=== 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: 9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/01 job id: 50887603 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/02 job id: 50887604 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/03 job id: 50887605 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/04 job id: 50887606 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/05 job id: 50887607 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/06 job id: 50887608 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/07 job id: 50887609 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/08 job id: 50887610 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/09 job id: 50887611 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/10 job id: 50887612 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/11 job id: 50887613 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/12 job id: 50887614 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/13 job id: 50887615 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/14 job id: 50887616 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/15 job id: 50887617 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/16 job id: 50887618 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/17 job id: 50887619 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/18 job id: 50887620 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/19 job id: 50887621 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/20 job id: 50887622 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/21 job id: 50887623 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/22 job id: 50887624 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/23 job id: 50887625 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/24 job id: 50887626 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/25 job id: 50887627 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/26 job id: 50887628 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/27 job id: 50887629 --- 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/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 172669. 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 0.019 0.034 134.148 134.149 farming_run 1 2.0 133.632 133.635 134.117 134.121 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.486774E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.025 114.555 114.556 qs_energies 1 2.0 0.000 0.000 114.371 114.372 mp2_main 1 3.0 0.000 0.000 112.270 112.271 mp2_gpw_main 1 4.0 0.027 0.033 111.255 111.256 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.399 92.727 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.739 55.068 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.167 41.145 46.199 get_2c_integrals 1 6.0 0.023 0.028 37.057 37.485 integrate_v_rspace 273 8.0 0.440 0.455 24.711 29.535 pw_transfer 6555 10.6 0.379 0.390 26.896 27.380 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.546 26.038 grid_integrate_task_list 273 9.0 20.600 25.908 20.600 25.908 fft_wrap_pw1pw2_100 2178 12.4 0.030 0.031 23.092 23.571 compute_2c_integrals 1 7.0 0.002 0.002 18.961 18.977 rpa_ri_compute_en 1 5.0 0.019 0.019 18.750 18.964 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.598 18.705 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.434 18.597 18.704 cp_fm_cholesky_decompose 12 8.2 18.010 18.426 18.010 18.426 cholesky_decomp 1 7.0 0.000 0.000 16.914 17.335 fft3d_s 5443 13.4 16.129 16.474 16.150 16.496 ao_to_mo_and_store_B_mult_1 272 7.0 10.769 15.315 10.769 15.315 calculate_wavefunction 272 8.0 5.395 5.543 12.340 12.947 rpa_num_int 1 6.0 0.000 0.006 10.650 10.659 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.627 10.630 calc_mat_Q 8 8.0 0.000 0.000 9.487 9.574 contract_S_to_Q 8 9.0 0.000 0.000 8.910 9.000 calc_potential_gpw 544 9.5 0.004 0.005 8.212 8.698 parallel_gemm_fm 14 9.1 0.000 0.000 8.492 8.600 parallel_gemm_fm_cosma 14 10.1 8.492 8.600 8.492 8.600 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.073 8.316 potential_pw2rs 545 10.0 0.106 0.108 7.523 8.226 create_integ_mat 1 6.0 0.015 0.028 7.679 7.688 collocate_single_gaussian 272 10.0 0.038 0.040 7.326 7.586 array2fm 1 7.0 0.000 0.000 6.672 7.122 pw_scatter_s 2720 13.7 4.353 4.483 4.353 4.483 pw_gather_s 2722 13.2 3.501 3.724 3.501 3.724 array2fm_buffer_send 1 8.0 2.914 3.098 2.914 3.098 pw_poisson_solve 545 10.5 1.123 1.169 2.178 2.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.254457, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.037 421.391 421.394 farming_run 1 2.0 420.429 420.434 421.348 421.353 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.243050E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 216.731 216.732 qs_energies 1 2.0 0.007 0.028 216.498 216.511 scf_env_do_scf 1 3.0 0.000 0.000 114.575 114.575 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.659 113.665 rebuild_ks_matrix 4 6.0 0.000 0.000 113.658 113.664 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.658 113.664 hfx_ks_matrix 4 8.0 0.001 0.001 113.316 113.320 integrate_four_center 4 9.0 0.152 0.456 113.315 113.319 integrate_four_center_main 4 10.0 0.106 0.696 102.033 104.671 integrate_four_center_bin 268 11.0 101.927 104.637 101.927 104.637 mp2_main 1 3.0 0.006 0.038 101.628 101.646 mp2_gpw_main 1 4.0 0.062 0.131 100.715 100.731 init_scf_loop 1 4.0 0.000 0.000 96.591 96.591 mp2_ri_gpw_compute_in 1 5.0 0.066 0.095 73.909 74.929 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.692 54.713 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.162 41.311 46.473 integrate_v_rspace 95 8.0 0.395 0.559 27.792 32.837 pw_transfer 2240 10.6 0.143 0.170 29.512 29.910 ao_to_mo_and_store_B_mult_1 91 7.0 10.715 29.152 10.715 29.152 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.508 28.905 mp2_ri_gpw_compute_en 1 5.0 0.057 0.072 26.577 28.545 grid_integrate_task_list 95 9.0 23.220 28.452 23.220 28.452 fft_wrap_pw1pw2_100 730 12.4 0.013 0.013 26.261 26.626 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.896 24.896 24.906 get_2c_integrals 1 6.0 0.004 0.020 20.115 20.150 compute_2c_integrals 1 7.0 0.003 0.012 19.093 19.101 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.671 18.958 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.844 18.670 18.956 fft3d_s 1823 13.4 18.480 18.823 18.494 18.836 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.982 17.982 calculate_wavefunction 91 8.0 2.012 2.042 9.612 9.818 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.581 8.746 9.258 potential_pw2rs 186 10.0 0.033 0.034 8.412 9.031 local_gemm 172 8.0 8.188 8.680 8.188 8.680 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.183 8.600 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.530 7.928 8.197 calc_potential_gpw 182 9.5 0.002 0.002 7.932 8.138 collocate_single_gaussian 91 10.0 0.016 0.018 7.777 8.023 integrate_four_center_load 4 10.0 0.000 0.000 6.792 6.796 hfx_load_balance 1 11.0 0.000 0.000 6.792 6.796 mp2_ri_gpw_compute_en_ener 172 7.0 6.332 6.432 6.332 6.432 mp_sendrecv_dm3 2068 8.0 5.950 6.248 5.950 6.248 mp_sync 37 10.5 3.240 6.154 3.240 6.154 pw_gather_s 912 13.2 4.461 4.983 4.461 4.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.723188, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1502.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.419008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 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 0.016 0.027 55.256 55.258 qs_mol_dyn_low 1 2.0 0.004 0.015 54.885 54.905 qs_forces 11 3.9 0.006 0.021 54.801 54.811 qs_energies 11 4.9 0.001 0.002 53.324 53.349 scf_env_do_scf 11 5.9 0.000 0.001 47.172 47.172 scf_env_do_scf_inner_loop 108 6.5 0.004 0.009 45.101 45.102 dbcsr_multiply_generic 2286 12.5 0.094 0.098 35.006 35.533 qs_scf_new_mos 108 7.5 0.000 0.001 35.209 35.494 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.209 35.494 ot_scf_mini 108 9.5 0.002 0.002 33.512 33.708 multiply_cannon 2286 13.5 0.179 0.188 27.180 28.804 multiply_cannon_loop 2286 14.5 1.851 1.954 26.545 28.192 velocity_verlet 10 3.0 0.001 0.001 26.780 26.781 ot_mini 108 10.5 0.001 0.001 20.535 20.796 qs_ot_get_derivative 108 11.5 0.005 0.017 17.499 17.671 mp_waitall_1 245248 16.5 9.010 15.505 9.010 15.505 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.519 13.806 multiply_cannon_multrec 54864 15.5 3.595 5.656 7.802 11.236 qs_ot_get_p 119 10.4 0.001 0.001 8.221 8.522 rebuild_ks_matrix 119 8.3 0.000 0.000 7.779 7.913 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.778 7.912 mp_sum_l 7287 12.8 5.920 7.741 5.920 7.741 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.846 6.966 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.373 6.823 multiply_cannon_sync_h2d 54864 15.5 5.102 6.700 5.102 6.700 dbcsr_mm_accdrv_process 76910 16.1 1.881 2.929 4.122 5.854 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.687 5.781 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.249 5.299 init_scf_run 11 5.9 0.000 0.001 4.815 4.815 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.814 4.815 sum_up_and_integrate 119 10.3 0.002 0.009 4.391 4.399 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.396 4.396 integrate_v_rspace 119 11.3 0.002 0.002 4.379 4.386 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.207 4.207 cp_fm_redistribute_end 50 14.0 2.147 4.184 2.152 4.186 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.047 4.154 calculate_rho_elec 119 8.7 0.011 0.016 4.046 4.153 cp_fm_diag_elpa_base 50 14.0 2.029 4.081 2.032 4.088 multiply_cannon_metrocomm1 54864 15.5 0.057 0.064 1.884 3.389 acc_transpose_blocks 54864 15.5 0.225 0.249 2.482 3.211 calculate_dm_sparse 119 9.5 0.000 0.000 3.014 3.144 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.943 3.137 apply_single 119 13.6 0.000 0.000 2.943 3.137 jit_kernel_multiply 13 15.8 2.172 2.830 2.172 2.830 ot_diis_step 108 11.5 0.006 0.006 2.735 2.735 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.560 2.636 calculate_first_density_matrix 1 7.0 0.000 0.001 2.604 2.605 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.355 2.357 density_rs2pw 119 9.7 0.004 0.004 2.087 2.192 wfi_extrapolate 11 7.9 0.001 0.001 2.143 2.143 grid_integrate_task_list 119 12.3 2.020 2.116 2.020 2.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.995 2.074 init_scf_loop 11 6.9 0.000 0.000 2.054 2.054 mp_sum_d 4137 12.0 1.387 1.971 1.387 1.971 potential_pw2rs 119 12.3 0.004 0.004 1.768 1.781 pw_transfer 1439 11.6 0.051 0.056 1.619 1.689 make_m2s 4572 13.5 0.053 0.056 1.594 1.641 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.544 1.616 make_images 4572 14.5 0.132 0.138 1.512 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.482 1.497 acc_transpose_blocks_kernels 54864 16.5 0.249 0.385 0.988 1.467 acc_transpose_blocks_sync 164592 16.5 1.226 1.458 1.226 1.458 transfer_rs2pw 487 10.6 0.005 0.006 1.356 1.455 mp_alltoall_d11v 2130 13.8 1.260 1.448 1.260 1.448 grid_collocate_task_list 119 9.7 1.357 1.431 1.357 1.431 mp_waitany 12084 13.8 1.282 1.416 1.282 1.416 fft3d_ps 1201 14.6 0.369 0.473 1.257 1.321 transfer_pw2rs 487 13.2 0.006 0.006 1.292 1.303 fft_wrap_pw1pw2_140 487 13.2 0.108 0.116 1.201 1.272 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.758 1.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.258000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.706048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 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 0.008 0.028 39.013 39.014 qs_mol_dyn_low 1 2.0 0.003 0.003 38.799 38.806 qs_forces 11 3.9 0.002 0.002 38.489 38.491 qs_energies 11 4.9 0.001 0.001 36.834 36.840 scf_env_do_scf 11 5.9 0.000 0.001 31.310 31.310 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.933 28.934 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.340 22.701 qs_scf_new_mos 108 7.5 0.001 0.001 20.422 20.651 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.422 20.651 ot_scf_mini 108 9.5 0.002 0.003 19.510 19.673 multiply_cannon 2286 13.5 0.205 0.213 17.363 18.820 velocity_verlet 10 3.0 0.014 0.039 18.316 18.322 multiply_cannon_loop 2286 14.5 1.196 1.253 16.224 17.632 ot_mini 108 10.5 0.001 0.001 12.031 12.257 mp_waitall_1 200699 16.5 5.643 10.631 5.643 10.631 qs_ot_get_derivative 108 11.5 0.001 0.001 9.677 9.842 multiply_cannon_multrec 27432 15.5 1.799 4.005 6.509 9.464 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.192 9.433 rebuild_ks_matrix 119 8.3 0.000 0.000 6.695 6.833 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.695 6.833 dbcsr_mm_accdrv_process 47894 16.0 3.701 5.852 4.632 6.800 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.892 6.017 qs_ot_get_p 119 10.4 0.001 0.001 4.605 4.826 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.719 4.521 init_scf_run 11 5.9 0.000 0.001 4.329 4.329 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.329 4.329 mp_sum_l 7287 12.8 1.972 3.845 1.972 3.845 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.945 3.666 apply_single 119 13.6 0.000 0.000 2.945 3.666 sum_up_and_integrate 119 10.3 0.001 0.001 3.620 3.641 integrate_v_rspace 119 11.3 0.002 0.002 3.606 3.626 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.550 3.592 calculate_rho_elec 119 8.7 0.021 0.024 3.550 3.592 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.955 2.972 make_m2s 4572 13.5 0.051 0.053 2.546 2.876 calculate_first_density_matrix 1 7.0 0.000 0.000 2.847 2.853 make_images 4572 14.5 0.205 0.246 2.456 2.790 calculate_dm_sparse 119 9.5 0.000 0.001 2.520 2.602 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.562 2.563 jit_kernel_multiply 11 16.2 0.870 2.558 0.870 2.558 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.286 2.371 init_scf_loop 11 6.9 0.000 0.000 2.357 2.358 ot_diis_step 108 11.5 0.011 0.011 2.306 2.307 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.255 2.255 cp_fm_redistribute_end 50 14.0 1.143 2.231 1.146 2.234 cp_fm_diag_elpa_base 50 14.0 1.056 2.147 1.084 2.187 multiply_cannon_sync_h2d 27432 15.5 1.663 2.128 1.663 2.128 acc_transpose_blocks 27432 15.5 0.115 0.120 1.629 1.985 density_rs2pw 119 9.7 0.004 0.004 1.848 1.965 grid_integrate_task_list 119 12.3 1.836 1.925 1.836 1.