=== 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: a11bcb3b8565e52344423df66485879857c7b754 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, 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 (19.04.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/01 job id: 46765294 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/02 job id: 46765295 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/03 job id: 46765296 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/04 job id: 46765297 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/05 job id: 46765298 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/06 job id: 46765299 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/07 job id: 46765300 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/08 job id: 46765301 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/09 job id: 46765302 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/10 job id: 46765303 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/11 job id: 46765304 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/12 job id: 46765305 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/13 job id: 46765306 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/14 job id: 46765307 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/15 job id: 46765308 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/16 job id: 46765309 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/17 job id: 46765310 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/18 job id: 46765311 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/19 job id: 46765312 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/20 job id: 46765313 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/21 job id: 46765314 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/22 job id: 46765315 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/23 job id: 46765317 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/24 job id: 46765319 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/25 job id: 46765320 --- 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/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/26 job id: 46765322 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/27 job id: 46765323 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.060 140.648 140.649 farming_run 1 2.0 138.784 138.787 140.580 140.581 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.470153E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 115.458 115.459 qs_energies 1 2.0 0.000 0.000 115.266 115.267 mp2_main 1 3.0 0.000 0.000 113.212 113.214 mp2_gpw_main 1 4.0 0.019 0.025 112.202 112.203 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.120 93.744 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.404 56.030 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.723 47.317 get_2c_integrals 1 6.0 0.008 0.009 36.830 37.544 integrate_v_rspace 273 8.0 0.435 0.448 25.120 30.347 pw_transfer 6555 10.6 0.373 0.380 27.354 27.995 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.038 26.680 grid_integrate_task_list 273 9.0 20.936 26.631 20.936 26.631 fft_wrap_pw1pw2_100 2178 12.4 1.168 1.235 23.560 24.204 compute_2c_integrals 1 7.0 0.002 0.003 19.227 19.228 rpa_ri_compute_en 1 5.0 0.019 0.020 18.971 19.139 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.803 18.954 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.420 18.800 18.951 cp_fm_cholesky_decompose 12 8.2 17.531 18.289 17.531 18.289 cholesky_decomp 1 7.0 0.000 0.000 16.446 17.205 fft3d_s 5443 13.4 16.144 16.774 16.167 16.795 ao_to_mo_and_store_B_mult_1 272 7.0 10.849 15.576 10.849 15.576 calculate_wavefunction 272 8.0 5.409 5.568 12.527 13.153 rpa_num_int 1 6.0 0.000 0.000 10.623 10.632 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.613 10.621 calc_mat_Q 8 8.0 0.000 0.000 9.484 9.566 contract_S_to_Q 8 9.0 0.000 0.000 8.904 8.989 parallel_gemm_fm 14 9.1 0.000 0.000 8.491 8.590 parallel_gemm_fm_cosma 14 10.1 8.491 8.590 8.491 8.590 calc_potential_gpw 544 9.5 0.005 0.006 8.229 8.562 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.171 8.410 potential_pw2rs 545 10.0 0.107 0.109 7.663 8.363 collocate_single_gaussian 272 10.0 0.039 0.041 7.440 7.720 create_integ_mat 1 6.0 0.015 0.027 7.629 7.638 array2fm 1 7.0 0.000 0.000 6.758 7.107 pw_scatter_s 2720 13.7 4.444 4.596 4.444 4.596 pw_gather_s 2722 13.2 3.878 4.214 3.878 4.214 array2fm_buffer_send 1 8.0 3.046 3.145 3.046 3.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.203242, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.037 402.844 402.846 farming_run 1 2.0 402.108 402.114 402.811 402.813 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227665E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 15763251. 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 0.009 0.031 210.448 210.449 qs_energies 1 2.0 0.000 0.000 210.235 210.248 scf_env_do_scf 1 3.0 0.000 0.000 106.719 106.719 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.844 105.852 rebuild_ks_matrix 4 6.0 0.000 0.000 105.843 105.850 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.843 105.850 hfx_ks_matrix 4 8.0 0.001 0.001 105.460 105.464 integrate_four_center 4 9.0 0.144 0.454 105.459 105.463 mp2_main 1 3.0 0.000 0.000 103.228 103.241 mp2_gpw_main 1 4.0 0.031 0.046 102.354 102.370 integrate_four_center_main 4 10.0 0.097 0.557 96.941 99.814 integrate_four_center_bin 269 11.0 96.844 99.451 96.844 99.451 init_scf_loop 1 4.0 0.000 0.000 92.353 92.353 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.053 76.362 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.648 55.956 mp2_eri_3c_integrate_gpw 91 7.0 0.142 0.159 42.071 47.249 integrate_v_rspace 95 8.0 0.399 0.573 28.464 33.481 pw_transfer 2240 10.6 0.144 0.166 29.955 30.351 ao_to_mo_and_store_B_mult_1 91 7.0 10.902 29.527 10.902 29.527 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.952 29.370 mp2_ri_gpw_compute_en 1 5.0 0.060 0.073 27.154 29.182 grid_integrate_task_list 95 9.0 23.762 28.976 23.762 28.976 fft_wrap_pw1pw2_100 730 12.4 1.296 1.448 26.649 27.016 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.908 25.193 25.204 get_2c_integrals 1 6.0 0.000 0.000 20.300 20.340 compute_2c_integrals 1 7.0 0.002 0.003 19.283 19.288 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.940 19.153 mp2_eri_2c_integrate_gpw 1 9.0 1.722 1.858 18.939 19.152 fft3d_s 1823 13.4 18.440 18.663 18.453 18.677 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.364 14.364 calculate_wavefunction 91 8.0 2.013 2.042 9.730 9.918 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.584 8.860 9.487 potential_pw2rs 186 10.0 0.034 0.036 8.653 9.194 local_gemm 172 8.0 8.303 8.922 8.303 8.922 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.527 8.109 8.676 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.306 8.643 calc_potential_gpw 182 9.5 0.002 0.002 7.945 8.235 collocate_single_gaussian 91 10.0 0.017 0.025 7.916 8.154 mp_sendrecv_dm3 2068 8.0 6.128 6.671 6.128 6.671 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.445 6.342 6.445 mp_sync 37 10.5 3.197 5.869 3.197 5.869 pw_gather_s 912 13.2 4.889 5.445 4.889 5.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.349800, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 448.544768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 1324047. 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 0.057 0.117 65.924 65.931 qs_mol_dyn_low 1 2.0 0.003 0.011 64.043 64.691 qs_forces 11 3.9 0.002 0.003 63.930 63.936 qs_energies 11 4.9 0.001 0.001 62.394 62.409 scf_env_do_scf 11 5.9 0.000 0.001 45.984 45.984 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.755 43.755 dbcsr_multiply_generic 2286 12.5 0.089 0.093 42.017 42.462 multiply_cannon 2286 13.5 0.185 0.203 34.933 36.572 multiply_cannon_loop 2286 14.5 1.480 1.566 34.157 35.888 qs_scf_new_mos 108 7.5 0.000 0.001 33.028 33.296 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.027 33.295 ot_scf_mini 108 9.5 0.002 0.002 31.435 31.603 velocity_verlet 10 3.0 0.001 0.002 26.806 26.807 mp_waitall_1 245248 16.5 12.737 19.167 12.737 19.167 ot_mini 108 10.5 0.001 0.001 18.758 18.996 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 9.954 17.716 qs_ot_get_derivative 108 11.5 0.004 0.030 15.817 16.001 init_scf_run 11 5.9 0.000 0.001 14.839 14.839 scf_env_initial_rho_setup 11 6.9 0.002 0.003 14.839 14.839 calculate_first_density_matrix 1 7.0 0.000 0.000 12.491 12.495 multiply_cannon_multrec 54864 15.5 4.294 6.644 7.504 10.902 qs_ot_get_p 119 10.4 0.001 0.001 8.181 8.483 rebuild_ks_matrix 119 8.3 0.000 0.000 8.255 8.415 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.255 8.414 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.973 7.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.267 7.414 calculate_dm_sparse 119 9.5 0.000 0.001 7.238 7.363 acc_transpose_blocks 54864 15.5 0.226 0.249 6.500 7.245 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.082 7.151 multiply_cannon_sync_h2d 54864 15.5 6.020 6.905 6.020 6.905 multiply_cannon_metrocomm1 54864 15.5 0.054 0.059 1.974 6.659 mp_sum_l 7287 12.8 4.928 6.629 4.928 6.629 acc_transpose_blocks_kernels 54864 16.5 0.247 0.390 5.542 6.093 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.646 6.093 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 5.814 5.839 jit_kernel_transpose 5 15.5 5.295 5.717 5.295 5.717 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.458 5.507 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.907 5.004 sum_up_and_integrate 119 10.3 0.013 0.015 4.799 4.805 integrate_v_rspace 119 11.3 0.002 0.003 4.786 4.794 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.697 4.698 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.549 4.655 calculate_rho_elec 119 8.7 0.012 0.017 4.548 4.654 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.441 4.442 dbcsr_mm_accdrv_process 76910 16.1 1.146 1.839 3.130 4.419 cp_fm_redistribute_end 50 14.0 2.264 4.404 2.273 4.408 cp_fm_diag_elpa_base 50 14.0 2.128 4.279 2.132 4.284 rs_pw_transfer 974 11.9 0.012 0.013 3.031 3.176 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.932 3.134 apply_single 119 13.6 0.000 0.000 2.932 3.134 ot_diis_step 108 11.5 0.006 0.006 2.703 2.704 density_rs2pw 119 9.7 0.004 0.005 2.481 2.589 jit_kernel_multiply 13 15.8 1.921 2.501 1.921 2.501 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.337 2.406 wfi_extrapolate 11 7.9 0.001 0.001 2.283 2.283 init_scf_loop 11 6.9 0.000 0.000 2.207 2.207 grid_integrate_task_list 119 12.3 2.023 2.128 2.023 2.128 potential_pw2rs 119 12.3 0.004 0.004 2.087 2.108 pw_transfer 1439 11.6 0.052 0.056 2.025 2.107 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.949 2.032 make_m2s 4572 13.5 0.053 0.056 1.894 1.952 mp_sum_d 4135 12.0 1.297 1.930 1.297 1.930 make_images 4572 14.5 0.133 0.139 1.812 1.869 fft3d_ps 1201 14.6 0.370 0.479 1.593 1.673 mp_alltoall_d11v 2130 13.8 1.420 1.597 1.420 1.597 fft_wrap_pw1pw2_140 487 13.2 0.183 0.198 1.494 1.574 mp_waitany 12084 13.8 1.316 1.513 1.316 1.513 grid_collocate_task_list 119 9.7 1.375 1.474 1.375 1.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=65.931000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 486.182912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 708689. 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 0.050 0.110 42.600 42.600 qs_mol_dyn_low 1 2.0 0.003 0.003 42.255 42.265 qs_forces 11 3.9 0.002 0.003 41.035 41.036 qs_energies 11 4.9 0.001 0.001 39.315 39.319 scf_env_do_scf 11 5.9 0.000 0.001 32.843 32.844 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.296 30.296 dbcsr_multiply_generic 2286 12.5 0.121 0.147 21.860 22.230 qs_scf_new_mos 108 7.5 0.001 0.001 20.768 21.020 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.768 21.020 ot_scf_mini 108 9.5 0.002 0.003 19.866 20.046 velocity_verlet 10 3.0 0.001 0.002 19.834 19.859 multiply_cannon 2286 13.5 0.210 0.220 16.714 18.252 multiply_cannon_loop 2286 14.5 0.897 0.964 15.560 16.936 ot_mini 108 10.5 0.001 0.001 12.161 12.401 mp_waitall_1 200699 16.5 5.977 11.308 5.977 11.308 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.407 9.921 qs_ot_get_derivative 108 11.5 0.001 0.001 9.686 9.867 multiply_cannon_multrec 27432 15.5 1.970 4.555 5.909 8.962 rebuild_ks_matrix 119 8.3 0.000 0.000 7.404 7.543 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.404 7.543 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.541 6.668 dbcsr_mm_accdrv_process 47894 16.0 3.117 5.446 3.869 5.942 init_scf_run 11 5.9 0.000 0.001 5.247 5.247 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.246 5.247 qs_ot_get_p 119 10.4 0.001 0.001 4.784 5.019 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.928 4.772 sum_up_and_integrate 119 10.3 0.025 0.028 4.375 4.379 integrate_v_rspace 119 11.3 0.002 0.003 4.350 4.358 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.060 4.135 apply_single 119 13.6 0.000 0.000 3.060 4.135 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.990 4.030 calculate_rho_elec 119 8.7 0.021 0.024 3.989 4.029 mp_sum_l 7287 12.8 2.055 4.013 2.055 4.013 calculate_first_density_matrix 1 7.0 0.000 0.000 3.656 3.658 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.278 3.296 rs_pw_transfer 974 11.9 0.010 0.011 2.745 3.206 make_m2s 4572 13.5 0.052 0.054 2.618 2.863 multiply_cannon_sync_h2d 27432 15.5 2.210 2.854 2.210 2.854 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.837 2.837 make_images 4572 14.5 0.201 0.237 2.533 2.777 density_rs2pw 119 9.7 0.004 0.004 2.254 2.711 init_scf_loop 11 6.9 0.000 0.000 2.525 2.525 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.476 