278 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.359 2.237 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.395 2.205 mp_alltoall_i22 627 13.8 1.369 2.151 1.369 2.151 grid_integrate_task_list 119 12.3 2.062 2.080 2.062 2.080 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.048 2.050 ot_diis_step 108 11.5 0.014 0.014 2.039 2.039 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.931 1.970 mp_sum_l 7207 12.9 1.122 1.832 1.122 1.832 jit_kernel_multiply 10 15.4 0.905 1.805 0.905 1.805 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.790 1.791 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.770 1.770 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.685 1.734 rs_pw_transfer 974 11.9 0.009 0.010 1.657 1.700 potential_pw2rs 119 12.3 0.014 0.015 1.573 1.587 cp_fm_cholesky_invert 11 10.9 1.582 1.585 1.582 1.585 make_images_data 4572 15.5 0.043 0.046 1.156 1.555 grid_collocate_task_list 119 9.7 1.493 1.514 1.493 1.514 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.172 1.481 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.471 1.472 wfi_extrapolate 11 7.9 0.001 0.001 1.470 1.470 cp_fm_diag_elpa_base 50 14.0 1.331 1.381 1.470 1.470 make_basis_sm 11 9.8 0.001 0.001 1.465 1.466 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.335 1.363 parallel_gemm_fm 81 9.0 0.000 0.000 1.285 1.319 parallel_gemm_fm_cosma 81 10.0 1.285 1.319 1.285 1.319 mp_alltoall_d11v 2130 13.8 1.143 1.183 1.143 1.183 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.151 1.168 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.129 1.144 acc_transpose_blocks 9144 15.5 0.038 0.038 1.053 1.086 multiply_cannon_sync_h2d 9144 15.5 1.045 1.050 1.045 1.050 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.039 yz_to_x 606 15.1 0.459 0.469 0.990 1.002 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.952 0.964 qs_create_task_list 11 7.9 0.000 0.000 0.941 0.954 generate_qs_task_list 11 8.9 0.370 0.389 0.941 0.954 mp_alltoall_z22v 1201 16.6 0.888 0.924 0.888 0.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.793000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.727273, yerr=12.792689 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.714944E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.039 0.060 84.500 84.501 qs_mol_dyn_low 1 2.0 0.003 0.003 84.083 84.093 qs_forces 11 3.9 0.007 0.015 84.016 84.018 qs_energies 11 4.9 0.001 0.001 81.158 81.178 scf_env_do_scf 11 5.9 0.000 0.001 71.120 71.122 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.689 65.689 dbcsr_multiply_generic 2055 12.4 0.107 0.109 52.951 53.231 qs_scf_new_mos 99 7.5 0.000 0.001 48.270 48.405 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.270 48.404 ot_scf_mini 99 9.5 0.002 0.002 45.835 45.959 multiply_cannon 2055 13.4 0.176 0.194 43.334 44.278 multiply_cannon_loop 2055 14.4 1.551 1.590 42.432 43.354 velocity_verlet 10 3.0 0.001 0.001 41.991 41.992 ot_mini 99 10.5 0.001 0.001 27.625 27.739 qs_ot_get_derivative 99 11.5 0.001 0.001 20.803 20.922 multiply_cannon_multrec 49320 15.4 12.197 12.978 17.632 18.528 rebuild_ks_matrix 110 8.3 0.000 0.001 14.593 14.715 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.592 14.715 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.802 12.918 mp_waitall_1 220248 16.4 11.458 12.384 11.458 12.384 multiply_cannon_sync_h2d 49320 15.4 9.980 10.691 9.980 10.691 qs_ot_get_p 110 10.4 0.001 0.001 9.589 9.695 multiply_cannon_metrocomm3 49320 15.4 0.079 0.086 7.034 8.163 init_scf_run 11 5.9 0.000 0.001 7.975 7.975 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.975 7.975 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.208 7.758 apply_single 110 13.6 0.000 0.001 7.208 7.758 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.897 7.470 sum_up_and_integrate 110 10.3 0.036 0.043 7.090 7.103 integrate_v_rspace 110 11.3 0.003 0.003 7.053 7.074 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.824 6.897 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.631 6.758 calculate_rho_elec 110 8.6 0.021 0.026 6.630 6.757 ot_diis_step 99 11.5 0.005 0.006 6.560 6.561 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.395 6.424 mp_sum_l 6514 12.8 5.338 6.076 5.338 6.076 dbcsr_mm_accdrv_process 87628 16.1 2.069 2.184 5.314 5.639 init_scf_loop 11 6.9 0.000 0.000 5.402 5.403 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.358 5.359 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.900 4.901 cp_fm_redistribute_end 48 14.0 4.278 4.877 4.281 4.878 cp_fm_diag_elpa_base 48 14.0 0.590 4.735 0.594 4.762 calculate_dm_sparse 110 9.5 0.000 0.001 4.196 4.306 rs_pw_transfer 902 11.9 0.012 0.013 3.669 4.285 wfi_extrapolate 11 7.9 0.001 0.001 4.029 4.029 density_rs2pw 110 9.6 0.004 0.005 3.422 3.977 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.919 3.923 make_m2s 4110 13.4 0.061 0.065 3.818 3.918 calculate_first_density_matrix 1 7.0 0.000 0.000 3.859 3.862 make_images 4110 14.4 0.177 0.190 3.722 3.823 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.629 3.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.566 3.618 grid_integrate_task_list 110 12.3 3.233 3.404 3.233 3.404 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.338 3.396 pw_transfer 1331 11.6 0.055 0.066 3.225 3.297 prepare_preconditioner 11 7.9 0.000 0.000 3.249 3.272 make_preconditioner 11 8.9 0.000 0.000 3.249 3.272 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.137 3.211 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.031 3.083 jit_kernel_multiply 13 15.9 2.962 3.024 2.962 3.024 fft_wrap_pw1pw2_140 451 13.1 0.172 0.192 2.673 2.757 fft3d_ps 1111 14.6 0.774 0.874 2.664 2.727 potential_pw2rs 110 12.3 0.006 0.007 2.621 2.639 acc_transpose_blocks 49320 15.4 0.226 0.236 2.539 2.613 mp_alltoall_d11v 2046 13.8 2.080 2.475 2.080 2.475 mp_waitany 14300 13.8 1.830 2.434 1.830 2.434 grid_collocate_task_list 110 9.6 2.082 2.319 2.082 2.319 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.033 2.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.930 1.944 mp_sum_d 3881 11.9 1.420 1.930 1.420 1.930 make_images_data 4110 15.4 0.042 0.045 1.740 1.895 hybrid_alltoall_any 4261 16.3 0.082 0.477 1.504 1.803 cp_fm_cholesky_invert 11 10.9 1.742 1.746 1.742 1.746 parallel_gemm_fm 81 9.0 0.000 0.000 1.722 1.726 parallel_gemm_fm_cosma 81 10.0 1.721 1.726 1.721 1.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.501000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.296242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 588.230656E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2462973. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.028 0.092 71.370 71.371 qs_mol_dyn_low 1 2.0 0.005 0.018 71.002 71.011 qs_forces 11 3.9 0.006 0.031 70.908 70.909 qs_energies 11 4.9 0.004 0.026 67.101 67.109 scf_env_do_scf 11 5.9 0.001 0.004 56.337 56.340 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.697 48.699 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.856 38.058 velocity_verlet 10 3.0 0.002 0.007 36.017 36.020 qs_scf_new_mos 99 7.5 0.001 0.001 32.311 32.429 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.310 32.429 multiply_cannon 2055 13.4 0.230 0.246 31.020 32.302 ot_scf_mini 99 9.5 0.003 0.005 30.661 30.769 multiply_cannon_loop 2055 14.4 0.931 0.959 29.632 30.583 ot_mini 99 10.5 0.001 0.001 18.011 18.132 multiply_cannon_multrec 24660 15.4 7.624 9.116 13.784 15.193 rebuild_ks_matrix 110 8.3 0.000 0.001 14.308 14.445 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.034 14.307 14.444 qs_ot_get_derivative 99 11.5 0.001 0.001 12.250 12.359 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.219 12.346 mp_waitall_1 176588 16.5 7.615 10.302 7.615 10.302 multiply_cannon_sync_h2d 24660 15.4 6.960 8.234 6.960 8.234 init_scf_run 11 5.9 0.000 0.001 8.010 8.011 scf_env_initial_rho_setup 11 6.9 0.001 0.003 8.010 8.011 init_scf_loop 11 6.9 0.000 0.002 7.604 7.608 multiply_cannon_metrocomm3 24660 15.4 0.071 0.074 5.102 7.418 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.485 7.240 apply_single 110 13.6 0.000 0.001 6.484 7.240 sum_up_and_integrate 110 10.3 0.052 0.063 7.022 7.089 integrate_v_rspace 110 11.3 0.002 0.004 6.970 7.039 dbcsr_mm_accdrv_process 52282 16.1 4.617 5.381 6.001 6.280 qs_ot_get_p 110 10.4 0.001 0.004 6.124 6.266 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.216 6.226 calculate_rho_elec 110 8.6 0.039 0.048 6.215 6.226 ot_diis_step 99 11.5 0.010 0.011 5.714 5.714 prepare_preconditioner 11 7.9 0.000 0.000 5.468 5.489 make_preconditioner 11 8.9 0.000 0.001 5.468 5.489 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.616 5.326 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.062 5.220 make_m2s 4110 13.4 0.057 0.060 4.425 4.955 make_images 4110 14.4 0.398 0.442 4.316 4.845 calculate_first_density_matrix 1 7.0 0.061 0.484 4.387 4.390 pw_transfer 1331 11.6 0.066 0.073 4.132 4.329 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.216 4.236 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.025 4.225 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.158 4.220 density_rs2pw 