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.898 1.900 pw_transfer 1439 11.6 0.064 0.069 1.808 1.838 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.717 1.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.694 1.734 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.714 1.725 make_images_data 4572 15.5 0.047 0.054 1.213 1.545 prepare_preconditioner 11 7.9 0.000 0.000 1.462 1.488 make_preconditioner 11 8.9 0.000 0.000 1.462 1.488 wfi_extrapolate 11 7.9 0.001 0.001 1.421 1.421 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.366 1.420 fft_wrap_pw1pw2_140 487 13.2 0.128 0.133 1.379 1.412 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.076 1.404 potential_pw2rs 119 12.3 0.006 0.006 1.377 1.382 fft3d_ps 1201 14.6 0.522 0.577 1.348 1.373 grid_collocate_task_list 119 9.7 1.287 1.339 1.287 1.339 mp_alltoall_d11v 2130 13.8 1.175 1.311 1.175 1.311 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.191 1.241 transfer_rs2pw 487 10.6 0.005 0.005 1.040 1.153 mp_sum_d 4137 12.0 0.551 1.009 0.551 1.009 acc_transpose_blocks_sync 82296 16.5 0.825 0.952 0.825 0.952 mp_allgather_i34 2286 14.5 0.562 0.943 0.562 0.943 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.940 0.942 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.664 0.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.860 0.873 transfer_pw2rs 487 13.2 0.004 0.005 0.855 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.014000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.636364, yerr=1.553455 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.457664E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.024 32.184 32.195 qs_mol_dyn_low 1 2.0 0.003 0.003 31.875 31.883 qs_forces 11 3.9 0.002 0.003 31.759 31.760 qs_energies 11 4.9 0.002 0.002 30.213 30.214 scf_env_do_scf 11 5.9 0.000 0.001 25.495 25.495 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.100 23.100 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.923 17.000 qs_scf_new_mos 108 7.5 0.001 0.001 15.446 15.462 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.445 15.461 velocity_verlet 10 3.0 0.001 0.001 15.126 15.129 ot_scf_mini 108 9.5 0.002 0.002 14.686 14.702 multiply_cannon 2286 13.5 0.191 0.195 13.681 14.487 multiply_cannon_loop 2286 14.5 0.864 0.891 12.891 13.736 ot_mini 108 10.5 0.001 0.001 9.044 9.061 qs_ot_get_derivative 108 11.5 0.001 0.001 7.558 7.573 multiply_cannon_multrec 18288 15.5 1.872 2.914 6.991 7.365 rebuild_ks_matrix 119 8.3 0.000 0.000 5.945 5.961 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.944 5.960 dbcsr_mm_accdrv_process 38222 16.0 4.947 5.793 5.026 5.861 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.235 5.250 init_scf_run 11 5.9 0.000 0.001 3.554 3.554 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.554 3.554 sum_up_and_integrate 119 10.3 0.001 0.001 3.531 3.537 integrate_v_rspace 119 11.3 0.002 0.003 3.519 3.526 mp_waitall_1 158411 16.6 2.591 3.498 2.591 3.498 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.812 3.444 qs_ot_get_p 119 10.4 0.001 0.001 3.404 3.424 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.252 3.259 calculate_rho_elec 119 8.7 0.031 0.032 3.252 3.258 init_scf_loop 11 6.9 0.000 0.000 2.375 2.376 calculate_first_density_matrix 1 7.0 0.000 0.000 2.344 2.345 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.042 2.335 apply_single 119 13.6 0.000 0.000 2.042 2.335 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.285 2.292 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.432 2.167 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.994 1.995 make_m2s 4572 13.5 0.044 0.045 1.794 1.949 grid_integrate_task_list 119 12.3 1.807 1.889 1.807 1.889 density_rs2pw 119 9.7 0.004 0.004 1.820 1.885 calculate_dm_sparse 119 9.5 0.000 0.001 1.857 1.864 make_images 4572 14.5 0.191 0.204 1.708 1.862 pw_transfer 1439 11.6 0.065 0.068 1.841 1.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.762 1.763 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.748 1.760 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.740 1.752 cp_fm_diag_elpa_base 50 14.0 1.716 1.731 1.738 1.750 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.697 1.704 acc_transpose_blocks 18288 15.5 0.080 0.082 1.631 1.664 prepare_preconditioner 11 7.9 0.000 0.000 1.629 1.632 make_preconditioner 11 8.9 0.000 0.000 1.629 1.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.579 1.586 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.490 1.573 ot_diis_step 108 11.5 0.011 0.011 1.473 1.473 fft_wrap_pw1pw2_140 487 13.2 0.178 0.182 1.414 1.425 mp_sum_l 7287 12.8 1.075 1.385 1.075 1.385 potential_pw2rs 119 12.3 0.007 0.008 1.330 1.335 fft3d_ps 1201 14.6 0.547 0.566 1.303 1.314 grid_collocate_task_list 119 9.7 1.233 1.297 1.233 1.297 multiply_cannon_sync_h2d 18288 15.5 1.037 1.209 1.037 1.209 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.194 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.159 1.159 transfer_rs2pw 487 10.6 0.005 0.005 0.966 1.051 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.936 0.951 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.940 0.940 make_images_data 4572 15.5 0.047 0.051 0.768 0.937 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.672 0.866 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 0.790 0.812 transfer_pw2rs 487 13.2 0.004 0.004 0.806 0.810 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 acc_transpose_blocks_sync 54864 16.5 0.742 0.770 0.742 0.770 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.675 0.743 mp_alltoall_d11v 2130 13.8 0.658 0.733 0.658 0.733 cp_fm_cholesky_invert 11 10.9 0.697 0.701 0.697 0.701 mp_alltoall_z22v 1201 16.6 0.606 0.672 0.606 0.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.195000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 554.545152E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.041 36.419 36.420 qs_mol_dyn_low 1 2.0 0.003 0.004 36.209 36.216 qs_forces 11 3.9 0.005 0.007 36.145 36.152 qs_energies 11 4.9 0.002 0.004 34.443 34.454 scf_env_do_scf 11 5.9 0.000 0.001 28.989 28.990 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.660 25.660 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.968 20.064 velocity_verlet 10 3.0 0.001 0.002 18.287 18.289 qs_scf_new_mos 108 7.5 0.001 0.001 17.634 17.689 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.633 17.689 multiply_cannon 2286 13.5 0.216 0.227 16.344 16.675 ot_scf_mini 108 9.5 0.002 0.003 16.598 16.656 multiply_cannon_loop 2286 14.5 1.558 1.632 15.412 15.734 ot_mini 108 10.5 0.001 0.001 10.217 10.279 multiply_cannon_multrec 27432 15.5 2.455 3.138 9.106 9.387 qs_ot_get_derivative 108 11.5 0.001 0.001 8.360 8.417 dbcsr_mm_accdrv_process 47916 15.9 6.160 7.483 6.551 7.916 rebuild_ks_matrix 119 8.3 0.000 0.000 6.276 6.331 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.275 6.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.557 5.605 init_scf_run 11 5.9 0.000 0.001 3.912 3.913 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.912 3.912 qs_ot_get_p 119 10.4 0.001 0.001 3.533 3.611 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.056 3.507 sum_up_and_integrate 119 10.3 0.001 0.001 3.447 3.456 integrate_v_rspace 119 11.3 0.002 0.003 3.436 3.445 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.346 3.384 calculate_rho_elec 119 8.7 0.040 0.046 3.346 3.383 init_scf_loop 11 6.9 0.000 0.000 3.310 3.310 acc_transpose_blocks 27432 15.5 0.121 0.125 2.571 2.760 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.152 2.598 apply_single 119 13.6 0.000 0.000 2.151 2.598 calculate_first_density_matrix 1 7.0 0.000 0.000 2.534 2.539 prepare_preconditioner 11 7.9 0.000 0.000 2.492 2.500 make_preconditioner 11 8.9 0.000 0.000 2.492 2.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.097 2.426 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.305 2.334 make_m2s 4572 13.5 0.054 0.055 2.219 2.333 calculate_dm_sparse 119 9.5 0.000 0.000 2.234 2.288 make_images 4572 14.5 0.273 0.333 2.111 2.223 mp_waitall_1 137007 16.6 1.690 2.162 1.690 2.162 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.146 2.157 pw_transfer 1439 11.6 0.064 0.068 1.962 1.997 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.946 1.947 grid_integrate_task_list 119 12.3 1.828 1.910 1.828 1.910 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.870 1.909 density_rs2pw 119 9.7 0.004 0.004 1.773 1.874 ot_diis_step 108 11.5 0.012 0.012 1.819 1.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.782 1.795 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.744 1.744 acc_transpose_blocks_sync 82296 16.5 1.470 1.612 1.470 1.612 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.558 1.598 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.511 1.521 cp_fm_diag_elpa_base 50 14.0 1.477 1.495 1.509 1.518 mp_sum_l 7287 12.8 0.970 1.484 0.970 1.484 fft3d_ps 1201 14.6 0.591 0.649 1.358 1.382 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.335 grid_collocate_task_list 119 9.7 1.248 1.332 1.248 1.332 wfi_extrapolate 11 7.9 0.001 0.001 1.322 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.287 1.303 potential_pw2rs 119 12.3 0.008 0.009 1.256 1.260 qs_energies_init_hamiltonians 11 5.9 0.023 0.061 1.190 1.199 cp_fm_upper_to_full 72 14.2 0.827 1.175 0.827 1.175 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.718 1.170 jit_kernel_multiply 6 16.2 0.321 1.076 0.321 1.076 dbcsr_complete_redistribute 329 12.2 0.117 0.143 0.741 1.021 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.952 0.999 transfer_rs2pw 487 10.6 0.004 0.005 0.867 0.981 mp_alltoall_d11v 2130 13.8 0.737 0.947 0.737 0.947 make_images_data 4572 15.5 0.048 0.053 0.819 0.938 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.687 0.888 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.875 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.555 0.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.420000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.363636, yerr=1.720081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 628.051968E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.441 30.099 30.157 qs_mol_dyn_low 1 2.0 0.004 0.028 29.216 29.495 qs_forces 11 3.9 0.021 0.024 29.119 29.121 qs_energies 11 4.9 0.002 0.003 27.386 27.392 scf_env_do_scf 11 5.9 0.000 0.001 22.255 22.255 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.742 19.742 velocity_verlet 10 3.0 0.022 0.025 14.892 14.895 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.320 13.392 qs_scf_new_mos 108 7.5 0.001 0.001 11.987 12.009 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.987 12.008 ot_scf_mini 108 9.5 0.002 0.002 11.265 11.295 multiply_cannon 2286 13.5 0.224 0.228 10.699 11.174 multiply_cannon_loop 2286 14.5 0.645 0.667 9.790 9.974 ot_mini 108 10.5 0.001 0.001 6.417 6.455 multiply_cannon_multrec 9144 15.5 1.677 1.932 5.967 6.266 rebuild_ks_matrix 119 8.3 0.000 0.000 5.810 5.832 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.809 5.832 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.168 5.188 qs_ot_get_derivative 108 11.5 0.001 0.001 5.115 5.145 dbcsr_mm_accdrv_process 12550 15.8 3.559 4.243 4.182 4.286 init_scf_run 11 5.9 0.000 0.001 3.618 3.618 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.618 3.618 sum_up_and_integrate 119 10.3 0.001 0.001 3.458 3.463 integrate_v_rspace 119 11.3 0.003 0.003 3.448 3.453 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.427 3.434 calculate_rho_elec 119 8.7 0.060 0.061 3.426 3.434 qs_ot_get_p 119 10.4 0.001 0.001 2.949 2.986 init_scf_loop 11 6.9 0.000 0.000 2.492 2.492 calculate_first_density_matrix 1 7.0 0.000 0.000 2.474 2.474 pw_transfer 1439 11.6 0.065 0.067 2.142 2.152 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.049 2.061 calculate_dm_sparse 119 9.5 0.000 0.000 1.969 1.987 grid_integrate_task_list 119 12.3 1.861 1.960 1.861 1.960 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.945 1.948 mp_waitall_1 115863 16.7 1.463 1.940 1.463 1.940 make_m2s 4572 13.5 0.033 0.034 1.798 1.937 density_rs2pw 119 9.7 0.003 0.003 1.800 1.847 make_images 4572 14.5 0.269 0.301 1.708 1.846 prepare_preconditioner 11 7.9 0.000 0.000 1.762 1.767 make_preconditioner 11 8.9 0.000 0.000 1.762 1.767 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.739 1.761 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.693 1.708 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.696 1.