2.476 cp_fm_redistribute_end 50 14.0 1.254 2.441 1.260 2.446 ot_diis_step 108 11.5 0.011 0.011 2.428 2.429 cp_fm_diag_elpa_base 50 14.0 1.150 2.327 1.181 2.375 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.174 2.267 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.212 2.215 calculate_dm_sparse 119 9.5 0.000 0.001 2.094 2.169 pw_transfer 1439 11.6 0.065 0.069 2.032 2.066 potential_pw2rs 119 12.3 0.006 0.006 2.036 2.050 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.940 1.976 grid_integrate_task_list 119 12.3 1.876 1.974 1.876 1.974 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.774 1.816 jit_kernel_multiply 8 16.1 0.700 1.766 0.700 1.766 prepare_preconditioner 11 7.9 0.000 0.000 1.601 1.628 make_preconditioner 11 8.9 0.000 0.000 1.601 1.628 acc_transpose_blocks 27432 15.5 0.109 0.115 1.278 1.589 make_images_data 4572 15.5 0.045 0.052 1.163 1.581 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.496 1.555 wfi_extrapolate 11 7.9 0.001 0.001 1.532 1.532 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.006 1.520 fft3d_ps 1201 14.6 0.518 0.573 1.490 1.518 fft_wrap_pw1pw2_140 487 13.2 0.202 0.214 1.476 1.513 parallel_gemm_fm 81 9.0 0.000 0.000 1.497 1.501 parallel_gemm_fm_cosma 81 10.0 1.497 1.501 1.497 1.501 grid_collocate_task_list 119 9.7 1.287 1.425 1.287 1.425 make_basis_sm 11 9.8 0.001 0.003 1.392 1.394 mp_alltoall_d11v 2130 13.8 1.250 1.351 1.250 1.351 mp_allgather_i34 2286 14.5 0.591 1.348 0.591 1.348 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.305 mp_sum_dm 438 4.9 1.214 1.266 1.214 1.266 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.187 1.233 md_write_output 11 3.9 0.023 1.083 0.027 1.206 update_particle_set 20 4.0 0.000 0.000 1.178 1.203 md_output 10 3.0 0.000 0.000 0.028 1.165 mp_sum_d 4135 12.0 0.578 1.026 0.578 1.026 mp_waitany 5720 13.7 0.519 1.002 0.519 1.002 rs_pw_transfer_RS2PW_140 130 11.5 0.142 0.151 0.547 0.988 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.746 0.980 qs_energies_init_hamiltonians 11 5.9 0.010 0.017 0.945 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.935 make_images_sizes 4572 15.5 0.005 0.005 0.641 0.905 mp_alltoall_i44 4572 16.5 0.636 0.901 0.636 0.901 dbcsr_complete_redistribute 329 12.2 0.154 0.318 0.813 0.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.600000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 516.997120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607886. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.043 35.476 35.476 qs_mol_dyn_low 1 2.0 0.003 0.005 35.165 35.174 qs_forces 11 3.9 0.003 0.003 34.318 34.319 qs_energies 11 4.9 0.001 0.001 32.717 32.720 scf_env_do_scf 11 5.9 0.000 0.001 26.324 26.324 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.766 23.766 dbcsr_multiply_generic 2286 12.5 0.094 0.096 17.177 17.262 velocity_verlet 10 3.0 0.001 0.001 16.366 16.392 qs_scf_new_mos 108 7.5 0.001 0.001 15.133 15.154 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.132 15.153 multiply_cannon 2286 13.5 0.194 0.202 13.963 14.707 ot_scf_mini 108 9.5 0.002 0.002 14.420 14.430 multiply_cannon_loop 2286 14.5 0.632 0.660 13.166 13.941 ot_mini 108 10.5 0.001 0.001 8.958 8.970 qs_ot_get_derivative 108 11.5 0.001 0.001 7.475 7.485 multiply_cannon_multrec 18288 15.5 1.977 2.920 7.103 7.472 rebuild_ks_matrix 119 8.3 0.000 0.000 6.458 6.477 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.458 6.476 dbcsr_mm_accdrv_process 38222 16.0 4.157 5.904 5.040 6.086 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.688 5.704 init_scf_run 11 5.9 0.000 0.001 5.245 5.245 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.244 5.244 mp_waitall_1 158411 16.6 2.971 4.290 2.971 4.290 sum_up_and_integrate 119 10.3 0.032 0.033 4.039 4.045 calculate_first_density_matrix 1 7.0 0.000 0.000 4.041 4.043 integrate_v_rspace 119 11.3 0.003 0.003 4.007 4.017 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.031 3.634 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.500 3.511 calculate_rho_elec 119 8.7 0.031 0.032 3.499 3.510 qs_ot_get_p 119 10.4 0.001 0.001 3.294 3.310 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.776 2.999 init_scf_loop 11 6.9 0.000 0.000 2.542 2.542 rs_pw_transfer 974 11.9 0.009 0.010 2.268 2.533 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.009 2.329 apply_single 119 13.6 0.000 0.000 2.009 2.328 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.254 2.258 jit_kernel_multiply 10 16.3 0.831 2.227 0.831 2.227 density_rs2pw 119 9.7 0.004 0.004 1.970 2.218 calculate_dm_sparse 119 9.5 0.000 0.001 2.183 2.197 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.993 1.993 make_m2s 4572 13.5 0.044 0.046 1.793 1.945 grid_integrate_task_list 119 12.3 1.836 1.939 1.836 1.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.925 1.927 pw_transfer 1439 11.6 0.065 0.070 1.913 1.922 make_images 4572 14.5 0.191 0.202 1.710 1.861 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.821 1.833 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.796 1.802 prepare_preconditioner 11 7.9 0.000 0.000 1.749 1.752 make_preconditioner 11 8.9 0.000 0.000 1.749 1.752 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.738 1.747 potential_pw2rs 119 12.3 0.007 0.008 1.738 1.746 cp_fm_diag_elpa_base 50 14.0 1.715 1.725 1.737 1.745 acc_transpose_blocks 18288 15.5 0.075 0.077 1.443 1.702 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.603 1.687 multiply_cannon_sync_h2d 18288 15.5 1.412 1.674 1.412 1.674 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.561 1.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.535 1.542 parallel_gemm_fm 81 9.0 0.000 0.000 1.535 1.539 parallel_gemm_fm_cosma 81 10.0 1.535 1.538 1.535 1.538 make_basis_sm 11 9.8 0.001 0.001 1.477 1.479 fft_wrap_pw1pw2_140 487 13.2 0.255 0.261 1.458 1.470 ot_diis_step 108 11.5 0.011 0.011 1.463 1.464 grid_collocate_task_list 119 9.7 1.261 1.411 1.261 1.411 mp_sum_l 7287 12.8 1.032 1.365 1.032 1.365 fft3d_ps 1201 14.6 0.528 0.544 1.305 1.315 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 1.005 1.263 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 jit_kernel_transpose 5 15.6 0.796 1.049 0.796 1.049 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.401 0.984 make_images_data 4572 15.5 0.045 0.049 0.804 0.978 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.935 0.936 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.688 0.890 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.866 0.889 mp_alltoall_d11v 2130 13.8 0.691 0.872 0.691 0.872 mp_sum_dm 438 4.9 0.821 0.850 0.821 0.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.842 md_write_output 11 3.9 0.026 0.817 0.028 0.834 update_particle_set 20 4.0 0.000 0.000 0.803 0.830 mp_waitany 9880 13.7 0.505 0.786 0.505 0.786 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.124 0.503 0.777 cp_fm_cholesky_invert 11 10.9 0.734 0.738 0.734 0.738 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.636 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.476000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.272727, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.004480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.134 0.248 46.973 46.992 qs_mol_dyn_low 1 2.0 0.013 0.045 45.790 45.797 qs_forces 11 3.9 0.068 0.188 42.240 42.265 qs_energies 11 4.9 0.002 0.008 40.428 40.495 scf_env_do_scf 11 5.9 0.001 0.001 28.934 28.935 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 25.485 25.485 dbcsr_multiply_generic 2286 12.5 0.096 0.100 23.192 23.325 velocity_verlet 10 3.0 0.001 0.002 21.516 21.660 multiply_cannon 2286 13.5 0.224 0.269 19.444 20.338 multiply_cannon_loop 2286 14.5 0.936 0.963 18.469 19.179 qs_scf_new_mos 108 7.5 0.001 0.001 16.698 16.754 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.697 16.753 ot_scf_mini 108 9.5 0.003 0.003 15.763 15.811 ot_mini 108 10.5 0.001 0.001 9.397 9.457 multiply_cannon_multrec 27432 15.5 2.335 3.098 8.582 9.032 init_scf_run 11 5.9 0.000 0.001 8.805 8.806 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.805 8.806 qs_ot_get_derivative 108 11.5 0.001 0.001 7.585 7.635 calculate_first_density_matrix 1 7.0 0.000 0.002 7.393 7.395 dbcsr_mm_accdrv_process 47916 15.9 5.293 6.922 6.153 7.364 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.913 6.914 rebuild_ks_matrix 119 8.3 0.000 0.000 6.792 6.854 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.791 6.854 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.646 6.666 acc_transpose_blocks 27432 15.5 0.110 0.112 6.431 6.465 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.032 6.089 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 5.809 5.835 jit_kernel_transpose 5 15.6 5.542 5.572 5.542 5.572 sum_up_and_integrate 119 10.3 0.036 0.038 4.003 4.011 integrate_v_rspace 119 11.3 0.003 0.003 3.967 3.976 qs_ot_get_p 119 10.4 0.001 0.001 3.756 3.830 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.696 3.723 calculate_rho_elec 119 8.7 0.040 0.046 3.696 3.723 rs_pw_transfer 974 11.9 0.009 0.010 3.119 3.572 mp_sum_dm 438 4.9 3.341 3.493 3.341 3.493 md_write_output 11 3.9 0.021 0.485 0.146 3.471 md_output 10 3.0 0.000 0.000 0.147 3.469 update_particle_set 20 4.0 0.000 0.000 3.323 3.467 init_scf_loop 11 6.9 0.001 0.003 3.427 3.427 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.819 3.231 write_trajectory 44 4.9 0.003 0.049 0.125 2.986 write_particle_coordinates 11 5.9 0.122 2.936 0.122 2.936 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.046 2.578 apply_single 119 13.6 0.000 0.000 2.046 2.578 prepare_preconditioner 11 7.9 0.000 0.001 2.561 2.570 make_preconditioner 11 8.9 0.000 0.001 2.561 2.570 mp_waitall_1 137007 16.6 1.941 2.517 1.941 2.517 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.163 2.489 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.420 2.484 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.438 2.446 make_m2s 4572 13.5 0.054 0.055 2.266 2.374 make_images 4572 14.5 0.270 0.332 2.161 2.267 density_rs2pw 119 9.7 0.004 0.004 2.057 2.243 pw_transfer 1439 11.6 0.065 0.070 2.098 2.137 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.115 2.115 calculate_dm_sparse 119 9.5 0.000 0.001 2.026 2.086 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.006 2.049 qs_env_update_s_mstruct 11 6.9 0.002 0.008 1.754 2.011 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.941 1.968 grid_integrate_task_list 119 12.3 1.867 1.942 1.867 1.942 rs_pw_transfer_RS2PW_140 130 11.5 0.117 0.141 1.438 1.881 mp_waitany 7280 13.7 1.425 1.875 1.425 1.875 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.834 1.843 cp_fm_diag_elpa_base 50 14.0 1.797 1.813 1.831 1.840 jit_kernel_multiply 10 16.1 0.799 1.818 0.799 1.818 mp_sum_l 7287 12.8 1.162 1.793 1.162 1.793 ot_diis_step 108 11.5 0.012 0.012 1.770 1.770 potential_pw2rs 119 12.3 0.009 0.010 1.683 1.689 fft_wrap_pw1pw2_140 487 13.2 0.289 0.302 1.617 1.662 calculate_rho_core 11 7.9 0.221 0.602 1.306 1.551 fft3d_ps 1201 14.6 0.555 0.608 1.444 1.468 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.842 1.439 grid_collocate_task_list 119 9.7 1.264 1.378 1.264 1.378 wfi_extrapolate 11 7.9 0.001 0.001 1.359 1.359 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.222 build_core_hamiltonian_matrix 11 6.9 0.059 0.323 0.572 1.170 cp_fm_upper_to_full 72 14.2 0.815 1.164 0.815 1.164 dbcsr_complete_redistribute 329 12.2 0.141 0.235 0.879 1.147 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.104 1.121 multiply_cannon_sync_h2d 27432 15.5 0.981 1.047 0.981 1.047 make_images_data 4572 15.5 0.045 0.048 0.874 0.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=46.992000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.545455, yerr=4.335134 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 598.196224E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.097 29.573 29.573 qs_mol_dyn_low 1 2.0 0.003 0.004 29.203 29.217 qs_forces 11 3.9 0.002 0.002 29.136 29.137 qs_energies 11 4.9 0.001 0.001 27.390 27.392 scf_env_do_scf 11 5.9 0.000 0.001 21.769 21.769 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.206 19.207 velocity_verlet 10 3.0 0.001 0.001 14.698 14.701 dbcsr_multiply_generic 2286 12.5 0.089 0.094 12.023 12.098 qs_scf_new_mos 108 7.5 0.001 0.001 10.947 10.969 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.946 10.969 ot_scf_mini 108 9.5 0.002 0.002 10.275 10.298 multiply_cannon 2286 13.5 0.230 0.236 9.507 9.976 multiply_cannon_loop 2286 14.5 0.328 0.340 8.603 8.780 rebuild_ks_matrix 119 8.3 0.000 0.000 6.097 6.117 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.096 6.117 multiply_cannon_multrec 9144 15.5 1.573 1.807 5.786 6.041 ot_mini 108 10.5 0.001 0.001 5.621 5.650 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.416 5.435 qs_ot_get_derivative 108 11.5 0.001 0.001 4.337 4.359 dbcsr_mm_accdrv_process 12550 15.8 3.388 3.820 4.113 4.215 init_scf_run 11 5.9 0.000 0.001 4.164 4.165 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.164 4.164 sum_up_and_integrate 119 10.3 0.038 0.042 3.790 3.794 integrate_v_rspace 119 11.3 0.003 0.003 3.752 3.756 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.606 3.611 calculate_rho_elec 119 8.7 0.060 0.061 3.606 3.611 calculate_first_density_matrix 1 7.0 0.000 0.000 2.988 2.989 qs_ot_get_p 119 10.4 0.001 0.001 2.885 2.926 init_scf_loop 11 6.9 0.000 0.000 2.542 2.543 pw_transfer 1439 11.6 0.066 0.069 2.075 2.089 density_rs2pw 119 9.7 0.004 0.004 1.897 2.039 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.981 1.996 grid_integrate_task_list 119 12.3 1.900 1.976 1.900 1.976 mp_waitall_1 115863 16.7 1.416 1.962 1.416 1.962 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.935 1.938 make_m2s 4572 13.5 0.035 0.036 1.735 1.892 rs_pw_transfer 974 11.9 0.008 0.009 1.764 1.890 make_images 4572 14.5 0.268 0.302 1.648 1.804 prepare_preconditioner 11 7.9 0.000 0.000 1.791 1.795 make_preconditioner 11 8.9 0.000 0.000 1.791 1.795 calculate_dm_sparse 119 9.5 0.000 0.000 1.757 1.774 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.737 1.