=== 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: 96ef312a72745dc8e373ed79515b61e51a8022ea ################# 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/01 job id: 50876898 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/02 job id: 50876899 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/03 job id: 50876900 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/04 job id: 50876901 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/05 job id: 50876902 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/06 job id: 50876903 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/07 job id: 50876904 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/08 job id: 50876905 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/09 job id: 50876906 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/10 job id: 50876907 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/11 job id: 50876908 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/12 job id: 50876909 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/13 job id: 50876910 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/14 job id: 50876911 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/15 job id: 50876912 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/16 job id: 50876913 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/17 job id: 50876914 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/18 job id: 50876915 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/19 job id: 50876916 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/20 job id: 50876917 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/21 job id: 50876918 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/22 job id: 50876919 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/23 job id: 50876920 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/24 job id: 50876922 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/25 job id: 50876923 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/26 job id: 50876924 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/27 job id: 50876925 --- 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/96ef312a72745dc8e373ed79515b61e51a8022ea_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 0.081 0.204 135.917 135.918 farming_run 1 2.0 134.668 134.686 135.785 135.875 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.498014E+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 0.006 0.027 115.515 115.515 qs_energies 1 2.0 0.000 0.000 115.297 115.299 mp2_main 1 3.0 0.000 0.000 112.906 112.908 mp2_gpw_main 1 4.0 0.027 0.034 111.830 111.832 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.809 93.251 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.584 55.027 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.161 40.986 46.148 get_2c_integrals 1 6.0 0.132 0.154 37.307 38.052 integrate_v_rspace 273 8.0 0.440 0.452 24.597 29.541 pw_transfer 6555 10.6 0.378 0.391 26.790 27.362 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.422 26.031 grid_integrate_task_list 273 9.0 20.506 25.909 20.506 25.909 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.030 22.946 23.524 compute_2c_integrals 1 7.0 0.002 0.002 19.063 19.186 rpa_ri_compute_en 1 5.0 0.019 0.021 18.917 19.032 cp_fm_cholesky_decompose 12 8.2 18.140 18.874 18.140 18.874 compute_2c_integrals_loop_lm 1 8.0 0.001 0.004 18.588 18.764 mp2_eri_2c_integrate_gpw 1 9.0 2.397 2.432 18.587 18.763 cholesky_decomp 1 7.0 0.000 0.000 16.956 17.683 fft3d_s 5443 13.4 16.122 16.442 16.144 16.464 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.314 10.771 15.314 calculate_wavefunction 272 8.0 5.397 5.543 12.305 12.974 rpa_num_int 1 6.0 0.000 0.001 10.734 10.743 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.705 10.723 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.595 contract_S_to_Q 8 9.0 0.000 0.000 8.909 9.017 calc_potential_gpw 544 9.5 0.004 0.005 8.203 8.663 parallel_gemm_fm 14 9.1 0.000 0.000 8.506 8.602 parallel_gemm_fm_cosma 14 10.1 8.506 8.602 8.506 8.602 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.054 8.290 potential_pw2rs 545 10.0 0.106 0.108 7.486 8.096 create_integ_mat 1 6.0 0.016 0.028 7.648 7.656 collocate_single_gaussian 272 10.0 0.038 0.041 7.324 7.627 array2fm 1 7.0 0.000 0.000 6.709 7.070 pw_scatter_s 2720 13.7 4.334 4.431 4.334 4.431 pw_gather_s 2722 13.2 3.434 3.793 3.434 3.793 array2fm_buffer_send 1 8.0 2.946 3.152 2.946 3.152 pw_poisson_solve 545 10.5 1.120 1.203 2.194 2.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.831631, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 0.027 0.039 426.832 426.834 farming_run 1 2.0 425.538 425.542 426.768 426.771 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.241911E+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 4854402. 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 0.019 0.039 217.590 217.591 qs_energies 1 2.0 0.000 0.000 217.328 217.334 scf_env_do_scf 1 3.0 0.000 0.000 114.596 114.596 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.412 113.419 rebuild_ks_matrix 4 6.0 0.000 0.000 113.410 113.417 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 113.410 113.417 hfx_ks_matrix 4 8.0 0.001 0.001 112.973 112.977 integrate_four_center 4 9.0 0.154 0.466 112.972 112.976 integrate_four_center_main 4 10.0 0.101 0.416 101.961 104.767 integrate_four_center_bin 264 11.0 101.860 104.736 101.860 104.736 mp2_main 1 3.0 0.003 0.011 102.399 102.406 mp2_gpw_main 1 4.0 0.048 0.070 100.911 100.923 init_scf_loop 1 4.0 0.000 0.000 96.539 96.539 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.995 75.128 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.751 54.874 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 41.385 46.608 integrate_v_rspace 95 8.0 0.397 0.566 27.939 32.985 pw_transfer 2240 10.6 0.146 0.164 29.493 29.903 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.513 28.936 grid_integrate_task_list 95 9.0 23.271 28.522 23.271 28.522 mp2_ri_gpw_compute_en 1 5.0 0.055 0.064 26.759 28.396 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.240 26.686 ao_to_mo_and_store_B_mult_1 91 7.0 10.689 26.395 10.689 26.395 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.852 1.905 24.981 24.992 get_2c_integrals 1 6.0 0.000 0.001 20.071 20.173 compute_2c_integrals 1 7.0 0.002 0.003 19.047 19.051 fft3d_s 1823 13.4 18.474 18.945 18.488 18.959 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.554 18.888 mp2_eri_2c_integrate_gpw 1 9.0 1.751 1.913 18.553 18.887 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.049 18.049 calculate_wavefunction 91 8.0 2.013 2.045 9.622 9.823 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.763 9.344 potential_pw2rs 186 10.0 0.033 0.034 8.509 9.103 local_gemm 172 8.0 8.207 8.784 8.207 8.784 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.111 8.423 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.511 7.983 8.308 calc_potential_gpw 182 9.5 0.002 0.002 7.856 8.101 collocate_single_gaussian 91 10.0 0.016 0.020 7.718 7.908 mp_sync 37 10.5 2.989 7.860 2.989 7.860 integrate_four_center_load 4 10.0 0.000 0.000 6.784 6.788 hfx_load_balance 1 11.0 0.000 0.000 6.784 6.788 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.426 6.338 6.426 mp_sendrecv_dm3 2068 8.0 6.032 6.339 6.032 6.339 pw_gather_s 912 13.2 4.505 5.009 4.505 5.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.912427, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 453.677056E+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 1281356. 