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.654 1.682 acc_transpose_blocks 9144 15.5 0.043 0.045 1.616 1.657 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.534 1.535 fft3d_ps 1201 14.6 0.651 0.664 1.402 1.413 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.389 1.396 cp_fm_diag_elpa_base 50 14.0 1.361 1.378 1.387 1.395 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.371 1.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.373 1.380 grid_collocate_task_list 119 9.7 1.296 1.378 1.296 1.378 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.343 1.348 jit_kernel_multiply 8 15.3 0.582 1.336 0.582 1.336 ot_diis_step 108 11.5 0.013 0.013 1.290 1.290 potential_pw2rs 119 12.3 0.010 0.010 1.271 1.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.230 1.261 apply_single 119 13.6 0.000 0.000 1.230 1.261 qs_energies_init_hamiltonians 11 5.9 0.014 0.016 1.241 1.246 wfi_extrapolate 11 7.9 0.001 0.001 1.092 1.093 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.775 0.965 mp_alltoall_d11v 2130 13.8 0.810 0.949 0.810 0.949 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.931 make_images_data 4572 15.5 0.042 0.046 0.767 0.926 cp_fm_cholesky_invert 11 10.9 0.883 0.886 0.883 0.886 acc_transpose_blocks_kernels 9144 16.5 0.118 0.122 0.833 0.885 transfer_rs2pw 487 10.6 0.004 0.004 0.789 0.851 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.835 0.840 mp_sum_l 7287 12.8 0.652 0.786 0.652 0.786 jit_kernel_transpose 5 15.6 0.715 0.768 0.715 0.768 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.763 0.766 acc_transpose_blocks_sync 27432 16.5 0.724 0.764 0.724 0.764 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.373 0.721 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.673 0.718 mp_alltoall_z22v 1201 16.6 0.625 0.661 0.625 0.661 transfer_pw2rs 487 13.2 0.003 0.004 0.659 0.661 mp_allgather_i34 2286 14.5 0.224 0.634 0.224 0.634 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.157000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.000000, yerr=4.112067 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 770.932736E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.059 43.396 43.397 qs_mol_dyn_low 1 2.0 0.003 0.003 43.035 43.042 qs_forces 11 3.9 0.003 0.005 42.798 42.799 qs_energies 11 4.9 0.001 0.001 40.778 40.781 scf_env_do_scf 11 5.9 0.001 0.001 34.895 34.895 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.150 27.152 velocity_verlet 10 3.0 0.002 0.002 24.316 24.344 dbcsr_multiply_generic 2286 12.5 0.102 0.104 19.002 19.252 qs_scf_new_mos 108 7.5 0.001 0.001 17.367 17.470 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.366 17.470 ot_scf_mini 108 9.5 0.002 0.002 16.197 16.300 multiply_cannon 2286 13.5 0.298 0.306 15.002 16.016 multiply_cannon_loop 2286 14.5 0.866 0.889 13.749 14.798 ot_mini 108 10.5 0.001 0.001 9.830 9.943 multiply_cannon_multrec 9144 15.5 3.371 4.636 8.826 8.939 qs_ot_get_derivative 108 11.5 0.001 0.001 7.777 7.879 init_scf_loop 11 6.9 0.000 0.000 7.714 7.719 rebuild_ks_matrix 119 8.3 0.000 0.000 7.327 7.466 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.326 7.465 prepare_preconditioner 11 7.9 0.000 0.000 6.730 6.743 make_preconditioner 11 8.9 0.000 0.000 6.730 6.743 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.608 6.734 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.305 6.624 dbcsr_mm_accdrv_process 12550 15.8 4.677 6.583 5.322 6.617 cp_fm_upper_to_full 72 14.2 3.152 4.525 3.152 4.525 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.257 4.266 calculate_rho_elec 119 8.7 0.117 0.120 4.257 4.266 sum_up_and_integrate 119 10.3 0.001 0.001 3.891 3.898 integrate_v_rspace 119 11.3 0.003 0.004 3.881 3.888 init_scf_run 11 5.9 0.000 0.001 3.715 3.715 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.714 3.715 qs_ot_get_p 119 10.4 0.001 0.001 3.424 3.559 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.784 3.251 mp_waitall_1 94719 16.7 2.188 3.231 2.188 3.231 pw_transfer 1439 11.6 0.068 0.069 3.096 3.101 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.998 3.002 dbcsr_complete_redistribute 329 12.2 0.279 0.284 1.970 2.759 fft_wrap_pw1pw2_140 487 13.2 0.650 0.652 2.548 2.555 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.267 2.536 apply_single 119 13.6 0.000 0.000 2.267 2.536 make_m2s 4572 13.5 0.037 0.037 2.371 2.518 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.660 2.450 make_images 4572 14.5 0.354 0.388 2.251 2.398 calculate_dm_sparse 119 9.5 0.000 0.000 2.308 2.325 density_rs2pw 119 9.7 0.003 0.003 2.278 2.289 mp_alltoall_i22 627 13.8 1.401 2.232 1.401 2.232 calculate_first_density_matrix 1 7.0 0.000 0.000 2.211 2.213 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.245 2.208 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.419 2.203 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.079 2.130 grid_integrate_task_list 119 12.3 2.067 2.096 2.067 2.096 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.036 2.038 ot_diis_step 108 11.5 0.014 0.014 2.029 2.029 fft3d_ps 1201 14.6 0.868 0.885 1.919 1.925 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.881 1.882 acc_transpose_blocks 9144 15.5 0.045 0.046 1.841 1.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.806 1.807 mp_sum_l 7287 12.8 1.084 1.757 1.084 1.757 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.685 1.685 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.549 1.590 grid_collocate_task_list 119 9.7 1.512 1.523 1.512 1.523 wfi_extrapolate 11 7.9 0.001 0.001 1.434 1.434 potential_pw2rs 119 12.3 0.014 0.014 1.411 1.413 cp_fm_cholesky_invert 11 10.9 1.399 1.403 1.399 1.403 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.388 1.388 cp_fm_diag_elpa_base 50 14.0 1.242 1.294 1.386 1.386 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.328 1.346 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.304 1.313 hybrid_alltoall_any 4725 16.4 0.091 0.149 1.067 1.279 mp_alltoall_d11v 2130 13.8 1.199 1.235 1.199 1.235 qs_env_update_s_mstruct 11 6.9 0.036 0.042 1.187 1.208 make_images_data 4572 15.5 0.047 0.050 0.989 1.177 acc_transpose_blocks_sync 27432 16.5 1.108 1.143 1.108 1.143 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.997 1.046 qs_create_task_list 11 7.9 0.035 0.037 0.964 0.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.930 0.944 generate_qs_task_list 11 8.9 0.368 0.387 0.928 0.936 mp_alltoall_z22v 1201 16.6 0.913 0.935 0.913 0.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.397000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=723.272727, yerr=13.625450 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.607872E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1592512. 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 0.016 0.024 83.221 83.223 qs_mol_dyn_low 1 2.0 0.003 0.003 82.579 82.590 qs_forces 11 3.9 0.005 0.007 82.474 82.479 qs_energies 11 4.9 0.001 0.002 79.606 79.622 scf_env_do_scf 11 5.9 0.000 0.001 70.072 70.074 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.515 64.515 dbcsr_multiply_generic 2055 12.4 0.107 0.110 50.996 51.243 qs_scf_new_mos 99 7.5 0.000 0.001 47.332 47.490 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.331 47.489 ot_scf_mini 99 9.5 0.002 0.002 44.949 45.059 multiply_cannon 2055 13.4 0.176 0.180 42.511 43.601 velocity_verlet 10 3.0 0.001 0.002 43.382 43.383 multiply_cannon_loop 2055 14.4 1.811 1.862 41.558 42.696 ot_mini 99 10.5 0.001 0.001 26.633 26.769 qs_ot_get_derivative 99 11.5 0.001 0.001 19.778 19.916 multiply_cannon_multrec 49320 15.4 11.388 12.080 17.455 18.312 rebuild_ks_matrix 110 8.3 0.000 0.001 14.556 14.722 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.556 14.722 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.767 12.920 mp_waitall_1 220248 16.4 10.612 11.378 10.612 11.378 multiply_cannon_sync_h2d 49320 15.4 9.623 10.367 9.623 10.367 qs_ot_get_p 110 10.4 0.001 0.001 9.860 10.027 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.528 8.118 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.279 7.800 apply_single 110 13.6 0.000 0.001 7.278 7.800 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.398 7.587 init_scf_run 11 5.9 0.000 0.001 7.428 7.428 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.428 7.428 sum_up_and_integrate 110 10.3 0.002 0.003 7.050 7.063 integrate_v_rspace 110 11.3 0.002 0.003 7.025 7.044 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.627 6.657 ot_diis_step 99 11.5 0.005 0.006 6.625 6.626 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.417 6.598 calculate_rho_elec 110 8.6 0.020 0.024 6.417 6.598 dbcsr_mm_accdrv_process 87628 16.1 3.104 3.229 5.937 6.240 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.799 5.800 init_scf_loop 11 6.9 0.000 0.000 5.528 5.529 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.422 5.509 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.228 5.256 cp_fm_diag_elpa_base 48 14.0 5.214 5.244 5.226 5.254 mp_sum_l 6594 12.7 4.080 5.117 4.080 5.117 wfi_extrapolate 11 7.9 0.001 0.001 4.096 4.096 make_m2s 4110 13.4 0.061 0.064 3.952 4.060 make_images 4110 14.4 0.178 0.190 3.856 3.969 calculate_dm_sparse 110 9.5 0.001 0.001 3.838 3.935 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.312 3.750 density_rs2pw 110 9.6 0.004 0.004 3.446 3.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.533 3.537 grid_integrate_task_list 110 12.3 3.259 3.409 3.259 3.409 prepare_preconditioner 11 7.9 0.000 0.000 3.361 3.376 make_preconditioner 11 8.9 0.000 0.000 3.361 3.376 calculate_first_density_matrix 1 7.0 0.000 0.000 3.226 3.234 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.185 3.227 pw_transfer 1331 11.6 0.055 0.069 3.136 3.222 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.154 3.194 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.137 3.175 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.048 3.136 acc_transpose_blocks 49320 15.4 0.211 0.220 2.655 2.745 fft_wrap_pw1pw2_140 451 13.1 0.309 0.330 2.588 2.680 potential_pw2rs 110 12.3 0.005 0.006 2.636 2.658 jit_kernel_multiply 13 15.9 2.546 2.580 2.546 2.580 mp_alltoall_d11v 2046 13.8 2.110 2.543 2.110 2.543 fft3d_ps 1111 14.6 0.798 0.887 2.342 2.399 grid_collocate_task_list 110 9.6 2.156 2.301 2.156 2.301 mp_waitany 14300 13.8 1.817 2.137 1.817 2.137 transfer_rs2pw 451 10.6 0.005 0.006 1.949 2.100 mp_sum_d 3891 11.9 1.406 2.020 1.406 2.020 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.933 parallel_gemm_fm 81 9.0 0.000 0.000 1.890 1.896 parallel_gemm_fm_cosma 81 10.0 1.890 1.896 1.890 1.896 make_images_data 4110 15.4 0.044 0.048 1.774 1.892 cp_fm_cholesky_invert 11 10.9 1.884 1.888 1.884 1.888 transfer_pw2rs 451 13.1 0.006 0.007 1.796 1.810 hybrid_alltoall_any 4261 16.3 0.084 0.487 1.561 1.808 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.728 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.223000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.842999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 593.907712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2063149. 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 0.040 0.099 69.167 69.172 qs_mol_dyn_low 1 2.0 0.005 0.014 68.772 68.804 qs_forces 11 3.9 0.008 0.024 68.686 68.694 qs_energies 11 4.9 0.003 0.010 65.348 65.365 scf_env_do_scf 11 5.9 0.003 0.023 56.589 56.593 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.969 48.970 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.035 38.208 velocity_verlet 10 3.0 0.001 0.003 35.831 35.832 qs_scf_new_mos 99 7.5 0.001 0.001 33.132 33.294 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.131 33.294 multiply_cannon 2055 13.4 0.222 0.244 31.445 32.596 ot_scf_mini 99 9.5 0.003 0.004 31.464 31.622 multiply_cannon_loop 2055 14.4 1.166 1.202 30.173 31.170 ot_mini 99 10.5 0.001 0.002 18.380 18.536 multiply_cannon_multrec 24660 15.4 6.939 8.299 14.148 15.472 rebuild_ks_matrix 110 8.3 0.000 0.000 13.598 13.723 qs_ks_build_kohn_sham_matrix 110 9.3 0.021 0.065 13.598 13.722 qs_ot_get_derivative 99 11.5 0.002 0.015 12.586 12.742 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.977 12.098 mp_waitall_1 176588 16.5 7.477 9.833 7.477 9.833 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.121 7.989 init_scf_loop 11 6.9 0.019 0.085 7.583 7.587 multiply_cannon_sync_h2d 24660 15.4 6.401 7.472 6.401 7.472 dbcsr_mm_accdrv_process 52282 16.1 5.482 6.437 7.040 7.387 apply_preconditioner_dbcsr 110 12.6 0.000 0.002 6.498 7.148 apply_single 110 13.6 0.000 0.001 6.498 7.148 qs_ot_get_p 110 10.4 0.001 0.002 6.434 6.587 sum_up_and_integrate 110 10.3 0.002 0.004 6.368 6.378 integrate_v_rspace 110 11.3 0.002 0.003 6.341 6.353 init_scf_run 11 5.9 0.000 0.001 6.281 6.281 scf_env_initial_rho_setup 11 6.9 0.002 0.015 6.280 6.281 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.901 5.911 calculate_rho_elec 110 8.6 0.040 0.049 5.901 5.910 ot_diis_step 99 11.5 0.010 0.012 5.747 5.747 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.715 5.464 prepare_preconditioner 11 7.9 0.000 0.000 5.419 5.447 make_preconditioner 11 8.9 0.000 0.001 5.419 5.447 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.026 5.185 make_m2s 4110 13.4 0.057 0.060 4.196 4.656 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.557 4.578 make_images 4110 14.4 0.408 0.466 4.085 4.540 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.095 4.095 pw_transfer 1331 11.6 0.065 0.073 3.559 3.697 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.452 3.595 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.534 3.544 cp_fm_diag_elpa_base 48 14.0 3.487 3.504 3.530 3.541 wfi_extrapolate 11 7.9 0.001 0.002 3.510 3.510 density_rs2pw 110 9.6 0.004 0.004 3.144 3.400 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.373 3.375 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.275 3.352 grid_integrate_task_list 110 12.3 3.159 3.341 3.159 3.341 fft_wrap_pw1pw2_140 451 13.1 0.357 0.376 2.957 3.102 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.041 3.090 calculate_dm_sparse 110 9.5 0.001 0.001 2.993 3.028 fft3d_ps 1111 14.6 1.117 1.349 2.544 2.700 make_images_data 4110 15.4 0.049 0.054 2.248 2.688 hybrid_alltoall_any 4261 16.3 0.105 0.452 1.928 2.688 calculate_first_density_matrix 1 7.0 0.001 0.003 2.663 2.667 cp_fm_cholesky_invert 11 10.9 2.539 2.548 2.539 2.548 mp_sum_l 6594 12.7 1.818 2.360 1.818 2.360 grid_collocate_task_list 110 9.6 2.167 2.321 2.167 2.321 potential_pw2rs 110 12.3 0.008 0.008 2.261 2.271 acc_transpose_blocks 24660 15.4 0.118 0.123 1.997 2.048 jit_kernel_multiply 10 16.3 1.185 2.022 1.185 2.022 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.960 1.981 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.965 1.978 mp_alltoall_d11v 2046 13.8 1.696 1.858 1.696 1.