110 9.6 0.004 0.004 3.341 3.810 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.795 3.795 fft_wrap_pw1pw2_140 451 13.1 0.203 0.221 3.506 3.714 rs_pw_transfer 902 11.9 0.012 0.014 3.145 3.671 fft3d_ps 1111 14.6 1.100 1.313 3.376 3.561 wfi_extrapolate 11 7.9 0.001 0.002 3.518 3.518 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.403 3.448 grid_integrate_task_list 110 12.3 3.162 3.339 3.162 3.339 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.279 3.281 cp_fm_redistribute_end 48 14.0 2.454 3.256 2.456 3.257 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.139 3.205 cp_fm_diag_elpa_base 48 14.0 0.767 3.117 0.798 3.204 make_images_data 4110 15.4 0.046 0.050 2.442 2.988 calculate_dm_sparse 110 9.5 0.001 0.001 2.947 2.979 potential_pw2rs 110 12.3 0.008 0.008 2.806 2.873 hybrid_alltoall_any 4261 16.3 0.102 0.440 2.115 2.822 cp_fm_cholesky_invert 11 10.9 2.500 2.507 2.500 2.507 grid_collocate_task_list 110 9.6 2.057 2.475 2.057 2.475 mp_sum_l 6514 12.8 1.827 2.401 1.827 2.401 make_basis_sm 11 9.8 0.013 0.025 2.243 2.305 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.254 2.266 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 2.147 2.149 mp_alltoall_d11v 2046 13.8 1.689 2.037 1.689 2.037 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.927 1.948 mp_waitany 10164 13.8 1.432 1.921 1.432 1.921 jit_kernel_multiply 10 16.4 1.037 1.913 1.037 1.913 dbcsr_complete_redistribute 325 12.2 0.271 0.334 1.589 1.910 mp_allgather_i34 2055 14.4 0.680 1.892 0.680 1.892 parallel_gemm_fm 81 9.0 0.000 0.000 1.762 1.769 parallel_gemm_fm_cosma 81 10.0 1.762 1.769 1.762 1.769 mp_alltoall_z22v 1111 16.6 1.602 1.763 1.602 1.763 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.810 1.719 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.217 1.169 1.697 acc_transpose_blocks 24660 15.4 0.110 0.114 1.625 1.692 mp_irecv_dv 57340 16.2 0.686 1.604 0.686 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.529 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.374 1.481 yz_to_x 451 15.2 0.270 0.308 1.336 1.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.371000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.454545, yerr=7.643936 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 659.816448E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 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 0.082 59.893 59.913 qs_mol_dyn_low 1 2.0 0.003 0.003 59.462 59.471 qs_forces 11 3.9 0.002 0.006 59.399 59.400 qs_energies 11 4.9 0.010 0.047 56.169 56.188 scf_env_do_scf 11 5.9 0.001 0.001 47.860 47.860 scf_env_do_scf_inner_loop 99 6.5 0.004 0.021 39.557 39.557 velocity_verlet 10 3.0 0.001 0.001 32.185 32.186 dbcsr_multiply_generic 2055 12.4 0.107 0.109 28.367 28.612 qs_scf_new_mos 99 7.5 0.001 0.001 24.433 24.527 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.432 24.526 multiply_cannon 2055 13.4 0.211 0.221 22.297 23.435 ot_scf_mini 99 9.5 0.003 0.004 23.222 23.350 multiply_cannon_loop 2055 14.4 0.618 0.633 21.102 22.179 ot_mini 99 10.5 0.001 0.001 13.270 13.400 rebuild_ks_matrix 110 8.3 0.000 0.000 12.424 12.548 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.423 12.548 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.956 11.073 multiply_cannon_multrec 16440 15.4 3.953 5.165 9.727 10.954 mp_waitall_1 139946 16.5 6.684 9.882 6.684 9.882 qs_ot_get_derivative 99 11.5 0.001 0.001 8.886 9.014 init_scf_loop 11 6.9 0.002 0.013 8.270 8.271 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.156 7.241 sum_up_and_integrate 110 10.3 0.060 0.061 6.559 6.573 integrate_v_rspace 110 11.3 0.003 0.004 6.499 6.514 prepare_preconditioner 11 7.9 0.000 0.000 6.475 6.493 make_preconditioner 11 8.9 0.000 0.002 6.475 6.493 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.835 6.192 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.029 6.042 calculate_rho_elec 110 8.6 0.058 0.059 6.028 6.041 init_scf_run 11 5.9 0.000 0.001 5.882 5.883 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.882 5.882 dbcsr_mm_accdrv_process 34862 16.1 4.689 5.349 5.627 5.804 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.919 5.386 apply_single 110 13.6 0.000 0.000 4.919 5.386 qs_ot_get_p 110 10.4 0.001 0.001 5.174 5.320 make_m2s 4110 13.4 0.050 0.051 4.043 4.426 ot_diis_step 99 11.5 0.010 0.011 4.358 4.358 make_images 4110 14.4 0.392 0.507 3.929 4.311 multiply_cannon_sync_h2d 16440 15.4 3.744 4.272 3.744 4.272 density_rs2pw 110 9.6 0.004 0.005 3.104 4.268 rs_pw_transfer 902 11.9 0.010 0.011 2.633 3.777 pw_transfer 1331 11.6 0.066 0.072 3.725 3.744 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.070 3.701 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.618 3.639 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.620 3.624 grid_integrate_task_list 110 12.3 3.183 3.379 3.183 3.379 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.287 3.287 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.137 3.161 wfi_extrapolate 11 7.9 0.001 0.001 2.932 2.932 fft3d_ps 1111 14.6 1.085 1.095 2.917 2.930 calculate_first_density_matrix 1 7.0 0.000 0.003 2.865 2.866 calculate_dm_sparse 110 9.5 0.001 0.001 2.804 2.838 make_images_data 4110 15.4 0.043 0.048 2.332 2.814 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.784 2.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.782 2.783 cp_fm_redistribute_end 48 14.0 1.741 2.762 1.743 2.763 cp_fm_diag_elpa_base 48 14.0 0.960 2.616 1.017 2.719 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.039 2.681 cp_fm_cholesky_invert 11 10.9 2.535 2.541 2.535 2.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.490 grid_collocate_task_list 110 9.6 2.082 2.474 2.082 2.474 mp_waitany 17072 13.8 1.183 2.383 1.183 2.383 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.871 2.361 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.268 2.338 potential_pw2rs 110 12.3 0.010 0.010 2.286 2.297 mp_irecv_dv 48980 15.7 0.803 2.237 0.803 2.237 mp_alltoall_d11v 2046 13.8 1.660 2.158 1.660 2.158 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.179 0.905 2.051 mp_sum_l 6514 12.8 1.434 2.034 1.434 2.034 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.976 1.977 dbcsr_complete_redistribute 325 12.2 0.342 0.368 1.409 1.848 cp_fm_upper_to_full 70 13.6 1.378 1.834 1.378 1.834 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.772 1.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.688 mp_allgather_i34 2055 14.4 0.509 1.526 0.509 1.526 cp_fm_cholesky_decompose 22 10.9 1.506 1.522 1.506 1.522 acc_transpose_blocks 16440 15.4 0.071 0.073 1.500 1.516 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.460 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.940 1.383 rs_gather_matrices 110 12.3 0.233 0.264 0.951 1.378 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.245 1.253 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.913000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.181818, yerr=8.589827 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 722.468864E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 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 0.389 0.472 67.023 67.025 qs_mol_dyn_low 1 2.0 0.003 0.003 66.142 66.151 qs_forces 11 3.9 0.002 0.002 65.971 65.972 qs_energies 11 4.9 0.011 0.013 62.459 62.466 scf_env_do_scf 11 5.9 0.001 0.002 52.791 52.794 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.226 41.227 velocity_verlet 10 3.0 0.001 0.001 36.918 36.924 dbcsr_multiply_generic 2055 12.4 0.114 0.117 30.457 30.678 qs_scf_new_mos 99 7.5 0.001 0.001 25.986 26.085 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.985 26.084 multiply_cannon 2055 13.4 0.240 0.261 23.558 24.815 ot_scf_mini 99 9.5 0.003 0.004 24.371 24.479 multiply_cannon_loop 2055 14.4 0.890 0.909 22.163 22.717 multiply_cannon_multrec 24660 15.4 4.254 6.865 13.042 14.278 ot_mini 99 10.5 0.001 0.001 14.105 14.234 rebuild_ks_matrix 110 8.3 0.000 0.000 12.437 12.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.437 12.539 init_scf_loop 11 6.9 0.001 0.003 11.524 11.525 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.921 11.011 qs_ot_get_derivative 99 11.5 0.001 0.001 9.990 10.107 prepare_preconditioner 11 7.9 0.000 0.000 9.622 9.638 make_preconditioner 11 8.9 0.000 0.001 9.622 9.638 dbcsr_mm_accdrv_process 52304 16.0 6.964 8.538 8.643 9.