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 0.011 0.030 57.526 57.532 qs_mol_dyn_low 1 2.0 0.003 0.005 57.058 57.066 qs_forces 11 3.9 0.002 0.003 56.989 56.990 qs_energies 11 4.9 0.011 0.021 55.512 55.525 scf_env_do_scf 11 5.9 0.000 0.001 49.176 49.176 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 47.026 47.027 qs_scf_new_mos 108 7.5 0.000 0.001 36.640 36.950 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.640 36.950 dbcsr_multiply_generic 2286 12.5 0.095 0.098 36.203 36.637 ot_scf_mini 108 9.5 0.002 0.002 34.868 35.088 multiply_cannon 2286 13.5 0.179 0.187 27.465 28.956 multiply_cannon_loop 2286 14.5 1.864 1.966 26.742 28.256 velocity_verlet 10 3.0 0.002 0.002 26.708 26.709 ot_mini 108 10.5 0.001 0.001 21.316 21.582 qs_ot_get_derivative 108 11.5 0.001 0.001 18.221 18.393 mp_waitall_1 245248 16.5 9.584 15.783 9.584 15.783 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.696 13.578 multiply_cannon_multrec 54864 15.5 3.557 5.507 7.805 11.150 qs_ot_get_p 119 10.4 0.001 0.001 8.631 8.916 mp_sum_l 7287 12.8 6.547 8.222 6.547 8.222 rebuild_ks_matrix 119 8.3 0.000 0.000 8.055 8.204 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.016 8.055 8.204 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.111 7.246 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.631 6.745 multiply_cannon_sync_h2d 54864 15.5 5.043 6.232 5.043 6.232 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.777 6.203 dbcsr_mm_accdrv_process 76910 16.1 1.897 2.945 4.162 5.892 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.635 5.682 init_scf_run 11 5.9 0.000 0.001 4.982 4.982 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.982 4.982 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.625 4.625 sum_up_and_integrate 119 10.3 0.001 0.003 4.570 4.577 integrate_v_rspace 119 11.3 0.002 0.003 4.559 4.567 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.331 4.477 calculate_rho_elec 119 8.7 0.011 0.016 4.331 4.476 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.384 4.384 cp_fm_redistribute_end 50 14.0 2.236 4.355 2.242 4.358 cp_fm_diag_elpa_base 50 14.0 2.110 4.245 2.114 4.252 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 2.133 3.308 calculate_dm_sparse 119 9.5 0.000 0.000 3.120 3.298 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.978 3.164 apply_single 119 13.6 0.000 0.000 2.977 3.164 jit_kernel_multiply 13 15.8 2.197 2.853 2.197 2.853 acc_transpose_blocks 54864 15.5 0.226 0.249 2.296 2.826 ot_diis_step 108 11.5 0.006 0.006 2.798 2.799 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.689 2.754 calculate_first_density_matrix 1 7.0 0.000 0.002 2.672 2.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.411 2.413 density_rs2pw 119 9.7 0.004 0.004 2.291 2.386 wfi_extrapolate 11 7.9 0.001 0.001 2.239 2.239 init_scf_loop 11 6.9 0.001 0.004 2.129 2.130 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.025 2.113 grid_integrate_task_list 119 12.3 2.020 2.109 2.020 2.109 mp_sum_d 4137 12.0 1.431 2.096 1.431 2.096 potential_pw2rs 119 12.3 0.004 0.004 1.927 1.941 make_m2s 4572 13.5 0.054 0.056 1.875 1.931 pw_transfer 1439 11.6 0.051 0.056 1.835 1.914 make_images 4572 14.5 0.132 0.138 1.793 1.848 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.760 1.840 mp_alltoall_d11v 2130 13.8 1.454 1.617 1.454 1.617 transfer_rs2pw 487 10.6 0.005 0.006 1.451 1.553 fft3d_ps 1201 14.6 0.373 0.478 1.474 1.551 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.515 1.547 acc_transpose_blocks_sync 164592 16.5 1.239 1.469 1.239 1.469 grid_collocate_task_list 119 9.7 1.345 1.439 1.345 1.439 fft_wrap_pw1pw2_140 487 13.2 0.108 0.114 1.359 1.437 transfer_pw2rs 487 13.2 0.006 0.006 1.388 1.399 mp_waitany 12084 13.8 1.285 1.375 1.285 1.375 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.815 1.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.532000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 491.347968E+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 683782. 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 0.019 0.053 48.400 48.401 qs_mol_dyn_low 1 2.0 0.003 0.008 48.072 48.082 qs_forces 11 3.9 0.002 0.002 48.005 48.007 qs_energies 11 4.9 0.004 0.013 46.138 46.145 scf_env_do_scf 11 5.9 0.001 0.001 40.378 40.378 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 36.784 36.785 dbcsr_multiply_generic 2286 12.5 0.104 0.107 24.689 25.597 qs_scf_new_mos 108 7.5 0.001 0.001 25.007 25.243 qs_scf_loop_do_ot 108 8.5 0.001 0.001 25.006 25.242 ot_scf_mini 108 9.5 0.003 0.004 23.925 24.085 velocity_verlet 10 3.0 0.001 0.002 23.495 23.496 multiply_cannon 2286 13.5 0.208 0.215 17.565 20.491 multiply_cannon_loop 2286 14.5 1.196 1.260 15.971 19.101 ot_mini 108 10.5 0.001 0.001 14.411 14.650 mp_waitall_1 200699 16.5 7.576 12.856 7.576 12.856 qs_ot_get_derivative 108 11.5 0.001 0.001 11.778 11.937 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.390 10.043 multiply_cannon_multrec 27432 15.5 1.770 4.081 6.120 8.936 rebuild_ks_matrix 119 8.3 0.000 0.000 8.585 8.828 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.039 8.584 8.827 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.624 7.846 qs_ot_get_p 119 10.4 0.001 0.002 6.266 6.488 dbcsr_mm_accdrv_process 47894 16.0 3.484 5.785 4.274 6.374 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.425 5.478 calculate_rho_elec 119 8.7 0.021 0.024 5.424 5.477 sum_up_and_integrate 119 10.3 0.001 0.002 5.304 5.309 integrate_v_rspace 119 11.3 0.002 0.002 5.290 5.294 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.324 5.237 mp_sum_l 7287 12.8 3.033 5.166 3.033 5.166 init_scf_run 11 5.9 0.000 0.001 4.329 4.329 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.328 4.329 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.145 4.300 apply_single 119 13.6 0.000 0.000 3.145 4.300 make_m2s 4572 13.5 0.052 0.053 3.658 4.006 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.958 3.995 make_images 4572 14.5 0.205 0.243 3.568 3.913 density_rs2pw 119 9.7 0.004 0.004 3.647 3.748 init_scf_loop 11 6.9 0.001 0.003 3.546 3.547 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.486 3.487 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.993 3.079 potential_pw2rs 119 12.3 0.006 0.006 2.980 3.035 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.902 2.904 cp_fm_redistribute_end 50 14.0 1.464 2.843 1.472 2.846 cp_fm_diag_elpa_base 50 14.0 1.327 2.686 1.361 2.749 transfer_rs2pw 487 10.6 0.005 0.005 2.506 2.674 pw_transfer 1439 11.6 0.065 0.068 2.527 2.596 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.436 2.507 ot_diis_step 108 11.5 0.011 0.011 2.502 2.503 prepare_preconditioner 11 7.9 0.000 0.000 2.394 2.421 make_preconditioner 11 8.9 0.000 0.001 2.394 2.421 calculate_first_density_matrix 1 7.0 0.001 0.002 2.341 2.343 calculate_dm_sparse 119 9.5 0.000 0.001 2.248 2.331 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.270 2.330 make_images_data 4572 15.5 0.048 0.054 1.586 2.251 transfer_pw2rs 487 13.2 0.004 0.005 2.198 2.241 fft3d_ps 1201 14.6 0.527 0.586 2.070 2.135 multiply_cannon_sync_h2d 27432 15.5 1.572 2.091 1.572 2.091 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.022 2.025 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.392 1.970 acc_transpose_blocks 27432 15.5 0.115 0.122 1.622 1.959 grid_integrate_task_list 119 12.3 1.817 1.936 1.817 1.936 mp_alltoall_d11v 2130 13.8 1.708 1.909 1.708 1.909 wfi_extrapolate 11 7.9 0.001 0.001 1.877 1.877 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.690 1.754 jit_kernel_multiply 9 16.3 0.730 1.737 0.730 1.737 fft_wrap_pw1pw2_140 487 13.2 0.126 0.131 1.644 1.707 mp_sendrecv_dv 16779 12.7 1.612 1.671 1.612 1.671 make_images_sizes 4572 15.5 0.005 0.005 1.210 1.628 mp_alltoall_i44 4572 16.5 1.206 1.624 1.206 1.624 mp_allgather_i34 2286 14.5 1.016 1.594 1.016 1.594 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.448 1.485 grid_collocate_task_list 119 9.7 1.297 1.355 1.297 1.355 mp_sum_d 4137 12.0 0.880 1.333 0.880 1.333 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.306 1.319 mp_alltoall_z22v 1201 16.6 1.132 1.276 1.132 1.276 cp_fm_cholesky_invert 11 10.9 1.200 1.208 1.200 1.208 transfer_rs2pw_50 119 11.7 0.182 0.192 1.008 1.116 qs_energies_init_hamiltonians 11 5.9 0.016 0.022 1.049 1.051 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.027 1.047 dbcsr_complete_redistribute 329 12.2 0.076 0.094 0.967 1.046 transfer_pw2rs_50 119 14.3 0.176 0.191 0.928 1.018 dbcsr_dot_sd 1205 11.9 0.071 0.088 0.566 0.