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.800 1.814 transfer_rs2pw 451 10.6 0.006 0.007 1.496 1.687 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.753 1.675 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.560 1.568 mp_irecv_dv 57340 16.2 0.630 1.560 0.630 1.560 mp_allgather_i34 2055 14.4 0.558 1.536 0.558 1.536 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.492 mp_waitany 10164 13.8 1.236 1.416 1.236 1.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.172000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.363636, yerr=7.351842 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.685440E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.124 0.184 60.762 60.764 qs_mol_dyn_low 1 2.0 0.003 0.007 59.917 59.994 qs_forces 11 3.9 0.003 0.005 59.560 59.568 qs_energies 11 4.9 0.002 0.007 56.403 56.413 scf_env_do_scf 11 5.9 0.001 0.001 47.949 47.949 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.603 39.604 velocity_verlet 10 3.0 0.027 0.056 32.399 32.407 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.748 29.016 qs_scf_new_mos 99 7.5 0.001 0.001 25.065 25.163 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.064 25.162 ot_scf_mini 99 9.5 0.002 0.003 23.817 23.931 multiply_cannon 2055 13.4 0.211 0.221 22.644 23.830 multiply_cannon_loop 2055 14.4 0.818 0.854 21.474 22.423 ot_mini 99 10.5 0.001 0.001 13.688 13.800 rebuild_ks_matrix 110 8.3 0.000 0.000 12.070 12.206 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.018 12.070 12.206 multiply_cannon_multrec 16440 15.4 3.647 4.801 9.935 11.145 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.624 10.748 mp_waitall_1 139946 16.5 6.794 9.826 6.794 9.826 qs_ot_get_derivative 99 11.5 0.001 0.001 9.279 9.388 init_scf_loop 11 6.9 0.001 0.003 8.312 8.313 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.316 7.193 prepare_preconditioner 11 7.9 0.000 0.000 6.557 6.573 make_preconditioner 11 8.9 0.000 0.002 6.557 6.573 dbcsr_mm_accdrv_process 34862 16.1 5.496 5.856 6.133 6.274 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.911 6.264 sum_up_and_integrate 110 10.3 0.002 0.003 6.222 6.236 integrate_v_rspace 110 11.3 0.003 0.003 6.196 6.211 init_scf_run 11 5.9 0.000 0.001 6.057 6.058 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.057 6.057 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.706 5.716 calculate_rho_elec 110 8.6 0.059 0.059 5.705 5.716 qs_ot_get_p 110 10.4 0.001 0.001 5.307 5.447 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.876 5.423 apply_single 110 13.6 0.000 0.000 4.876 5.422 make_m2s 4110 13.4 0.049 0.051 4.089 4.473 ot_diis_step 99 11.5 0.010 0.011 4.379 4.380 make_images 4110 14.4 0.398 0.522 3.973 4.356 multiply_cannon_sync_h2d 16440 15.4 3.265 3.910 3.265 3.910 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.251 3.844 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.697 3.701 pw_transfer 1331 11.6 0.065 0.071 3.577 3.584 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.471 3.480 grid_integrate_task_list 110 12.3 3.187 3.394 3.187 3.394 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.328 3.329 density_rs2pw 110 9.6 0.004 0.004 2.943 3.187 calculate_first_density_matrix 1 7.0 0.000 0.002 3.027 3.029 fft_wrap_pw1pw2_140 451 13.1 0.462 0.473 3.014 3.028 wfi_extrapolate 11 7.9 0.001 0.001 2.940 2.940 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.834 2.836 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.800 2.809 cp_fm_diag_elpa_base 48 14.0 2.735 2.767 2.798 2.807 make_images_data 4110 15.4 0.046 0.051 2.274 2.749 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.056 2.678 calculate_dm_sparse 110 9.5 0.001 0.001 2.574 2.605 cp_fm_cholesky_invert 11 10.9 2.576 2.582 2.576 2.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.496 2.540 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.429 2.497 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.884 2.419 fft3d_ps 1111 14.6 1.116 1.128 2.386 2.401 grid_collocate_task_list 110 9.6 2.210 2.347 2.210 2.347 mp_irecv_dv 48980 15.7 0.811 2.289 0.811 2.289 mp_sum_l 6594 12.7 1.425 2.086 1.425 2.086 potential_pw2rs 110 12.3 0.010 0.011 2.074 2.078 mp_alltoall_d11v 2046 13.8 1.758 2.059 1.758 2.059 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.962 1.971 dbcsr_complete_redistribute 325 12.2 0.320 0.357 1.363 1.818 cp_fm_upper_to_full 70 14.2 1.394 1.759 1.394 1.759 acc_transpose_blocks 16440 15.4 0.079 0.082 1.643 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.653 cp_fm_cholesky_decompose 22 10.9 1.529 1.548 1.529 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.510 1.523 transfer_rs2pw 451 10.6 0.005 0.006 1.255 1.500 mp_allgather_i34 2055 14.4 0.483 1.463 0.483 1.463 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.357 1.449 mp_waitany 17072 13.8 1.125 1.427 1.125 1.427 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.952 1.398 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.269 1.280 make_basis_sm 11 9.8 0.001 0.007 1.265 1.268 parallel_gemm_fm 81 9.0 0.000 0.000 1.259 1.261 parallel_gemm_fm_cosma 81 10.0 1.259 1.261 1.259 1.261 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.764000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.818182, yerr=7.802309 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 740.978688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.030 67.522 67.524 qs_mol_dyn_low 1 2.0 0.003 0.004 67.217 67.239 qs_forces 11 3.9 0.003 0.003 65.059 65.061 qs_energies 11 4.9 0.001 0.002 61.676 61.681 scf_env_do_scf 11 5.9 0.000 0.001 53.061 53.064 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.534 41.534 velocity_verlet 10 3.0 0.001 0.002 38.963 39.052 dbcsr_multiply_generic 2055 12.4 0.118 0.122 30.463 30.672 qs_scf_new_mos 99 7.5 0.001 0.001 27.074 27.190 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.074 27.189 ot_scf_mini 99 9.5 0.002 0.003 25.388 25.491 multiply_cannon 2055 13.4 0.236 0.250 23.622 24.851 multiply_cannon_loop 2055 14.4 1.398 1.489 22.134 22.778 ot_mini 99 10.5 0.001 0.001 14.660 14.790 multiply_cannon_multrec 24660 15.4 4.065 6.691 13.031 14.183 rebuild_ks_matrix 110 8.3 0.000 0.000 11.857 11.961 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.857 11.961 init_scf_loop 11 6.9 0.000 0.000 11.490 11.490 qs_ot_get_derivative 99 11.5 0.001 0.002 10.495 10.603 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.468 10.559 prepare_preconditioner 11 7.9 0.000 0.000 9.774 9.789 make_preconditioner 11 8.9 0.000 0.000 9.774 9.789 dbcsr_mm_accdrv_process 52304 16.0 7.847 9.230 8.810 9.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.970 9.448 sum_up_and_integrate 110 10.3 0.001 0.002 6.158 6.169 integrate_v_rspace 110 11.3 0.003 0.003 6.132 6.142 mp_waitall_1 121746 16.5 3.979 5.929 3.979 5.929 init_scf_run 11 5.9 0.000 0.001 5.880 5.880 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.880 5.880 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.776 5.791 calculate_rho_elec 110 8.6 0.077 0.081 5.775 5.790 qs_ot_get_p 110 10.4 0.001 0.001 5.626 5.767 make_m2s 4110 13.4 0.059 0.061 5.200 5.639 make_images 4110 14.4 0.577 0.698 5.059 5.494 cp_fm_upper_to_full 70 14.2 3.377 4.894 3.377 4.894 ot_diis_step 99 11.5 0.011 0.011 4.129 4.129 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.124 apply_single 110 13.6 0.000 0.000 4.046 4.124 dbcsr_complete_redistribute 325 12.2 0.423 0.485 2.681 3.834 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.803 3.822 pw_transfer 1331 11.6 0.065 0.073 3.717 3.750 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.610 3.647 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.471 3.513 grid_integrate_task_list 110 12.3 3.264 3.425 3.264 3.425 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.315 3.375 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.171 3.308 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.248 3.249 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.363 3.176 fft_wrap_pw1pw2_140 451 13.1 0.532 0.548 3.125 3.165 calculate_dm_sparse 110 9.5 0.001 0.001 3.047 3.080 density_rs2pw 110 9.6 0.004 0.004 2.913 3.074 wfi_extrapolate 11 7.9 0.001 0.001 2.979 2.979 hybrid_alltoall_any 4261 16.3 0.122 0.457 2.214 2.964 make_images_data 4110 15.4 0.049 0.052 2.516 2.948 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.795 2.918 mp_alltoall_i22 605 13.7 1.659 2.872 1.659 2.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.829 2.831 calculate_first_density_matrix 1 7.0 0.000 0.001 2.805 2.809 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.730 2.739 cp_fm_diag_elpa_base 48 14.0 2.584 2.645 2.728 2.738 acc_transpose_blocks 24660 15.4 0.114 0.116 2.458 2.625 cp_fm_cholesky_invert 11 10.9 2.574 2.582 2.574 2.582 multiply_cannon_sync_h2d 24660 15.4 2.350 2.524 2.350 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.440 2.481 fft3d_ps 1111 14.6 1.154 1.196 2.434 2.447 grid_collocate_task_list 110 9.6 2.259 2.447 2.259 2.447 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.250 2.252 mp_sum_dm 438 4.9 2.084 2.189 2.084 2.189 md_output 10 3.0 0.000 0.000 0.093 2.157 md_write_output 11 3.9 0.089 2.125 0.091 2.144 update_particle_set 20 4.0 0.000 0.000 2.050 2.140 potential_pw2rs 110 12.3 0.012 0.013 1.992 1.997 mp_alltoall_d11v 2046 13.8 1.759 1.949 1.759 1.949 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.787 1.818 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.713 cp_fm_cholesky_decompose 22 10.9 1.609 1.657 1.609 1.657 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.602 1.620 mp_sum_l 6594 12.7 0.948 1.580 0.948 1.580 mp_allgather_i34 2055 14.4 0.583 1.556 0.583 1.556 acc_transpose_blocks_sync 73980 16.4 1.380 1.550 1.380 1.550 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.854 1.457 jit_kernel_multiply 8 15.7 0.611 1.422 0.611 1.422 mp_irecv_dv 62702 16.1 0.751 1.378 0.751 1.378 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.524000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.363636, yerr=11.443718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 861.437952E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.057 55.040 55.048 qs_mol_dyn_low 1 2.0 0.003 0.003 54.774 54.785 qs_forces 11 3.9 0.020 0.038 54.685 54.692 qs_energies 11 4.9 0.001 0.001 51.045 51.071 scf_env_do_scf 11 5.9 0.000 0.001 42.642 42.642 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.134 35.135 velocity_verlet 10 3.0 0.001 0.002 30.716 30.720 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.348 23.606 qs_scf_new_mos 99 7.5 0.001 0.001 20.695 20.768 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.694 20.768 ot_scf_mini 99 9.5 0.002 0.002 19.455 19.506 multiply_cannon 2055 13.4 0.236 0.250 17.961 19.113 multiply_cannon_loop 2055 14.4 0.605 0.626 16.752 17.173 rebuild_ks_matrix 110 8.3 0.000 0.000 11.542 11.606 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.542 11.606 ot_mini 99 10.5 0.001 0.001 11.064 11.119 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.265 10.319 multiply_cannon_multrec 8220 15.4 3.205 4.770 7.630 8.812 mp_waitall_1 103326 16.6 6.047 7.639 6.047 7.639 init_scf_loop 11 6.9 0.000 0.000 7.463 7.464 qs_ot_get_derivative 99 11.5 0.001 0.001 7.302 7.356 sum_up_and_integrate 110 10.3 0.001 0.002 6.162 6.172 integrate_v_rspace 110 11.3 0.003 0.003 6.135 6.146 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.937 5.950 calculate_rho_elec 110 8.6 0.112 0.113 5.937 5.949 prepare_preconditioner 11 7.9 0.000 0.000 5.834 5.841 make_preconditioner 11 8.9 0.000 0.000 5.834 5.841 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.431 5.504 dbcsr_mm_accdrv_process 17442 15.9 3.208 4.340 4.285 5.267 init_scf_run 11 5.9 0.000 0.001 5.187 5.187 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.186 5.187 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.274 4.558 qs_ot_get_p 110 10.4 0.001 0.001 4.488 4.553 make_m2s 4110 13.4 0.037 0.039 4.054 4.248 make_images 4110 14.4 0.647 0.704 3.924 4.113 pw_transfer 1331 11.6 0.065 0.070 4.041 4.050 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.934 3.946 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.797 3.861 apply_single 110 13.6 0.000 0.000 3.797 3.860 ot_diis_step 99 11.5 0.012 0.012 3.741 3.741 grid_integrate_task_list 110 12.3 3.382 3.536 3.382 3.536 fft_wrap_pw1pw2_140 451 13.1 0.718 0.730 3.430 3.446 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.164 3.168 density_rs2pw 110 9.6 0.004 0.004 2.973 3.141 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.830 2.830 cp_fm_cholesky_invert 11 10.9 2.815 2.819 2.815 2.819 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.770 2.793 wfi_extrapolate 11 7.9 0.001 0.001 2.665 2.665 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.520 grid_collocate_task_list 110 9.6 2.369 2.516 2.369 2.516 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.509 2.511 fft3d_ps 1111 14.6 1.290 1.296 2.481 2.501 multiply_cannon_sync_h2d 8220 15.4 2.347 2.488 2.347 2.488 hybrid_alltoall_any 4261 16.3 0.199 0.853 2.099 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.428 make_images_data 4110 15.4 0.041 0.046 2.141 2.398 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.337 2.344 cp_fm_diag_elpa_base 48 14.0 2.282 2.303 2.335 2.