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.912 9.317 init_scf_run 11 5.9 0.000 0.001 6.862 6.863 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.862 6.863 sum_up_and_integrate 110 10.3 0.068 0.071 6.650 6.672 integrate_v_rspace 110 11.3 0.003 0.003 6.581 6.602 mp_waitall_1 121746 16.5 4.443 6.225 4.443 6.225 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.188 6.197 calculate_rho_elec 110 8.6 0.077 0.081 6.187 6.197 make_m2s 4110 13.4 0.060 0.063 5.360 5.722 make_images 4110 14.4 0.577 0.699 5.220 5.579 qs_ot_get_p 110 10.4 0.010 0.074 5.389 5.541 cp_fm_upper_to_full 70 13.8 3.179 4.518 3.179 4.518 ot_diis_step 99 11.5 0.011 0.011 4.076 4.077 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.991 4.064 apply_single 110 13.6 0.000 0.000 3.991 4.063 pw_transfer 1331 11.6 0.066 0.075 3.986 4.021 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.879 3.918 calculate_first_density_matrix 1 7.0 0.000 0.003 3.792 3.794 dbcsr_complete_redistribute 325 12.2 0.423 0.478 2.615 3.713 density_rs2pw 110 9.6 0.004 0.005 3.097 3.606 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.544 3.558 grid_integrate_task_list 110 12.3 3.282 3.452 3.282 3.452 fft_wrap_pw1pw2_140 451 13.1 0.203 0.212 3.354 3.398 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.328 3.380 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.367 3.369 calculate_dm_sparse 110 9.5 0.001 0.001 3.320 3.351 multiply_cannon_sync_h2d 24660 15.4 3.200 3.307 3.200 3.307 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.537 3.300 fft3d_ps 1111 14.6 1.093 1.126 3.170 3.204 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.059 3.150 hybrid_alltoall_any 4261 16.3 0.119 0.456 2.287 3.142 make_images_data 4110 15.4 0.046 0.050 2.746 3.124 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.076 3.077 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.005 3.040 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.941 3.001 wfi_extrapolate 11 7.9 0.001 0.001 2.977 2.977 rs_pw_transfer 902 11.9 0.010 0.011 2.408 2.931 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.701 2.775 mp_alltoall_i22 605 13.7 1.585 2.708 1.585 2.708 cp_fm_cholesky_invert 11 10.9 2.594 2.603 2.594 2.603 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.563 2.565 cp_fm_redistribute_end 48 14.0 1.284 2.541 1.285 2.542 cp_fm_diag_elpa_base 48 14.0 1.178 2.408 1.254 2.517 jit_kernel_multiply 11 15.8 1.351 2.455 1.351 2.455 grid_collocate_task_list 110 9.6 2.180 2.454 2.180 2.454 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.306 2.308 potential_pw2rs 110 12.3 0.013 0.013 2.281 2.298 acc_transpose_blocks 24660 15.4 0.105 0.107 2.137 2.218 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.944 1.964 mp_alltoall_d11v 2046 13.8 1.735 1.959 1.735 1.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.761 1.771 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.711 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.675 1.709 cp_fm_cholesky_decompose 22 10.9 1.636 1.695 1.636 1.695 mp_allgather_i34 2055 14.4 0.502 1.649 0.502 1.649 acc_transpose_blocks_kernels 24660 16.4 0.303 0.308 1.494 1.571 mp_waitany 13376 13.8 1.079 1.557 1.079 1.557 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.831 1.537 mp_sum_l 6514 12.8 0.890 1.517 0.890 1.517 mp_irecv_dv 62702 16.1 0.732 1.460 0.732 1.460 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.181 0.846 1.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.025000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.636364, yerr=6.064393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 832.806912E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 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 0.074 0.159 57.658 57.690 qs_mol_dyn_low 1 2.0 0.003 0.003 56.364 56.940 qs_forces 11 3.9 0.008 0.009 56.118 56.118 qs_energies 11 4.9 0.001 0.001 52.415 52.420 scf_env_do_scf 11 5.9 0.000 0.001 42.667 42.668 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.040 35.040 velocity_verlet 10 3.0 0.001 0.001 31.084 31.097 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.769 22.896 qs_scf_new_mos 99 7.5 0.001 0.001 19.860 19.903 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.859 19.902 multiply_cannon 2055 13.4 0.248 0.268 17.463 18.839 ot_scf_mini 99 9.5 0.002 0.002 18.634 18.669 multiply_cannon_loop 2055 14.4 0.321 0.334 16.099 16.403 rebuild_ks_matrix 110 8.3 0.000 0.000 12.013 12.046 qs_ks_build_kohn_sham_matrix 110 9.3 0.022 0.024 12.013 12.046 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.672 10.702 ot_mini 99 10.5 0.001 0.001 10.174 10.209 multiply_cannon_multrec 8220 15.4 3.377 4.696 7.808 8.744 init_scf_loop 11 6.9 0.000 0.000 7.580 7.581 mp_waitall_1 103326 16.6 5.727 7.298 5.727 7.298 sum_up_and_integrate 110 10.3 0.079 0.081 6.546 6.557 init_scf_run 11 5.9 0.000 0.001 6.490 6.490 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.490 6.490 integrate_v_rspace 110 11.3 0.003 0.003 6.467 6.478 qs_ot_get_derivative 99 11.5 0.001 0.001 6.398 6.432 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.364 6.380 calculate_rho_elec 110 8.6 0.115 0.116 6.364 6.379 prepare_preconditioner 11 7.9 0.000 0.000 5.872 5.880 make_preconditioner 11 8.9 0.000 0.000 5.871 5.880 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.464 5.538 dbcsr_mm_accdrv_process 17442 15.9 2.953 4.294 4.303 5.415 qs_ot_get_p 110 10.4 0.001 0.001 4.601 4.629 make_m2s 4110 13.4 0.038 0.040 4.098 4.348 make_images 4110 14.4 0.638 0.698 3.969 4.222 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.920 4.216 pw_transfer 1331 11.6 0.067 0.074 4.136 4.155 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.027 4.050 ot_diis_step 99 11.5 0.012 0.012 3.756 3.756 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.682 3.709 apply_single 110 13.6 0.000 0.000 3.682 3.709 calculate_first_density_matrix 1 7.0 0.000 0.000 3.665 3.666 fft_wrap_pw1pw2_140 451 13.1 0.215 0.219 3.505 3.533 grid_integrate_task_list 110 12.3 3.360 3.511 3.360 3.511 density_rs2pw 110 9.6 0.004 0.004 3.117 3.487 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.321 3.324 fft3d_ps 1111 14.6 1.143 1.178 3.272 3.297 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.026 3.026 multiply_cannon_sync_h2d 8220 15.4 2.923 2.995 2.923 2.995 cp_fm_cholesky_invert 11 10.9 2.811 2.814 2.811 2.814 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.812 2.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.728 2.729 wfi_extrapolate 11 7.9 0.001 0.001 2.728 2.728 make_images_data 4110 15.4 0.038 0.043 2.331 2.707 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.258 2.676 rs_pw_transfer 902 11.9 0.010 0.010 2.243 2.642 calculate_dm_sparse 110 9.5 0.001 0.001 2.549 2.589 grid_collocate_task_list 110 9.6 2.276 2.538 2.276 2.538 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.533 2.534 cp_fm_redistribute_end 48 14.0 0.644 2.511 0.648 2.511 cp_fm_diag_elpa_base 48 14.0 1.689 2.333 1.857 2.485 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.312 2.324 potential_pw2rs 110 12.3 0.015 0.016 2.097 2.101 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.992 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.825 1.840 mp_alltoall_d11v 2046 13.8 1.559 1.807 1.559 1.807 make_basis_sm 11 9.8 0.028 0.037 1.773 1.775 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.645 1.760 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.715 1.734 parallel_gemm_fm 81 9.0 0.000 0.000 1.684 1.688 parallel_gemm_fm_cosma 81 10.0 1.684 1.688 1.684 1.688 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.675 1.678 cp_fm_cholesky_decompose 22 10.9 1.650 1.671 1.650 1.671 jit_kernel_multiply 8 15.9 1.044 1.660 1.044 1.660 mp_allgather_i34 2055 14.4 0.497 1.579 0.497 1.579 dbcsr_complete_redistribute 325 12.2 0.601 0.647 1.484 1.575 mp_waitany 9240 13.8 1.137 1.572 1.137 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.535 qs_create_task_list 11 7.9 0.019 0.021 1.242 1.346 generate_qs_task_list 11 8.9 0.378 0.445 1.223 1.327 rs_pw_transfer_RS2PW_140 121 11.5 0.191 0.264 0.911 1.312 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.690000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=784.000000, yerr=12.533591 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 1.404457E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. 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 0.068 0.139 87.288 87.289 qs_mol_dyn_low 1 2.0 0.003 0.003 86.608 86.617 qs_forces 11 3.9 0.002 0.002 86.539 86.545 qs_energies 11 4.9 0.001 0.001 82.327 82.333 scf_env_do_scf 11 5.9 0.001 0.001 72.307 72.307 velocity_verlet 10 3.0 0.001 0.001 55.442 55.448 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.097 44.098 dbcsr_multiply_generic 2055 12.4 0.119 0.120 28.418 28.494 init_scf_loop 11 6.9 0.000 0.000 28.138 28.139 prepare_preconditioner 11 7.9 0.000 0.000 26.033 26.040 make_preconditioner 11 8.9 0.000 0.000 26.033 26.039 qs_scf_new_mos 99 7.5 0.001 0.001 25.762 25.805 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.762 25.804 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.178 25.500 ot_scf_mini 99 9.5 0.002 0.002 24.017 24.040 multiply_cannon 2055 13.4 0.349 0.373 21.378 22.157 multiply_cannon_loop 2055 14.4 0.342 0.345 19.605 20.144 cp_fm_upper_to_full 70 14.2 12.736 18.303 12.736 18.303 rebuild_ks_matrix 110 8.3 0.001 0.001 14.262 14.291 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.262 14.290 ot_mini 99 10.5 0.001 0.001 13.554 13.579 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.925 12.950 dbcsr_complete_redistribute 325 12.2 1.025 1.047 7.399 10.567 multiply_cannon_multrec 8220 15.4 4.400 4.599 9.591 9.703 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.390 9.558 qs_ot_get_derivative 99 11.5 0.001 0.001 9.102 9.126 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.840 