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.680 1.705 fft_wrap_pw1pw2_140 487 13.2 0.366 0.376 1.614 1.630 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.534 1.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.546 1.547 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.457 1.465 cp_fm_diag_elpa_base 50 14.0 1.428 1.447 1.455 1.463 jit_kernel_multiply 8 16.0 0.687 1.462 0.687 1.462 potential_pw2rs 119 12.3 0.010 0.011 1.450 1.456 grid_collocate_task_list 119 9.7 1.309 1.424 1.309 1.424 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.377 1.386 fft3d_ps 1201 14.6 0.561 0.572 1.312 1.323 ot_diis_step 108 11.5 0.012 0.013 1.270 1.271 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.246 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.185 parallel_gemm_fm 81 9.0 0.000 0.000 1.125 1.177 parallel_gemm_fm_cosma 81 10.0 1.124 1.177 1.124 1.177 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.144 1.162 apply_single 119 13.6 0.000 0.000 1.144 1.162 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.138 1.150 make_basis_sm 11 9.8 0.001 0.001 1.132 1.133 wfi_extrapolate 11 7.9 0.001 0.001 1.128 1.128 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.760 1.000 make_images_data 4572 15.5 0.039 0.043 0.800 0.992 acc_transpose_blocks 9144 15.5 0.038 0.039 0.903 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.921 cp_fm_cholesky_invert 11 10.9 0.886 0.889 0.886 0.889 mp_alltoall_d11v 2130 13.8 0.777 0.886 0.777 0.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.806 multiply_cannon_sync_h2d 9144 15.5 0.705 0.780 0.705 0.780 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.700 0.751 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.723 0.733 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.664 0.680 mp_alltoall_z22v 1201 16.6 0.625 0.653 0.625 0.653 mp_allgather_i34 2286 14.5 0.237 0.638 0.237 0.638 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.313 0.628 qs_create_task_list 11 7.9 0.015 0.031 0.573 0.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.573000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.545455, yerr=4.207431 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 752.533504E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 47.734 47.735 qs_mol_dyn_low 1 2.0 0.003 0.004 47.498 47.504 qs_forces 11 3.9 0.002 0.002 46.612 46.613 qs_energies 11 4.9 0.001 0.001 44.617 44.620 scf_env_do_scf 11 5.9 0.001 0.001 33.999 33.999 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.052 26.052 velocity_verlet 10 3.0 0.001 0.001 24.259 24.368 dbcsr_multiply_generic 2286 12.5 0.095 0.098 22.668 22.910 multiply_cannon 2286 13.5 0.301 0.311 18.791 19.825 multiply_cannon_loop 2286 14.5 0.342 0.349 17.537 18.614 qs_scf_new_mos 108 7.5 0.001 0.001 16.205 16.303 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.204 16.302 ot_scf_mini 108 9.5 0.002 0.002 15.114 15.214 ot_mini 108 10.5 0.001 0.001 9.111 9.225 init_scf_run 11 5.9 0.000 0.001 8.599 8.599 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.599 8.599 multiply_cannon_multrec 9144 15.5 3.361 4.695 8.492 8.593 init_scf_loop 11 6.9 0.000 0.000 7.919 7.919 rebuild_ks_matrix 119 8.3 0.000 0.000 7.365 7.508 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.365 7.508 qs_ot_get_derivative 108 11.5 0.001 0.001 7.095 7.196 calculate_first_density_matrix 1 7.0 0.000 0.000 7.118 7.120 prepare_preconditioner 11 7.9 0.000 0.000 6.955 6.969 make_preconditioner 11 8.9 0.000 0.000 6.955 6.969 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.516 6.848 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.827 6.828 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.639 6.769 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.569 6.614 acc_transpose_blocks 9144 15.5 0.038 0.039 6.057 6.331 dbcsr_mm_accdrv_process 12550 15.8 4.142 5.596 5.009 6.315 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 5.807 6.081 jit_kernel_transpose 5 15.6 5.690 5.964 5.690 5.964 cp_fm_upper_to_full 72 14.2 3.219 4.604 3.219 4.604 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.287 4.295 calculate_rho_elec 119 8.7 0.118 0.121 4.287 4.294 sum_up_and_integrate 119 10.3 0.065 0.066 4.083 4.089 integrate_v_rspace 119 11.3 0.004 0.004 4.018 4.025 mp_waitall_1 94719 16.7 2.467 3.720 2.467 3.720 qs_ot_get_p 119 10.4 0.001 0.001 3.270 3.407 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.770 3.255 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.232 2.833 apply_single 119 13.6 0.000 0.000 2.232 2.833 dbcsr_complete_redistribute 329 12.2 0.289 0.292 1.985 2.793 pw_transfer 1439 11.6 0.069 0.069 2.666 2.672 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.447 2.613 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.567 2.573 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.669 2.476 make_m2s 4572 13.5 0.038 0.038 2.259 2.425 make_images 4572 14.5 0.355 0.387 2.141 2.306 mp_alltoall_i22 627 13.8 1.410 2.259 1.410 2.259 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.234 density_rs2pw 119 9.7 0.004 0.004 2.178 2.191 calculate_dm_sparse 119 9.5 0.000 0.000 2.158 2.175 fft_wrap_pw1pw2_140 487 13.2 0.618 0.622 2.158 2.164 grid_integrate_task_list 119 12.3 2.124 2.146 2.124 2.146 mp_sum_l 7287 12.8 1.075 2.108 1.075 2.108 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.006 2.007 ot_diis_step 108 11.5 0.014 0.015 1.990 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.751 1.752 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.733 1.734 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.680 1.730 fft3d_ps 1201 14.6 0.594 0.606 1.583 1.588 grid_collocate_task_list 119 9.7 1.518 1.530 1.518 1.530 cp_fm_cholesky_invert 11 10.9 1.500 1.503 1.500 1.503 rs_pw_transfer 974 11.9 0.009 0.009 1.474 1.489 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.433 1.433 cp_fm_diag_elpa_base 50 14.0 1.282 1.335 1.431 1.431 wfi_extrapolate 11 7.9 0.001 0.001 1.412 1.413 potential_pw2rs 119 12.3 0.014 0.014 1.403 1.406 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.115 1.315 make_images_data 4572 15.5 0.043 0.046 1.045 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.214 1.223 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.140 1.160 mp_alltoall_d11v 2130 13.8 1.102 1.140 1.102 1.140 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.086 1.106 multiply_cannon_sync_h2d 9144 15.5 1.043 1.047 1.043 1.047 jit_kernel_multiply 6 15.7 0.840 1.016 0.840 1.016 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.982 1.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.735000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=709.454545, yerr=13.950325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 496.971776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1975108. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.225 88.206 88.208 qs_mol_dyn_low 1 2.0 0.003 0.005 87.692 87.703 qs_forces 11 3.9 0.006 0.024 87.124 87.125 qs_energies 11 4.9 0.002 0.008 84.199 84.217 scf_env_do_scf 11 5.9 0.001 0.002 69.733 69.736 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 64.149 64.150 dbcsr_multiply_generic 2055 12.4 0.102 0.105 55.555 55.761 multiply_cannon 2055 13.4 0.178 0.185 47.421 48.486 multiply_cannon_loop 2055 14.4 1.511 1.552 46.437 47.561 qs_scf_new_mos 99 7.5 0.000 0.001 46.490 46.612 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.489 46.611 ot_scf_mini 99 9.5 0.002 0.003 44.139 44.267 velocity_verlet 10 3.0 0.005 0.032 43.150 43.156 ot_mini 99 10.5 0.001 0.001 25.906 26.026 qs_ot_get_derivative 99 11.5 0.001 0.002 19.155 19.273 multiply_cannon_multrec 49320 15.4 12.556 13.240 17.341 18.119 rebuild_ks_matrix 110 8.3 0.000 0.000 14.798 14.921 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 14.797 14.920 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.953 13.067 init_scf_run 11 5.9 0.000 0.001 12.332 12.332 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.332 12.332 mp_waitall_1 220248 16.4 10.442 11.710 10.442 11.710 multiply_cannon_sync_h2d 49320 15.4 10.472 11.082 10.472 11.082 qs_ot_get_p 110 10.4 0.003 0.013 9.881 10.004 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.001 9.009 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.656 8.696 calculate_first_density_matrix 1 7.0 0.000 0.002 8.121 8.127 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.192 7.724 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.207 7.679 apply_single 110 13.6 0.000 0.001 7.206 7.679 acc_transpose_blocks 49320 15.4 0.206 0.216 7.459 7.647 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 6.250 7.472 sum_up_and_integrate 110 10.3 0.037 0.044 7.291 7.307 integrate_v_rspace 110 11.3 0.003 0.004 7.254 7.279 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.783 6.912 calculate_rho_elec 110 8.6 0.022 0.026 6.782 6.911 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.787 6.822 acc_transpose_blocks_kernels 49320 16.4 0.413 0.437 6.462 6.631 ot_diis_step 99 11.5 0.006 0.007 6.562 6.562 jit_kernel_transpose 5 15.6 6.049 6.221 6.049 6.221 cp_dbcsr_syevd 48 12.0 0.003 0.005 5.895 5.895 init_scf_loop 11 6.9 0.001 0.009 5.558 5.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.331 5.398 cp_fm_diag_elpa 48 13.0 0.000 0.002 5.316 5.337 cp_fm_diag_elpa_base 48 14.0 5.300 5.321 5.312 5.333 dbcsr_mm_accdrv_process 87628 16.1 1.851 1.959 4.659 4.956 mp_sum_l 6594 12.7 3.918 4.837 3.918 4.837 rs_pw_transfer 902 11.9 0.012 0.014 3.853 4.480 density_rs2pw 110 9.6 0.006 0.015 3.493 4.173 wfi_extrapolate 11 7.9 0.003 0.014 4.099 4.099 make_m2s 4110 13.4 0.060 0.065 3.826 3.958 make_images 4110 14.4 0.177 0.191 3.731 3.865 calculate_dm_sparse 110 9.5 0.001 0.001 3.724 3.853 grid_integrate_task_list 110 12.3 3.333 3.509 3.333 3.509 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.301 3.501 prepare_preconditioner 11 7.9 0.000 0.000 3.339 3.365 make_preconditioner 11 8.9 0.000 0.001 3.339 3.365 pw_transfer 1331 11.6 0.055 0.067 3.185 3.264 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.096 3.178 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.116 3.177 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.095 3.154 fft_wrap_pw1pw2_140 451 13.1 0.454 0.501 2.625 2.727 potential_pw2rs 110 12.3 0.006 0.007 2.695 2.720 mp_alltoall_d11v 2046 13.8 2.097 2.681 2.097 2.681 mp_waitany 14300 13.8 1.859 2.624 1.859 2.624 jit_kernel_multiply 13 15.9 2.534 2.575 2.534 2.575 grid_collocate_task_list 110 9.6 2.175 2.415 2.175 2.415 fft3d_ps 1111 14.6 0.786 0.884 2.287 2.335 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.977 1.996 mp_sum_d 3889 11.9 1.420 1.949 1.420 1.949 make_images_data 4110 15.4 0.043 0.046 1.762 1.887 cp_fm_cholesky_invert 11 10.9 1.834 1.838 1.834 1.838 hybrid_alltoall_any 4261 16.3 0.082 0.486 1.534 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.208000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=471.636364, yerr=2.185603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.424896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2334102. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.069 75.694 75.697 qs_mol_dyn_low 1 2.0 0.005 0.018 75.292 75.300 qs_forces 11 3.9 0.025 0.073 74.574 74.575 qs_energies 11 4.9 0.034 0.091 71.066 71.079 scf_env_do_scf 11 5.9 0.002 0.014 57.121 57.125 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.518 49.519 dbcsr_multiply_generic 2055 12.4 0.112 0.116 42.702 42.902 multiply_cannon 2055 13.4 0.222 0.253 36.148 37.477 velocity_verlet 10 3.0 0.001 0.002 36.877 36.892 multiply_cannon_loop 2055 14.4 0.919 0.947 34.873 35.891 qs_scf_new_mos 99 7.5 0.001 0.001 32.934 33.078 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.933 33.078 ot_scf_mini 99 9.5 0.003 0.004 31.289 31.423 ot_mini 99 10.5 0.001 0.001 18.131 18.270 multiply_cannon_multrec 24660 15.4 7.622 9.509 13.823 15.966 rebuild_ks_matrix 110 8.3 0.000 0.000 14.009 14.156 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.016 14.009 14.156 qs_ot_get_derivative 99 11.5 0.001 0.002 12.339 12.476 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.256 12.384 init_scf_run 11 5.9 0.000 0.001 11.435 11.436 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.435 11.436 mp_waitall_1 176588 16.5 7.673 10.012 7.673 10.012 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.524 8.525 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 8.181 8.237 multiply_cannon_sync_h2d 24660 15.4 6.998 7.975 6.998 7.975 multiply_cannon_metrocomm3 24660 15.4 0.068 0.073 5.185 7.905 calculate_first_density_matrix 1 7.0 0.001 0.004 7.801 7.803 init_scf_loop 11 6.9 0.004 0.015 7.566 7.569 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.493 7.144 apply_single 110 13.6 0.000 0.001 6.493 7.144 sum_up_and_integrate 110 10.3 0.054 0.063 6.862 6.884 acc_transpose_blocks 24660 15.4 0.111 0.115 6.663 6.831 integrate_v_rspace 110 11.3 0.003 0.003 6.808 6.829 qs_ot_get_p 110 10.4 0.001 0.001 6.572 6.704 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.366 6.375 calculate_rho_elec 110 8.6 0.040 0.048 6.365 6.374 dbcsr_mm_accdrv_process 52282 16.1 4.591 5.425 6.040 6.318 acc_transpose_blocks_kernels 24660 16.4 0.309 0.320 6.042 6.203 jit_kernel_transpose 5 15.6 5.733 5.895 5.733 5.895 ot_diis_step 99 11.5 0.010 0.010 5.746 5.746 prepare_preconditioner 11 7.9 0.000 0.000 5.476 5.497 make_preconditioner 11 8.9 0.000 0.001 5.476 5.497 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.638 5.343 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.072 5.235 qs_ot_p2m_diag 48 11.0 0.031 0.051 4.636 4.657 make_m2s 4110 13.4 0.056 0.059 4.150 4.634 make_images 4110 14.4 0.399 0.443 4.048 4.529 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.201 4.204 density_rs2pw 110 9.6 0.005 0.005 3.411 4.035 pw_transfer 1331 11.6 0.067 0.074 3.738 3.887 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.630 3.778 rs_pw_transfer 902 11.9 0.012 0.014 3.116 3.778 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.631 3.644 cp_fm_diag_elpa_base 48 14.0 3.583 3.598 3.628 3.643 wfi_extrapolate 11 7.9 0.001 0.001 3.535 3.535 grid_integrate_task_list 110 12.3 3.241 3.462 3.241 3.462 fft_wrap_pw1pw2_140 451 13.1 0.524 0.543 3.091 3.238 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.162 3.233 calculate_dm_sparse 110 9.5 0.001 0.001 2.945 2.973 make_images_data 4110 15.4 0.048 0.056 2.315 2.792 fft3d_ps 1111 14.6 1.102 1.324 2.598 2.766 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.040 2.719 grid_collocate_task_list 110 9.6 2.118 2.551 2.118 2.551 cp_fm_cholesky_invert 11 10.9 2.531 2.538 2.531 2.538 potential_pw2rs 110 12.3 0.008 0.009 2.518 2.537 mp_sum_l 6594 12.7 1.799 2.390 1.799 2.390 mp_alltoall_d11v 2046 13.8 1.795 2.118 1.795 2.118 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.939 1.959 jit_kernel_multiply 10 16.1 1.104 1.942 1.104 1.