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=48.401000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.818182, yerr=2.587318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 523.292672E+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 0.010 0.028 32.498 32.499 qs_mol_dyn_low 1 2.0 0.003 0.006 32.223 32.233 qs_forces 11 3.9 0.005 0.026 32.147 32.148 qs_energies 11 4.9 0.001 0.002 30.581 30.584 scf_env_do_scf 11 5.9 0.000 0.001 25.877 25.877 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.353 23.353 dbcsr_multiply_generic 2286 12.5 0.095 0.098 17.117 17.174 qs_scf_new_mos 108 7.5 0.001 0.001 15.629 15.643 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.628 15.642 velocity_verlet 10 3.0 0.001 0.001 15.395 15.397 ot_scf_mini 108 9.5 0.002 0.002 14.864 14.880 multiply_cannon 2286 13.5 0.191 0.196 13.740 14.493 multiply_cannon_loop 2286 14.5 0.868 0.926 12.955 13.715 ot_mini 108 10.5 0.001 0.001 9.204 9.219 qs_ot_get_derivative 108 11.5 0.001 0.001 7.700 7.716 multiply_cannon_multrec 18288 15.5 1.858 2.906 6.972 7.283 rebuild_ks_matrix 119 8.3 0.000 0.000 6.031 6.049 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 6.031 6.048 dbcsr_mm_accdrv_process 38222 16.0 4.960 5.747 5.021 5.816 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.342 5.362 mp_waitall_1 158411 16.6 2.663 3.762 2.663 3.762 sum_up_and_integrate 119 10.3 0.001 0.003 3.581 3.587 integrate_v_rspace 119 11.3 0.002 0.003 3.568 3.575 init_scf_run 11 5.9 0.000 0.001 3.520 3.520 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.520 3.520 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.851 3.470 qs_ot_get_p 119 10.4 0.001 0.001 3.415 3.443 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.280 3.311 calculate_rho_elec 119 8.7 0.031 0.031 3.280 3.311 init_scf_loop 11 6.9 0.000 0.000 2.507 2.507 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.449 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.032 2.351 apply_single 119 13.6 0.000 0.000 2.032 2.350 calculate_first_density_matrix 1 7.0 0.000 0.003 2.273 2.275 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.259 2.266 make_m2s 4572 13.5 0.044 0.045 1.845 1.991 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.959 1.960 density_rs2pw 119 9.7 0.004 0.004 1.840 1.947 grid_integrate_task_list 119 12.3 1.807 1.923 1.807 1.923 calculate_dm_sparse 119 9.5 0.000 0.001 1.900 1.908 make_images 4572 14.5 0.192 0.207 1.759 1.904 pw_transfer 1439 11.6 0.065 0.068 1.864 1.874 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.771 1.784 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.755 1.767 acc_transpose_blocks 18288 15.5 0.080 0.082 1.660 1.746 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.705 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.695 1.715 make_preconditioner 11 8.9 0.000 0.000 1.695 1.715 cp_fm_diag_elpa_base 50 14.0 1.682 1.693 1.704 1.714 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.686 1.688 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.550 1.646 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.521 1.528 mp_sum_l 7287 12.8 1.142 1.524 1.142 1.524 ot_diis_step 108 11.5 0.011 0.012 1.489 1.489 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.426 1.437 potential_pw2rs 119 12.3 0.007 0.008 1.364 1.368 fft3d_ps 1201 14.6 0.546 0.568 1.326 1.335 grid_collocate_task_list 119 9.7 1.233 1.308 1.233 1.308 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.232 1.236 wfi_extrapolate 11 7.9 0.001 0.001 1.193 1.193 multiply_cannon_sync_h2d 18288 15.5 1.035 1.168 1.035 1.168 transfer_rs2pw 487 10.6 0.005 0.005 0.991 1.106 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.959 0.974 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.960 0.961 make_images_data 4572 15.5 0.047 0.051 0.770 0.935 mp_alltoall_d11v 2130 13.8 0.700 0.871 0.700 0.871 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.669 0.865 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.810 0.839 transfer_pw2rs 487 13.2 0.004 0.004 0.832 0.836 acc_transpose_blocks_sync 54864 16.5 0.751 0.834 0.751 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.782 0.784 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.669 0.728 cp_fm_cholesky_invert 11 10.9 0.708 0.712 0.708 0.712 mp_alltoall_z22v 1201 16.6 0.631 0.699 0.631 0.699 multiply_cannon_metrocomm1 18288 15.5 0.030 0.030 0.452 0.659 mp_waitany 9880 13.7 0.539 0.655 0.539 0.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.499000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.363636, yerr=1.298442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.977600E+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 0.009 0.030 37.402 37.402 qs_mol_dyn_low 1 2.0 0.003 0.004 37.226 37.235 qs_forces 11 3.9 0.003 0.003 37.163 37.164 qs_energies 11 4.9 0.001 0.001 35.436 35.441 scf_env_do_scf 11 5.9 0.000 0.001 30.370 30.371 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.851 26.852 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.191 20.320 velocity_verlet 10 3.0 0.001 0.002 19.023 19.025 qs_scf_new_mos 108 7.5 0.001 0.001 18.479 18.528 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.478 18.527 ot_scf_mini 108 9.5 0.002 0.003 17.420 17.479 multiply_cannon 2286 13.5 0.217 0.226 16.190 16.574 multiply_cannon_loop 2286 14.5 1.554 1.642 15.184 15.507 ot_mini 108 10.5 0.001 0.001 10.588 10.653 multiply_cannon_multrec 27432 15.5 2.453 3.117 9.028 9.308 qs_ot_get_derivative 108 11.5 0.001 0.001 8.703 8.762 dbcsr_mm_accdrv_process 47916 15.9 6.056 7.711 6.475 7.803 rebuild_ks_matrix 119 8.3 0.000 0.000 6.491 6.540 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.490 6.540 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.755 5.799 qs_ot_get_p 119 10.4 0.001 0.001 3.937 4.019 init_scf_run 11 5.9 0.000 0.001 3.714 3.714 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.714 3.714 sum_up_and_integrate 119 10.3 0.001 0.001 3.627 3.635 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.164 3.629 integrate_v_rspace 119 11.3 0.002 0.003 3.616 3.623 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.527 3.562 calculate_rho_elec 119 8.7 0.040 0.046 3.526 3.562 init_scf_loop 11 6.9 0.000 0.000 3.496 3.496 prepare_preconditioner 11 7.9 0.000 0.000 2.642 2.649 make_preconditioner 11 8.9 0.000 0.000 2.642 2.649 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.191 2.610 apply_single 119 13.6 0.000 0.000 2.191 2.610 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.242 2.574 acc_transpose_blocks 27432 15.5 0.121 0.124 2.460 2.556 make_m2s 4572 13.5 0.054 0.056 2.430 2.540 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.423 2.455 make_images 4572 14.5 0.273 0.337 2.322 2.430 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.413 2.422 mp_waitall_1 137007 16.6 1.834 2.313 1.834 2.313 calculate_first_density_matrix 1 7.0 0.000 0.000 2.277 2.279 calculate_dm_sparse 119 9.5 0.000 0.000 2.165 2.222 density_rs2pw 119 9.7 0.003 0.004 1.962 2.126 pw_transfer 1439 11.6 0.065 0.069 2.088 2.126 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.996 2.037 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.997 1.997 grid_integrate_task_list 119 12.3 1.836 1.906 1.836 1.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.862 1.863 ot_diis_step 108 11.5 0.012 0.012 1.845 1.846 mp_sum_l 7287 12.8 1.138 1.808 1.138 1.808 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.718 1.726 cp_fm_diag_elpa_base 50 14.0 1.681 1.699 1.715 1.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.669 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.616 1.660 acc_transpose_blocks_sync 82296 16.5 1.474 1.562 1.474 1.562 fft3d_ps 1201 14.6 0.592 0.654 1.485 1.508 jit_kernel_multiply 8 16.5 0.349 1.423 0.349 1.423 potential_pw2rs 119 12.3 0.008 0.009 1.417 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.381 1.381 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.322 1.335 grid_collocate_task_list 119 9.7 1.247 1.327 1.247 1.327 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.252 cp_fm_upper_to_full 72 14.2 0.839 1.192 0.839 1.192 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.683 1.185 transfer_rs2pw 487 10.6 0.004 0.005 0.984 1.149 dbcsr_complete_redistribute 329 12.2 0.139 0.172 0.840 1.128 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.075 1.076 make_images_data 4572 15.5 0.048 0.053 0.891 1.016 hybrid_alltoall_any 4725 16.4 0.066 0.154 0.756 0.937 mp_alltoall_d11v 2130 13.8 0.798 0.916 0.798 0.916 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.618 0.895 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.875 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.837 0.