8.972 mp_alltoall_i22 605 13.7 5.446 8.626 5.446 8.626 mp_waitall_1 84994 16.7 7.249 8.027 7.249 8.027 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.744 7.779 calculate_rho_elec 110 8.6 0.227 0.227 7.743 7.779 sum_up_and_integrate 110 10.3 0.150 0.152 7.379 7.394 integrate_v_rspace 110 11.3 0.004 0.004 7.229 7.242 init_scf_run 11 5.9 0.000 0.001 5.743 5.743 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.743 5.743 make_m2s 4110 13.4 0.043 0.044 5.133 5.618 make_images 4110 14.4 0.883 0.935 4.945 5.429 qs_ot_get_p 110 10.4 0.001 0.001 5.352 5.378 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.833 5.374 apply_single 110 13.6 0.000 0.000 4.832 5.374 dbcsr_mm_accdrv_process 11614 15.7 3.404 3.641 5.050 5.327 pw_transfer 1331 11.6 0.075 0.075 5.290 5.297 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.174 5.180 cp_fm_cholesky_invert 11 10.9 4.999 5.003 4.999 5.003 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.589 4.950 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.518 4.529 ot_diis_step 99 11.5 0.015 0.016 4.437 4.437 fft3d_ps 1111 14.6 1.293 1.299 4.369 4.377 multiply_cannon_sync_h2d 8220 15.4 3.949 3.958 3.949 3.958 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.873 3.880 density_rs2pw 110 9.6 0.004 0.004 3.778 3.807 grid_integrate_task_list 110 12.3 3.657 3.718 3.657 3.718 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.690 3.695 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.876 3.581 hybrid_alltoall_any 4261 16.3 0.257 0.553 2.800 3.516 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.438 3.439 make_images_data 4110 15.4 0.042 0.045 2.779 3.376 wfi_extrapolate 11 7.9 0.001 0.001 3.271 3.271 calculate_dm_sparse 110 9.5 0.001 0.001 3.144 3.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.862 2.866 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.856 2.856 cp_fm_diag_elpa_base 48 14.0 2.325 2.525 2.854 2.854 grid_collocate_task_list 110 9.6 2.631 2.650 2.631 2.650 potential_pw2rs 110 12.3 0.021 0.022 2.512 2.514 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.389 calculate_first_density_matrix 1 7.0 0.000 0.000 2.348 2.349 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.316 2.330 qs_env_update_s_mstruct 11 6.9 0.007 0.008 2.218 2.288 rs_pw_transfer 902 11.9 0.011 0.011 2.233 2.249 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.196 mp_alltoall_d11v 2046 13.8 1.910 1.993 1.910 1.993 cp_fm_cholesky_decompose 22 10.9 1.932 1.955 1.932 1.955 qs_create_task_list 11 7.9 0.000 0.001 1.897 1.947 generate_qs_task_list 11 8.9 0.734 0.789 1.897 1.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.941 1.944 jit_kernel_multiply 10 15.1 1.447 1.777 1.447 1.777 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.723 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.289000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1264.181818, yerr=57.102489 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.121600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520949. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.102 0.146 205.658 205.674 qs_mol_dyn_low 1 2.0 0.003 0.006 204.209 204.222 qs_forces 11 3.9 0.005 0.017 204.125 204.128 qs_energies 11 4.9 0.004 0.020 198.342 198.360 scf_env_do_scf 11 5.9 0.001 0.002 181.228 181.232 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.718 160.720 dbcsr_multiply_generic 2507 12.6 0.178 0.185 124.023 124.752 velocity_verlet 10 3.0 0.001 0.001 122.993 122.994 qs_scf_new_mos 117 7.6 0.001 0.001 120.516 120.764 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.515 120.764 ot_scf_mini 117 9.6 0.003 0.004 113.895 114.090 multiply_cannon 2507 13.6 0.236 0.249 100.778 103.129 multiply_cannon_loop 2507 14.6 2.116 2.189 98.683 100.749 ot_mini 117 10.6 0.001 0.001 65.235 65.479 multiply_cannon_multrec 60168 15.6 33.197 35.468 41.607 43.441 qs_ot_get_derivative 117 11.6 0.001 0.001 40.456 40.684 rebuild_ks_matrix 128 8.3 0.001 0.001 33.934 34.439 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.933 34.438 mp_waitall_1 267128 16.5 28.196 31.498 28.196 31.498 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.427 30.895 multiply_cannon_sync_h2d 60168 15.6 27.257 29.585 27.257 29.585 qs_ot_get_p 128 10.4 0.001 0.002 26.738 27.064 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.287 24.854 apply_single 128 13.6 0.001 0.001 24.287 24.853 ot_diis_step 117 11.6 0.007 0.008 24.418 24.419 init_scf_loop 11 6.9 0.001 0.002 20.433 20.437 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.032 20.109 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 18.848 19.057 multiply_cannon_metrocomm3 60168 15.6 0.115 0.122 15.692 18.206 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.457 17.458 prepare_preconditioner 11 7.9 0.000 0.001 15.764 15.813 make_preconditioner 11 8.9 0.000 0.001 15.764 15.812 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.003 15.182 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.630 14.636 cp_fm_redistribute_end 83 14.4 11.573 14.566 11.583 14.569 sum_up_and_integrate 128 10.3 0.090 0.109 14.342 14.366 cp_fm_diag_elpa_base 83 14.4 2.944 14.216 2.974 14.320 integrate_v_rspace 128 11.3 0.003 0.004 14.252 14.280 make_m2s 5014 13.6 0.105 0.112 13.789 14.257 make_images 5014 14.6 0.400 0.420 13.607 14.085 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.794 13.918 calculate_rho_elec 128 8.7 0.045 0.064 13.793 13.917 init_scf_run 11 5.9 0.000 0.001 12.718 12.719 scf_env_initial_rho_setup 11 6.9 0.003 0.004 12.718 12.719 density_rs2pw 128 9.7 0.006 0.007 7.160 10.327 mp_sum_l 7870 13.0 8.211 9.626 8.211 9.626 cp_fm_cholesky_invert 11 10.9 8.996 9.004 8.996 9.004 wfi_extrapolate 11 7.9 0.001 0.001 9.004 9.004 rs_pw_transfer 1046 11.9 0.016 0.018 5.751 8.934 calculate_dm_sparse 128 9.5 0.001 0.001 8.484 8.603 dbcsr_mm_accdrv_process 124484 16.2 3.287 3.439 7.975 8.519 pw_transfer 1547 11.6 0.076 0.099 8.152 8.403 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.945 8.203 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.908 8.010 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.890 8.000 make_images_data 5014 15.6 0.065 0.072 6.775 7.712 grid_integrate_task_list 128 12.3 7.039 7.529 7.039 7.529 multiply_cannon_metrocomm1 60168 15.6 0.086 0.092 6.054 7.504 fft_wrap_pw1pw2_140 523 13.2 0.447 0.519 6.934 7.241 hybrid_alltoall_any 5200 16.5 0.290 2.244 5.958 7.141 fft3d_ps 1291 14.7 2.103 2.875 6.732 6.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.873 6.882 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.001 0.011 5.952 6.131 mp_waitany 16020 13.9 2.777 5.942 2.777 5.942 grid_collocate_task_list 128 9.7 4.556 5.778 4.556 5.778 mp_alltoall_d11v 2415 14.1 4.217 5.523 4.217 5.523 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.299 2.254 5.444 potential_pw2rs 128 12.3 0.009 0.011 4.721 4.749 cp_fm_cholesky_decompose 22 10.9 4.579 4.597 4.579 4.597 mp_sum_d 4457 12.1 3.730 4.477 3.730 4.477 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.674000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.957961 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 832.614400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57145. MP_Allreduce 11262 946. MP_Sync 168 MP_Alltoall 1955 5616370. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.144 189.921 189.922 qs_mol_dyn_low 1 2.0 0.003 0.004 189.366 189.379 qs_forces 11 3.9 0.003 0.003 189.285 189.288 qs_energies 11 4.9 0.003 0.016 182.429 182.440 scf_env_do_scf 11 5.9 0.001 0.002 165.294 165.304 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.509 132.511 velocity_verlet 10 3.0 0.001 0.001 118.275 118.276 dbcsr_multiply_generic 2485 12.5 0.185 0.191 96.715 97.866 qs_scf_new_mos 116 7.6 0.001 0.001 92.519 92.967 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.518 92.966 ot_scf_mini 116 9.6 0.004 0.004 87.781 88.352 multiply_cannon 2485 13.5 0.475 0.524 77.157 81.496 multiply_cannon_loop 2485 14.5 1.250 1.292 73.909 76.453 ot_mini 116 10.6 0.001 0.001 49.079 49.590 mp_waitall_1 212858 16.6 24.078 37.855 24.078 37.855 multiply_cannon_multrec 29820 15.5 22.016 27.136 31.715 37.075 rebuild_ks_matrix 127 8.3 0.001 0.001 33.442 33.912 qs_ks_build_kohn_sham_matrix 127 9.3 0.138 0.225 33.441 33.911 init_scf_loop 11 6.9 0.000 0.001 32.696 32.701 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.065 30.509 prepare_preconditioner 11 7.9 0.000 0.000 28.132 28.301 make_preconditioner 11 8.9 0.016 0.129 28.132 28.301 qs_ot_get_derivative 116 11.6 0.001 0.002 27.396 27.967 multiply_cannon_metrocomm3 29820 15.5 0.095 0.102 15.507 27.