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.913 1.927 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.905 1.914 mp_waitany 10164 13.8 1.277 1.880 1.277 1.880 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.800 1.751 rs_pw_transfer_RS2PW_140 121 11.5 0.212 0.226 1.024 1.677 mp_irecv_dv 57340 16.2 0.673 1.631 0.673 1.631 mp_allgather_i34 2055 14.4 0.604 1.610 0.604 1.610 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.543 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=75.697000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.636364, yerr=7.437897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 656.437248E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.167 66.387 66.388 qs_mol_dyn_low 1 2.0 0.003 0.003 65.866 65.875 qs_forces 11 3.9 0.003 0.003 65.377 65.378 qs_energies 11 4.9 0.002 0.007 62.096 62.100 scf_env_do_scf 11 5.9 0.001 0.001 49.046 49.046 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.578 40.578 dbcsr_multiply_generic 2055 12.4 0.104 0.109 33.708 33.974 velocity_verlet 10 3.0 0.001 0.001 33.107 33.122 multiply_cannon 2055 13.4 0.212 0.228 27.474 28.715 multiply_cannon_loop 2055 14.4 0.615 0.627 26.279 27.226 qs_scf_new_mos 99 7.5 0.001 0.001 25.225 25.311 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.225 25.310 ot_scf_mini 99 9.5 0.006 0.028 24.006 24.131 ot_mini 99 10.5 0.001 0.001 13.812 13.936 rebuild_ks_matrix 110 8.3 0.000 0.000 12.593 12.718 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.592 12.718 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.058 11.167 multiply_cannon_multrec 16440 15.4 3.932 4.823 9.751 10.823 init_scf_run 11 5.9 0.000 0.001 10.632 10.633 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.632 10.632 mp_waitall_1 139946 16.5 7.256 10.194 7.256 10.194 qs_ot_get_derivative 99 11.5 0.001 0.001 9.413 9.535 init_scf_loop 11 6.9 0.001 0.005 8.433 8.434 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 7.892 7.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.542 7.589 calculate_first_density_matrix 1 7.0 0.000 0.002 7.550 7.551 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.587 7.295 sum_up_and_integrate 110 10.3 0.061 0.062 6.741 6.755 integrate_v_rspace 110 11.3 0.003 0.003 6.680 6.693 prepare_preconditioner 11 7.9 0.000 0.000 6.641 6.659 make_preconditioner 11 8.9 0.000 0.001 6.641 6.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.972 6.316 acc_transpose_blocks 16440 15.4 0.074 0.077 6.260 6.314 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.139 6.147 calculate_rho_elec 110 8.6 0.059 0.059 6.138 6.146 dbcsr_mm_accdrv_process 34862 16.1 4.448 5.455 5.671 6.062 acc_transpose_blocks_kernels 16440 16.4 0.209 0.213 5.833 5.888 jit_kernel_transpose 5 15.6 5.624 5.679 5.624 5.679 qs_ot_get_p 110 10.4 0.003 0.017 5.404 5.525 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.927 5.359 apply_single 110 13.6 0.000 0.000 4.927 5.358 make_m2s 4110 13.4 0.051 0.062 4.145 4.531 make_images 4110 14.4 0.392 0.506 4.034 4.422 density_rs2pw 110 9.6 0.005 0.005 3.171 4.391 ot_diis_step 99 11.5 0.010 0.011 4.373 4.373 multiply_cannon_sync_h2d 16440 15.4 3.658 4.205 3.658 4.205 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.507 4.184 rs_pw_transfer 902 11.9 0.011 0.012 2.759 3.963 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.805 3.809 pw_transfer 1331 11.6 0.066 0.072 3.691 3.702 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.584 3.595 grid_integrate_task_list 110 12.3 3.267 3.514 3.267 3.514 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.470 3.470 fft_wrap_pw1pw2_140 451 13.1 0.641 0.648 3.085 3.100 wfi_extrapolate 11 7.9 0.001 0.001 2.995 2.995 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.936 2.945 cp_fm_diag_elpa_base 48 14.0 2.868 2.900 2.935 2.943 make_images_data 4110 15.4 0.044 0.048 2.381 2.898 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.121 2.843 jit_kernel_multiply 12 16.3 0.843 2.664 0.843 2.664 cp_fm_cholesky_invert 11 10.9 2.579 2.585 2.579 2.585 calculate_dm_sparse 110 9.5 0.001 0.001 2.513 2.543 grid_collocate_task_list 110 9.6 2.139 2.522 2.139 2.522 mp_waitany 17072 13.8 1.194 2.448 1.194 2.448 fft3d_ps 1111 14.6 1.088 1.097 2.402 2.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.003 2.309 2.374 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.874 2.361 potential_pw2rs 110 12.3 0.011 0.012 2.322 2.329 mp_irecv_dv 48980 15.7 0.805 2.238 0.805 2.238 mp_alltoall_d11v 2046 13.8 1.861 2.233 1.861 2.233 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.183 0.924 2.129 mp_sum_l 6594 12.7 1.508 2.095 1.508 2.095 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.974 1.976 dbcsr_complete_redistribute 325 12.2 0.334 0.374 1.438 1.877 cp_fm_upper_to_full 70 14.2 1.386 1.752 1.386 1.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.732 1.749 mp_allgather_i34 2055 14.4 0.518 1.592 0.518 1.592 cp_fm_cholesky_decompose 22 10.9 1.555 1.572 1.555 1.572 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.490 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.476 1.488 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.992 1.429 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=66.388000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=619.181818, yerr=9.271836 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 727.396352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.100 66.423 66.423 qs_mol_dyn_low 1 2.0 0.003 0.004 66.047 66.056 qs_forces 11 3.9 0.003 0.003 65.763 65.763 qs_energies 11 4.9 0.001 0.001 62.251 62.255 scf_env_do_scf 11 5.9 0.000 0.001 53.376 53.379 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.790 41.791 velocity_verlet 10 3.0 0.001 0.001 37.281 37.292 dbcsr_multiply_generic 2055 12.4 0.111 0.116 29.460 29.628 qs_scf_new_mos 99 7.5 0.001 0.001 26.472 26.578 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.472 26.577 ot_scf_mini 99 9.5 0.002 0.003 24.863 24.955 multiply_cannon 2055 13.4 0.247 0.268 22.578 23.619 multiply_cannon_loop 2055 14.4 0.883 0.899 21.008 21.526 ot_mini 99 10.5 0.001 0.001 14.022 14.137 multiply_cannon_multrec 24660 15.4 4.230 6.741 12.706 13.901 rebuild_ks_matrix 110 8.3 0.000 0.000 12.414 12.535 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.413 12.534 init_scf_loop 11 6.9 0.000 0.000 11.544 11.545 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.907 11.019 qs_ot_get_derivative 99 11.5 0.001 0.001 9.851 9.950 prepare_preconditioner 11 7.9 0.000 0.000 9.785 9.800 make_preconditioner 11 8.9 0.000 0.000 9.785 9.800 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.024 9.457 dbcsr_mm_accdrv_process 52304 16.0 7.086 8.469 8.329 9.298 sum_up_and_integrate 110 10.3 0.068 0.071 6.684 6.699 integrate_v_rspace 110 11.3 0.003 0.003 6.616 6.629 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.274 6.284 calculate_rho_elec 110 8.6 0.078 0.082 6.274 6.284 mp_waitall_1 121746 16.5 4.187 6.278 4.187 6.278 init_scf_run 11 5.9 0.000 0.001 6.121 6.122 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.121 6.121 qs_ot_get_p 110 10.4 0.001 0.001 5.815 5.953 make_m2s 4110 13.4 0.059 0.060 5.296 5.800 make_images 4110 14.4 0.578 0.701 5.159 5.660 cp_fm_upper_to_full 70 14.2 3.343 4.823 3.343 4.823 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.078 4.177 apply_single 110 13.6 0.000 0.000 4.078 4.176 ot_diis_step 99 11.5 0.011 0.011 4.135 4.135 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.058 4.073 pw_transfer 1331 11.6 0.066 0.076 3.788 3.821 dbcsr_complete_redistribute 325 12.2 0.417 0.460 2.629 3.769 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.682 3.719 density_rs2pw 110 9.6 0.004 0.005 3.116 3.635 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.578 3.578 grid_integrate_task_list 110 12.3 3.363 3.529 3.363 3.529 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.265 3.339 multiply_cannon_sync_h2d 24660 15.4 3.196 3.309 3.196 3.309 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.107 3.225 fft_wrap_pw1pw2_140 451 13.1 0.673 0.695 3.150 3.187 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.431 3.151 make_images_data 4110 15.4 0.046 0.050 2.646 3.147 hybrid_alltoall_any 4261 16.3 0.120 0.454 2.261 3.141 calculate_first_density_matrix 1 7.0 0.000 0.000 3.048 3.051 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.994 3.046 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.031 3.043 cp_fm_diag_elpa_base 48 14.0 2.878 2.940 3.028 3.041 calculate_dm_sparse 110 9.5 0.001 0.001 2.972 3.001 rs_pw_transfer 902 11.9 0.010 0.011 2.460 2.982 wfi_extrapolate 11 7.9 0.001 0.001 2.980 2.980 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.751 2.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.777 2.778 mp_alltoall_i22 605 13.7 1.610 2.746 1.610 2.746 cp_fm_cholesky_invert 11 10.9 2.666 2.673 2.666 2.673 grid_collocate_task_list 110 9.6 2.242 2.490 2.242 2.490 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.473 fft3d_ps 1111 14.6 1.083 1.117 2.436 2.451 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.263 2.264 jit_kernel_multiply 12 15.7 0.915 2.254 0.915 2.254 potential_pw2rs 110 12.3 0.013 0.013 2.202 2.209 mp_alltoall_d11v 2046 13.8 1.790 2.111 1.790 2.111 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.749 1.760 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.685 1.722 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.708 cp_fm_cholesky_decompose 22 10.9 1.634 1.683 1.634 1.683 mp_allgather_i34 2055 14.4 0.662 1.659 0.662 1.659 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.600 1.617 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.841 1.593 mp_waitany 13376 13.8 1.055 1.564 1.055 1.564 mp_sum_l 6594 12.7 0.898 1.544 0.898 1.544 acc_transpose_blocks 24660 15.4 0.103 0.105 1.493 1.518 mp_irecv_dv 62702 16.1 0.742 1.514 0.742 1.514 make_basis_sm 11 9.8 0.001 0.001 1.361 1.364 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.183 0.812 1.346 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.423000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.909091, yerr=7.959609 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 825.065472E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 56.537 56.538 qs_mol_dyn_low 1 2.0 0.003 0.003 56.269 56.302 qs_forces 11 3.9 0.003 0.004 56.179 56.180 qs_energies 11 4.9 0.001 0.001 52.446 52.451 scf_env_do_scf 11 5.9 0.000 0.001 44.097 44.097 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 36.384 36.385 velocity_verlet 10 3.0 0.001 0.002 31.164 31.169 dbcsr_multiply_generic 2055 12.4 0.102 0.107 23.754 23.953 qs_scf_new_mos 99 7.5 0.001 0.001 21.249 21.302 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.248 21.302 ot_scf_mini 99 9.5 0.002 0.002 20.028 20.059 multiply_cannon 2055 13.4 0.243 0.260 18.075 19.502 multiply_cannon_loop 2055 14.4 0.319 0.331 16.737 17.375 rebuild_ks_matrix 110 8.3 0.000 0.000 12.023 12.063 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.023 12.063 ot_mini 99 10.5 0.001 0.001 11.411 11.439 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.639 10.676 multiply_cannon_multrec 8220 15.4 3.275 5.354 7.474 9.436 mp_waitall_1 103326 16.6 6.678 8.794 6.678 8.794 init_scf_loop 11 6.9 0.000 0.000 7.667 7.667 qs_ot_get_derivative 99 11.5 0.001 0.001 7.570 7.600 sum_up_and_integrate 110 10.3 0.079 0.081 6.629 6.640 integrate_v_rspace 110 11.3 0.003 0.003 6.549 6.561 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.334 6.352 calculate_rho_elec 110 8.6 0.115 0.116 6.334 6.352 prepare_preconditioner 11 7.9 0.000 0.000 6.002 6.008 make_preconditioner 11 8.9 0.000 0.000 6.002 6.007 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.608 5.