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.832 transfer_pw2rs 487 13.2 0.004 0.004 0.814 0.817 mp_alltoall_z22v 1201 16.6 0.775 0.805 0.775 0.805 cp_fm_cholesky_invert 11 10.9 0.793 0.796 0.793 0.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.402000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.727273, yerr=2.799055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 633.987072E+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 0.025 0.044 28.195 28.196 qs_mol_dyn_low 1 2.0 0.003 0.003 27.961 27.971 qs_forces 11 3.9 0.002 0.003 27.902 27.902 qs_energies 11 4.9 0.001 0.001 26.207 26.209 scf_env_do_scf 11 5.9 0.000 0.001 21.560 21.560 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.140 19.141 velocity_verlet 10 3.0 0.002 0.002 14.535 14.538 dbcsr_multiply_generic 2286 12.5 0.094 0.098 12.807 12.883 qs_scf_new_mos 108 7.5 0.001 0.001 11.533 11.557 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.533 11.557 ot_scf_mini 108 9.5 0.002 0.002 10.829 10.856 multiply_cannon 2286 13.5 0.223 0.229 10.297 10.698 multiply_cannon_loop 2286 14.5 0.642 0.665 9.420 9.592 ot_mini 108 10.5 0.001 0.001 6.169 6.202 multiply_cannon_multrec 9144 15.5 1.655 1.839 5.939 6.169 rebuild_ks_matrix 119 8.3 0.000 0.000 5.723 5.742 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.723 5.742 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.091 5.109 qs_ot_get_derivative 108 11.5 0.001 0.001 4.885 4.913 dbcsr_mm_accdrv_process 12550 15.8 3.511 4.224 4.176 4.268 sum_up_and_integrate 119 10.3 0.001 0.001 3.389 3.393 integrate_v_rspace 119 11.3 0.003 0.003 3.378 3.383 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.345 3.351 calculate_rho_elec 119 8.7 0.060 0.061 3.344 3.350 init_scf_run 11 5.9 0.000 0.001 3.218 3.218 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.217 3.218 qs_ot_get_p 119 10.4 0.001 0.001 2.792 2.826 init_scf_loop 11 6.9 0.000 0.000 2.401 2.401 calculate_first_density_matrix 1 7.0 0.000 0.000 2.102 2.103 pw_transfer 1439 11.6 0.065 0.068 2.094 2.102 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.001 2.012 grid_integrate_task_list 119 12.3 1.861 1.919 1.861 1.919 make_m2s 4572 13.5 0.033 0.034 1.731 1.846 density_rs2pw 119 9.7 0.003 0.003 1.730 1.823 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.810 1.813 make_images 4572 14.5 0.269 0.299 1.641 1.755 calculate_dm_sparse 119 9.5 0.000 0.000 1.735 1.752 mp_waitall_1 115863 16.7 1.237 1.746 1.237 1.746 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.690 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.680 1.684 make_preconditioner 11 8.9 0.000 0.000 1.680 1.684 fft_wrap_pw1pw2_140 487 13.2 0.325 0.331 1.670 1.681 jit_kernel_multiply 9 15.4 0.624 1.611 0.624 1.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.573 1.599 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.571 1.571 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.528 1.529 acc_transpose_blocks 9144 15.5 0.043 0.044 1.427 1.471 grid_collocate_task_list 119 9.7 1.295 1.377 1.295 1.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.366 1.374 fft3d_ps 1201 14.6 0.651 0.663 1.353 1.363 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.330 1.344 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.294 1.302 cp_fm_diag_elpa_base 50 14.0 1.268 1.285 1.293 1.301 ot_diis_step 108 11.5 0.013 0.013 1.272 1.272 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.221 1.256 apply_single 119 13.6 0.000 0.000 1.221 1.256 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.222 potential_pw2rs 119 12.3 0.010 0.011 1.210 1.210 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.125 1.131 wfi_extrapolate 11 7.9 0.001 0.001 1.065 1.065 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.927 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.750 0.918 make_images_data 4572 15.5 0.042 0.046 0.732 0.864 mp_alltoall_d11v 2130 13.8 0.779 0.852 0.779 0.852 cp_fm_cholesky_invert 11 10.9 0.833 0.836 0.833 0.836 transfer_rs2pw 487 10.6 0.004 0.004 0.743 0.835 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.811 0.817 acc_transpose_blocks_sync 27432 16.5 0.721 0.765 0.721 0.765 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.751 0.754 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.666 0.717 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.337 0.674 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.648 0.654 transfer_pw2rs 487 13.2 0.003 0.004 0.616 0.618 mp_alltoall_z22v 1201 16.6 0.576 0.609 0.576 0.609 mp_allgather_i34 2286 14.5 0.193 0.589 0.193 0.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.196000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=602.090909, yerr=4.962671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 765.935616E+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 0.010 0.028 44.391 44.392 qs_mol_dyn_low 1 2.0 0.003 0.003 44.179 44.188 qs_forces 11 3.9 0.002 0.002 44.079 44.079 qs_energies 11 4.9 0.001 0.001 42.014 42.017 scf_env_do_scf 11 5.9 0.001 0.001 36.029 36.029 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.099 28.100 velocity_verlet 10 3.0 0.006 0.007 24.864 24.874 dbcsr_multiply_generic 2286 12.5 0.102 0.103 19.379 19.616 qs_scf_new_mos 108 7.5 0.001 0.001 18.002 18.105 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.002 18.105 ot_scf_mini 108 9.5 0.002 0.002 16.807 16.911 multiply_cannon 2286 13.5 0.297 0.310 15.078 16.075 multiply_cannon_loop 2286 14.5 0.872 0.888 13.795 14.838 ot_mini 108 10.5 0.001 0.001 10.133 10.249 multiply_cannon_multrec 9144 15.5 3.442 4.689 8.865 8.939 qs_ot_get_derivative 108 11.5 0.001 0.001 8.056 8.160 init_scf_loop 11 6.9 0.000 0.000 7.896 7.898 rebuild_ks_matrix 119 8.3 0.000 0.000 7.510 7.652 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.510 7.651 prepare_preconditioner 11 7.9 0.000 0.000 6.898 6.910 make_preconditioner 11 8.9 0.000 0.000 6.898 6.910 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.748 6.877 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.430 6.792 dbcsr_mm_accdrv_process 12550 15.8 4.824 6.565 5.291 6.599 cp_fm_upper_to_full 72 14.2 3.165 4.585 3.165 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.464 4.471 calculate_rho_elec 119 8.7 0.117 0.120 4.464 4.471 sum_up_and_integrate 119 10.3 0.001 0.001 4.041 4.047 integrate_v_rspace 119 11.3 0.003 0.003 4.030 4.037 init_scf_run 11 5.9 0.000 0.001 3.900 3.900 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.899 3.900 qs_ot_get_p 119 10.4 0.001 0.001 3.678 3.815 mp_waitall_1 94719 16.7 2.328 3.446 2.328 3.446 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.851 3.323 pw_transfer 1439 11.6 0.068 0.069 3.210 3.215 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.111 3.115 dbcsr_complete_redistribute 329 12.2 0.279 0.282 2.041 2.865 make_m2s 4572 13.5 0.037 0.037 2.575 2.724 fft_wrap_pw1pw2_140 487 13.2 0.648 0.651 2.606 2.613 make_images 4572 14.5 0.354 0.388 2.455 2.605 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.718 2.541 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.281 2.522 apply_single 119 13.6 0.000 0.000 2.281 2.522 density_rs2pw 119 9.7 0.003 0.003 2.438 2.470 calculate_first_density_matrix 1 7.0 0.000 0.000 2.358 2.359 mp_alltoall_i22 627 13.8 1.456 2.318 1.456 2.318 calculate_dm_sparse 119 9.5 0.000 0.000 2.289 2.303 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.463 2.279 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.231 2.229 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.175 2.228 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.204 2.206 grid_integrate_task_list 119 12.3 2.112 2.178 2.112 2.178 ot_diis_step 108 11.5 0.014 0.015 2.049 2.050 fft3d_ps 1201 14.6 0.870 0.888 2.033 2.036 acc_transpose_blocks 9144 15.5 0.045 0.047 1.849 1.880 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.840 1.841 mp_sum_l 7287 12.8 1.176 1.836 1.176 1.