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.831 27.460 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.810 22.868 apply_single 127 13.6 0.001 0.001 21.810 22.868 multiply_cannon_sync_h2d 29820 15.5 19.333 21.916 19.333 21.916 ot_diis_step 116 11.6 0.014 0.015 21.521 21.523 qs_ot_get_p 127 10.4 0.001 0.002 20.606 21.180 cp_fm_cholesky_invert 11 10.9 16.379 16.391 16.379 16.391 qs_ot_p2m_diag 82 11.4 0.184 0.212 15.894 15.927 make_m2s 4970 13.5 0.089 0.094 14.029 15.673 make_images 4970 14.5 1.146 1.339 13.820 15.463 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.777 14.778 sum_up_and_integrate 127 10.3 0.115 0.130 14.724 14.757 integrate_v_rspace 127 11.3 0.003 0.004 14.608 14.642 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.060 14.088 calculate_rho_elec 127 8.7 0.087 0.105 14.060 14.087 init_scf_run 11 5.9 0.000 0.001 12.334 12.336 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.334 12.336 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.760 11.769 cp_fm_redistribute_end 82 14.4 6.869 11.705 6.881 11.706 cp_fm_diag_elpa_base 82 14.4 4.596 11.214 4.807 11.573 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.708 11.151 make_images_data 4970 15.5 0.063 0.071 8.519 10.456 multiply_cannon_metrocomm4 27335 15.5 0.095 0.111 3.653 10.366 mp_irecv_dv 68888 16.3 3.462 9.980 3.462 9.980 density_rs2pw 127 9.7 0.006 0.006 7.470 9.792 dbcsr_mm_accdrv_process 61726 16.2 4.652 5.275 9.164 9.746 hybrid_alltoall_any 5155 16.4 0.338 1.491 7.267 9.724 pw_transfer 1535 11.6 0.086 0.109 8.986 9.065 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.762 8.846 wfi_extrapolate 11 7.9 0.001 0.001 8.371 8.372 rs_pw_transfer 1038 11.9 0.014 0.016 5.624 7.968 fft_wrap_pw1pw2_140 519 13.2 0.469 0.525 7.750 7.867 grid_integrate_task_list 127 12.3 7.133 7.588 7.133 7.588 fft3d_ps 1281 14.7 2.727 2.901 7.144 7.186 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.147 6.931 cp_fm_cholesky_decompose 22 10.9 6.732 6.816 6.732 6.816 calculate_dm_sparse 127 9.5 0.001 0.001 6.615 6.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.702 6.710 mp_sum_l 7804 13.0 3.836 5.864 3.836 5.864 grid_collocate_task_list 127 9.7 4.669 5.819 4.669 5.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.343 5.468 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.228 5.287 potential_pw2rs 127 12.3 0.014 0.016 5.026 5.043 mp_allgather_i34 2485 14.5 1.826 4.855 1.826 4.855 mp_waitany 11660 13.9 2.459 4.785 2.459 4.785 mp_alltoall_d11v 2401 14.1 4.053 4.653 4.053 4.653 dbcsr_complete_redistribute 393 12.7 0.881 1.364 3.546 4.389 rs_pw_transfer_RS2PW_140 138 11.5 0.347 0.374 2.067 4.375 mp_sum_d 4446 12.1 2.623 3.932 2.623 3.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.780 3.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.922000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 950.497280E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.173 0.332 176.620 176.630 qs_mol_dyn_low 1 2.0 0.003 0.004 175.679 175.692 qs_forces 11 3.9 0.021 0.058 175.542 175.545 qs_energies 11 4.9 0.007 0.050 168.777 168.804 scf_env_do_scf 11 5.9 0.005 0.076 152.833 152.834 scf_env_do_scf_inner_loop 116 6.6 0.003 0.011 117.880 117.881 velocity_verlet 10 3.0 0.001 0.001 112.009 112.012 dbcsr_multiply_generic 2485 12.5 0.180 0.184 81.799 83.041 qs_scf_new_mos 116 7.6 0.001 0.001 79.659 80.018 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.658 80.017 ot_scf_mini 116 9.6 0.004 0.006 75.567 75.994 multiply_cannon 2485 13.5 0.495 0.516 61.956 67.061 multiply_cannon_loop 2485 14.5 0.849 0.874 58.897 61.980 ot_mini 116 10.6 0.001 0.001 42.672 43.110 init_scf_loop 11 6.9 0.001 0.003 34.782 34.783 mp_waitall_1 169034 16.6 24.972 34.407 24.972 34.407 rebuild_ks_matrix 127 8.3 0.001 0.001 30.860 31.321 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.020 30.859 31.320 prepare_preconditioner 11 7.9 0.000 0.000 30.688 30.736 make_preconditioner 11 8.9 0.000 0.001 30.688 30.736 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.287 29.780 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.705 28.127 multiply_cannon_multrec 19880 15.5 13.420 16.318 22.364 25.274 multiply_cannon_metrocomm3 19880 15.5 0.059 0.064 15.195 24.650 qs_ot_get_derivative 116 11.6 0.002 0.002 23.105 23.544 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.853 20.967 apply_single 127 13.6 0.001 0.001 19.853 20.967 ot_diis_step 116 11.6 0.018 0.019 19.470 19.470 qs_ot_get_p 127 10.4 0.001 0.002 18.428 18.954 multiply_cannon_sync_h2d 19880 15.5 14.189 15.811 14.189 15.811 make_m2s 4970 13.5 0.081 0.088 14.694 15.653 make_images 4970 14.5 1.163 1.252 14.461 15.417 sum_up_and_integrate 127 10.3 0.133 0.145 14.717 14.754 integrate_v_rspace 127 11.3 0.004 0.005 14.583 14.623 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.344 14.377 calculate_rho_elec 127 8.7 0.131 0.146 14.344 14.376 cp_fm_cholesky_invert 11 10.9 14.192 14.201 14.192 14.201 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.179 14.188 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.199 13.200 init_scf_run 11 5.9 0.003 0.027 10.835 10.835 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.831 10.835 make_images_data 4970 15.5 0.060 0.068 9.108 10.516 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.259 10.259 cp_fm_redistribute_end 82 14.4 3.880 10.205 3.893 10.207 cp_fm_diag_elpa_base 82 14.4 5.894 9.633 6.294 10.088 hybrid_alltoall_any 5155 16.4 0.431 1.970 7.949 9.809 density_rs2pw 127 9.7 0.006 0.006 7.248 9.375 pw_transfer 1535 11.6 0.086 0.109 9.213 9.338 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.839 9.158 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.990 9.123 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.399 9.067 mp_irecv_dv 49801 16.2 3.280 8.823 3.280 8.823 dbcsr_mm_accdrv_process 41158 16.2 4.519 5.514 8.397 8.642 fft_wrap_pw1pw2_140 519 13.2 0.474 0.518 7.989 8.125 grid_integrate_task_list 127 12.3 7.202 7.637 7.202 7.637 cp_fm_upper_to_full 104 14.5 5.805 7.528 5.805 7.528 wfi_extrapolate 11 7.9 0.001 0.001 7.448 7.448 fft3d_ps 1281 14.7 2.645 2.872 7.268 7.367 cp_fm_cholesky_decompose 22 10.9 7.252 7.287 7.252 7.287 rs_pw_transfer 1038 11.9 0.013 0.014 5.031 7.137 dbcsr_complete_redistribute 393 12.7 1.167 1.197 4.450 6.226 calculate_dm_sparse 127 9.5 0.001 0.001 5.947 6.055 grid_collocate_task_list 127 9.7 4.856 5.767 4.856 5.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.536 5.544 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.609 5.465 mp_sum_l 7804 13.0 3.400 5.303 3.400 5.303 mp_alltoall_d11v 2401 14.1 4.266 5.086 4.266 5.086 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.296 5.075 potential_pw2rs 127 12.3 0.020 0.022 4.850 4.877 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.652 4.792 mp_allgather_i34 2485 14.5 1.595 4.586 1.595 4.586 mp_waitany 11660 13.9 2.304 4.414 2.304 4.414 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.381 4.091 mp_alltoall_i22 712 14.1 2.048 3.991 2.048 3.991 rs_pw_transfer_RS2PW_140 138 11.5 0.326 0.345 1.897 3.979 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.938 3.973 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.780 3.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.661 3.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.630000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=899.090909, yerr=17.164734 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.154839E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 1067. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 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 0.106 0.476 188.471 188.841 qs_mol_dyn_low 1 2.0 0.003 0.003 187.267 187.280 qs_forces 11 3.9 0.003 0.003 187.144 187.153 qs_energies 11 4.9 0.001 0.002 179.939 179.950 scf_env_do_scf 11 5.9 0.001 0.001 163.045 163.051 velocity_verlet 10 3.0 0.001 0.001 124.330 124.334 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.300 117.302 dbcsr_multiply_generic 2507 12.6 0.189 0.196 79.832 80.637 qs_scf_new_mos 117 7.6 0.001 0.001 79.186 79.471 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.185 79.470 ot_scf_mini 117 9.6 0.003 0.004 74.505 74.768 multiply_cannon 2507 13.6 0.578 0.636 54.987 58.385 multiply_cannon_loop 2507 14.6 1.186 1.222 51.203 53.081 init_scf_loop 11 6.9 0.000 0.000 45.626 45.626 ot_mini 117 10.6 0.001 0.001 42.449 42.701 prepare_preconditioner 11 7.9 0.000 0.000 41.489 41.519 make_preconditioner 11 8.9 0.000 0.000 41.489 41.519 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.275 40.115 multiply_cannon_multrec 30084 15.6 14.273 19.468 26.234 31.155 rebuild_ks_matrix 128 8.3 0.001 0.001 30.173 30.470 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.172 30.470 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.180 27.436 mp_waitall_1 147882 16.7 17.449 26.579 17.449 26.579 qs_ot_get_derivative 117 11.6 0.002 0.002 22.881 23.145 make_m2s 5014 13.6 0.096 0.098 20.553 21.682 make_images 5014 14.6 1.950 2.445 20.248 21.376 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.023 19.528 apply_single 128 13.6 0.001 0.001 19.022 19.528 ot_diis_step 117 11.6 0.017 0.018 19.447 19.449 qs_ot_get_p 128 10.4 0.001 0.001 18.296 18.576 cp_fm_cholesky_invert 11 10.9 15.957 15.966 15.957 15.966 cp_fm_upper_to_full 105 14.7 10.780 15.904 10.780 15.904 multiply_cannon_metrocomm3 30084 15.6 0.046 0.048 6.414 15.546 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.970 15.003 calculate_rho_elec 128 8.7 0.175 0.190 14.970 15.002 sum_up_and_integrate 128 10.3 0.141 0.154 14.831 