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.589 5.660 init_scf_run 11 5.9 0.000 0.001 5.229 5.229 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.229 5.229 dbcsr_mm_accdrv_process 17442 15.9 2.805 3.908 4.069 4.960 qs_ot_get_p 110 10.4 0.001 0.001 4.726 4.774 make_m2s 4110 13.4 0.039 0.040 4.107 4.366 make_images 4110 14.4 0.639 0.697 3.980 4.235 pw_transfer 1331 11.6 0.067 0.072 3.902 3.915 ot_diis_step 99 11.5 0.012 0.012 3.811 3.811 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.794 3.811 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.735 3.792 apply_single 110 13.6 0.000 0.000 3.734 3.792 grid_integrate_task_list 110 12.3 3.453 3.670 3.453 3.670 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.401 3.405 density_rs2pw 110 9.6 0.004 0.005 3.019 3.394 fft_wrap_pw1pw2_140 451 13.1 0.837 0.849 3.286 3.313 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.102 3.102 multiply_cannon_sync_h2d 8220 15.4 2.917 3.022 2.917 3.022 cp_fm_cholesky_invert 11 10.9 2.926 2.930 2.926 2.930 wfi_extrapolate 11 7.9 0.001 0.001 2.750 2.750 make_images_data 4110 15.4 0.039 0.044 2.344 2.748 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.671 2.672 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.271 2.635 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.584 2.591 grid_collocate_task_list 110 9.6 2.334 2.590 2.334 2.590 cp_fm_diag_elpa_base 48 14.0 2.527 2.549 2.582 2.590 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.581 2.583 rs_pw_transfer 902 11.9 0.010 0.011 2.185 2.580 calculate_dm_sparse 110 9.5 0.001 0.001 2.439 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.384 2.385 fft3d_ps 1111 14.6 1.134 1.149 2.333 2.347 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.236 2.248 potential_pw2rs 110 12.3 0.015 0.016 2.056 2.061 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.893 2.035 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.984 mp_alltoall_d11v 2046 13.8 1.629 1.959 1.629 1.959 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.850 1.872 mp_sum_l 6594 12.7 0.878 1.772 0.878 1.772 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.753 1.768 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.713 1.717 cp_fm_cholesky_decompose 22 10.9 1.641 1.654 1.641 1.654 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.509 1.627 mp_allgather_i34 2055 14.4 0.527 1.591 0.527 1.591 dbcsr_complete_redistribute 325 12.2 0.558 0.575 1.409 1.504 mp_waitany 9240 13.8 1.058 1.470 1.058 1.470 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.418 1.430 jit_kernel_multiply 8 15.7 0.963 1.419 0.963 1.419 qs_create_task_list 11 7.9 0.001 0.001 1.229 1.331 generate_qs_task_list 11 8.9 0.375 0.444 1.228 1.331 rs_gather_matrices 110 12.3 0.325 0.367 0.951 1.264 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.167 0.798 1.193 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.121 1.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.538000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=776.090909, yerr=10.833068 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 1.365606E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 94.486 94.486 qs_mol_dyn_low 1 2.0 0.003 0.004 94.152 94.160 qs_forces 11 3.9 0.003 0.003 94.058 94.059 qs_energies 11 4.9 0.001 0.001 89.784 89.787 scf_env_do_scf 11 5.9 0.001 0.001 74.325 74.325 velocity_verlet 10 3.0 0.001 0.001 56.464 56.474 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 45.399 45.401 dbcsr_multiply_generic 2055 12.4 0.116 0.121 34.890 34.936 init_scf_loop 11 6.9 0.000 0.000 28.854 28.857 multiply_cannon 2055 13.4 0.345 0.375 27.869 28.594 qs_scf_new_mos 99 7.5 0.001 0.001 27.082 27.115 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.081 27.114 prepare_preconditioner 11 7.9 0.000 0.000 26.829 26.837 make_preconditioner 11 8.9 0.000 0.000 26.829 26.837 multiply_cannon_loop 2055 14.4 0.339 0.344 26.094 26.534 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.747 26.319 ot_scf_mini 99 9.5 0.002 0.002 25.330 25.356 cp_fm_upper_to_full 70 14.2 12.981 18.814 12.981 18.814 ot_mini 99 10.5 0.001 0.001 14.562 14.583 rebuild_ks_matrix 110 8.3 0.000 0.001 14.209 14.247 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.208 14.247 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.827 12.862 multiply_cannon_multrec 8220 15.4 5.484 5.683 11.084 11.184 init_scf_run 11 5.9 0.000 0.001 11.138 11.138 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.137 11.137 dbcsr_complete_redistribute 325 12.2 1.025 1.049 7.643 11.079 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.624 10.050 qs_ot_get_derivative 99 11.5 0.001 0.001 10.018 10.043 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.067 9.448 mp_alltoall_i22 605 13.7 5.699 9.138 5.699 9.138 mp_waitall_1 84994 16.7 7.586 8.352 7.586 8.352 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.791 7.796 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.743 7.782 calculate_rho_elec 110 8.6 0.227 0.228 7.743 7.782 calculate_first_density_matrix 1 7.0 0.000 0.000 7.692 7.693 sum_up_and_integrate 110 10.3 0.150 0.151 7.409 7.423 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.307 7.314 integrate_v_rspace 110 11.3 0.004 0.004 7.258 7.273 acc_transpose_blocks 8220 15.4 0.035 0.035 5.829 5.847 make_m2s 4110 13.4 0.043 0.043 5.336 5.846 dbcsr_mm_accdrv_process 11614 15.7 3.340 3.692 5.457 5.728 make_images 4110 14.4 0.879 0.920 5.152 5.662 acc_transpose_blocks_kernels 8220 16.4 0.109 0.111 5.591 5.605 qs_ot_get_p 110 10.4 0.001 0.001 5.585 5.601 jit_kernel_transpose 5 15.6 5.482 5.494 5.482 5.494 cp_fm_cholesky_invert 11 10.9 5.263 5.268 5.263 5.268 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.651 5.104 apply_single 110 13.6 0.000 0.000 4.651 5.104 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.700 5.041 pw_transfer 1331 11.6 0.075 0.076 5.012 5.027 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.894 4.909 ot_diis_step 99 11.5 0.015 0.015 4.523 4.524 fft_wrap_pw1pw2_140 451 13.1 1.340 1.346 4.274 4.288 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.064 4.072 multiply_cannon_sync_h2d 8220 15.4 3.948 3.952 3.948 3.952 grid_integrate_task_list 110 12.3 3.797 3.861 3.797 3.861 density_rs2pw 110 9.6 0.004 0.005 3.715 3.756 hybrid_alltoall_any 4261 16.3 0.257 0.556 2.968 3.723 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.716 3.717 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.629 3.630 calculate_dm_sparse 110 9.5 0.001 0.001 3.593 3.610 make_images_data 4110 15.4 0.042 0.045 2.986 3.606 wfi_extrapolate 11 7.9 0.001 0.001 3.327 3.327 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.824 3.273 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.035 3.035 cp_fm_diag_elpa_base 48 14.0 2.481 2.692 3.032 3.033 fft3d_ps 1111 14.6 1.301 1.311 2.818 2.835 grid_collocate_task_list 110 9.6 2.693 2.706 2.693 2.706 potential_pw2rs 110 12.3 0.022 0.022 2.397 2.403 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.350 2.361 rs_pw_transfer 902 11.9 0.011 0.011 2.302 2.336 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.236 2.293 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.095 2.234 jit_kernel_multiply 10 15.2 1.921 2.212 1.921 2.212 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.137 2.152 cp_fm_cholesky_decompose 22 10.9 1.999 2.040 1.999 2.040 mp_alltoall_d11v 2046 13.8 1.894 2.009 1.894 2.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.983 1.986 qs_create_task_list 11 7.9 0.001 0.001 1.896 1.945 generate_qs_task_list 11 8.9 0.733 0.786 1.895 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.486000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1228.545455, yerr=62.707639 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.777536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2368970. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.059 209.325 209.327 qs_mol_dyn_low 1 2.0 0.003 0.004 208.219 208.233 qs_forces 11 3.9 0.005 0.007 208.107 208.108 qs_energies 11 4.9 0.002 0.008 202.452 202.460 scf_env_do_scf 11 5.9 0.001 0.002 185.657 185.662 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 165.305 165.307 dbcsr_multiply_generic 2507 12.6 0.172 0.175 125.167 125.829 qs_scf_new_mos 117 7.6 0.001 0.001 124.943 125.167 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.942 125.166 velocity_verlet 10 3.0 0.001 0.001 123.327 123.328 ot_scf_mini 117 9.6 0.003 0.004 118.307 118.537 multiply_cannon 2507 13.6 0.241 0.251 101.952 104.040 multiply_cannon_loop 2507 14.6 2.078 2.124 99.794 101.933 ot_mini 117 10.6 0.001 0.001 66.278 66.518 multiply_cannon_multrec 60168 15.6 33.411 34.967 41.546 43.101 qs_ot_get_derivative 117 11.6 0.001 0.001 41.483 41.685 rebuild_ks_matrix 128 8.3 0.001 0.001 33.990 34.542 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.990 34.541 mp_waitall_1 267128 16.5 28.826 32.522 28.826 32.522 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.462 30.950 qs_ot_get_p 128 10.4 0.001 0.001 30.024 30.270 multiply_cannon_sync_h2d 60168 15.6 27.583 29.418 27.583 29.418 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.382 25.067 apply_single 128 13.6 0.001 0.001 24.382 25.066 ot_diis_step 117 11.6 0.008 0.008 24.562 24.563 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.449 23.504 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.842 20.842 init_scf_loop 11 6.9 0.001 0.005 20.280 20.281 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.616 19.786 multiply_cannon_metrocomm3 60168 15.6 0.118 0.122 16.335 18.598 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.875 17.909 cp_fm_diag_elpa_base 83 14.4 17.802 17.843 17.871 17.906 prepare_preconditioner 11 7.9 0.000 0.001 15.684 15.741 make_preconditioner 11 8.9 0.000 0.001 15.684 15.741 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.933 15.151 sum_up_and_integrate 128 10.3 0.090 0.110 14.478 14.498 integrate_v_rspace 128 11.3 0.004 0.005 14.387 14.406 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.866 14.038 calculate_rho_elec 128 8.7 0.046 0.065 13.865 14.037 make_m2s 5014 13.6 0.102 0.112 13.588 13.936 make_images 5014 14.6 0.407 0.423 13.415 13.772 init_scf_run 11 5.9 0.000 0.001 12.583 12.584 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.583 12.584 density_rs2pw 128 9.7 0.007 0.008 7.077 10.316 mp_sum_l 7950 12.9 8.362 9.756 8.362 9.756 rs_pw_transfer 1046 11.9 0.017 0.019 5.867 9.090 wfi_extrapolate 11 7.9 0.001 0.001 9.066 9.066 cp_fm_cholesky_invert 11 10.9 9.000 9.007 9.000 9.007 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.575 8.706 calculate_dm_sparse 128 9.5 0.001 0.001 8.444 8.524 dbcsr_mm_accdrv_process 124484 16.2 3.080 3.202 7.697 8.198 multiply_cannon_metrocomm1 60168 15.6 0.092 0.096 6.070 8.137 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.873 7.964 pw_transfer 1547 11.6 0.075 0.089 7.679 7.847 grid_integrate_task_list 128 12.3 7.277 7.840 7.277 7.840 make_images_data 5014 15.6 0.071 0.076 6.699 7.685 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.476 7.645 hybrid_alltoall_any 5200 16.5 0.291 2.266 5.854 7.128 fft_wrap_pw1pw2_140 523 13.2 1.281 1.324 6.575 6.749 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.680 6.692 mp_waitany 16020 13.9 2.740 6.057 2.740 6.057 grid_collocate_task_list 128 9.7 4.707 5.979 4.707 5.979 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.833 5.957 mp_alltoall_d11v 2415 14.1 4.253 5.576 4.253 5.576 fft3d_ps 1291 14.7 2.134 2.781 5.231 5.550 rs_pw_transfer_RS2PW_140 139 11.5 0.337 0.368 2.228 5.458 potential_pw2rs 128 12.3 0.009 0.011 4.585 4.612 cp_fm_cholesky_decompose 22 10.9 4.539 4.552 4.539 4.552 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.327000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.909091, yerr=7.153569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 833.380352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3994 58335. MP_Allreduce 11063 960. MP_Sync 87 MP_Alltoall 1969 5578381. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.171 194.338 194.340 qs_mol_dyn_low 1 2.0 0.003 0.004 193.685 193.696 qs_forces 11 3.9 0.004 0.005 193.594 193.596 qs_energies 11 4.9 0.004 0.024 186.604 186.617 scf_env_do_scf 11 5.9 0.003 0.017 169.868 169.878 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 136.846 136.849 velocity_verlet 10 3.0 0.001 0.002 120.683 120.685 dbcsr_multiply_generic 2507 12.6 0.184 0.189 98.285 99.572 qs_scf_new_mos 117 7.6 0.001 0.001 96.183 96.769 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.182 96.769 ot_scf_mini 117 9.6 0.004 0.004 91.382 92.020 multiply_cannon 2507 13.6 0.475 0.528 78.067 82.271 multiply_cannon_loop 2507 14.6 1.241 1.276 75.065 77.734 ot_mini 117 10.6 0.001 0.001 50.963 51.512 mp_waitall_1 214728 16.6 25.841 40.179 25.841 40.179 multiply_cannon_multrec 30084 15.6 21.816 25.904 31.581 37.022 rebuild_ks_matrix 128 8.3 0.001 0.001 33.773 34.349 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.023 33.772 34.348 init_scf_loop 11 6.9 0.003 0.022 32.920 32.921 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.297 30.820 multiply_cannon_metrocomm3 30084 15.6 0.092 0.099 16.852 30.053 qs_ot_get_derivative 117 11.6 0.001 0.002 28.918 29.534 prepare_preconditioner 11 7.9 0.000 0.000 28.478 28.544 make_preconditioner 11 8.9 0.000 0.001 28.478 28.544 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.156 27.719 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.068 23.295 apply_single 128 13.6 0.001 0.001 22.067 23.295 qs_ot_get_p 128 10.4 0.001 0.002 22.313 23.008 ot_diis_step 117 11.6 0.014 0.015 21.865 21.868 multiply_cannon_sync_h2d 30084 15.6 19.085 21.154 19.085 21.154 qs_ot_p2m_diag 83 11.4 0.188 0.217 17.422 17.461 cp_fm_cholesky_invert 11 10.9 16.559 16.572 16.559 16.572 cp_dbcsr_syevd 83 12.4 0.006 0.008 16.266 16.267 make_m2s 5014 13.6 0.089 0.095 14.429 15.674 make_images 5014 14.6 1.158 1.351 14.233 15.481 sum_up_and_integrate 128 10.3 0.117 0.134 15.175 15.209 integrate_v_rspace 128 11.3 0.004 0.004 15.058 15.096 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.382 14.414 calculate_rho_elec 128 8.7 0.089 0.106 14.381 14.413 cp_fm_diag_elpa 83 13.4 0.001 0.002 13.033 13.067 cp_fm_diag_elpa_base 83 14.4 12.772 12.875 13.025 13.056 init_scf_run 11 5.9 0.000 0.001 11.893 11.895 scf_env_initial_rho_setup 11 6.9 0.006 0.046 11.893 11.895 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.004 11.442 multiply_cannon_metrocomm4 27577 15.6 0.103 0.117 3.835 10.740 dbcsr_mm_accdrv_process 62242 16.2 4.636 6.032 9.231 10.566 make_images_data 5014 15.6 0.068 0.075 8.809 10.354 mp_irecv_dv 69486 16.3 3.634 10.334 3.634 10.334 density_rs2pw 128 9.7 0.007 0.008 7.609 10.306 hybrid_alltoall_any 5200 16.5 0.345 1.511 7.452 9.554 pw_transfer 1547 11.6 0.086 0.102 8.890 8.951 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.665 8.729 rs_pw_transfer 1046 11.9 0.015 0.016 5.972 8.674 wfi_extrapolate 11 7.9 0.004 0.023 8.476 8.476 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.215 7.978 grid_integrate_task_list 128 12.3 7.405 7.964 7.405 7.964 fft_wrap_pw1pw2_140 523 13.2 1.340 1.360 7.606 7.691 cp_fm_cholesky_decompose 22 10.9 6.913 6.979 6.913 6.979 calculate_dm_sparse 128 9.5 0.001 0.001 6.503 6.652 mp_sum_l 7950 12.9 4.079 6.205 4.079 6.205 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.140 6.153 grid_collocate_task_list 128 9.7 4.829 6.029 4.829 6.029 fft3d_ps 1291 14.7 2.812 2.992 5.962 6.018 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.342 5.