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.821 1.823 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.784 1.785 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.564 1.608 grid_collocate_task_list 119 9.7 1.544 1.596 1.544 1.596 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.517 cp_fm_diag_elpa_base 50 14.0 1.366 1.424 1.514 1.514 cp_fm_cholesky_invert 11 10.9 1.489 1.493 1.489 1.493 potential_pw2rs 119 12.3 0.014 0.014 1.477 1.478 wfi_extrapolate 11 7.9 0.001 0.001 1.464 1.464 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.143 1.388 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.351 1.369 mp_alltoall_d11v 2130 13.8 1.281 1.332 1.281 1.332 make_images_data 4572 15.5 0.047 0.050 1.092 1.318 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.273 acc_transpose_blocks_sync 27432 16.5 1.118 1.143 1.118 1.143 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.112 1.129 mp_alltoall_z22v 1201 16.6 1.026 1.053 1.026 1.053 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.994 1.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.974 0.987 qs_create_task_list 11 7.9 0.001 0.001 0.932 0.943 generate_qs_task_list 11 8.9 0.368 0.388 0.932 0.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.392000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=718.000000, yerr=13.341664 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 87.535 87.536 qs_mol_dyn_low 1 2.0 0.003 0.003 87.277 87.289 qs_forces 11 3.9 0.003 0.003 87.185 87.186 qs_energies 11 4.9 0.001 0.002 84.237 84.260 scf_env_do_scf 11 5.9 0.000 0.001 75.145 75.147 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 69.070 69.071 dbcsr_multiply_generic 2055 12.4 0.108 0.111 52.730 53.085 qs_scf_new_mos 99 7.5 0.000 0.001 50.580 50.686 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.579 50.686 ot_scf_mini 99 9.5 0.002 0.002 48.097 48.212 velocity_verlet 10 3.0 0.001 0.002 45.567 45.568 multiply_cannon 2055 13.4 0.176 0.180 43.018 44.116 multiply_cannon_loop 2055 14.4 1.828 1.869 41.884 42.975 ot_mini 99 10.5 0.001 0.001 27.873 27.973 qs_ot_get_derivative 99 11.5 0.001 0.001 20.916 21.013 multiply_cannon_multrec 49320 15.4 11.222 11.949 17.349 18.124 rebuild_ks_matrix 110 8.3 0.000 0.001 15.320 15.473 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 15.320 15.472 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.488 13.635 mp_waitall_1 220248 16.4 11.763 12.884 11.763 12.884 qs_ot_get_p 110 10.4 0.001 0.001 11.371 11.491 multiply_cannon_sync_h2d 49320 15.4 9.441 10.034 9.441 10.034 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.944 8.472 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 6.743 8.118 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.933 7.977 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.287 7.733 apply_single 110 13.6 0.000 0.001 7.286 7.733 sum_up_and_integrate 110 10.3 0.002 0.003 7.632 7.647 integrate_v_rspace 110 11.3 0.002 0.003 7.607 7.628 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.113 7.250 calculate_rho_elec 110 8.6 0.020 0.024 7.112 7.249 init_scf_run 11 5.9 0.000 0.001 6.981 6.982 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.981 6.982 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.881 6.882 ot_diis_step 99 11.5 0.005 0.006 6.688 6.688 dbcsr_mm_accdrv_process 87628 16.1 3.171 3.268 5.998 6.279 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.190 6.218 cp_fm_diag_elpa_base 48 14.0 6.169 6.199 6.187 6.215 init_scf_loop 11 6.9 0.000 0.000 6.042 6.042 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.895 5.946 mp_sum_l 6594 12.7 4.638 5.538 4.638 5.538 make_m2s 4110 13.4 0.061 0.066 4.610 4.749 make_images 4110 14.4 0.179 0.192 4.512 4.655 wfi_extrapolate 11 7.9 0.001 0.001 4.269 4.269 density_rs2pw 110 9.6 0.004 0.004 4.037 4.199 calculate_dm_sparse 110 9.5 0.001 0.001 3.855 3.960 pw_transfer 1331 11.6 0.055 0.065 3.736 3.850 prepare_preconditioner 11 7.9 0.000 0.000 3.772 3.802 make_preconditioner 11 8.9 0.000 0.000 3.772 3.802 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.648 3.758 multiply_cannon_metrocomm1 49320 15.4 0.066 0.070 2.620 3.714 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.603 3.607 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.527 3.576 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.492 3.548 grid_integrate_task_list 110 12.3 3.234 3.399 3.234 3.399 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.214 3.261 potential_pw2rs 110 12.3 0.005 0.006 3.142 3.180 fft_wrap_pw1pw2_140 451 13.1 0.308 0.332 2.963 3.071 fft3d_ps 1111 14.6 0.800 0.893 2.944 3.045 mp_alltoall_d11v 2046 13.8 2.453 2.856 2.453 2.856 acc_transpose_blocks 49320 15.4 0.215 0.223 2.611 2.686 calculate_first_density_matrix 1 7.0 0.000 0.000 2.603 2.607 jit_kernel_multiply 13 15.9 2.538 2.561 2.538 2.561 transfer_rs2pw 451 10.6 0.005 0.006 2.252 2.427 grid_collocate_task_list 110 9.6 2.157 2.250 2.157 2.250 make_images_data 4110 15.4 0.044 0.049 1.992 2.170 transfer_pw2rs 451 13.1 0.006 0.007 2.112 2.148 cp_fm_cholesky_invert 11 10.9 2.064 2.068 2.064 2.068 mp_sum_d 3891 11.9 1.484 2.057 1.484 2.057 hybrid_alltoall_any 4261 16.3 0.083 0.479 1.727 2.043 mp_waitany 14300 13.8 1.863 2.028 1.863 2.028 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.966 1.987 mp_alltoall_z22v 1111 16.6 1.768 1.946 1.768 1.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=87.536000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 593.932288E+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 2224323. 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 0.015 0.033 74.271 74.273 qs_mol_dyn_low 1 2.0 0.003 0.004 73.951 73.966 qs_forces 11 3.9 0.003 0.003 73.875 73.875 qs_energies 11 4.9 0.004 0.012 70.446 70.449 scf_env_do_scf 11 5.9 0.001 0.001 61.181 61.184 scf_env_do_scf_inner_loop 99 6.5 0.012 0.044 52.795 52.796 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.982 39.138 velocity_verlet 10 3.0 0.002 0.003 38.289 38.290 qs_scf_new_mos 99 7.5 0.001 0.001 35.702 35.850 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.701 35.850 ot_scf_mini 99 9.5 0.003 0.004 33.986 34.125 multiply_cannon 2055 13.4 0.222 0.244 31.575 32.676 multiply_cannon_loop 2055 14.4 1.164 1.193 30.160 31.162 ot_mini 99 10.5 0.001 0.001 19.074 19.215 multiply_cannon_multrec 24660 15.4 6.954 8.346 14.078 15.673 rebuild_ks_matrix 110 8.3 0.000 0.000 14.242 14.393 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.242 14.393 qs_ot_get_derivative 99 11.5 0.001 0.001 13.244 13.387 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.533 12.666 mp_waitall_1 176588 16.5 8.120 10.657 8.120 10.657 init_scf_loop 11 6.9 0.001 0.003 8.341 8.342 qs_ot_get_p 110 10.4 0.001 0.002 8.127 8.294 multiply_cannon_metrocomm3 24660 15.4 0.073 0.077 5.147 7.942 multiply_cannon_sync_h2d 24660 15.4 6.390 7.608 6.390 7.608 dbcsr_mm_accdrv_process 52282 16.1 5.560 6.397 6.955 7.292 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.489 7.222 apply_single 110 13.6 0.000 0.001 6.489 7.222 sum_up_and_integrate 110 10.3 0.001 0.003 6.964 6.974 integrate_v_rspace 110 11.3 0.002 0.003 6.937 6.948 init_scf_run 11 5.9 0.000 0.001 6.767 6.767 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.766 6.767 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.614 6.632 calculate_rho_elec 110 8.6 0.039 0.047 6.613 6.632 prepare_preconditioner 11 7.9 0.000 0.000 6.247 6.276 make_preconditioner 11 8.9 0.000 0.002 6.247 6.276 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.799 5.980 qs_ot_p2m_diag 48 11.0 0.036 0.102 5.905 5.926 ot_diis_step 99 11.5 0.010 0.010 5.777 5.778 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.944 5.629 cp_dbcsr_syevd 48 12.0 0.003 0.005 5.341 5.341 make_m2s 4110 13.4 0.057 0.059 4.789 5.265 make_images 4110 14.4 0.405 0.457 4.676 5.147 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.631 4.644 cp_fm_diag_elpa_base 48 14.0 4.566 4.604 4.626 4.639 pw_transfer 1331 11.6 0.066 0.074 4.102 4.274 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.995 4.169 density_rs2pw 110 9.6 0.004 0.004 3.768 3.949 wfi_extrapolate 11 7.9 0.001 0.001 3.756 3.756 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.553 3.622 fft_wrap_pw1pw2_140 451 13.1 0.357 0.381 3.200 3.373 grid_integrate_task_list 110 12.3 3.153 3.342 3.153 3.