14.859 integrate_v_rspace 128 11.3 0.004 0.004 14.690 14.722 qs_ot_p2m_diag 83 11.4 0.342 0.390 14.157 14.208 make_images_data 5014 15.6 0.063 0.068 11.070 12.951 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.898 12.899 multiply_cannon_sync_h2d 30084 15.6 11.731 12.717 11.731 12.717 dbcsr_complete_redistribute 395 12.7 1.502 1.641 8.723 12.360 dbcsr_mm_accdrv_process 62264 16.2 7.333 8.155 11.534 12.107 hybrid_alltoall_any 5200 16.5 0.520 2.165 9.745 11.975 init_scf_run 11 5.9 0.000 0.001 11.026 11.027 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.026 11.027 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.375 11.009 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.858 9.860 cp_fm_redistribute_end 83 14.4 1.707 9.793 1.721 9.800 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.192 9.759 cp_fm_diag_elpa_base 83 14.4 7.481 9.190 8.048 9.675 pw_transfer 1547 11.6 0.088 0.109 9.524 9.591 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.380 9.571 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.296 9.370 mp_alltoall_i22 716 14.1 5.375 8.986 5.375 8.986 density_rs2pw 128 9.7 0.006 0.006 7.133 8.590 fft_wrap_pw1pw2_140 523 13.2 0.485 0.493 8.265 8.351 grid_integrate_task_list 128 12.3 7.503 7.961 7.503 7.961 wfi_extrapolate 11 7.9 0.001 0.001 7.660 7.660 fft3d_ps 1291 14.7 2.782 2.863 7.520 7.583 cp_fm_cholesky_decompose 22 10.9 7.452 7.549 7.452 7.549 multiply_cannon_metrocomm4 25070 15.6 0.075 0.086 2.713 7.009 mp_irecv_dv 76098 16.2 2.570 6.744 2.570 6.744 calculate_dm_sparse 128 9.5 0.001 0.001 6.371 6.458 rs_pw_transfer 1046 11.9 0.013 0.014 4.599 6.121 mp_alltoall_d11v 2415 14.1 5.196 5.908 5.196 5.908 grid_collocate_task_list 128 9.7 5.070 5.779 5.070 5.779 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.435 5.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.613 4.707 potential_pw2rs 128 12.3 0.022 0.023 4.614 4.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.382 4.476 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 4.445 4.448 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.165 4.257 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.841000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1090.090909, yerr=18.603297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.496412E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350154512 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.073 174.301 174.309 qs_mol_dyn_low 1 2.0 0.003 0.003 173.716 173.727 qs_forces 11 3.9 0.025 0.030 173.583 173.589 qs_energies 11 4.9 0.002 0.002 166.071 166.077 scf_env_do_scf 11 5.9 0.001 0.001 147.979 147.981 velocity_verlet 10 3.0 0.001 0.001 113.496 113.502 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 112.020 112.021 dbcsr_multiply_generic 2507 12.6 0.183 0.189 73.415 73.936 qs_scf_new_mos 117 7.6 0.001 0.001 73.669 73.807 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.668 73.807 ot_scf_mini 117 9.6 0.004 0.004 69.305 69.384 multiply_cannon 2507 13.6 0.589 0.632 54.195 58.310 multiply_cannon_loop 2507 14.6 0.445 0.456 49.568 50.557 ot_mini 117 10.6 0.001 0.001 39.259 39.339 init_scf_loop 11 6.9 0.000 0.000 35.813 35.814 mp_waitall_1 125778 16.7 25.133 32.483 25.133 32.483 prepare_preconditioner 11 7.9 0.000 0.000 31.620 31.654 make_preconditioner 11 8.9 0.000 0.000 31.620 31.654 rebuild_ks_matrix 128 8.3 0.001 0.001 29.867 30.000 qs_ks_build_kohn_sham_matrix 128 9.3 0.249 0.250 29.866 29.999 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.514 29.764 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.170 27.293 multiply_cannon_multrec 10028 15.6 10.490 14.252 17.876 20.860 ot_diis_step 117 11.6 0.020 0.020 19.898 19.899 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.449 19.786 apply_single 128 13.6 0.001 0.001 19.449 19.785 qs_ot_get_derivative 117 11.6 0.002 0.002 19.296 19.383 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.279 19.330 cp_fm_cholesky_invert 11 10.9 18.124 18.130 18.124 18.130 make_m2s 5014 13.6 0.066 0.070 15.663 18.094 make_images 5014 14.6 2.284 2.646 15.357 17.784 qs_ot_get_p 128 10.4 0.001 0.001 17.358 17.471 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.282 15.350 calculate_rho_elec 128 8.7 0.259 0.268 15.282 15.349 sum_up_and_integrate 128 10.3 0.180 0.190 14.958 15.005 integrate_v_rspace 128 11.3 0.004 0.004 14.778 14.834 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.632 13.647 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.501 12.503 multiply_cannon_sync_h2d 10028 15.6 11.652 12.105 11.652 12.105 make_images_data 5014 15.6 0.052 0.059 9.492 11.903 hybrid_alltoall_any 5200 16.5 0.844 3.841 9.375 11.699 init_scf_run 11 5.9 0.000 0.001 10.852 10.853 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.852 10.852 pw_transfer 1547 11.6 0.086 0.094 9.571 9.608 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.480 9.492 cp_fm_diag_elpa_base 83 14.4 9.224 9.313 9.469 9.481 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.347 9.389 fft_wrap_pw1pw2_140 523 13.2 0.497 0.519 8.245 8.291 grid_integrate_task_list 128 12.3 7.761 8.189 7.761 8.189 cp_fm_cholesky_decompose 22 10.9 7.874 7.979 7.874 7.979 density_rs2pw 128 9.7 0.005 0.006 6.894 7.780 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.720 7.770 dbcsr_mm_accdrv_process 20762 16.1 2.620 3.424 7.013 7.668 fft3d_ps 1291 14.7 2.701 2.777 7.509 7.525 wfi_extrapolate 11 7.9 0.001 0.001 7.357 7.357 mp_allgather_i34 2507 14.6 2.808 7.239 2.808 7.239 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.142 7.063 calculate_dm_sparse 128 9.5 0.001 0.001 6.143 6.236 grid_collocate_task_list 128 9.7 5.395 6.052 5.395 6.052 qs_energies_init_hamiltonians 11 5.9 0.414 0.417 5.726 5.727 mp_alltoall_d11v 2415 14.1 4.813 5.554 4.813 5.554 dbcsr_complete_redistribute 395 12.7 2.123 2.176 5.075 5.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.181 5.195 rs_pw_transfer 1046 11.9 0.012 0.013 4.115 5.012 potential_pw2rs 128 12.3 0.027 0.028 4.494 4.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.132 4.169 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.864 3.895 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.842 mp_irecv_dv 28860 15.9 1.827 3.826 1.827 3.826 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.689 3.709 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.323 3.636 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.601 3.611 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.498 3.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.309000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1400.818182, yerr=42.445298 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.931929E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 59156. MP_Allreduce 11000 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.200 0.218 297.064 297.078 qs_mol_dyn_low 1 2.0 0.003 0.003 296.162 296.174 qs_forces 11 3.9 0.003 0.003 296.060 296.065 qs_energies 11 4.9 0.011 0.017 286.735 286.749 scf_env_do_scf 11 5.9 0.001 0.001 264.644 264.660 velocity_verlet 10 3.0 0.001 0.001 212.641 212.650 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 139.799 139.801 init_scf_loop 11 6.9 0.000 0.000 124.593 124.596 prepare_preconditioner 11 7.9 0.000 0.000 119.457 119.486 make_preconditioner 11 8.9 0.000 0.000 119.457 119.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.215 116.639 qs_scf_new_mos 116 7.6 0.001 0.001 89.765 89.878 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.765 89.877 ot_scf_mini 116 9.6 0.004 0.004 85.091 85.174 dbcsr_multiply_generic 2485 12.5 0.212 0.219 83.073 83.604 cp_fm_upper_to_full 104 14.8 53.162 76.415 53.162 76.415 multiply_cannon 2485 13.5 0.697 0.752 59.103 59.523 multiply_cannon_loop 2485 14.5 0.471 0.478 55.553 56.867 ot_mini 116 10.6 0.001 0.001 44.665 44.755 dbcsr_complete_redistribute 393 12.7 3.978 4.001 29.798 43.080 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.440 39.792 rebuild_ks_matrix 127 8.3 0.001 0.001 37.823 37.854 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 37.822 37.854 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.196 37.492 mp_alltoall_i22 712 14.1 22.038 35.380 22.038 35.380 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.844 34.881 cp_fm_cholesky_invert 11 10.9 33.056 33.062 33.056 33.062 mp_waitall_1 102768 16.8 28.008 31.831 28.008 31.831 qs_ot_get_p 127 10.4 0.001 0.001 25.126 25.216 qs_ot_get_derivative 116 11.6 0.002 0.002 24.881 24.963 qs_ot_p2m_diag 82 11.4 0.868 0.872 21.188 21.215 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.416 20.430 calculate_rho_elec 127 8.7 0.478 0.479 20.415 20.429 make_m2s 4970 13.5 0.075 0.078 19.419 20.353 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 18.807 19.962 make_images 4970 14.5 3.706 3.863 18.943 19.879 ot_diis_step 116 11.6 0.022 0.022 19.751 19.752 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.508 19.508 sum_up_and_integrate 127 10.3 0.318 0.320 19.145 19.219 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.055 19.193 apply_single 127 13.6 0.001 0.001 19.055 19.193 multiply_cannon_multrec 9940 15.5 10.708 13.012 18.589 19.099 integrate_v_rspace 127 11.3 0.004 0.004 18.826 18.900 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.388 16.389 cp_fm_diag_elpa_base 82 14.4 12.020 13.667 16.385 16.385 multiply_cannon_sync_h2d 9940 15.5 15.533 15.543 15.533 15.543 hybrid_alltoall_any 5155 16.4 1.291 3.010 10.591 12.444 pw_transfer 1535 11.6 0.092 0.092 12.177 12.182 init_scf_run 11 5.9 0.000 0.001 12.167 12.168 scf_env_initial_rho_setup 11 6.9 0.013 0.020 12.167 12.167 make_images_data 4970 15.5 0.059 0.064 10.341 11.996 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.944 11.949 fft_wrap_pw1pw2_140 519 13.2 0.538 0.541 10.628 10.638 dbcsr_mm_accdrv_process 20590 16.0 4.170 6.263 7.645 10.139 fft3d_ps 1281 14.7 2.721 2.736 9.993 10.002 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.185 9.258 wfi_extrapolate 11 7.9 0.001 0.001 8.958 8.958 cp_fm_cholesky_decompose 22 10.9 8.757 8.866 8.757 8.866 grid_integrate_task_list 127 12.3 8.514 8.702 8.514 8.702 mp_alltoall_d11v 2401 14.1 7.645 8.660 7.645 8.660 density_rs2pw 127 9.7 0.005 0.005 8.194 8.337 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.941 7.943 calculate_dm_sparse 127 9.5 0.001 0.001 6.617 6.687 grid_collocate_task_list 127 9.7 6.313 6.342 6.313 6.