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.245 5.406 mp_waitany 11748 13.9 2.524 5.306 2.524 5.306 mp_alltoall_d11v 2415 14.1 4.251 5.175 4.251 5.175 potential_pw2rs 128 12.3 0.016 0.019 5.132 5.160 mp_allgather_i34 2507 14.6 1.583 4.830 1.583 4.830 rs_pw_transfer_RS2PW_140 139 11.5 0.356 0.382 2.124 4.822 mp_sum_d 4466 12.1 2.912 4.360 2.912 4.360 dbcsr_complete_redistribute 395 12.7 0.781 0.891 3.345 4.204 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.868 3.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.340000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.090909, yerr=2.429162 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 941.305856E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890820720 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57635. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.067 178.636 178.637 qs_mol_dyn_low 1 2.0 0.003 0.006 178.119 178.132 qs_forces 11 3.9 0.005 0.007 177.720 177.726 qs_energies 11 4.9 0.002 0.007 170.969 170.977 scf_env_do_scf 11 5.9 0.001 0.002 155.393 155.394 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 120.226 120.227 velocity_verlet 10 3.0 0.001 0.001 114.866 114.878 dbcsr_multiply_generic 2529 12.6 0.177 0.182 82.047 83.117 qs_scf_new_mos 118 7.6 0.001 0.001 81.155 81.496 qs_scf_loop_do_ot 118 8.6 0.001 0.002 81.154 81.495 ot_scf_mini 118 9.6 0.004 0.004 76.978 77.349 multiply_cannon 2529 13.6 0.505 0.535 62.229 66.092 multiply_cannon_loop 2529 14.6 0.862 0.895 59.226 61.916 ot_mini 118 10.6 0.001 0.001 42.878 43.256 init_scf_loop 11 6.9 0.001 0.003 35.067 35.068 mp_waitall_1 172006 16.6 25.676 34.938 25.676 34.938 rebuild_ks_matrix 129 8.3 0.001 0.001 31.392 31.887 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 31.391 31.887 prepare_preconditioner 11 7.9 0.000 0.000 31.009 31.053 make_preconditioner 11 8.9 0.000 0.001 31.009 31.053 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.659 30.059 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.223 28.685 multiply_cannon_multrec 20232 15.6 13.444 17.229 22.079 25.777 multiply_cannon_metrocomm3 20232 15.6 0.059 0.063 15.883 25.428 qs_ot_get_derivative 118 11.6 0.002 0.002 22.843 23.215 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.037 20.993 apply_single 129 13.6 0.001 0.001 20.037 20.993 ot_diis_step 118 11.6 0.018 0.019 19.934 19.935 qs_ot_get_p 129 10.4 0.001 0.001 19.360 19.817 multiply_cannon_sync_h2d 20232 15.6 14.254 15.806 14.254 15.806 make_m2s 5058 13.6 0.081 0.086 14.922 15.629 make_images 5058 14.6 1.179 1.266 14.697 15.408 sum_up_and_integrate 129 10.3 0.136 0.147 15.107 15.139 qs_ot_p2m_diag 84 11.4 0.269 0.276 15.107 15.116 integrate_v_rspace 129 11.3 0.004 0.005 14.971 14.997 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.671 14.705 calculate_rho_elec 129 8.7 0.133 0.149 14.670 14.704 cp_fm_cholesky_invert 11 10.9 14.538 14.547 14.538 14.547 cp_dbcsr_syevd 84 12.4 0.006 0.006 14.086 14.088 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.949 10.962 cp_fm_diag_elpa_base 84 14.4 10.529 10.692 10.946 10.959 init_scf_run 11 5.9 0.000 0.001 10.489 10.489 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.488 10.489 make_images_data 5058 15.6 0.063 0.069 9.222 10.363 multiply_cannon_metrocomm4 17703 15.6 0.062 0.072 3.578 9.566 hybrid_alltoall_any 5245 16.5 0.440 2.017 8.114 9.543 density_rs2pw 129 9.7 0.006 0.007 7.337 9.328 mp_irecv_dv 50659 16.2 3.455 9.314 3.455 9.314 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.007 9.292 pw_transfer 1559 11.6 0.087 0.106 9.009 9.113 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 8.782 8.892 dbcsr_mm_accdrv_process 41846 16.2 4.398 5.032 8.092 8.191 grid_integrate_task_list 129 12.3 7.565 8.067 7.565 8.067 fft_wrap_pw1pw2_140 527 13.2 1.432 1.458 7.764 7.882 wfi_extrapolate 11 7.9 0.001 0.001 7.483 7.483 rs_pw_transfer 1054 12.0 0.014 0.015 5.374 7.368 cp_fm_cholesky_decompose 22 10.9 7.326 7.362 7.326 7.362 cp_fm_upper_to_full 106 14.8 5.746 7.243 5.746 7.243 dbcsr_complete_redistribute 397 12.7 1.172 1.199 4.491 6.282 grid_collocate_task_list 129 9.7 5.060 5.968 5.060 5.968 fft3d_ps 1301 14.7 2.761 2.988 5.868 5.953 calculate_dm_sparse 129 9.5 0.001 0.001 5.818 5.886 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.391 5.398 mp_alltoall_d11v 2429 14.1 4.354 5.312 4.354 5.312 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.639 5.245 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.292 5.078 potential_pw2rs 129 12.3 0.021 0.024 4.833 4.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.495 4.621 mp_waitany 11836 13.9 2.453 4.542 2.453 4.542 mp_sum_l 8016 12.9 3.076 4.507 3.076 4.507 mp_allgather_i34 2529 14.6 1.518 4.230 1.518 4.230 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.066 4.098 transfer_fm_to_dbcsr 11 9.9 0.019 0.029 2.330 4.035 rs_pw_transfer_RS2PW_140 140 11.5 0.335 0.355 2.005 3.995 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.817 3.820 mp_alltoall_i22 720 14.1 1.927 3.769 1.927 3.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.672 3.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.637000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.363636, yerr=9.078173 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 1.147228E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57954. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.070 196.189 196.191 qs_mol_dyn_low 1 2.0 0.003 0.004 195.736 195.749 qs_forces 11 3.9 0.004 0.005 195.627 195.631 qs_energies 11 4.9 0.006 0.041 188.303 188.314 scf_env_do_scf 11 5.9 0.001 0.001 165.880 165.889 velocity_verlet 10 3.0 0.003 0.016 125.914 125.917 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.061 119.062 dbcsr_multiply_generic 2507 12.6 0.183 0.189 85.120 85.854 qs_scf_new_mos 117 7.6 0.001 0.001 81.026 81.369 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.025 81.368 ot_scf_mini 117 9.6 0.003 0.004 76.533 76.828 multiply_cannon 2507 13.6 0.558 0.595 60.479 63.101 multiply_cannon_loop 2507 14.6 1.177 1.209 56.732 58.625 init_scf_loop 11 6.9 0.001 0.004 46.695 46.698 ot_mini 117 10.6 0.001 0.001 42.867 43.150 prepare_preconditioner 11 7.9 0.000 0.000 42.533 42.559 make_preconditioner 11 8.9 0.000 0.001 42.532 42.559 make_full_inverse_cholesky 11 9.9 0.013 0.023 36.087 41.059 rebuild_ks_matrix 128 8.3 0.001 0.001 30.385 30.698 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 30.384 30.697 multiply_cannon_multrec 30084 15.6 14.053 19.369 25.948 30.676 mp_waitall_1 147882 16.7 17.578 27.873 17.578 27.873 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.279 27.558 qs_ot_get_derivative 117 11.6 0.001 0.002 23.121 23.421 make_m2s 5014 13.6 0.096 0.101 20.236 21.223 make_images 5014 14.6 1.957 2.295 19.942 20.933 qs_ot_get_p 128 10.4 0.001 0.002 19.963 20.259 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.154 19.734 apply_single 128 13.6 0.001 0.001 19.153 19.734 ot_diis_step 117 11.6 0.017 0.020 19.627 19.628 cp_fm_upper_to_full 105 14.8 11.409 16.783 11.409 16.783 init_scf_run 11 5.9 0.000 0.001 16.374 16.375 scf_env_initial_rho_setup 11 6.9 0.001 0.001 16.374 16.375 cp_fm_cholesky_invert 11 10.9 16.338 16.346 16.338 16.346 qs_ot_p2m_diag 83 11.4 0.343 0.391 15.833 15.887 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.713 15.886 sum_up_and_integrate 128 10.3 0.141 0.153 15.181 15.214 integrate_v_rspace 128 11.3 0.004 0.004 15.040 15.077 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.820 14.845 calculate_rho_elec 128 8.7 0.176 0.192 14.819 14.845 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.558 14.559 multiply_cannon_sync_h2d 30084 15.6 11.765 13.015 11.765 13.015 dbcsr_complete_redistribute 395 12.7 1.517 1.652 9.093 12.835 make_images_data 5014 15.6 0.065 0.070 10.870 12.709 dbcsr_mm_accdrv_process 62264 16.2 7.385 8.381 11.467 12.090 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.442 11.452 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.695 11.448 cp_fm_diag_elpa_base 83 14.4 10.458 10.778 11.436 11.447 hybrid_alltoall_any 5200 16.5 0.529 2.224 9.848 11.099 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.481 10.565 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.422 10.085 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.339 9.563 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 9.382 9.465 pw_transfer 1547 11.6 0.087 0.103 9.340 9.429 mp_alltoall_i22 716 14.1 5.569 9.270 5.569 9.270 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.113 9.209 density_rs2pw 128 9.7 0.006 0.007 7.132 8.552 calculate_first_density_matrix 1 7.0 0.000 0.003 8.305 8.309 fft_wrap_pw1pw2_140 523 13.2 1.568 1.597 8.079 8.184 grid_integrate_task_list 128 12.3 7.687 8.176 7.687 8.176 wfi_extrapolate 11 7.9 0.001 0.001 7.863 7.863 cp_fm_cholesky_decompose 22 10.9 7.630 7.745 7.630 7.745 acc_transpose_blocks 30084 15.6 0.133 0.136 7.310 7.666 multiply_cannon_metrocomm4 25070 15.6 0.079 0.092 2.816 7.657 mp_irecv_dv 76098 16.2 2.669 7.380 2.669 7.380 acc_transpose_blocks_kernels 30084 16.6 0.384 0.391 6.448 6.807 calculate_dm_sparse 128 9.5 0.001 0.001 6.592 6.696 jit_kernel_transpose 5 15.6 6.064 6.424 6.064 6.424 rs_pw_transfer 1046 11.9 0.014 0.015 4.831 6.339 fft3d_ps 1291 14.7 2.840 2.914 6.023 6.084 mp_alltoall_d11v 2415 14.1 5.050 6.066 5.050 6.066 grid_collocate_task_list 128 9.7 5.198 5.954 5.198 5.954 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.877 4.945 potential_pw2rs 128 12.3 0.023 0.023 4.709 4.735 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.468 4.469 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.182 4.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=196.191000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.909091, yerr=17.987140 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.501090E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 181.895 181.895 qs_mol_dyn_low 1 2.0 0.003 0.003 181.471 181.483 qs_forces 11 3.9 0.004 0.005 181.321 181.326 qs_energies 11 4.9 0.001 0.001 173.680 173.688 scf_env_do_scf 11 5.9 0.001 0.001 150.861 150.869 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 115.183 115.184 velocity_verlet 10 3.0 0.001 0.001 115.088 115.095 dbcsr_multiply_generic 2527 12.6 0.179 0.193 81.515 82.101 qs_scf_new_mos 118 7.6 0.001 0.001 76.066 76.164 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.065 76.164 ot_scf_mini 118 9.6 0.004 0.004 71.588 71.695 multiply_cannon 2527 13.6 0.591 0.634 61.542 65.669 multiply_cannon_loop 2527 14.6 0.446 0.459 56.918 58.072 ot_mini 118 10.6 0.001 0.001 41.270 41.367 init_scf_loop 11 6.9 0.000 0.000 35.532 35.533 mp_waitall_1 126780 16.7 26.500 34.430 26.500 34.430 prepare_preconditioner 11 7.9 0.000 0.000 31.528 31.553 make_preconditioner 11 8.9 0.000 0.000 31.528 31.553 rebuild_ks_matrix 129 8.3 0.001 0.001 30.358 30.483 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 30.357 30.483 make_full_inverse_cholesky 11 9.9 0.016 0.029 29.372 29.624 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.530 27.644 multiply_cannon_multrec 10108 15.6 10.504 15.367 18.906 22.680 qs_ot_get_derivative 118 11.6 0.002 0.002 20.911 21.017 ot_diis_step 118 11.6 0.020 0.021 20.293 20.294 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.817 20.254 apply_single 129 13.6 0.001 0.001 19.816 20.254 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 12.993 20.121 make_m2s 5054 13.6 0.069 0.072 16.323 18.643 make_images 5054 14.6 2.329 2.809 16.020 18.332 cp_fm_cholesky_invert 11 10.9 17.960 17.966 17.960 17.966 qs_ot_get_p 129 10.4 0.001 0.001 17.255 17.391 init_scf_run 11 5.9 0.000 0.001 15.994 15.994 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.994 15.994 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.528 15.560 calculate_rho_elec 129 8.7 0.261 0.272 15.527 15.559 sum_up_and_integrate 129 10.3 0.181 0.192 15.364 15.408 integrate_v_rspace 129 11.3 0.004 0.004 15.183 15.235 qs_ot_p2m_diag 83 11.4 0.496 0.502 13.456 13.473 make_images_data 5054 15.6 0.055 0.062 9.889 12.586 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.331 12.333 hybrid_alltoall_any 5240 16.5 0.822 3.839 9.706 12.270 multiply_cannon_sync_h2d 10108 15.6 11.632 12.055 11.632 12.