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.312 3.314 fft3d_ps 1111 14.6 1.120 1.362 3.092 3.279 calculate_dm_sparse 110 9.5 0.001 0.001 3.151 3.186 make_images_data 4110 15.4 0.049 0.053 2.504 3.014 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.916 2.965 hybrid_alltoall_any 4261 16.3 0.104 0.449 2.191 2.950 cp_fm_cholesky_invert 11 10.9 2.920 2.929 2.920 2.929 calculate_first_density_matrix 1 7.0 0.001 0.006 2.906 2.909 potential_pw2rs 110 12.3 0.008 0.009 2.749 2.783 mp_sum_l 6594 12.7 2.002 2.649 2.002 2.649 mp_alltoall_d11v 2046 13.8 2.131 2.450 2.131 2.450 grid_collocate_task_list 110 9.6 2.171 2.339 2.171 2.339 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.042 2.065 transfer_rs2pw 451 10.6 0.006 0.007 1.799 2.023 acc_transpose_blocks 24660 15.4 0.118 0.121 1.987 2.017 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.932 1.932 jit_kernel_multiply 9 16.3 1.022 1.890 1.022 1.890 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.871 1.888 mp_allgather_i34 2055 14.4 0.725 1.789 0.725 1.789 mp_alltoall_z22v 1111 16.6 1.682 1.782 1.682 1.782 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.685 1.697 dbcsr_complete_redistribute 325 12.2 0.225 0.287 1.347 1.622 cp_fm_cholesky_decompose 22 10.9 1.592 1.612 1.592 1.612 transfer_pw2rs 451 13.1 0.006 0.006 1.572 1.605 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.753 1.604 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.495 mp_irecv_dv 57340 16.2 0.631 1.489 0.631 1.489 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.273000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.090909, yerr=6.584480 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 668.278784E+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 0.008 0.027 60.791 60.792 qs_mol_dyn_low 1 2.0 0.003 0.004 60.527 60.538 qs_forces 11 3.9 0.016 0.024 60.433 60.442 qs_energies 11 4.9 0.001 0.002 57.251 57.274 scf_env_do_scf 11 5.9 0.000 0.001 48.949 48.949 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.302 40.302 velocity_verlet 10 3.0 0.003 0.006 32.743 32.745 dbcsr_multiply_generic 2055 12.4 0.109 0.114 28.981 29.192 qs_scf_new_mos 99 7.5 0.001 0.001 25.522 25.635 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.521 25.634 ot_scf_mini 99 9.5 0.002 0.003 24.264 24.390 multiply_cannon 2055 13.4 0.211 0.218 22.676 23.878 multiply_cannon_loop 2055 14.4 0.820 0.842 21.478 22.506 ot_mini 99 10.5 0.001 0.001 13.751 13.881 rebuild_ks_matrix 110 8.3 0.000 0.000 12.199 12.331 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.198 12.330 multiply_cannon_multrec 16440 15.4 3.720 4.987 9.983 11.201 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.741 10.859 mp_waitall_1 139946 16.5 6.920 9.955 6.920 9.955 qs_ot_get_derivative 99 11.5 0.001 0.001 9.362 9.489 init_scf_loop 11 6.9 0.000 0.000 8.612 8.613 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.250 7.110 prepare_preconditioner 11 7.9 0.000 0.000 6.861 6.877 make_preconditioner 11 8.9 0.000 0.000 6.861 6.877 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.182 6.535 sum_up_and_integrate 110 10.3 0.001 0.002 6.334 6.349 integrate_v_rspace 110 11.3 0.003 0.003 6.308 6.323 dbcsr_mm_accdrv_process 34862 16.1 5.437 5.865 6.108 6.286 init_scf_run 11 5.9 0.000 0.001 5.895 5.895 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.894 5.894 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.837 5.845 calculate_rho_elec 110 8.6 0.059 0.059 5.836 5.845 qs_ot_get_p 110 10.4 0.001 0.001 5.647 5.785 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.879 5.366 apply_single 110 13.6 0.000 0.000 4.879 5.365 make_m2s 4110 13.4 0.049 0.050 4.262 4.648 make_images 4110 14.4 0.398 0.519 4.145 4.529 ot_diis_step 99 11.5 0.011 0.011 4.361 4.361 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.966 3.968 multiply_cannon_sync_h2d 16440 15.4 3.273 3.861 3.273 3.861 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.237 3.816 pw_transfer 1331 11.6 0.065 0.071 3.684 3.693 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.578 3.588 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.586 3.586 grid_integrate_task_list 110 12.3 3.216 3.377 3.216 3.377 density_rs2pw 110 9.6 0.004 0.004 3.063 3.252 fft_wrap_pw1pw2_140 451 13.1 0.460 0.470 3.073 3.085 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.028 3.038 cp_fm_diag_elpa_base 48 14.0 2.960 2.991 3.026 3.036 wfi_extrapolate 11 7.9 0.001 0.001 2.986 2.986 make_images_data 4110 15.4 0.046 0.050 2.374 2.857 calculate_first_density_matrix 1 7.0 0.000 0.000 2.815 2.816 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.805 2.806 hybrid_alltoall_any 4261 16.3 0.107 0.378 2.089 2.784 cp_fm_cholesky_invert 11 10.9 2.728 2.734 2.728 2.734 calculate_dm_sparse 110 9.5 0.001 0.001 2.678 2.712 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.464 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.479 2.522 fft3d_ps 1111 14.6 1.119 1.134 2.495 2.506 grid_collocate_task_list 110 9.6 2.220 2.433 2.220 2.433 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.862 2.369 mp_irecv_dv 48980 15.7 0.789 2.233 0.789 2.233 potential_pw2rs 110 12.3 0.010 0.011 2.170 2.174 mp_sum_l 6594 12.7 1.413 2.112 1.413 2.112 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.969 1.990 dbcsr_complete_redistribute 325 12.2 0.323 0.373 1.490 1.936 mp_alltoall_d11v 2046 13.8 1.763 1.913 1.763 1.913 cp_fm_upper_to_full 70 14.2 1.420 1.788 1.420 1.788 acc_transpose_blocks 16440 15.4 0.080 0.082 1.629 1.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.652 1.666 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.607 1.618 cp_fm_cholesky_decompose 22 10.9 1.577 1.593 1.577 1.593 mp_allgather_i34 2055 14.4 0.511 1.560 0.511 1.560 transfer_rs2pw 451 10.6 0.005 0.006 1.323 1.510 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.362 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.982 1.429 mp_waitany 17072 13.8 1.135 1.353 1.135 1.353 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.291 1.301 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.792000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.636364, yerr=9.865207 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 739.078144E+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 0.011 0.032 70.035 70.037 qs_mol_dyn_low 1 2.0 0.003 0.003 69.792 69.804 qs_forces 11 3.9 0.003 0.003 69.719 69.719 qs_energies 11 4.9 0.001 0.002 66.264 66.268 scf_env_do_scf 11 5.9 0.000 0.001 57.796 57.798 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 45.073 45.074 velocity_verlet 10 3.0 0.001 0.002 39.355 39.357 dbcsr_multiply_generic 2055 12.4 0.119 0.123 31.672 31.936 qs_scf_new_mos 99 7.5 0.001 0.001 29.844 29.959 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.843 29.958 ot_scf_mini 99 9.5 0.003 0.003 28.116 28.210 multiply_cannon 2055 13.4 0.235 0.255 23.981 25.147 multiply_cannon_loop 2055 14.4 1.439 1.497 22.468 23.140 ot_mini 99 10.5 0.001 0.001 15.833 15.949 multiply_cannon_multrec 24660 15.4 4.054 6.719 13.047 14.082 init_scf_loop 11 6.9 0.000 0.000 12.674 12.674 rebuild_ks_matrix 110 8.3 0.000 0.000 12.314 12.416 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.314 12.415 qs_ot_get_derivative 99 11.5 0.001 0.001 11.567 11.664 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.884 10.973 prepare_preconditioner 11 7.9 0.000 0.000 10.866 10.881 make_preconditioner 11 8.9 0.000 0.000 10.865 10.881 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.037 10.519 dbcsr_mm_accdrv_process 52304 16.0 7.766 9.248 8.836 9.766 qs_ot_get_p 110 10.4 0.001 0.001 7.077 7.214 mp_waitall_1 121746 16.5 4.773 6.805 4.773 6.805 sum_up_and_integrate 110 10.3 0.001 0.002 6.523 6.537 integrate_v_rspace 110 11.3 0.003 0.003 6.497 6.510 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.198 6.208 calculate_rho_elec 110 8.6 0.077 0.081 6.198 6.208 make_m2s 4110 13.4 0.059 0.061 5.850 6.065 make_images 4110 14.4 0.574 0.697 5.708 5.918 init_scf_run 11 5.9 0.000 0.001 5.687 5.687 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.687 5.687 cp_fm_upper_to_full 70 14.2 3.445 4.984 3.445 4.984 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.949 4.965 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.348 4.349 ot_diis_step 99 11.5 0.011 0.011 4.223 4.223 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.113 4.196 apply_single 110 13.6 0.000 0.000 4.113 4.196 pw_transfer 1331 11.6 0.066 0.075 4.031 4.070 dbcsr_complete_redistribute 325 12.2 0.412 0.452 2.877 4.025 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.924 3.966 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.767 3.823 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.672 3.682 cp_fm_diag_elpa_base 48 14.0 3.509 3.573 3.670 3.680 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.582 3.635 grid_integrate_task_list 110 12.3 3.292 3.430 3.292 3.430 density_rs2pw 110 9.6 0.004 0.004 3.282 3.418 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.253 3.403 fft_wrap_pw1pw2_140 451 13.1 0.532 0.546 3.281 3.326 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.564 3.287 cp_fm_cholesky_invert 11 10.9 3.237 3.248 3.237 3.248 make_images_data 4110 15.4 0.049 0.052 2.886 3.185 wfi_extrapolate 11 7.9 0.001 0.001 3.134 3.134 calculate_dm_sparse 110 9.5 0.001 0.001 3.047 3.080 hybrid_alltoall_any 4261 16.3 0.122 0.457 2.443 3.032 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.820 2.946 mp_alltoall_i22 605 13.7 1.707 2.916 1.707 2.