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.149 2.179 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.980 2.021 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.989 potential_pw2rs 110 12.3 0.015 0.015 1.976 1.979 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.875 1.907 mp_alltoall_d11v 2046 13.8 1.552 1.798 1.552 1.798 qs_env_update_s_mstruct 11 6.9 0.040 0.048 1.572 1.686 cp_fm_cholesky_decompose 22 10.9 1.611 1.623 1.611 1.623 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.604 1.614 dbcsr_complete_redistribute 325 12.2 0.554 0.578 1.385 1.465 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.450 acc_transpose_blocks 8220 15.4 0.041 0.042 1.371 1.403 mp_allgather_i34 2055 14.4 0.381 1.388 0.381 1.388 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.482 1.375 mp_irecv_dv 24056 15.7 0.456 1.330 0.456 1.330 qs_create_task_list 11 7.9 0.002 0.003 1.205 1.305 generate_qs_task_list 11 8.9 0.373 0.441 1.203 1.302 transfer_rs2pw 451 10.6 0.005 0.005 1.095 1.289 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.828 1.262 mp_waitany 9240 13.8 1.039 1.242 1.039 1.242 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.109 1.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.048000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=809.181818, yerr=12.028891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 1.388573E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.122 88.320 88.322 qs_mol_dyn_low 1 2.0 0.003 0.004 87.599 87.728 qs_forces 11 3.9 0.003 0.003 87.483 87.483 qs_energies 11 4.9 0.006 0.007 83.329 83.330 scf_env_do_scf 11 5.9 0.000 0.001 72.855 72.855 velocity_verlet 10 3.0 0.002 0.002 55.893 55.904 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.745 44.747 dbcsr_multiply_generic 2055 12.4 0.120 0.124 29.766 29.933 init_scf_loop 11 6.9 0.000 0.000 28.040 28.043 qs_scf_new_mos 99 7.5 0.001 0.001 26.838 26.930 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.837 26.929 prepare_preconditioner 11 7.9 0.000 0.000 26.009 26.024 make_preconditioner 11 8.9 0.000 0.000 26.009 26.024 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.263 25.480 ot_scf_mini 99 9.5 0.002 0.002 25.045 25.135 multiply_cannon 2055 13.4 0.332 0.350 22.643 23.440 multiply_cannon_loop 2055 14.4 0.829 0.839 20.802 21.227 cp_fm_upper_to_full 70 14.2 12.708 18.175 12.708 18.175 ot_mini 99 10.5 0.001 0.001 14.269 14.358 rebuild_ks_matrix 110 8.3 0.000 0.001 13.987 14.076 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.986 14.075 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.690 12.770 dbcsr_complete_redistribute 325 12.2 1.005 1.043 7.283 10.493 multiply_cannon_multrec 8220 15.4 4.045 4.206 9.799 9.924 qs_ot_get_derivative 99 11.5 0.001 0.001 9.456 9.545 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.281 9.503 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.730 8.925 mp_waitall_1 84994 16.7 7.649 8.640 7.649 8.640 mp_alltoall_i22 605 13.7 5.366 8.566 5.366 8.566 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.428 7.481 calculate_rho_elec 110 8.6 0.222 0.223 7.428 7.480 sum_up_and_integrate 110 10.3 0.002 0.002 6.892 6.908 integrate_v_rspace 110 11.3 0.003 0.003 6.864 6.880 init_scf_run 11 5.9 0.000 0.001 6.177 6.177 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.177 6.177 make_m2s 4110 13.4 0.042 0.043 5.345 5.905 dbcsr_mm_accdrv_process 11614 15.7 4.059 4.510 5.604 5.838 pw_transfer 1331 11.6 0.074 0.075 5.763 5.770 make_images 4110 14.4 0.882 0.939 5.157 5.716 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.646 5.653 qs_ot_get_p 110 10.4 0.001 0.001 5.478 5.579 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.950 5.457 apply_single 110 13.6 0.000 0.000 4.949 5.457 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.965 5.231 cp_fm_cholesky_invert 11 10.9 5.083 5.087 5.083 5.087 fft_wrap_pw1pw2_140 451 13.1 1.367 1.372 4.977 4.983 ot_diis_step 99 11.5 0.015 0.016 4.797 4.797 density_rs2pw 110 9.6 0.004 0.004 3.898 3.915 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.903 3.911 grid_integrate_task_list 110 12.3 3.677 3.723 3.677 3.723 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.692 3.692 hybrid_alltoall_any 4261 16.3 0.263 0.569 2.888 3.622 make_images_data 4110 15.4 0.045 0.048 2.851 3.523 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.986 3.464 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.390 3.391 fft3d_ps 1111 14.6 1.882 1.890 3.352 3.359 wfi_extrapolate 11 7.9 0.001 0.001 3.332 3.332 calculate_dm_sparse 110 9.5 0.001 0.001 3.200 3.231 multiply_cannon_sync_h2d 8220 15.4 3.115 3.139 3.115 3.139 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.911 2.916 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.817 2.818 cp_fm_diag_elpa_base 48 14.0 2.279 2.471 2.815 2.815 calculate_first_density_matrix 1 7.0 0.000 0.000 2.726 2.730 grid_collocate_task_list 110 9.6 2.667 2.703 2.667 2.703 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.645 2.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.407 2.436 potential_pw2rs 110 12.3 0.021 0.021 2.420 2.429 qs_env_update_s_mstruct 11 6.9 0.013 0.018 2.202 2.256 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.111 2.204 mp_alltoall_d11v 2046 13.8 2.023 2.081 2.023 2.081 cp_fm_cholesky_decompose 22 10.9 1.941 1.960 1.941 1.960 qs_create_task_list 11 7.9 0.000 0.001 1.877 1.922 generate_qs_task_list 11 8.9 0.732 0.786 1.877 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.898 1.907 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.752 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.322000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1251.909091, yerr=60.426446 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 634.318848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 57717. MP_Allreduce 11092 796. MP_Sync 86 MP_Alltoall 2210 2528394. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 205.023 205.025 qs_mol_dyn_low 1 2.0 0.003 0.004 204.552 204.566 qs_forces 11 3.9 0.005 0.005 204.392 204.393 qs_energies 11 4.9 0.001 0.002 198.792 198.810 scf_env_do_scf 11 5.9 0.001 0.001 181.786 181.790 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 161.499 161.501 dbcsr_multiply_generic 2485 12.5 0.175 0.180 125.446 126.483 qs_scf_new_mos 116 7.6 0.001 0.001 122.840 123.127 qs_scf_loop_do_ot 116 8.6 0.001 0.001 122.839 123.126 velocity_verlet 10 3.0 0.002 0.004 121.133 121.134 ot_scf_mini 116 9.6 0.003 0.003 116.186 116.447 multiply_cannon 2485 13.5 0.234 0.250 101.637 104.143 multiply_cannon_loop 2485 14.5 2.401 2.489 99.563 101.853 ot_mini 116 10.6 0.001 0.001 66.192 66.447 multiply_cannon_multrec 59640 15.5 31.525 34.268 41.393 43.157 qs_ot_get_derivative 116 11.6 0.001 0.002 41.588 41.823 rebuild_ks_matrix 127 8.3 0.001 0.001 33.174 33.756 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.037 33.173 33.756 mp_waitall_1 264810 16.5 28.896 32.272 28.896 32.272 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.752 30.286 multiply_cannon_sync_h2d 59640 15.5 26.194 28.897 26.194 28.897 qs_ot_get_p 127 10.4 0.001 0.001 28.006 28.467 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.157 25.126 apply_single 127 13.6 0.001 0.001 24.157 25.126 ot_diis_step 116 11.6 0.007 0.008 24.354 24.355 qs_ot_p2m_diag 82 11.4 0.078 0.090 21.480 21.570 init_scf_loop 11 6.9 0.000 0.001 20.215 20.216 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.680 19.879 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.840 18.840 multiply_cannon_metrocomm3 59640 15.5 0.116 0.123 16.082 18.075 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.970 16.005 cp_fm_diag_elpa_base 82 14.4 15.905 15.951 15.966 16.000 prepare_preconditioner 11 7.9 0.000 0.000 15.584 15.643 make_preconditioner 11 8.9 0.000 0.000 15.584 15.643 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.829 14.993 make_m2s 4970 13.5 0.104 0.112 13.638 14.098 make_images 4970 14.5 0.394 0.413 13.455 13.923 sum_up_and_integrate 127 10.3 0.002 0.004 13.710 13.735 integrate_v_rspace 127 11.3 0.003 0.004 13.652 13.678 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.814 12.972 calculate_rho_elec 127 8.7 0.045 0.063 12.814 12.971 init_scf_run 11 5.9 0.000 0.001 12.886 12.887 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.886 12.886 mp_sum_l 7884 12.9 8.915 10.318 8.915 10.318 dbcsr_mm_accdrv_process 123452 16.2 4.754 4.983 9.433 10.019 wfi_extrapolate 11 7.9 0.001 0.001 9.019 9.019 cp_fm_cholesky_invert 11 10.9 8.860 8.868 8.860 8.868 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.479 8.668 multiply_cannon_metrocomm1 59640 15.5 0.095 0.100 6.519 8.556 calculate_dm_sparse 127 9.5 0.001 0.001 8.460 8.547 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.977 8.129 make_images_data 4970 15.5 0.066 0.073 6.679 7.640 pw_transfer 1535 11.6 0.073 0.089 7.431 7.611 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.231 7.415 grid_integrate_task_list 127 12.3 7.028 7.413 7.028 7.413 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.267 7.276 hybrid_alltoall_any 5155 16.4 0.289 2.236 5.885 7.239 density_rs2pw 127 9.7 0.006 0.006 6.526 7.165 fft_wrap_pw1pw2_140 519 13.2 0.852 0.888 6.397 6.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.385 6.564 fft3d_ps 1281 14.7 2.179 2.819 5.250 5.565 grid_collocate_task_list 127 9.7 4.800 5.232 4.800 5.232 mp_alltoall_d11v 2401 14.1 4.196 5.223 4.196 5.223 cp_fm_cholesky_decompose 22 10.9 4.562 4.575 4.562 4.575 acc_transpose_blocks 59640 15.5 0.288 0.304 3.802 4.398 potential_pw2rs 127 12.3 0.008 0.010 4.298 4.319 mp_sum_d 4455 12.1 3.502 4.272 3.502 4.272 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.025000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.273386 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 841.641984E+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 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5359083. 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 0.010 0.027 187.973 187.973 qs_mol_dyn_low 1 2.0 0.003 0.004 187.485 187.642 qs_forces 11 3.9 0.004 0.005 187.394 187.399 qs_energies 11 4.9 0.001 0.002 180.652 180.666 scf_env_do_scf 11 5.9 0.001 0.001 164.040 164.050 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.703 131.705 velocity_verlet 10 3.0 0.001 0.002 117.949 117.951 dbcsr_multiply_generic 2507 12.6 0.186 0.192 97.427 98.730 qs_scf_new_mos 117 7.6 0.001 0.001 93.420 93.939 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.419 93.938 ot_scf_mini 117 9.6 0.004 0.004 88.625 89.250 multiply_cannon 2507 13.6 0.503 0.560 77.505 82.198 multiply_cannon_loop 2507 14.6 1.571 1.638 74.172 76.800 ot_mini 117 10.6 0.001 0.001 50.074 50.628 mp_waitall_1 214728 16.6 23.888 39.445 23.888 39.445 multiply_cannon_multrec 30084 15.6 21.067 26.634 31.774 37.886 rebuild_ks_matrix 128 8.3 0.001 0.001 32.111 32.847 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 32.110 32.847 init_scf_loop 11 6.9 0.000 0.000 32.250 32.251 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.861 29.536 multiply_cannon_metrocomm3 30084 15.6 0.098 0.104 15.662 29.278 qs_ot_get_derivative 117 11.6 0.001 0.002 28.129 28.755 prepare_preconditioner 11 7.9 0.000 0.000 27.961 28.036 make_preconditioner 11 8.9 0.000 0.000 27.961 28.036 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.697 27.243 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.050 23.275 apply_single 128 13.6 0.001 0.001 22.049 23.275 ot_diis_step 117 11.6 0.014 0.015 21.769 21.771 qs_ot_get_p 128 10.4 0.001 0.001 20.619 21.419 multiply_cannon_sync_h2d 30084 15.6 18.079 20.976 18.079 20.976 cp_fm_cholesky_invert 11 10.9 16.311 16.323 16.311 16.323 qs_ot_p2m_diag 83 11.4 0.188 0.217 15.729 15.763 make_m2s 5014 13.6 0.086 0.091 13.978 15.661 make_images 5014 14.6 1.169 1.369 13.763 15.446 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.533 14.534 sum_up_and_integrate 128 10.3 0.002 0.004 13.880 13.912 integrate_v_rspace 128 11.3 0.003 0.003 13.820 13.853 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.246 13.281 calculate_rho_elec 128 8.7 0.087 0.103 13.245 13.281 init_scf_run 11 5.9 0.000 0.001 11.819 11.821 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.819 11.821 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.156 11.605 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.414 11.446 cp_fm_diag_elpa_base 83 14.4 11.163 11.270 11.408 11.435 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.807 11.142 dbcsr_mm_accdrv_process 62242 16.2 5.500 6.264 10.152 10.921 mp_irecv_dv 69486 16.3 3.608 10.738 3.608 10.738 make_images_data 5014 15.6 0.067 0.076 8.213 10.268 hybrid_alltoall_any 5200 16.5 0.349 1.529 6.908 9.736 pw_transfer 1547 11.6 0.084 0.097 8.464 8.532 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.240 8.311 wfi_extrapolate 11 7.9 0.001 0.001 8.278 8.278 grid_integrate_task_list 128 12.3 7.162 7.588 7.162 7.588 density_rs2pw 128 9.7 0.006 0.006 6.950 7.574 fft_wrap_pw1pw2_140 523 13.2 0.932 0.955 7.320 7.413 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.256 7.071 cp_fm_cholesky_decompose 22 10.9 6.812 6.893 6.812 6.893 calculate_dm_sparse 128 9.5 0.001 0.001 6.445 6.