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.139 10.164 pw_transfer 1559 11.6 0.087 0.096 9.520 9.546 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.332 9.343 cp_fm_diag_elpa_base 83 14.4 9.098 9.165 9.328 9.339 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.294 9.326 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 9.050 9.103 dbcsr_mm_accdrv_process 20926 16.1 2.943 4.375 8.026 8.704 grid_integrate_task_list 129 12.3 8.037 8.487 8.037 8.487 fft_wrap_pw1pw2_140 527 13.2 1.939 1.968 8.236 8.271 calculate_first_density_matrix 1 7.0 0.000 0.000 8.223 8.227 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.885 7.955 cp_fm_cholesky_decompose 22 10.9 7.830 7.943 7.830 7.943 density_rs2pw 129 9.7 0.006 0.006 6.897 7.838 wfi_extrapolate 11 7.9 0.001 0.001 7.551 7.551 multiply_cannon_metrocomm1 10108 15.6 0.028 0.030 4.401 7.295 mp_allgather_i34 2527 14.6 2.824 7.008 2.824 7.008 calculate_dm_sparse 129 9.5 0.001 0.001 6.612 6.699 acc_transpose_blocks 10108 15.6 0.046 0.047 6.346 6.424 grid_collocate_task_list 129 9.7 5.550 6.242 5.550 6.242 acc_transpose_blocks_kernels 10108 16.6 0.136 0.140 6.000 6.074 mp_alltoall_d11v 2423 14.1 5.184 5.995 5.184 5.995 jit_kernel_transpose 5 15.6 5.864 5.937 5.864 5.937 fft3d_ps 1301 14.7 2.781 2.852 5.727 5.760 dbcsr_complete_redistribute 395 12.7 2.131 2.211 5.224 5.607 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.297 5.298 multiply_cannon_metrocomm4 7581 15.6 0.023 0.027 1.896 5.140 rs_pw_transfer 1054 12.0 0.013 0.013 4.124 5.083 mp_irecv_dv 29086 15.9 1.862 5.060 1.862 5.060 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.684 4.743 potential_pw2rs 129 12.3 0.027 0.028 4.430 4.443 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.566 3.857 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.769 3.807 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.740 3.752 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.423 3.720 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.560 3.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=181.895000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1396.181818, yerr=52.243866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/20/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 11.606413E+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 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.112800E+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 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59070. MP_Allreduce 11095 1504. MP_Sync 86 MP_Alltoall 1700 36954382. 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 0.047 0.059 299.355 299.356 qs_mol_dyn_low 1 2.0 0.003 0.003 298.715 298.727 qs_forces 11 3.9 0.005 0.005 297.840 297.846 qs_energies 11 4.9 0.001 0.001 288.478 288.490 scf_env_do_scf 11 5.9 0.001 0.001 266.017 266.028 velocity_verlet 10 3.0 0.001 0.001 214.660 214.764 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 140.290 140.293 init_scf_loop 11 6.9 0.000 0.000 125.471 125.477 prepare_preconditioner 11 7.9 0.000 0.000 120.424 120.445 make_preconditioner 11 8.9 0.000 0.000 120.424 120.445 make_full_inverse_cholesky 11 9.9 0.036 0.039 95.759 117.588 qs_scf_new_mos 116 7.6 0.001 0.001 89.819 89.923 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.818 89.922 ot_scf_mini 116 9.6 0.003 0.004 85.031 85.064 dbcsr_multiply_generic 2485 12.5 0.207 0.218 82.775 83.369 cp_fm_upper_to_full 104 14.8 53.411 77.110 53.411 77.110 multiply_cannon 2485 13.5 0.705 0.752 58.086 58.766 multiply_cannon_loop 2485 14.5 0.465 0.475 54.497 55.862 dbcsr_complete_redistribute 393 12.7 4.018 4.073 30.336 44.115 ot_mini 116 10.6 0.001 0.001 43.954 43.994 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.960 40.731 rebuild_ks_matrix 127 8.3 0.001 0.001 38.257 38.314 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 38.256 38.313 transfer_fm_to_dbcsr 11 9.9 0.030 0.034 24.619 38.126 mp_alltoall_i22 712 14.1 22.429 36.250 22.429 36.250 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.168 35.222 cp_fm_cholesky_invert 11 10.9 33.261 33.267 33.261 33.267 mp_waitall_1 102768 16.8 28.211 31.732 28.211 31.732 qs_ot_get_p 127 10.4 0.001 0.001 25.636 25.729 qs_ot_get_derivative 116 11.6 0.002 0.002 23.910 23.943 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.598 21.625 make_m2s 4970 13.5 0.077 0.079 20.068 20.940 make_images 4970 14.5 3.742 3.937 19.598 20.473 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.361 20.380 calculate_rho_elec 127 8.7 0.480 0.481 20.360 20.379 ot_diis_step 116 11.6 0.022 0.023 20.007 20.007 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.893 19.894 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 18.612 19.837 sum_up_and_integrate 127 10.3 0.319 0.320 19.746 19.821 integrate_v_rspace 127 11.3 0.004 0.004 19.427 19.501 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.257 19.477 apply_single 127 13.6 0.001 0.001 19.257 19.477 multiply_cannon_multrec 9940 15.5 10.420 12.248 17.742 17.829 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.736 16.737 cp_fm_diag_elpa_base 82 14.4 12.348 13.975 16.731 16.731 multiply_cannon_sync_h2d 9940 15.5 15.536 15.547 15.536 15.547 hybrid_alltoall_any 5155 16.4 1.296 3.030 10.956 12.614 make_images_data 4970 15.5 0.062 0.065 10.846 12.563 init_scf_run 11 5.9 0.000 0.001 12.453 12.454 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.452 12.454 pw_transfer 1535 11.6 0.093 0.094 11.279 11.287 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 11.045 11.055 fft_wrap_pw1pw2_140 519 13.2 3.050 3.066 9.825 9.832 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.311 9.344 mp_alltoall_d11v 2401 14.1 8.298 9.235 8.298 9.235 wfi_extrapolate 11 7.9 0.001 0.001 9.117 9.117 grid_integrate_task_list 127 12.3 8.784 8.985 8.784 8.985 cp_fm_cholesky_decompose 22 10.9 8.851 8.882 8.851 8.882 dbcsr_mm_accdrv_process 20590 16.0 3.758 5.659 7.085 8.860 density_rs2pw 127 9.7 0.006 0.006 7.894 8.054 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 7.968 7.969 calculate_dm_sparse 127 9.5 0.001 0.001 6.759 6.860 grid_collocate_task_list 127 9.7 6.429 6.496 6.429 6.496 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.300 6.360 rs_scatter_matrices 138 9.7 3.677 4.634 6.011 6.220 fft3d_ps 1281 14.7 2.820 2.827 6.158 6.170 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.040 6.065 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.356000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2818.000000, yerr=157.097538 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 1.255973E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 43789858. 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 0.066 0.075 94.452 94.453 qs_energies 1 2.0 0.000 0.000 93.813 93.826 ls_scf 1 3.0 0.000 0.000 92.902 92.916 dbcsr_multiply_generic 111 6.7 0.015 0.016 81.812 81.981 multiply_cannon 111 7.7 0.017 0.020 61.722 66.179 multiply_cannon_loop 111 8.7 0.208 0.223 58.284 63.031 ls_scf_main 1 4.0 0.000 0.000 52.376 52.376 density_matrix_trs4 2 5.0 0.002 0.003 46.800 46.871 ls_scf_init_scf 1 4.0 0.000 0.000 37.483 37.484 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.351 36.403 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 34.235 34.255 mp_waitall_1 11031 10.9 23.041 29.266 23.041 29.266 multiply_cannon_multrec 2664 9.7 8.259 8.995 15.574 17.335 multiply_cannon_sync_h2d 2664 9.7 13.682 15.511 13.682 15.511 make_m2s 222 7.7 0.009 0.012 13.099 13.574 make_images 222 8.7 0.102 0.113 13.078 13.553 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.781 13.529 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.903 13.065 compute_matrix_preconditioner 1 6.0 0.000 0.001 10.126 10.130 mp_sum_l 887 5.1 6.454 8.355 6.454 8.355 make_images_data 222 9.7 0.005 0.006 7.700 8.280 hybrid_alltoall_any 227 10.6 0.219 1.856 6.550 8.001 dbcsr_mm_accdrv_process 4760 10.4 0.502 0.623 6.932 7.913 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.646 7.236 dbcsr_mm_accdrv_process_sort 4760 11.4 6.230 7.142 6.230 7.142 calculate_norms 4752 9.8 5.535 6.280 5.535 6.280 acc_transpose_blocks 2664 9.7 0.019 0.021 5.440 5.753 acc_transpose_blocks_kernels 2664 10.7 0.036 0.041 5.334 5.639 jit_kernel_transpose 1 13.0 5.298 5.602 5.298 5.602 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.169 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.076 4.009 mp_irecv_dv 6231 10.9 2.060 3.977 2.060 3.977 make_images_sizes 222 9.7 0.000 0.000 0.740 3.741 mp_alltoall_i44 222 10.7 0.740 3.741 0.740 3.741 arnoldi_extremal 4 6.8 0.000 0.000 3.237 3.262 arnoldi_normal_ev 4 7.8 0.001 0.004 3.237 3.262 build_subspace 16 8.4 0.009 0.012 3.142 3.146 ls_scf_post 1 4.0 0.000 0.000 3.044 3.057 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.905 dbcsr_special_finalize 555 9.7 0.005 0.006 2.295 2.705 dbcsr_merge_single_wm 555 10.7 0.465 0.588 2.287 2.696 make_images_pack 222 9.7 2.214 2.638 2.216 2.640 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.311 2.548 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.452 2.066 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.355 2.441 dbcsr_sort_data 658 11.4 2.080 2.420 2.080 2.420 buffer_matrices_ensure_size 222 8.7 1.780 2.101 1.780 2.101 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=94.453000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 2.098479E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.078 100.053 100.055 qs_energies 1 2.0 0.000 0.001 99.440 99.448 ls_scf 1 3.0 0.000 0.001 98.095 98.098 dbcsr_multiply_generic 111 6.7 0.016 0.016 83.840 84.225 multiply_cannon 111 7.7 0.029 0.042 60.065 64.880 multiply_cannon_loop 111 8.7 0.117 0.124 56.784 61.414 ls_scf_main 1 4.0 0.000 0.002 54.983 54.991 density_matrix_trs4 2 5.0 0.003 0.019 49.252 49.434 ls_scf_init_scf 1 4.0 0.000 0.002 39.667 39.668 ls_scf_init_matrix_S 1 5.0 0.000 0.001 38.183 38.275 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 35.796 35.809 mp_waitall_1 9105 10.9 23.123 34.344 23.123 34.344 multiply_cannon_multrec 1332 9.7 13.190 16.315 22.446 26.696 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 13.438 25.080 make_m2s 222 7.7 0.007 0.008 15.365 16.097 make_images 222 8.7 1.585 1.968 15.334 16.068 mp_sum_l 887 5.1 7.093 11.243 7.093 11.243 dbcsr_mm_accdrv_process 4041 10.4 0.272 0.452 8.855 10.439 dbcsr_mm_accdrv_process_sort 4041 11.4 8.448 10.002 8.448 10.002 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.954 9.957 make_images_data 222 9.7 0.004 0.005 8.814 9.769 compute_matrix_preconditioner 1 6.0 0.000 0.001 9.406 9.411 hybrid_alltoall_any 227 10.6 0.525 2.469 8.247 9.235 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.235 7.491 mp_irecv_dv 3311 11.0 3.215 7.428 3.215 7.428 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.510 7.415 calculate_norms 2376 9.8 6.017 6.741 6.017 6.741 multiply_cannon_sync_h2d 1332 9.7 4.749 5.734 4.749 5.734 acc_transpose_blocks 1332 9.7 0.007 0.008 5.015 5.235 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 4.966 5.187 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.969 5.168 jit_kernel_transpose 1 13.0 4.948 5.168 4.948 5.168 arnoldi_extremal 4 6.8 0.000 0.000 4.618 4.636 arnoldi_normal_ev 4 7.8 0.001 0.004 4.618 4.636 build_subspace 16 8.4 0.014 0.021 4.365 4.367 ls_scf_post 1 4.0 0.000 0.001 3.445 3.450 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.126 3.363 ls_scf_store_result 1 5.0 0.000 0.000 3.163 3.266 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.227 2.741 3.229 ls_scf_dm_to_ks 2 5.0 0.001 0.012 2.635 2.709 make_images_pack 222 9.7 2.034 2.456 2.036 2.458 mp_allgather_i34 111 8.7 0.974 2.413 0.974 2.413 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.101 2.102 rebuild_ks_matrix 3 7.3 0.000 0.000 2.087 2.089 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.019 2.087 2.088 dbcsr_sort_data 436 11.2 1.830 2.031 1.830 2.031 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=100.055000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1704.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 2.754847E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.085 94.871 94.872 qs_energies 1 2.0 0.000 0.000 94.187 94.193 ls_scf 1 3.0 0.000 0.000 92.791 92.798 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.384 77.673 ls_scf_main 1 4.0 0.000 0.000 58.207 58.212 multiply_cannon 111 7.7 0.045 0.107 53.233 57.734 multiply_cannon_loop 111 8.7 0.100 0.106 49.671 53.823 density_matrix_trs4 2 5.0 0.002 0.003 52.183 52.356 mp_waitall_1 7281 11.0 24.154 34.103 24.154 34.103 ls_scf_init_scf 1 4.0 0.000 0.000 31.011 31.013 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.774 29.881 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.293 27.304 multiply_cannon_multrec 888 9.7 12.742 15.147 21.310 24.292 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.027 23.206 make_m2s 222 7.7 0.007 0.008 17.106 18.382 make_images 222 8.7 1.978 2.278 17.068 18.342 hybrid_alltoall_any 227 10.6 0.624 2.886 9.543 10.833 make_images_data 222 9.7 0.004 0.005 9.819 10.823 dbcsr_mm_accdrv_process 3754 10.4 0.256 0.467 8.094 9.312 mp_sum_l 887 5.1 5.480 9.287 5.480 9.287 dbcsr_mm_accdrv_process_sort 3754 11.4 7.712 8.898 7.712 8.898 multiply_cannon_sync_h2d 888 9.7 6.044 7.397 6.044 7.397 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.183 7.323 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.938 6.748 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.468 6.622 mp_irecv_dv 2335 11.1 2.454 6.586 2.454 6.586 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.961 5.149 arnoldi_extremal 4 6.8 0.000 0.000 5.071 5.088 arnoldi_normal_ev 4 7.8 0.001 0.005 5.071 5.088 build_subspace 16 8.4 0.014 0.020 4.764 4.770 calculate_norms 1584 9.8 4.250 4.581 4.250 4.581 mp_allgather_i34 111 8.7 1.397 3.923 1.397 3.923 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.423 3.766 dbcsr_matrix_vector_mult_local 304 10.0 3.014 3.600 3.016 3.602 ls_scf_post 1 4.0 0.000 0.000 3.573 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.313 3.401 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.889 2.982 dbcsr_sort_data 325 11.1 1.892 2.174 1.892 2.174 dbcsr_data_release 9322 10.9 1.336 2.136 1.336 2.136 make_images_pack 222 9.7 1.823 2.107 1.826 2.110 make_images_sizes 222 9.7 0.000 0.000 0.922 2.079 mp_alltoall_i44 222 10.7 0.921 2.079 0.921 2.079 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.926 1.928 rebuild_ks_matrix 3 7.3 0.000 0.000 1.908 1.909 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.908 1.