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.855 2.857 fft3d_ps 1111 14.6 1.153 1.193 2.751 2.767 acc_transpose_blocks 24660 15.4 0.114 0.117 2.467 2.562 multiply_cannon_sync_h2d 24660 15.4 2.364 2.489 2.364 2.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.482 calculate_first_density_matrix 1 7.0 0.000 0.000 2.453 2.455 grid_collocate_task_list 110 9.6 2.265 2.407 2.265 2.407 potential_pw2rs 110 12.3 0.012 0.013 2.295 2.303 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.256 2.257 mp_alltoall_d11v 2046 13.8 1.979 2.129 1.979 2.129 cp_fm_cholesky_decompose 22 10.9 1.958 2.004 1.958 2.004 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.896 1.926 mp_allgather_i34 2055 14.4 0.600 1.743 0.600 1.743 mp_sum_l 6594 12.7 1.149 1.725 1.149 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.686 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.574 1.588 acc_transpose_blocks_sync 73980 16.4 1.444 1.552 1.444 1.552 transfer_rs2pw 451 10.6 0.005 0.006 1.346 1.519 mp_alltoall_z22v 1111 16.6 1.382 1.455 1.382 1.455 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.850 1.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=70.037000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.545455, yerr=9.670608 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 871.878656E+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 0.011 0.033 57.834 57.835 qs_mol_dyn_low 1 2.0 0.003 0.003 57.584 57.594 qs_forces 11 3.9 0.003 0.003 57.511 57.511 qs_energies 11 4.9 0.001 0.001 53.776 53.780 scf_env_do_scf 11 5.9 0.000 0.001 45.359 45.360 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 37.140 37.141 velocity_verlet 10 3.0 0.001 0.002 32.562 32.565 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.780 23.953 qs_scf_new_mos 99 7.5 0.001 0.001 22.071 22.116 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.071 22.115 ot_scf_mini 99 9.5 0.002 0.002 20.774 20.803 multiply_cannon 2055 13.4 0.237 0.251 17.738 19.170 multiply_cannon_loop 2055 14.4 0.610 0.630 16.398 16.733 rebuild_ks_matrix 110 8.3 0.000 0.000 12.008 12.023 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.008 12.022 ot_mini 99 10.5 0.001 0.001 11.141 11.156 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.674 10.688 multiply_cannon_multrec 8220 15.4 3.160 4.319 7.585 8.531 init_scf_loop 11 6.9 0.000 0.000 8.165 8.167 mp_waitall_1 103326 16.6 6.095 7.739 6.095 7.739 qs_ot_get_derivative 99 11.5 0.001 0.001 7.410 7.437 sum_up_and_integrate 110 10.3 0.001 0.002 6.437 6.449 prepare_preconditioner 11 7.9 0.000 0.000 6.420 6.423 make_preconditioner 11 8.9 0.000 0.000 6.420 6.423 integrate_v_rspace 110 11.3 0.003 0.003 6.410 6.422 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.275 6.300 calculate_rho_elec 110 8.6 0.112 0.113 6.275 6.300 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.981 6.059 qs_ot_get_p 110 10.4 0.001 0.001 5.592 5.615 dbcsr_mm_accdrv_process 17442 15.9 3.181 4.678 4.285 5.205 init_scf_run 11 5.9 0.000 0.001 5.204 5.205 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.204 5.204 make_m2s 4110 13.4 0.037 0.039 4.504 4.779 make_images 4110 14.4 0.648 0.707 4.373 4.649 pw_transfer 1331 11.6 0.066 0.071 4.328 4.342 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.220 4.237 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.874 4.166 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.058 4.063 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.745 3.799 apply_single 110 13.6 0.000 0.000 3.745 3.798 ot_diis_step 99 11.5 0.012 0.012 3.702 3.703 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.690 3.690 fft_wrap_pw1pw2_140 451 13.1 0.718 0.738 3.593 3.608 grid_integrate_task_list 110 12.3 3.379 3.489 3.379 3.489 density_rs2pw 110 9.6 0.004 0.004 3.256 3.398 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.127 3.135 cp_fm_cholesky_invert 11 10.9 3.129 3.134 3.129 3.134 cp_fm_diag_elpa_base 48 14.0 3.068 3.091 3.125 3.133 wfi_extrapolate 11 7.9 0.001 0.001 2.861 2.861 fft3d_ps 1111 14.6 1.297 1.310 2.764 2.784 hybrid_alltoall_any 4261 16.3 0.199 0.850 2.313 2.781 make_images_data 4110 15.4 0.041 0.046 2.392 2.757 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.696 2.697 calculate_dm_sparse 110 9.5 0.001 0.001 2.541 2.570 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.533 2.534 grid_collocate_task_list 110 9.6 2.371 2.529 2.371 2.529 multiply_cannon_sync_h2d 8220 15.4 2.363 2.452 2.363 2.452 calculate_first_density_matrix 1 7.0 0.000 0.000 2.243 2.245 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.168 2.200 potential_pw2rs 110 12.3 0.015 0.015 2.191 2.195 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.145 2.152 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.095 2.110 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 2.019 mp_alltoall_d11v 2046 13.8 1.811 1.970 1.811 1.970 cp_fm_cholesky_decompose 22 10.9 1.825 1.841 1.825 1.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.669 1.671 mp_allgather_i34 2055 14.4 0.508 1.668 0.508 1.668 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.517 1.634 dbcsr_complete_redistribute 325 12.2 0.547 0.580 1.525 1.633 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.486 1.495 acc_transpose_blocks 8220 15.4 0.041 0.042 1.375 1.401 transfer_rs2pw 451 10.6 0.005 0.005 1.211 1.336 qs_create_task_list 11 7.9 0.001 0.001 1.225 1.324 generate_qs_task_list 11 8.9 0.376 0.444 1.224 1.324 mp_alltoall_z22v 1111 16.6 1.233 1.261 1.233 1.261 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.916 1.237 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.206 1.227 mp_waitany 9240 13.8 1.080 1.214 1.080 1.214 jit_kernel_multiply 7 15.5 0.783 1.168 0.783 1.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.835000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=819.636364, yerr=13.261515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 1.399353E+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 0.014 0.033 88.639 88.641 qs_mol_dyn_low 1 2.0 0.003 0.003 88.330 88.339 qs_forces 11 3.9 0.004 0.011 88.249 88.250 qs_energies 11 4.9 0.001 0.002 84.078 84.080 scf_env_do_scf 11 5.9 0.000 0.001 73.925 73.926 velocity_verlet 10 3.0 0.002 0.002 56.442 56.450 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.623 45.624 dbcsr_multiply_generic 2055 12.4 0.120 0.125 29.955 30.096 init_scf_loop 11 6.9 0.000 0.000 28.227 28.230 qs_scf_new_mos 99 7.5 0.001 0.001 27.419 27.506 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.419 27.506 prepare_preconditioner 11 7.9 0.000 0.000 26.190 26.204 make_preconditioner 11 8.9 0.000 0.000 26.190 26.204 ot_scf_mini 99 9.5 0.002 0.002 25.603 25.677 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.491 25.669 multiply_cannon 2055 13.4 0.329 0.358 22.705 23.357 multiply_cannon_loop 2055 14.4 0.823 0.832 20.878 21.255 cp_fm_upper_to_full 70 14.2 12.629 18.057 12.629 18.057 ot_mini 99 10.5 0.001 0.001 14.495 14.570 rebuild_ks_matrix 110 8.3 0.000 0.001 14.133 14.229 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.133 14.229 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.817 12.904 dbcsr_complete_redistribute 325 12.2 1.005 1.028 7.267 10.366 multiply_cannon_multrec 8220 15.4 4.054 4.203 9.798 9.951 qs_ot_get_derivative 99 11.5 0.001 0.001 9.708 9.783 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.250 9.356 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.684 8.759 mp_waitall_1 84994 16.7 7.721 8.622 7.721 8.622 mp_alltoall_i22 605 13.7 5.329 8.404 5.329 8.404 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.604 7.642 calculate_rho_elec 110 8.6 0.222 0.223 7.604 7.642 sum_up_and_integrate 110 10.3 0.002 0.002 7.067 7.084 integrate_v_rspace 110 11.3 0.003 0.003 7.039 7.056 make_m2s 4110 13.4 0.043 0.043 5.407 5.925 qs_ot_get_p 110 10.4 0.001 0.001 5.776 5.869 init_scf_run 11 5.9 0.000 0.001 5.863 5.863 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.863 5.863 pw_transfer 1331 11.6 0.074 0.075 5.847 5.853 dbcsr_mm_accdrv_process 11614 15.7 4.012 4.914 5.595 5.829 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.730 5.736 make_images 4110 14.4 0.881 0.939 5.217 5.