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.464 6.472 mp_sum_l 7950 12.9 4.139 6.275 4.139 6.275 fft3d_ps 1291 14.7 2.827 3.001 5.796 5.842 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.519 5.751 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.354 5.418 grid_collocate_task_list 128 9.7 5.005 5.405 5.005 5.405 mp_alltoall_d11v 2415 14.1 4.079 4.883 4.079 4.883 mp_allgather_i34 2507 14.6 1.857 4.880 1.857 4.880 potential_pw2rs 128 12.3 0.015 0.017 4.369 4.382 mp_sum_d 4467 12.1 2.638 3.953 2.638 3.953 dbcsr_complete_redistribute 395 12.7 0.780 0.881 3.094 3.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.973000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 977.448960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.109 176.436 176.438 qs_mol_dyn_low 1 2.0 0.003 0.005 175.831 175.870 qs_forces 11 3.9 0.006 0.024 175.530 175.533 qs_energies 11 4.9 0.007 0.028 168.912 168.937 scf_env_do_scf 11 5.9 0.002 0.008 152.817 152.820 scf_env_do_scf_inner_loop 118 6.6 0.010 0.034 117.980 117.980 velocity_verlet 10 3.0 0.002 0.002 113.066 113.075 dbcsr_multiply_generic 2527 12.6 0.184 0.191 82.881 84.123 qs_scf_new_mos 118 7.6 0.001 0.001 80.479 80.843 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.478 80.842 ot_scf_mini 118 9.6 0.004 0.004 76.184 76.608 multiply_cannon 2527 13.6 0.513 0.532 62.873 68.769 multiply_cannon_loop 2527 14.6 1.145 1.207 59.821 62.320 ot_mini 118 10.6 0.001 0.001 42.876 43.302 mp_waitall_1 171878 16.6 25.023 35.155 25.023 35.155 init_scf_loop 11 6.9 0.001 0.005 34.735 34.735 rebuild_ks_matrix 129 8.3 0.001 0.001 30.624 31.083 qs_ks_build_kohn_sham_matrix 129 9.3 0.020 0.051 30.624 31.082 prepare_preconditioner 11 7.9 0.000 0.000 30.689 30.740 make_preconditioner 11 8.9 0.001 0.008 30.689 30.740 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.349 29.727 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.572 27.997 multiply_cannon_multrec 20216 15.6 13.191 16.118 22.669 25.604 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.445 24.883 qs_ot_get_derivative 118 11.6 0.002 0.002 23.148 23.581 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.766 20.868 apply_single 129 13.6 0.001 0.001 19.766 20.868 ot_diis_step 118 11.6 0.018 0.024 19.617 19.619 qs_ot_get_p 129 10.4 0.007 0.024 18.441 18.942 make_m2s 5054 13.6 0.078 0.084 14.909 16.244 make_images 5054 14.6 1.160 1.248 14.671 15.991 multiply_cannon_sync_h2d 20216 15.6 13.703 15.553 13.703 15.553 cp_fm_cholesky_invert 11 10.9 14.247 14.256 14.247 14.256 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.097 14.112 sum_up_and_integrate 129 10.3 0.002 0.004 14.019 14.044 integrate_v_rspace 129 11.3 0.003 0.004 13.958 13.985 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.677 13.720 calculate_rho_elec 129 8.7 0.130 0.144 13.676 13.719 cp_dbcsr_syevd 83 12.4 0.008 0.015 13.052 13.058 make_images_data 5054 15.6 0.065 0.075 9.030 10.726 init_scf_run 11 5.9 0.001 0.002 10.650 10.650 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.650 10.650 hybrid_alltoall_any 5240 16.5 0.449 2.028 7.835 10.347 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.880 9.903 cp_fm_diag_elpa_base 83 14.4 9.471 9.634 9.876 9.899 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.060 9.385 multiply_cannon_metrocomm4 17689 15.6 0.066 0.075 3.485 9.251 dbcsr_mm_accdrv_process 41830 16.2 5.763 6.061 8.931 9.096 mp_irecv_dv 50625 16.2 3.356 8.989 3.356 8.989 pw_transfer 1559 11.6 0.085 0.102 8.861 8.965 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.637 8.748 fft_wrap_pw1pw2_140 527 13.2 1.075 1.116 7.687 7.804 grid_integrate_task_list 129 12.3 7.329 7.760 7.329 7.760 wfi_extrapolate 11 7.9 0.001 0.001 7.438 7.438 density_rs2pw 129 9.7 0.005 0.006 6.917 7.433 cp_fm_cholesky_decompose 22 10.9 7.323 7.387 7.323 7.387 cp_fm_upper_to_full 105 14.8 5.705 7.195 5.705 7.195 dbcsr_complete_redistribute 395 12.7 1.177 1.208 4.524 6.245 calculate_dm_sparse 129 9.5 0.001 0.001 5.964 6.065 fft3d_ps 1301 14.7 2.850 3.121 5.924 6.046 grid_collocate_task_list 129 9.7 5.192 5.741 5.192 5.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.497 5.506 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.748 5.394 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.388 5.113 mp_sum_l 8010 12.9 3.269 5.012 3.269 5.012 mp_allgather_i34 2527 14.6 1.540 4.975 1.540 4.975 mp_alltoall_d11v 2423 14.1 4.423 4.842 4.423 4.842 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.600 4.718 potential_pw2rs 129 12.3 0.020 0.022 4.353 4.363 qs_energies_init_hamiltonians 11 5.9 0.010 0.074 4.096 4.096 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.320 4.024 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.929 3.954 mp_alltoall_i22 718 14.1 1.945 3.781 1.945 3.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.553 3.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.438000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=925.363636, yerr=17.264113 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.154617E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57809. MP_Allreduce 11152 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.033 188.381 188.382 qs_mol_dyn_low 1 2.0 0.003 0.004 187.887 187.911 qs_forces 11 3.9 0.007 0.011 186.956 186.968 qs_energies 11 4.9 0.002 0.002 179.846 179.855 scf_env_do_scf 11 5.9 0.001 0.001 163.162 163.172 velocity_verlet 10 3.0 0.026 0.031 125.658 125.696 scf_env_do_scf_inner_loop 118 6.6 0.012 0.021 116.822 116.823 dbcsr_multiply_generic 2527 12.6 0.193 0.198 80.760 81.668 qs_scf_new_mos 118 7.6 0.001 0.001 80.330 80.672 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.329 80.671 ot_scf_mini 118 9.6 0.003 0.003 75.736 76.065 multiply_cannon 2527 13.6 0.556 0.593 56.388 59.395 multiply_cannon_loop 2527 14.6 1.876 1.956 52.499 54.493 init_scf_loop 11 6.9 0.000 0.000 46.218 46.219 ot_mini 118 10.6 0.001 0.001 43.213 43.536 prepare_preconditioner 11 7.9 0.000 0.000 42.138 42.162 make_preconditioner 11 8.9 0.000 0.000 42.138 42.162 make_full_inverse_cholesky 11 9.9 0.012 0.021 35.698 40.800 multiply_cannon_multrec 30324 15.6 13.602 18.868 26.668 31.537 rebuild_ks_matrix 129 8.3 0.001 0.001 29.435 29.736 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 29.434 29.735 mp_waitall_1 149060 16.7 16.910 26.936 16.910 26.936 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.578 26.849 qs_ot_get_derivative 118 11.6 0.002 0.002 23.281 23.612 make_m2s 5054 13.6 0.092 0.097 19.965 21.195 make_images 5054 14.6 1.964 2.199 19.654 20.881 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.320 20.081 apply_single 129 13.6 0.001 0.001 19.319 20.081 ot_diis_step 118 11.6 0.018 0.019 19.806 19.808 qs_ot_get_p 129 10.4 0.001 0.001 18.624 18.977 cp_fm_upper_to_full 105 14.8 11.668 17.172 11.668 17.172 cp_fm_cholesky_invert 11 10.9 15.943 15.952 15.943 15.952 multiply_cannon_metrocomm3 30324 15.6 0.049 0.051 6.458 15.006 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.491 14.543 sum_up_and_integrate 129 10.3 0.002 0.003 14.062 14.088 integrate_v_rspace 129 11.3 0.003 0.004 14.002 14.030 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.818 13.867 calculate_rho_elec 129 8.7 0.171 0.186 13.817 13.867 dbcsr_mm_accdrv_process 62756 16.2 8.659 9.398 12.632 13.233 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.089 13.091 dbcsr_complete_redistribute 395 12.7 1.505 1.627 8.992 12.831 make_images_data 5054 15.6 0.066 0.074 10.594 12.358 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.655 11.499 multiply_cannon_sync_h2d 30324 15.6 10.461 11.483 10.461 11.483 hybrid_alltoall_any 5240 16.5 0.525 2.205 9.644 11.214 init_scf_run 11 5.9 0.000 0.001 10.725 10.727 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.725 10.726 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.418 10.162 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.034 10.046 cp_fm_diag_elpa_base 83 14.4 9.062 9.392 10.027 10.042 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.685 9.928 mp_alltoall_i22 718 14.1 5.710 9.591 5.710 9.591 pw_transfer 1559 11.6 0.086 0.101 9.249 9.326 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.021 9.102 fft_wrap_pw1pw2_140 527 13.2 1.262 1.309 8.038 8.134 grid_integrate_task_list 129 12.3 7.583 7.916 7.583 7.916 wfi_extrapolate 11 7.9 0.001 0.001 7.564 7.564 multiply_cannon_metrocomm4 25270 15.6 0.080 0.090 2.865 7.516 cp_fm_cholesky_decompose 22 10.9 7.374 7.455 7.374 7.455 mp_irecv_dv 76685 16.2 2.712 7.244 2.712 7.244 density_rs2pw 129 9.7 0.005 0.006 6.843 7.085 calculate_dm_sparse 129 9.5 0.001 0.001 6.267 6.373 fft3d_ps 1301 14.7 3.023 3.085 6.039 6.091 grid_collocate_task_list 129 9.7 5.323 5.608 5.323 5.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.441 5.499 mp_alltoall_d11v 2423 14.1 4.686 5.190 4.686 5.190 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.549 4.642 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.468 4.471 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.341 4.452 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.275 4.341 potential_pw2rs 129 12.3 0.023 0.024 4.306 4.312 mp_allgather_i34 2527 14.6 1.960 3.937 1.960 3.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.382000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1097.727273, yerr=13.511550 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.560744E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108700816 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828148. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.068 171.356 171.363 qs_mol_dyn_low 1 2.0 0.003 0.004 170.895 170.914 qs_forces 11 3.9 0.004 0.004 170.799 170.801 qs_energies 11 4.9 0.002 0.002 163.294 163.301 scf_env_do_scf 11 5.9 0.001 0.001 146.032 146.047 velocity_verlet 10 3.0 0.002 0.002 111.766 111.770 scf_env_do_scf_inner_loop 116 6.6 0.015 0.019 110.017 110.019 qs_scf_new_mos 116 7.6 0.001 0.001 73.050 73.175 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.050 73.174 dbcsr_multiply_generic 2485 12.5 0.181 0.186 71.956 72.445 ot_scf_mini 116 9.6 0.003 0.004 68.678 68.790 multiply_cannon 2485 13.5 0.555 0.573 53.116 56.496 multiply_cannon_loop 2485 14.5 0.803 0.851 50.100 51.121 ot_mini 116 10.6 0.001 0.001 37.625 37.736 init_scf_loop 11 6.9 0.000 0.000 35.863 35.866 prepare_preconditioner 11 7.9 0.000 0.000 31.946 31.979 make_preconditioner 11 8.9 0.000 0.000 31.946 31.979 mp_waitall_1 124680 16.7 25.086 31.737 25.086 31.737 make_full_inverse_cholesky 11 9.9 0.012 0.025 29.893 30.159 rebuild_ks_matrix 127 8.3 0.001 0.001 29.138 29.279 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.137 29.279 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.558 26.688 multiply_cannon_multrec 9940 15.5 10.226 14.882 17.800 22.882 qs_ot_get_derivative 116 11.6 0.002 0.002 20.570 20.673 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.809 20.373 qs_ot_get_p 127 10.4 0.001 0.001 18.353 18.465 cp_fm_cholesky_invert 11 10.9 18.357 18.363 18.357 18.363 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.097 17.356 apply_single 127 13.6 0.001 0.001 17.096 17.356 ot_diis_step 116 11.6 0.020 0.020 16.984 16.984 make_m2s 4970 13.5 0.063 0.068 14.951 15.831 make_images 4970 14.5 2.143 2.553 14.646 15.532 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.531 14.548 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.223 14.251 calculate_rho_elec 127 8.7 0.251 0.262 14.222 14.251 sum_up_and_integrate 127 10.3 0.002 0.002 14.165 14.210 integrate_v_rspace 127 11.3 0.004 0.004 14.105 14.152 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.360 13.361 multiply_cannon_sync_h2d 9940 15.5 10.544 11.253 10.544 11.253 init_scf_run 11 5.9 0.000 0.001 10.460 10.460 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.459 10.460 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.426 10.435 cp_fm_diag_elpa_base 82 14.4 10.186 10.263 10.422 10.432 pw_transfer 1535 11.6 0.084 0.092 9.869 9.901 make_images_data 4970 15.5 0.054 0.065 8.664 9.893 hybrid_alltoall_any 5155 16.4 0.833 3.754 8.442 9.860 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.648 9.686 fft_wrap_pw1pw2_140 519 13.2 1.656 1.698 8.524 8.571 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.086 8.164 cp_fm_cholesky_decompose 22 10.9 8.001 8.105 8.001 8.105 grid_integrate_task_list 127 12.3 7.703 8.033 7.703 8.033 dbcsr_mm_accdrv_process 20590 16.1 3.269 4.489 7.236 7.969 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.338 7.730 density_rs2pw 127 9.7 0.005 0.005 6.954 7.348 wfi_extrapolate 11 7.9 0.001 0.001 7.345 7.345 fft3d_ps 1281 14.7 3.111 3.214 6.132 6.166 calculate_dm_sparse 127 9.5 0.001 0.001 6.014 6.104 multiply_cannon_metrocomm4 7455 15.5 0.025 0.029 1.848 5.926 grid_collocate_task_list 127 9.7 5.524 5.846 5.524 5.846 mp_irecv_dv 28618 15.9 1.809 5.832 1.809 5.832 dbcsr_complete_redistribute 393 12.7 2.066 2.129 5.220 5.575 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.242 5.242 mp_alltoall_d11v 2401 14.1 4.657 5.234 4.657 5.234 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.133 5.143 mp_allgather_i34 2485 14.5 1.218 4.754 1.218 4.754 potential_pw2rs 127 12.3 0.026 0.026 4.400 4.408 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.052 4.128 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.600 3.964 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.607 3.686 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.620 3.661 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.331 3.647 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.465 3.548 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.474 3.