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.198 6.270 rs_scatter_matrices 138 9.7 3.550 4.466 5.870 6.113 copy_dbcsr_to_fm 185 11.7 0.004 0.005 6.002 6.053 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.078000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2632.727273, yerr=153.179094 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 1.261724E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.027 0.045 86.017 86.018 qs_energies 1 2.0 0.000 0.001 85.500 85.510 ls_scf 1 3.0 0.000 0.000 84.178 84.209 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.568 72.701 multiply_cannon 111 7.7 0.018 0.021 55.904 57.185 multiply_cannon_loop 111 8.7 0.211 0.223 52.453 53.872 ls_scf_main 1 4.0 0.000 0.000 52.248 52.254 density_matrix_trs4 2 5.0 0.002 0.003 46.718 46.806 ls_scf_init_scf 1 4.0 0.000 0.000 28.889 28.912 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.079 27.149 mp_waitall_1 11031 10.9 22.513 25.519 22.513 25.519 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.977 25.000 multiply_cannon_multrec 2664 9.7 8.132 8.774 15.453 17.034 multiply_cannon_sync_h2d 2664 9.7 13.526 15.854 13.526 15.854 make_m2s 222 7.7 0.008 0.011 13.023 13.594 make_images 222 8.7 0.100 0.111 13.001 13.573 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.746 11.723 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.500 8.585 make_images_data 222 9.7 0.004 0.005 7.597 8.182 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.617 6.939 7.868 hybrid_alltoall_any 227 10.6 0.215 1.828 6.512 7.753 dbcsr_mm_accdrv_process_sort 4760 11.4 6.230 7.106 6.230 7.106 calculate_norms 4752 9.8 5.512 6.084 5.512 6.084 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.047 5.159 mp_sum_l 807 5.4 3.111 4.909 3.111 4.909 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.083 3.816 mp_irecv_dv 6231 10.9 2.065 3.790 2.065 3.790 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.321 3.711 make_images_sizes 222 9.7 0.000 0.000 0.661 3.360 mp_alltoall_i44 222 10.7 0.660 3.360 0.660 3.360 arnoldi_extremal 4 6.8 0.000 0.000 3.179 3.206 arnoldi_normal_ev 4 7.8 0.008 0.020 3.179 3.206 build_subspace 16 8.4 0.009 0.012 3.075 3.076 ls_scf_post 1 4.0 0.000 0.000 3.042 3.054 ls_scf_store_result 1 5.0 0.000 0.000 2.859 2.908 dbcsr_special_finalize 555 9.7 0.005 0.006 2.408 2.799 dbcsr_merge_single_wm 555 10.7 0.457 0.596 2.400 2.791 make_images_pack 222 9.7 2.210 2.628 2.211 2.630 dbcsr_matrix_vector_mult 304 9.0 0.005 0.012 2.312 2.547 dbcsr_sort_data 658 11.4 2.201 2.531 2.201 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.072 2.456 2.074 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.280 2.367 buffer_matrices_ensure_size 222 8.7 1.748 2.062 1.748 2.062 ls_scf_initial_guess 1 5.0 0.000 0.000 1.809 1.866 ls_scf_qs_atomic_guess 1 6.0 0.000 0.000 1.809 1.866 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.835 1.838 rebuild_ks_matrix 3 7.3 0.000 0.000 1.826 1.829 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.020 1.826 1.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.018000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 2.110616E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. 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 0.045 0.104 91.524 91.525 qs_energies 1 2.0 0.002 0.016 90.953 90.974 ls_scf 1 3.0 0.005 0.038 89.256 89.263 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.145 75.509 multiply_cannon 111 7.7 0.034 0.058 53.251 56.594 ls_scf_main 1 4.0 0.007 0.075 54.866 54.879 multiply_cannon_loop 111 8.7 0.116 0.123 50.003 52.701 density_matrix_trs4 2 5.0 0.016 0.125 49.101 49.286 ls_scf_init_scf 1 4.0 0.000 0.003 30.871 30.881 mp_waitall_1 9105 10.9 21.157 30.070 21.157 30.070 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.530 29.660 multiply_cannon_multrec 1332 9.7 13.118 17.087 22.317 27.496 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.135 27.149 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.866 21.248 make_m2s 222 7.7 0.006 0.007 15.251 15.948 make_images 222 8.7 1.575 1.927 15.221 15.919 dbcsr_mm_accdrv_process 4041 10.4 0.273 0.438 8.796 10.287 dbcsr_mm_accdrv_process_sort 4041 11.4 8.392 9.864 8.392 9.864 make_images_data 222 9.7 0.004 0.004 8.744 9.689 hybrid_alltoall_any 227 10.6 0.520 2.453 8.183 9.092 mp_sum_l 807 5.4 5.359 8.128 5.359 8.128 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.220 7.678 mp_irecv_dv 3311 11.0 3.201 7.629 3.201 7.629 calculate_norms 2376 9.8 5.991 6.827 5.991 6.827 multiply_cannon_sync_h2d 1332 9.7 4.831 6.374 4.831 6.374 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.017 6.310 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.993 5.195 arnoldi_extremal 4 6.8 0.000 0.000 4.608 4.631 arnoldi_normal_ev 4 7.8 0.009 0.018 4.608 4.631 build_subspace 16 8.4 0.014 0.021 4.345 4.347 ls_scf_post 1 4.0 0.001 0.009 3.514 3.525 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.124 3.370 ls_scf_store_result 1 5.0 0.000 0.000 3.228 3.337 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.236 2.742 3.238 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.181 2.796 ls_scf_dm_to_ks 2 5.0 0.000 0.013 2.621 2.711 make_images_pack 222 9.7 2.033 2.389 2.035 2.391 mp_allgather_i34 111 8.7 0.941 2.389 0.941 2.389 dbcsr_sort_data 436 11.2 1.840 2.068 1.840 2.068 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.994 2.010 rebuild_ks_matrix 3 7.3 0.000 0.000 1.981 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.041 1.981 1.997 dbcsr_data_new 4174 10.1 1.617 1.833 1.617 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.525000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1683.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 2.699592E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. 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 0.082 0.093 96.371 96.373 qs_energies 1 2.0 0.001 0.012 95.147 95.169 ls_scf 1 3.0 0.003 0.014 93.716 93.729 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.447 77.665 ls_scf_main 1 4.0 0.000 0.001 58.298 58.307 multiply_cannon 111 7.7 0.040 0.065 53.100 56.946 multiply_cannon_loop 111 8.7 0.100 0.106 49.573 53.863 density_matrix_trs4 2 5.0 0.008 0.061 51.997 52.237 mp_waitall_1 7281 11.0 24.289 34.529 24.289 34.529 ls_scf_init_scf 1 4.0 0.000 0.003 31.856 31.858 ls_scf_init_matrix_S 1 5.0 0.079 0.322 30.321 30.392 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.793 27.878 multiply_cannon_multrec 888 9.7 12.532 15.378 21.062 24.498 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.217 23.899 make_m2s 222 7.7 0.006 0.007 17.238 18.453 make_images 222 8.7 1.963 2.284 17.200 18.414 hybrid_alltoall_any 227 10.6 0.620 2.859 9.496 11.081 make_images_data 222 9.7 0.003 0.004 9.803 10.808 mp_sum_l 807 5.4 5.557 9.443 5.557 9.443 dbcsr_mm_accdrv_process 3754 10.4 0.257 0.440 8.044 9.307 dbcsr_mm_accdrv_process_sort 3754 11.4 7.670 8.867 7.670 8.867 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.314 7.591 multiply_cannon_sync_h2d 888 9.7 6.000 7.365 6.000 7.365 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.893 7.153 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.462 7.131 mp_irecv_dv 2335 11.1 2.447 7.085 2.447 7.085 arnoldi_extremal 4 6.8 0.000 0.000 5.288 5.378 arnoldi_normal_ev 4 7.8 0.001 0.005 5.288 5.378 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.933 5.189 build_subspace 16 8.4 0.014 0.020 4.752 4.758 calculate_norms 1584 9.8 4.312 4.672 4.312 4.672 mp_allgather_i34 111 8.7 1.369 3.859 1.369 3.859 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.421 3.772 dbcsr_matrix_vector_mult_local 304 10.0 3.016 3.613 3.018 3.615 ls_scf_post 1 4.0 0.003 0.013 3.559 3.576 ls_scf_store_result 1 5.0 0.000 0.000 3.296 3.385 ls_scf_dm_to_ks 2 5.0 0.076 0.330 3.135 3.218 make_images_sizes 222 9.7 0.000 0.000 1.119 2.290 mp_alltoall_i44 222 10.7 1.119 2.289 1.119 2.289 make_images_pack 222 9.7 1.816 2.130 1.819 2.133 dbcsr_sort_data 325 11.1 1.888 2.113 1.888 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.108 2.110 rebuild_ks_matrix 3 7.3 0.000 0.000 2.090 2.092 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.027 2.090 2.092 dbcsr_data_release 9322 10.9 1.317 1.951 1.317 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.373000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 3.358507E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. 