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.872000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2188.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 3.386913E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.073 97.712 97.715 qs_energies 1 2.0 0.000 0.001 96.853 96.858 ls_scf 1 3.0 0.006 0.017 95.038 95.044 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.526 78.778 ls_scf_main 1 4.0 0.004 0.020 58.912 58.920 multiply_cannon 111 7.7 0.062 0.204 51.841 55.817 density_matrix_trs4 2 5.0 0.006 0.012 52.730 52.819 multiply_cannon_loop 111 8.7 0.115 0.129 46.801 49.697 ls_scf_init_scf 1 4.0 0.001 0.003 32.843 32.845 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.485 31.563 mp_waitall_1 6369 11.0 22.872 30.450 22.872 30.450 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.970 28.982 multiply_cannon_multrec 1332 9.7 14.233 17.681 22.157 25.519 make_m2s 222 7.7 0.008 0.010 21.193 22.599 make_images 222 8.7 3.148 3.617 21.142 22.550 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.339 17.045 make_images_data 222 9.7 0.004 0.005 11.815 13.411 hybrid_alltoall_any 227 10.6 0.802 3.776 11.242 12.943 dbcsr_mm_accdrv_process 3641 10.4 0.208 0.405 7.560 9.086 dbcsr_mm_accdrv_process_sort 3641 11.4 7.181 8.664 7.181 8.664 mp_sum_l 887 5.1 3.920 6.698 3.920 6.698 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.084 6.120 mp_irecv_dv 3229 10.9 2.062 6.046 2.062 6.046 multiply_cannon_sync_h2d 1332 9.7 5.497 5.973 5.497 5.973 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.560 5.396 arnoldi_extremal 4 6.8 0.000 0.000 5.168 5.183 arnoldi_normal_ev 4 7.8 0.001 0.005 5.168 5.183 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.973 5.130 build_subspace 16 8.4 0.014 0.021 4.836 4.842 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.527 4.681 calculate_norms 2376 9.8 4.210 4.562 4.210 4.562 mp_allgather_i34 111 8.7 2.149 4.429 2.149 4.429 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.569 3.864 dbcsr_matrix_vector_mult_local 304 10.0 3.180 3.677 3.182 3.678 dbcsr_sort_data 658 11.4 3.071 3.417 3.071 3.417 ls_scf_post 1 4.0 0.001 0.002 3.276 3.281 dbcsr_special_finalize 555 9.7 0.006 0.007 2.823 3.214 dbcsr_merge_single_wm 555 10.7 0.538 0.661 2.814 3.206 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.078 3.135 ls_scf_store_result 1 5.0 0.000 0.000 3.003 3.079 dbcsr_data_release 10477 10.7 1.596 2.450 1.596 2.450 dbcsr_finalize 304 7.8 0.050 0.062 1.807 1.995 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.984 1.987 rebuild_ks_matrix 3 7.3 0.000 0.000 1.960 1.962 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.013 1.960 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.715000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 4.662055E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.088 99.416 99.417 qs_energies 1 2.0 0.000 0.000 98.540 98.545 ls_scf 1 3.0 0.000 0.000 96.588 96.592 dbcsr_multiply_generic 111 6.7 0.018 0.019 77.871 78.079 ls_scf_main 1 4.0 0.000 0.000 62.170 62.193 multiply_cannon 111 7.7 0.089 0.200 55.286 60.178 density_matrix_trs4 2 5.0 0.002 0.003 54.833 54.925 multiply_cannon_loop 111 8.7 0.069 0.077 50.685 52.366 mp_waitall_1 5436 11.0 26.373 32.287 26.373 32.287 ls_scf_init_scf 1 4.0 0.000 0.000 30.783 30.856 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.507 29.540 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.351 27.362 multiply_cannon_multrec 444 9.7 13.963 16.449 21.034 23.209 make_m2s 222 7.7 0.006 0.007 17.742 20.301 make_images 222 8.7 3.732 4.438 17.679 20.240 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.046 16.106 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.209 15.393 make_images_data 222 9.7 0.003 0.004 9.924 12.425 hybrid_alltoall_any 227 10.6 0.793 3.794 9.715 12.208 dbcsr_mm_accdrv_process 3003 10.4 0.147 0.373 6.769 7.890 multiply_cannon_sync_h2d 444 9.7 6.547 7.800 6.547 7.800 dbcsr_mm_accdrv_process_sort 3003 11.4 6.442 7.543 6.442 7.543 mp_allgather_i34 111 8.7 2.815 7.002 2.815 7.002 arnoldi_extremal 4 6.8 0.000 0.000 5.840 5.854 arnoldi_normal_ev 4 7.8 0.002 0.005 5.839 5.853 build_subspace 16 8.4 0.015 0.020 5.452 5.461 mp_sum_l 887 5.1 2.887 5.127 2.887 5.127 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.622 4.748 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.168 4.357 dbcsr_matrix_vector_mult_local 304 10.0 3.679 4.155 3.681 4.157 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.653 4.072 mp_irecv_dv 1241 11.2 1.633 4.047 1.633 4.047 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.957 3.831 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.647 3.749 calculate_norms 792 9.8 3.555 3.736 3.555 3.736 ls_scf_post 1 4.0 0.000 0.000 3.635 3.638 make_images_sizes 222 9.7 0.000 0.000 1.142 3.551 mp_alltoall_i44 222 10.7 1.141 3.551 1.141 3.551 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.461 dbcsr_finalize 304 7.8 0.062 0.078 2.210 2.288 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.123 2.195 rebuild_ks_matrix 3 7.3 0.000 0.000 2.091 2.163 qs_ks_build_kohn_sham_matrix 3 8.3 0.022 0.094 2.091 2.163 dbcsr_merge_all 275 8.9 0.474 0.521 2.058 2.125 dbcsr_data_release 10123 10.8 1.339 1.992 1.339 1.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.417000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3665.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_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 8.813433E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.124 0.144 108.522 108.523 qs_energies 1 2.0 0.000 0.000 106.988 107.002 ls_scf 1 3.0 0.000 0.000 104.070 104.083 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.571 77.686 ls_scf_main 1 4.0 0.000 0.000 66.155 66.157 density_matrix_trs4 2 5.0 0.002 0.003 56.853 56.909 multiply_cannon 111 7.7 0.146 0.204 49.827 51.381 multiply_cannon_loop 111 8.7 0.068 0.071 46.492 47.428 ls_scf_init_scf 1 4.0 0.000 0.000 34.232 34.233 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.645 32.660 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.796 29.802 mp_waitall_1 4527 11.1 22.315 25.732 22.315 25.732 make_m2s 222 7.7 0.007 0.007 24.103 24.943 make_images 222 8.7 4.599 4.961 23.994 24.833 multiply_cannon_multrec 444 9.7 18.005 18.814 22.632 23.263 hybrid_alltoall_any 227 10.6 1.665 3.631 13.126 15.928 make_images_data 222 9.7 0.004 0.004 13.324 15.582 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.415 10.974 multiply_cannon_sync_h2d 444 9.7 8.844 8.894 8.844 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.361 7.372 arnoldi_normal_ev 4 7.8 0.003 0.009 7.361 7.372 build_subspace 16 8.4 0.026 0.036 6.810 6.818 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.469 5.630 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.249 5.341 dbcsr_matrix_vector_mult_local 304 10.0 5.003 5.319 5.005 5.321 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.038 5.282 dbcsr_mm_accdrv_process 1814 10.4 0.233 0.324 4.453 4.587 dbcsr_mm_accdrv_process_sort 1814 11.4 4.152 4.282 4.152 4.282 ls_scf_post 1 4.0 0.000 0.000 3.683 3.696 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.445 make_images_sizes 222 9.7 0.000 0.000 1.420 3.428 mp_alltoall_i44 222 10.7 1.419 3.428 1.419 3.428 calculate_norms 792 9.8 3.243 3.278 3.243 3.278 dbcsr_finalize 304 7.8 0.082 0.090 3.092 3.141 dbcsr_merge_all 275 8.9 0.894 0.917 2.877 2.912 mp_allgather_i34 111 8.7 0.949 2.895 0.949 2.895 dbcsr_complete_redistribute 5 7.6 1.439 1.480 2.770 2.895 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 2.888 2.888 dbcsr_data_release 12724 10.6 2.329 2.851 2.329 2.851 matrix_ls_to_qs 2 6.0 0.000 0.000 2.422 2.568 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.510 2.511 dbcsr_sort_data 325 11.1 2.443 2.495 2.443 2.495 rebuild_ks_matrix 3 7.3 0.000 0.000 2.444 2.445 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.444 2.445 dbcsr_new_transposed 4 7.5 0.243 0.252 2.280 2.294 dbcsr_frobenius_norm 74 6.6 2.056 2.139 2.200 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.138 2.208 dbcsr_add_anytype 103 7.2 0.860 0.892 2.137 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.523000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6825.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a11bcb3b8565e52344423df66485879857c7b754_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 586.690560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 655497805. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.123 229.252 229.253 qs_mol_dyn_low 1 2.0 0.006 0.056 227.790 227.812 qs_forces 5 3.8 0.006 0.031 227.606 227.621 qs_energies 5 4.8 0.003 0.050 224.537 224.554 scf_env_do_scf 5 5.8 0.001 0.019 207.247 207.250 scf_env_do_scf_inner_loop 105 6.6 0.006 0.048 180.924 180.942 qs_scf_new_mos 105 7.6 0.000 0.001 139.941 140.193 qs_scf_loop_do_ot 105 8.6 0.001 0.018 139.941 140.192 dbcsr_multiply_generic 1445 12.2 0.121 0.129 133.571 133.949 ot_scf_mini 105 9.6 0.004 0.030 130.393 130.577 multiply_cannon 1445 13.2 0.276 0.293 115.143 116.949 multiply_cannon_loop 1445 14.2 2.384 2.507 113.387 114.616 velocity_verlet 4 3.0 0.001 0.015 106.180 106.181 ot_mini 105 10.6 0.001 0.021 59.029 59.159 multiply_cannon_multrec 69360 15.2 31.855 37.588 39.434 44.913 qs_ot_get_p 112 10.4 0.001 0.001 41.720 42.063 mp_waitall_1 488190 16.1 33.088 41.160 33.088 41.160 qs_ot_get_derivative 55 11.6 0.001 0.012 37.262 37.441 multiply_cannon_sync_h2d 69360 15.2 30.988 35.214 30.988 35.214 multiply_cannon_metrocomm3 69360 15.2 0.199 0.209 24.457 33.064 qs_ot_p2m_diag 40 11.0 0.020 0.031 30.846 30.941 rebuild_ks_matrix 110 8.4 0.000 0.000 29.354 29.538 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.028 29.354 29.538 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.946 27.947 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.903 27.068 init_scf_loop 7 6.6 0.002 0.042 26.287 26.291 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.133 23.353 apply_single 62 13.6 0.000 0.000 23.133 23.353 cp_fm_syevd 40 13.0 0.000 0.000 22.948 23.089 prepare_preconditioner 7 7.6 0.000 0.001 21.360 21.391 make_preconditioner 7 8.6 0.001 0.010 21.360 21.391 ot_new_cg_direction 55 11.6 0.001 0.037 21.096 21.096 cp_fm_redistribute_end 40 14.0 9.151 18.261 9.156 18.262 cp_fm_syevd_base 40 14.0 9.099 18.211 9.099 18.211 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.634 18.005 calculate_rho_elec 110 8.6 0.030 0.033 17.633 18.004 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.439 14.630 init_scf_run 5 5.8 0.000 0.003 14.465 14.466 scf_env_initial_rho_setup 5 6.8 0.002 0.003 14.465 14.466 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.314 14.378 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.593 13.711 rs_pw_transfer 690 11.5 0.011 0.013 11.666 12.983 density_rs2pw 110 9.6 0.006 0.008 11.613 12.971 pw_transfer 1645 12.4 0.084 0.110 12.366 12.606 mp_sum_l 4764 12.2 11.653 12.605 11.653 12.605 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 12.223 12.467 calculate_dm_sparse 110 9.5 0.000 0.001 11.167 11.338 qs_vxc_create 110 10.4 0.002 0.006 10.652 10.705 fft_wrap_pw1pw2_240 915 15.0 1.184 1.296 10.410 10.651 cp_fm_cholesky_invert 7 10.6 10.020 10.028 10.020 10.028 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.861 9.956 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 9.908 9.924 acc_transpose_blocks 69360 15.2 0.354 0.373 9.123 9.676 calculate_first_density_matrix 1 7.0 0.000 0.002 9.559 9.583 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.485 9.526 check_diag 80 13.5 8.588 8.858 9.345 9.487 fft3d_pb 915 16.0 2.369 2.654 8.493 8.766 sum_up_and_integrate 60 10.3 0.028 0.031 8.681 8.702 integrate_v_rspace 60 11.3 0.002 0.002 8.653 8.675 dbcsr_mm_accdrv_process 154766 15.8 3.968 4.156 7.447 8.289 multiply_cannon_metrocomm1 69360 15.2 0.096 0.101 4.236 7.990 acc_transpose_blocks_kernels 69360 16.2 0.850 0.904 7.264 7.720 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.393 7.649 xc_vxc_pw_create 60 11.3 0.039 0.051 7.160 7.213 make_m2s 2890 13.2 0.075 0.085 6.313 6.887 jit_kernel_transpose 5 15.0 6.414 6.847 6.414 6.847 make_full_single_inverse 7 9.6 0.001 0.002 6.783 6.814 make_images 2890 14.2 0.247 0.263 6.215 6.790 xc_pw_derive 510 13.4 0.005 0.007 6.352 6.453 mp_waitany 7680 13.5 4.539 5.864 4.539 5.864 mp_alltoall_z22v 2340 17.7 5.586 5.862 5.586 5.862 multiply_cannon_metrocomm4 67915 15.2 0.182 0.195 2.029 5.008 potential_pw2rs 60 12.3 0.003 0.003 4.866 4.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.253000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a11bcb3b8565e52344423df66485879857c7b754 Summary: empty Status: OK