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.363000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1480.727273, yerr=11.201387 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430451062784 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986258751488 0.0% 0.0% 100.0% flops 32 x 32 x 32 1986422374400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1991997444096 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 11613050155008 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 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 94.256601E+12 0.0% 0.0% 100.0% flops max/rank 11.789007E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806737984 0.0% 0.0% 100.0% number of processed stacks 1984336 0.0% 0.0% 100.0% average stack size 0.0 0.0 3430.2 marketing flops 145.674553E+12 ------------------------------------------------------------------------------- # multiplications 2540 max memory usage/rank 3.176980E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101600 MPI messages size (bytes): total size 1.145338E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.273011E+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 46088 35802578944 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592076688 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4035 58670. MP_Allreduce 11164 1500. MP_Sync 88 MP_Alltoall 1724 36993736. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.085 295.838 295.861 qs_mol_dyn_low 1 2.0 0.004 0.009 294.976 295.019 qs_forces 11 3.9 0.013 0.015 294.770 294.773 qs_energies 11 4.9 0.002 0.002 285.878 285.883 scf_env_do_scf 11 5.9 0.001 0.001 263.266 263.269 velocity_verlet 10 3.0 0.002 0.002 213.418 213.439 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 138.436 138.438 init_scf_loop 11 6.9 0.000 0.000 124.557 124.562 prepare_preconditioner 11 7.9 0.000 0.000 119.654 119.680 make_preconditioner 11 8.9 0.000 0.000 119.654 119.680 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.936 116.785 qs_scf_new_mos 118 7.6 0.001 0.001 93.014 93.145 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.014 93.144 ot_scf_mini 118 9.6 0.004 0.004 88.045 88.134 dbcsr_multiply_generic 2540 12.6 0.217 0.227 83.967 84.805 cp_fm_upper_to_full 106 14.8 53.310 76.006 53.310 76.006 multiply_cannon 2540 13.6 0.689 0.729 60.196 61.163 multiply_cannon_loop 2540 14.6 1.063 1.073 55.882 57.989 ot_mini 118 10.6 0.001 0.001 45.658 45.778 dbcsr_complete_redistribute 397 12.7 3.972 4.020 29.680 42.842 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.151 39.340 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.676 36.721 rebuild_ks_matrix 129 8.3 0.001 0.001 35.225 35.376 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 35.224 35.376 mp_alltoall_i22 720 14.1 21.459 34.496 21.459 34.496 mp_waitall_1 105020 16.8 28.935 34.314 28.935 34.314 cp_fm_cholesky_invert 11 10.9 33.556 33.563 33.556 33.563 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.689 32.826 qs_ot_get_p 129 10.4 0.001 0.001 27.009 27.130 qs_ot_get_derivative 118 11.6 0.002 0.002 25.135 25.221 qs_ot_p2m_diag 84 11.4 0.891 0.895 22.710 22.742 multiply_cannon_metrocomm3 10160 15.6 0.025 0.026 19.417 21.526 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.877 20.878 make_m2s 5080 13.6 0.074 0.077 18.920 20.621 ot_diis_step 118 11.6 0.022 0.023 20.477 20.478 make_images 5080 14.6 3.094 3.283 18.435 20.138 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.626 19.986 apply_single 129 13.6 0.001 0.001 19.625 19.986 multiply_cannon_multrec 10160 15.6 10.436 12.325 18.348 18.407 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.604 17.605 cp_fm_diag_elpa_base 84 14.4 13.087 14.678 17.600 17.600 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.471 17.485 calculate_rho_elec 129 8.7 0.481 0.482 17.471 17.484 sum_up_and_integrate 129 10.3 0.002 0.002 16.589 16.678 integrate_v_rspace 129 11.3 0.004 0.004 16.526 16.614 multiply_cannon_sync_h2d 10160 15.6 14.409 14.436 14.409 14.436 hybrid_alltoall_any 5267 16.5 1.318 3.070 11.029 13.561 pw_transfer 1559 11.6 0.096 0.097 13.327 13.337 make_images_data 5080 15.6 0.065 0.070 10.768 13.191 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.087 13.098 init_scf_run 11 5.9 0.000 0.001 12.230 12.231 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.230 12.230 fft_wrap_pw1pw2_140 527 13.2 3.241 3.280 11.642 11.657 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.009 10.076 dbcsr_mm_accdrv_process 20978 16.1 4.304 6.105 7.658 9.558 wfi_extrapolate 11 7.9 0.001 0.001 9.044 9.045 cp_fm_cholesky_decompose 22 10.9 8.975 8.998 8.975 8.998 grid_integrate_task_list 129 12.3 8.623 8.810 8.623 8.810 density_rs2pw 129 9.7 0.005 0.005 8.428 8.477 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.222 8.223 fft3d_ps 1301 14.7 4.000 4.015 7.512 7.561 mp_alltoall_d11v 2429 14.1 6.905 7.036 6.905 7.036 calculate_dm_sparse 129 9.5 0.001 0.001 6.849 6.907 grid_collocate_task_list 129 9.7 6.469 6.545 6.469 6.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.435 6.463 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.227 6.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.861000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2894.181818, yerr=155.392651 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 1.260863E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 34131831. 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 0.064 0.101 85.003 85.003 qs_energies 1 2.0 0.000 0.000 84.344 84.357 ls_scf 1 3.0 0.000 0.000 83.437 83.445 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.457 72.640 multiply_cannon 111 7.7 0.017 0.020 55.814 57.233 multiply_cannon_loop 111 8.7 0.229 0.246 52.444 54.019 ls_scf_main 1 4.0 0.000 0.000 52.190 52.190 density_matrix_trs4 2 5.0 0.002 0.003 46.641 46.725 ls_scf_init_scf 1 4.0 0.000 0.000 28.186 28.187 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.042 27.096 mp_waitall_1 11031 10.9 22.312 25.431 22.312 25.431 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.953 24.971 multiply_cannon_multrec 2664 9.7 8.144 8.941 15.592 17.419 multiply_cannon_sync_h2d 2664 9.7 13.557 16.006 13.557 16.006 make_m2s 222 7.7 0.009 0.011 12.983 13.407 make_images 222 8.7 0.099 0.108 12.961 13.385 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.693 12.711 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.402 8.057 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.706 7.065 8.050 make_images_data 222 9.7 0.004 0.005 7.544 8.039 hybrid_alltoall_any 227 10.6 0.215 1.847 6.539 7.688 dbcsr_mm_accdrv_process_sort 4760 11.4 6.272 7.170 6.272 7.170 calculate_norms 4752 9.8 5.501 6.283 5.501 6.283 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.188 mp_sum_l 887 5.1 3.125 4.541 3.125 4.541 make_images_sizes 222 9.7 0.000 0.000 0.680 3.777 mp_alltoall_i44 222 10.7 0.680 3.776 0.680 3.776 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.036 3.413 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.319 3.380 mp_irecv_dv 6231 10.9 2.020 3.379 2.020 3.379 arnoldi_extremal 4 6.8 0.000 0.000 3.187 3.211 arnoldi_normal_ev 4 7.8 0.001 0.003 3.187 3.211 build_subspace 16 8.4 0.009 0.012 3.097 3.098 ls_scf_post 1 4.0 0.000 0.000 3.061 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.880 2.917 dbcsr_special_finalize 555 9.7 0.005 0.006 2.405 2.793 dbcsr_merge_single_wm 555 10.7 0.458 0.584 2.397 2.784 make_images_pack 222 9.7 2.207 2.622 2.209 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.561 dbcsr_sort_data 658 11.4 2.196 2.548 2.196 2.548 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.457 2.069 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.312 2.398 buffer_matrices_ensure_size 222 8.7 1.753 2.065 1.753 2.065 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.786 1.788 rebuild_ks_matrix 3 7.3 0.000 0.000 1.777 1.778 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.777 1.778 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.003000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 2.168627E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.105 90.595 90.596 qs_energies 1 2.0 0.001 0.012 89.951 90.005 ls_scf 1 3.0 0.003 0.025 88.567 88.613 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.170 74.457 multiply_cannon 111 7.7 0.027 0.039 52.390 57.370 ls_scf_main 1 4.0 0.001 0.022 54.575 54.590 multiply_cannon_loop 111 8.7 0.136 0.147 49.600 53.438 density_matrix_trs4 2 5.0 0.003 0.013 48.833 49.098 ls_scf_init_scf 1 4.0 0.000 0.002 30.341 30.364 mp_waitall_1 9105 10.9 20.284 30.059 20.284 30.059 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.931 29.049 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.563 26.577 multiply_cannon_multrec 1332 9.7 13.423 16.527 22.782 26.557 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 10.937 20.890 make_m2s 222 7.7 0.006 0.008 15.197 15.738 make_images 222 8.7 1.373 1.694 15.167 15.708 dbcsr_mm_accdrv_process 4041 10.4 0.363 0.531 8.949 10.563 dbcsr_mm_accdrv_process_sort 4041 11.4 8.463 10.024 8.463 10.024 hybrid_alltoall_any 227 10.6 0.542 2.562 8.102 9.733 make_images_data 222 9.7 0.004 0.005 8.811 9.631 mp_sum_l 887 5.1 5.117 8.190 5.117 8.190 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.155 7.678 mp_irecv_dv 3311 11.0 3.135 7.615 3.135 7.615 calculate_norms 2376 9.8 6.082 6.701 6.082 6.701 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.887 6.492 multiply_cannon_sync_h2d 1332 9.7 4.945 6.141 4.945 6.141 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.942 5.181 arnoldi_extremal 4 6.8 0.000 0.000 4.628 4.652 arnoldi_normal_ev 4 7.8 0.003 0.013 4.628 4.651 build_subspace 16 8.4 0.014 0.021 4.371 4.374 ls_scf_post 1 4.0 0.002 0.017 3.648 3.689 ls_scf_store_result 1 5.0 0.000 0.000 3.388 3.465 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.156 3.381 dbcsr_matrix_vector_mult_local 304 10.0 2.763 3.246 2.766 3.248 mp_allgather_i34 111 8.7 0.772 3.021 0.772 3.021 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.618 2.690 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.187 2.544 dbcsr_data_new 4174 10.1 2.117 2.428 2.117 2.428 make_images_pack 222 9.7 1.824 2.131 1.827 2.133 dbcsr_sort_data 436 11.2 1.833 2.073 1.833 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.047 2.071 rebuild_ks_matrix 3 7.3 0.000 0.000 2.035 2.058 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.120 2.035 2.058 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.596000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 2.877649E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.042 92.685 92.686 qs_energies 1 2.0 0.000 0.000 92.140 92.143 ls_scf 1 3.0 0.000 0.000 90.301 90.306 dbcsr_multiply_generic 111 6.7 0.015 0.017 74.445 74.728 multiply_cannon 111 7.7 0.035 0.086 51.665 55.899 ls_scf_main 1 4.0 0.000 0.000 55.831 55.836 multiply_cannon_loop 111 8.7 0.116 0.131 48.919 52.569 density_matrix_trs4 2 5.0 0.002 0.003 49.973 50.117 mp_waitall_1 7281 11.0 22.997 32.590 22.997 32.590 ls_scf_init_scf 1 4.0 0.000 0.000 30.831 30.834 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.638 29.717 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.806 26.819 multiply_cannon_multrec 888 9.7 12.773 15.255 21.428 24.697 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.702 22.095 make_m2s 222 7.7 0.006 0.008 16.353 17.063 make_images 222 8.7 1.585 1.857 16.315 17.025 make_images_data 222 9.7 0.004 0.004 9.547 10.535 hybrid_alltoall_any 227 10.6 0.638 2.952 9.060 10.381 dbcsr_mm_accdrv_process 3754 10.4 0.318 0.496 8.169 9.422 dbcsr_mm_accdrv_process_sort 3754 11.4 7.732 8.926 7.732 8.926 mp_sum_l 887 5.1 4.730 7.640 4.730 7.640 multiply_cannon_sync_h2d 888 9.7 6.154 7.553 6.154 7.553 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.395 7.117 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.388 7.090 mp_irecv_dv 2335 11.1 2.380 7.063 2.380 7.063 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.672 6.076 arnoldi_extremal 4 6.8 0.000 0.000 5.200 5.218 arnoldi_normal_ev 4 7.8 0.046 0.063 5.200 5.218 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.697 4.906 build_subspace 16 8.4 0.014 0.020 4.847 4.852 calculate_norms 1584 9.8 4.359 4.752 4.359 4.752 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.469 3.821 mp_allgather_i34 111 8.7 0.864 3.760 0.864 3.760 ls_scf_post 1 4.0 0.000 0.000 3.639 3.643 dbcsr_matrix_vector_mult_local 304 10.0 3.045 3.636 3.047 3.638 ls_scf_store_result 1 5.0 0.000 0.000 3.371 3.