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 0.082 0.156 97.471 97.495 qs_energies 1 2.0 0.000 0.000 96.535 96.545 ls_scf 1 3.0 0.002 0.007 94.868 94.876 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.493 78.683 ls_scf_main 1 4.0 0.000 0.000 58.949 58.961 multiply_cannon 111 7.7 0.057 0.137 51.707 55.634 density_matrix_trs4 2 5.0 0.002 0.003 52.835 52.948 multiply_cannon_loop 111 8.7 0.114 0.125 46.711 49.479 ls_scf_init_scf 1 4.0 0.000 0.002 32.646 32.659 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.386 31.466 mp_waitall_1 6369 11.0 22.983 30.151 22.983 30.151 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.898 28.909 multiply_cannon_multrec 1332 9.7 14.167 17.029 22.006 24.508 make_m2s 222 7.7 0.006 0.007 21.157 22.574 make_images 222 8.7 3.143 3.599 21.107 22.526 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.510 16.960 make_images_data 222 9.7 0.004 0.004 11.841 13.407 hybrid_alltoall_any 227 10.6 0.797 3.834 11.225 13.046 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.412 7.478 9.012 dbcsr_mm_accdrv_process_sort 3641 11.4 7.106 8.595 7.106 8.595 mp_sum_l 807 5.4 4.041 7.051 4.041 7.051 multiply_cannon_sync_h2d 1332 9.7 5.522 6.277 5.522 6.277 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.077 5.905 mp_irecv_dv 3229 10.9 2.056 5.860 2.056 5.860 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.123 5.501 arnoldi_extremal 4 6.8 0.000 0.000 5.175 5.188 arnoldi_normal_ev 4 7.8 0.001 0.005 5.175 5.188 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.463 5.068 build_subspace 16 8.4 0.014 0.021 4.840 4.848 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.500 4.679 calculate_norms 2376 9.8 4.192 4.568 4.192 4.568 mp_allgather_i34 111 8.7 2.110 4.352 2.110 4.352 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.577 3.873 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.694 3.186 3.696 dbcsr_sort_data 658 11.4 3.099 3.532 3.099 3.532 ls_scf_post 1 4.0 0.000 0.000 3.271 3.281 dbcsr_special_finalize 555 9.7 0.006 0.007 2.837 3.252 dbcsr_merge_single_wm 555 10.7 0.538 0.664 2.828 3.244 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.049 3.109 ls_scf_store_result 1 5.0 0.000 0.000 2.977 3.050 dbcsr_data_release 10477 10.7 1.591 2.442 1.591 2.442 dbcsr_finalize 304 7.8 0.049 0.061 1.817 2.038 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.986 1.997 rebuild_ks_matrix 3 7.3 0.000 0.000 1.964 1.974 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.021 1.964 1.974 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.495000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 4.631577E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. 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 0.071 0.108 99.887 99.887 qs_energies 1 2.0 0.000 0.000 98.614 98.620 ls_scf 1 3.0 0.000 0.000 96.667 96.677 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.008 78.218 ls_scf_main 1 4.0 0.000 0.000 62.245 62.254 multiply_cannon 111 7.7 0.082 0.170 55.416 60.680 density_matrix_trs4 2 5.0 0.002 0.003 54.981 55.072 multiply_cannon_loop 111 8.7 0.070 0.078 50.996 52.561 mp_waitall_1 5436 11.0 26.791 32.666 26.791 32.666 ls_scf_init_scf 1 4.0 0.000 0.000 30.806 30.812 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.385 29.432 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.241 27.247 multiply_cannon_multrec 444 9.7 13.936 16.773 20.953 22.953 make_m2s 222 7.7 0.004 0.005 17.813 20.207 make_images 222 8.7 3.727 4.443 17.752 20.146 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.108 16.838 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.462 14.631 make_images_data 222 9.7 0.003 0.004 10.023 12.415 hybrid_alltoall_any 227 10.6 0.790 3.771 9.861 12.214 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.346 6.723 7.859 multiply_cannon_sync_h2d 444 9.7 6.584 7.774 6.584 7.774 dbcsr_mm_accdrv_process_sort 3003 11.4 6.410 7.513 6.410 7.513 mp_allgather_i34 111 8.7 2.654 7.029 2.654 7.029 arnoldi_extremal 4 6.8 0.000 0.000 5.738 5.748 arnoldi_normal_ev 4 7.8 0.002 0.005 5.738 5.748 build_subspace 16 8.4 0.015 0.020 5.350 5.364 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.567 4.713 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.677 4.515 mp_irecv_dv 1241 11.2 1.659 4.488 1.659 4.488 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.143 4.330 mp_sum_l 807 5.4 2.840 4.231 2.840 4.231 dbcsr_matrix_vector_mult_local 304 10.0 3.678 4.139 3.680 4.141 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.676 3.756 calculate_norms 792 9.8 3.558 3.665 3.558 3.665 ls_scf_post 1 4.0 0.000 0.000 3.615 3.622 make_images_sizes 222 9.7 0.000 0.000 1.144 3.505 mp_alltoall_i44 222 10.7 1.144 3.505 1.144 3.505 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.446 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.888 3.027 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.272 2.275 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.269 rebuild_ks_matrix 3 7.3 0.000 0.000 2.240 2.243 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.011 2.240 2.243 dbcsr_merge_all 275 8.9 0.474 0.524 2.051 2.102 dbcsr_data_release 10123 10.8 1.324 2.012 1.324 2.012 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.887000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3640.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5def639336615614009cc9e019e9532e6735d0e9_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 8.717570E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. 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 0.100 0.156 109.417 109.418 qs_energies 1 2.0 0.000 0.000 107.875 107.890 ls_scf 1 3.0 0.000 0.000 104.672 104.687 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.541 77.653 ls_scf_main 1 4.0 0.000 0.000 66.545 66.706 density_matrix_trs4 2 5.0 0.002 0.003 56.861 56.906 multiply_cannon 111 7.7 0.139 0.272 50.110 52.181 multiply_cannon_loop 111 8.7 0.067 0.070 46.603 47.472 ls_scf_init_scf 1 4.0 0.000 0.000 34.457 34.854 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.531 32.547 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.710 29.721 mp_waitall_1 4527 11.1 22.338 26.124 22.338 26.124 make_m2s 222 7.7 0.005 0.005 23.850 24.881 make_images 222 8.7 4.581 4.964 23.744 24.773 multiply_cannon_multrec 444 9.7 17.888 18.656 22.482 23.167 hybrid_alltoall_any 227 10.6 1.658 3.624 12.901 15.632 make_images_data 222 9.7 0.003 0.003 13.095 15.568 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.657 11.074 multiply_cannon_sync_h2d 444 9.7 8.849 8.899 8.849 8.899 arnoldi_extremal 4 6.8 0.000 0.000 7.288 7.303 arnoldi_normal_ev 4 7.8 0.003 0.009 7.288 7.303 build_subspace 16 8.4 0.026 0.037 6.736 6.745 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.416 5.561 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.255 5.341 dbcsr_matrix_vector_mult_local 304 10.0 5.007 5.302 5.010 5.305 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.014 5.259 dbcsr_mm_accdrv_process 1814 10.4 0.252 0.316 4.420 4.541 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.251 4.122 4.251 ls_scf_post 1 4.0 0.000 0.000 3.670 3.685 mp_allgather_i34 111 8.7 1.157 3.642 1.157 3.642 make_images_sizes 222 9.7 0.000 0.000 1.436 3.471 mp_alltoall_i44 222 10.7 1.436 3.471 1.436 3.471 ls_scf_store_result 1 5.0 0.000 0.000 3.416 3.426 calculate_norms 792 9.8 3.239 3.285 3.239 3.285 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.849 3.246 rebuild_ks_matrix 3 7.3 0.000 0.000 2.783 3.181 qs_ks_build_kohn_sham_matrix 3 8.3 0.188 0.584 2.783 3.181 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 3.172 3.172 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.159 dbcsr_merge_all 275 8.9 0.892 0.915 2.873 2.940 dbcsr_complete_redistribute 5 7.6 1.436 1.475 2.764 2.870 dbcsr_data_release 12724 10.6 2.332 2.842 2.332 2.842 matrix_ls_to_qs 2 6.0 0.000 0.000 2.416 2.538 dbcsr_sort_data 325 11.1 2.441 2.497 2.441 2.497 ls_scf_initial_guess 1 5.0 0.000 0.000 1.926 2.316 ls_scf_qs_atomic_guess 1 6.0 0.000 0.000 1.926 2.316 dbcsr_new_transposed 4 7.5 0.243 0.252 2.294 2.310 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.194 2.236 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.204 dbcsr_add_anytype 103 7.2 0.859 0.892 2.133 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.418000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6837.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5def639336615614009cc9e019e9532e6735d0e9 Summary: empty Status: OK