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.844 2.949 dbcsr_data_new 4116 9.9 2.104 2.452 2.104 2.452 dbcsr_sort_data 325 11.1 1.896 2.132 1.896 2.132 make_images_sizes 222 9.7 0.000 0.000 0.948 2.118 mp_alltoall_i44 222 10.7 0.947 2.118 0.947 2.118 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 make_images_pack 222 9.7 1.622 1.880 1.625 1.883 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.686000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2240.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 3.335578E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.053 98.264 98.266 qs_energies 1 2.0 0.000 0.000 97.641 97.647 ls_scf 1 3.0 0.000 0.000 95.714 95.721 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.899 79.206 ls_scf_main 1 4.0 0.000 0.000 59.304 59.305 multiply_cannon 111 7.7 0.053 0.125 51.913 56.625 density_matrix_trs4 2 5.0 0.002 0.005 53.155 53.285 multiply_cannon_loop 111 8.7 0.154 0.170 46.876 50.285 ls_scf_init_scf 1 4.0 0.000 0.000 33.129 33.131 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.920 31.990 mp_waitall_1 6369 11.0 22.921 29.151 22.921 29.151 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.990 29.002 multiply_cannon_multrec 1332 9.7 14.090 17.441 22.063 24.481 make_m2s 222 7.7 0.006 0.007 21.213 22.621 make_images 222 8.7 3.145 3.606 21.163 22.573 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.236 17.810 make_images_data 222 9.7 0.004 0.004 11.806 13.320 hybrid_alltoall_any 227 10.6 0.798 3.768 11.267 12.992 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.479 7.607 9.142 dbcsr_mm_accdrv_process_sort 3641 11.4 7.165 8.644 7.165 8.644 mp_sum_l 887 5.1 4.209 8.274 4.209 8.274 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.230 6.601 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.103 6.165 mp_irecv_dv 3229 10.9 2.079 6.120 2.079 6.120 multiply_cannon_sync_h2d 1332 9.7 5.447 5.960 5.447 5.960 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.712 5.384 arnoldi_extremal 4 6.8 0.000 0.000 5.249 5.264 arnoldi_normal_ev 4 7.8 0.018 0.022 5.249 5.264 build_subspace 16 8.4 0.014 0.021 4.896 4.906 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.571 4.750 calculate_norms 2376 9.8 4.182 4.520 4.182 4.520 mp_allgather_i34 111 8.7 2.126 4.417 2.126 4.417 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.602 3.900 dbcsr_matrix_vector_mult_local 304 10.0 3.213 3.714 3.215 3.716 dbcsr_sort_data 658 11.4 3.097 3.496 3.097 3.496 ls_scf_post 1 4.0 0.000 0.000 3.281 3.287 dbcsr_special_finalize 555 9.7 0.006 0.007 2.858 3.263 dbcsr_merge_single_wm 555 10.7 0.543 0.670 2.849 3.255 ls_scf_store_result 1 5.0 0.000 0.000 3.017 3.085 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.016 3.053 dbcsr_data_release 10477 10.7 1.583 2.403 1.583 2.403 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.266000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 4.729721E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.060 93.260 93.261 qs_energies 1 2.0 0.000 0.000 92.344 92.351 ls_scf 1 3.0 0.000 0.000 89.947 89.952 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.509 71.736 ls_scf_main 1 4.0 0.000 0.000 57.214 57.214 multiply_cannon 111 7.7 0.076 0.136 52.938 55.994 multiply_cannon_loop 111 8.7 0.088 0.093 50.322 52.008 density_matrix_trs4 2 5.0 0.002 0.003 50.239 50.329 ls_scf_init_scf 1 4.0 0.000 0.000 29.340 29.342 mp_waitall_1 5436 11.0 24.974 29.027 24.974 29.027 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.099 28.139 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.013 26.024 multiply_cannon_multrec 444 9.7 13.736 16.201 20.872 22.163 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.029 16.002 make_m2s 222 7.7 0.005 0.005 13.807 14.784 make_images 222 8.7 2.040 2.481 13.740 14.716 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.213 14.716 hybrid_alltoall_any 227 10.6 0.802 3.836 8.287 10.066 make_images_data 222 9.7 0.003 0.004 8.548 9.808 multiply_cannon_sync_h2d 444 9.7 6.733 8.168 6.733 8.168 dbcsr_mm_accdrv_process 3003 10.4 0.364 0.410 6.837 7.957 dbcsr_mm_accdrv_process_sort 3003 11.4 6.473 7.560 6.473 7.560 arnoldi_extremal 4 6.8 0.000 0.000 5.874 5.887 arnoldi_normal_ev 4 7.8 0.002 0.005 5.874 5.887 build_subspace 16 8.4 0.015 0.020 5.476 5.484 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.401 4.586 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.207 4.408 mp_sum_l 887 5.1 2.755 4.390 2.755 4.390 dbcsr_matrix_vector_mult_local 304 10.0 3.737 4.207 3.739 4.210 mp_allgather_i34 111 8.7 1.189 3.817 1.189 3.817 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.523 3.808 mp_irecv_dv 1241 11.2 1.508 3.772 1.508 3.772 calculate_norms 792 9.8 3.617 3.723 3.617 3.723 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.534 3.606 make_images_sizes 222 9.7 0.000 0.000 0.872 3.417 mp_alltoall_i44 222 10.7 0.871 3.416 0.871 3.416 ls_scf_post 1 4.0 0.000 0.000 3.393 3.399 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.969 3.296 ls_scf_store_result 1 5.0 0.000 0.000 3.182 3.218 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.381 2.383 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.278 dbcsr_data_new 4608 9.7 1.799 2.244 1.799 2.244 dbcsr_merge_all 275 8.9 0.478 0.524 2.054 2.124 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.030 rebuild_ks_matrix 3 7.3 0.000 0.000 1.998 1.999 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.998 1.999 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.261000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_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 8.826683E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.127 105.410 105.411 qs_energies 1 2.0 0.000 0.000 104.031 104.033 ls_scf 1 3.0 0.000 0.000 100.923 100.925 dbcsr_multiply_generic 111 6.7 0.023 0.028 74.371 74.506 ls_scf_main 1 4.0 0.000 0.000 63.636 63.636 density_matrix_trs4 2 5.0 0.002 0.003 54.588 54.638 multiply_cannon 111 7.7 0.188 0.217 48.311 50.390 multiply_cannon_loop 111 8.7 0.098 0.099 45.276 45.982 ls_scf_init_scf 1 4.0 0.000 0.000 33.505 33.507 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.952 31.978 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.258 29.266 mp_waitall_1 4527 11.1 21.437 25.198 21.437 25.198 make_m2s 222 7.7 0.005 0.005 22.409 23.505 make_images 222 8.7 3.568 3.876 22.301 23.397 multiply_cannon_multrec 444 9.7 17.843 18.581 22.524 23.268 hybrid_alltoall_any 227 10.6 1.650 3.612 12.605 15.311 make_images_data 222 9.7 0.004 0.004 12.805 14.958 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.051 10.340 multiply_cannon_sync_h2d 444 9.7 8.784 8.844 8.784 8.844 arnoldi_extremal 4 6.8 0.000 0.000 7.506 7.524 arnoldi_normal_ev 4 7.8 0.014 0.037 7.506 7.524 build_subspace 16 8.4 0.026 0.036 6.934 6.950 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.547 5.712 dbcsr_matrix_vector_mult_local 304 10.0 5.099 5.423 5.102 5.426 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.191 5.289 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.839 5.106 dbcsr_mm_accdrv_process 1814 10.4 0.291 0.360 4.496 4.625 dbcsr_mm_accdrv_process_sort 1814 11.4 4.158 4.295 4.158 4.295 ls_scf_post 1 4.0 0.000 0.000 3.782 3.784 make_images_sizes 222 9.7 0.000 0.000 1.470 3.599 mp_alltoall_i44 222 10.7 1.470 3.599 1.470 3.599 mp_allgather_i34 111 8.7 1.136 3.572 1.136 3.572 ls_scf_store_result 1 5.0 0.000 0.000 3.481 3.517 calculate_norms 792 9.8 3.236 3.290 3.236 3.290 dbcsr_finalize 304 7.8 0.082 0.089 3.065 3.164 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 3.077 3.077 dbcsr_merge_all 275 8.9 0.883 0.921 2.850 2.943 dbcsr_complete_redistribute 5 7.6 1.430 1.470 2.742 2.856 matrix_ls_to_qs 2 6.0 0.000 0.000 2.394 2.527 dbcsr_sort_data 325 11.1 2.435 2.500 2.435 2.500 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.463 2.465 rebuild_ks_matrix 3 7.3 0.000 0.000 2.397 2.399 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.397 2.399 dbcsr_new_transposed 4 7.5 0.242 0.250 2.281 2.293 dbcsr_data_new 6591 9.6 1.834 2.289 1.834 2.289 dbcsr_frobenius_norm 74 6.6 2.057 2.131 2.209 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.207 dbcsr_add_anytype 103 7.2 0.859 0.893 2.124 2.207 dbcsr_data_release 12724 10.6 1.983 2.190 1.983 2.190 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.411000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7038.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 594.071552E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4770526. 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 0.044 0.107 226.815 226.818 qs_mol_dyn_low 1 2.0 0.004 0.019 225.658 225.718 qs_forces 5 3.8 0.005 0.013 225.319 225.320 qs_energies 5 4.8 0.004 0.063 222.245 222.293 scf_env_do_scf 5 5.8 0.002 0.034 207.690 207.693 scf_env_do_scf_inner_loop 105 6.6 0.007 0.056 180.402 180.437 qs_scf_new_mos 105 7.6 0.000 0.001 141.046 141.225 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.045 141.224 dbcsr_multiply_generic 1445 12.2 0.127 0.136 133.490 134.014 ot_scf_mini 105 9.6 0.003 0.014 131.218 131.352 multiply_cannon 1445 13.2 0.273 0.283 114.502 116.660 multiply_cannon_loop 1445 14.2 2.876 3.048 112.758 113.988 velocity_verlet 4 3.0 0.029 0.134 105.901 105.905 ot_mini 105 10.6 0.002 0.046 59.850 59.984 multiply_cannon_multrec 69360 15.2 29.526 34.721 39.790 44.923 mp_waitall_1 488190 16.1 34.930 41.877 34.930 41.877 qs_ot_get_p 112 10.4 0.002 0.020 41.286 41.533 qs_ot_get_derivative 55 11.6 0.001 0.025 38.032 38.132 multiply_cannon_sync_h2d 69360 15.2 28.973 32.738 28.973 32.738 multiply_cannon_metrocomm3 69360 15.2 0.199 0.209 25.751 32.453 qs_ot_p2m_diag 40 11.0 0.022 0.073 30.376 30.480 rebuild_ks_matrix 110 8.4 0.000 0.000 28.867 29.065 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.067 28.866 29.065 init_scf_loop 7 6.6 0.002 0.043 27.257 27.261 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.158 27.160 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.427 26.606 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.096 23.361 apply_single 62 13.6 0.000 0.000 23.096 23.361 cp_fm_syevd 40 13.0 0.000 0.000 22.131 22.262 prepare_preconditioner 7 7.6 0.000 0.000 22.185 22.219 make_preconditioner 7 8.6 0.002 0.048 22.185 22.219 ot_new_cg_direction 55 11.6 0.003 0.059 21.138 21.139 cp_fm_redistribute_end 40 14.0 8.728 17.416 8.733 17.417 cp_fm_syevd_base 40 14.0 8.677 17.367 8.677 17.367 qs_rho_update_rho_low 110 7.6 0.001 0.010 16.496 16.833 calculate_rho_elec 110 8.6 0.030 0.033 16.495 16.833 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.780 15.019 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.266 14.328 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.462 13.557 mp_sum_l 4764 12.2 12.115 12.914 12.115 12.914 init_scf_run 5 5.8 0.000 0.002 11.912 11.913 scf_env_initial_rho_setup 5 6.8 0.003 0.052 11.912 11.913 pw_transfer 1645 12.4 0.079 0.098 11.716 11.899 calculate_dm_sparse 110 9.5 0.000 0.001 11.612 11.803 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.577 11.761 density_rs2pw 110 9.6 0.005 0.006 10.769 11.175 dbcsr_mm_accdrv_process 154766 15.8 6.362 6.544 10.132 11.118 qs_vxc_create 110 10.4 0.002 0.018 10.466 10.510 fft_wrap_pw1pw2_240 915 15.0 0.832 0.917 10.200 10.389 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.253 10.319 cp_fm_cholesky_invert 7 10.6 9.957 9.965 9.957 9.965 check_diag 80 13.5 8.684 8.935 9.410 9.541 fft3d_pb 915 16.0 2.406 2.644 8.378 8.595 sum_up_and_integrate 60 10.3 0.002 0.031 8.399 8.440 integrate_v_rspace 60 11.3 0.001 0.002 8.381 8.423 acc_transpose_blocks 69360 15.2 0.365 0.381 7.527 8.102 transfer_rs2pw 445 10.6 0.007 0.008 7.502 7.897 make_full_single_inverse 7 9.6 0.003 0.056 7.655 7.692 xc_rho_set_and_dset_create 110 12.4 0.076 0.130 7.373 7.620 multiply_cannon_metrocomm1 69360 15.2 0.097 0.103 4.740 7.175 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.120 7.137 xc_vxc_pw_create 60 11.3 0.039 0.086 7.008 7.053 calculate_first_density_matrix 1 7.0 0.000 0.003 7.036 7.052 make_m2s 2890 13.2 0.078 0.087 6.383 6.921 make_images 2890 14.2 0.237 0.257 6.276 6.813 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.675 6.731 xc_pw_derive 510 13.4 0.005 0.007 6.248 6.311 mp_alltoall_z22v 2340 17.7 5.378 5.666 5.378 5.666 acc_transpose_blocks_kernels 69360 16.2 0.855 0.897 4.711 5.211 mp_waitany 7680 13.5 4.500 5.063 4.500 5.063 multiply_cannon_metrocomm4 67915 15.2 0.185 0.200 2.022 4.728 potential_pw2rs 60 12.3 0.002 0.003 4.619 4.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=226.818000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9e76ba591fd0ff0e84c9765dbd9f1c86ae20f04d Summary: empty Status: OK