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.902 5.501 apply_single 110 13.6 0.000 0.000 4.902 5.501 cp_fm_cholesky_invert 11 10.9 5.351 5.355 5.351 5.355 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.004 5.341 fft_wrap_pw1pw2_140 451 13.1 1.366 1.369 5.033 5.039 ot_diis_step 99 11.5 0.015 0.016 4.760 4.760 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.131 4.139 density_rs2pw 110 9.6 0.004 0.004 4.015 4.073 grid_integrate_task_list 110 12.3 3.737 3.843 3.737 3.843 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.668 3.669 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.137 3.624 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.608 3.608 hybrid_alltoall_any 4261 16.3 0.264 0.575 2.868 3.605 make_images_data 4110 15.4 0.045 0.048 2.856 3.517 fft3d_ps 1111 14.6 1.881 1.889 3.437 3.444 wfi_extrapolate 11 7.9 0.001 0.001 3.386 3.386 calculate_dm_sparse 110 9.5 0.001 0.001 3.203 3.229 multiply_cannon_sync_h2d 8220 15.4 3.130 3.149 3.130 3.149 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.006 3.006 cp_fm_diag_elpa_base 48 14.0 2.469 2.660 3.005 3.005 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.931 2.935 grid_collocate_task_list 110 9.6 2.715 2.780 2.715 2.780 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.682 2.725 potential_pw2rs 110 12.3 0.021 0.021 2.469 2.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.411 2.444 calculate_first_density_matrix 1 7.0 0.000 0.000 2.359 2.360 mp_alltoall_d11v 2046 13.8 2.136 2.253 2.136 2.253 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.191 2.253 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.176 cp_fm_cholesky_decompose 22 10.9 1.993 2.003 1.993 2.003 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.915 1.925 qs_create_task_list 11 7.9 0.001 0.001 1.879 1.924 generate_qs_task_list 11 8.9 0.732 0.786 1.878 1.923 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.784 1.829 acc_transpose_blocks 8220 15.4 0.042 0.042 1.759 1.810 jit_kernel_multiply 10 15.3 1.384 1.790 1.384 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.641000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1256.818182, yerr=61.530212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 634.163200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+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 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 57774. MP_Allreduce 11082 797. MP_Sync 86 MP_Alltoall 2210 1571380. 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 0.043 0.099 218.750 218.752 qs_mol_dyn_low 1 2.0 0.008 0.036 217.784 217.797 qs_forces 11 3.9 0.005 0.008 217.620 217.622 qs_energies 11 4.9 0.005 0.022 211.843 211.858 scf_env_do_scf 11 5.9 0.001 0.001 194.771 194.775 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 172.480 172.482 qs_scf_new_mos 116 7.6 0.001 0.001 131.111 131.332 qs_scf_loop_do_ot 116 8.6 0.001 0.001 131.111 131.331 velocity_verlet 10 3.0 0.003 0.011 128.496 128.499 dbcsr_multiply_generic 2485 12.5 0.176 0.180 126.429 126.922 ot_scf_mini 116 9.6 0.003 0.004 124.413 124.625 multiply_cannon 2485 13.5 0.235 0.246 101.375 102.595 multiply_cannon_loop 2485 14.5 2.434 2.487 98.949 100.365 ot_mini 116 10.6 0.001 0.002 67.404 67.570 qs_ot_get_derivative 116 11.6 0.001 0.002 42.747 42.917 multiply_cannon_multrec 59640 15.5 31.420 33.298 41.338 42.813 qs_ot_get_p 127 10.4 0.001 0.002 34.891 35.102 rebuild_ks_matrix 127 8.3 0.001 0.001 34.559 34.794 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.021 34.558 34.794 mp_waitall_1 264810 16.5 29.891 32.221 29.891 32.221 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.007 31.212 qs_ot_p2m_diag 82 11.4 0.078 0.091 27.926 28.014 multiply_cannon_sync_h2d 59640 15.5 25.920 27.541 25.920 27.541 cp_dbcsr_syevd 82 12.4 0.005 0.005 25.088 25.090 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.211 25.055 apply_single 127 13.6 0.001 0.001 24.211 25.054 ot_diis_step 116 11.6 0.008 0.010 24.374 24.376 init_scf_loop 11 6.9 0.001 0.005 22.201 22.203 cp_fm_diag_elpa 82 13.4 0.000 0.000 21.652 21.679 cp_fm_diag_elpa_base 82 14.4 21.546 21.599 21.646 21.675 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 20.591 20.701 multiply_cannon_metrocomm3 59640 15.5 0.116 0.120 16.241 18.206 prepare_preconditioner 11 7.9 0.000 0.000 17.494 17.531 make_preconditioner 11 8.9 0.000 0.001 17.494 17.531 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.656 16.840 sum_up_and_integrate 127 10.3 0.003 0.005 14.956 14.976 integrate_v_rspace 127 11.3 0.003 0.004 14.897 14.920 make_m2s 4970 13.5 0.104 0.111 14.558 14.897 make_images 4970 14.5 0.398 0.418 14.375 14.721 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.453 14.608 calculate_rho_elec 127 8.7 0.045 0.064 14.453 14.607 init_scf_run 11 5.9 0.000 0.001 12.763 12.764 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.763 12.764 mp_sum_l 7884 12.9 9.263 10.591 9.263 10.591 dbcsr_mm_accdrv_process 123452 16.2 4.858 4.992 9.488 10.022 cp_fm_cholesky_invert 11 10.9 9.913 9.920 9.913 9.920 wfi_extrapolate 11 7.9 0.001 0.002 9.491 9.491 pw_transfer 1535 11.6 0.074 0.089 8.552 8.801 calculate_dm_sparse 127 9.5 0.001 0.001 8.552 8.629 density_rs2pw 127 9.7 0.006 0.006 8.003 8.627 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.351 8.604 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.368 8.524 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.069 8.199 make_images_data 4970 15.5 0.066 0.072 7.035 7.993 multiply_cannon_metrocomm1 59640 15.5 0.094 0.098 6.293 7.561 grid_integrate_task_list 127 12.3 6.982 7.484 6.982 7.484 hybrid_alltoall_any 5155 16.4 0.292 2.275 6.142 7.404 fft_wrap_pw1pw2_140 519 13.2 0.845 0.884 6.941 7.188 fft3d_ps 1281 14.7 2.186 2.833 6.385 6.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.720 6.729 mp_alltoall_d11v 2401 14.1 4.913 6.411 4.913 6.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.773 5.833 potential_pw2rs 127 12.3 0.008 0.010 5.385 5.447 grid_collocate_task_list 127 9.7 4.795 5.339 4.795 5.339 cp_fm_cholesky_decompose 22 10.9 5.248 5.262 5.248 5.262 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=218.752000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.067118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 841.621504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 5347915. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 191.328 191.330 qs_mol_dyn_low 1 2.0 0.003 0.004 190.852 190.865 qs_forces 11 3.9 0.005 0.015 190.726 190.727 qs_energies 11 4.9 0.010 0.045 183.941 183.951 scf_env_do_scf 11 5.9 0.001 0.002 167.272 167.283 scf_env_do_scf_inner_loop 116 6.6 0.004 0.011 133.918 133.920 velocity_verlet 10 3.0 0.001 0.002 119.461 119.464 dbcsr_multiply_generic 2485 12.5 0.185 0.191 97.129 98.495 qs_scf_new_mos 116 7.6 0.001 0.001 95.500 96.138 qs_scf_loop_do_ot 116 8.6 0.001 0.001 95.499 96.137 ot_scf_mini 116 9.6 0.004 0.005 90.707 91.395 multiply_cannon 2485 13.5 0.497 0.556 76.754 81.430 multiply_cannon_loop 2485 14.5 1.558 1.629 73.353 76.229 ot_mini 116 10.6 0.001 0.001 50.387 51.005 mp_waitall_1 212858 16.6 24.254 38.239 24.254 38.239 multiply_cannon_multrec 29820 15.5 20.741 25.840 31.338 36.709 init_scf_loop 11 6.9 0.003 0.021 33.257 33.259 rebuild_ks_matrix 127 8.3 0.001 0.001 32.157 32.817 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.029 32.157 32.816 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.889 29.483 qs_ot_get_derivative 116 11.6 0.004 0.021 28.343 29.023 prepare_preconditioner 11 7.9 0.000 0.000 28.890 28.953 make_preconditioner 11 8.9 0.000 0.001 28.890 28.952 multiply_cannon_metrocomm3 29820 15.5 0.097 0.103 15.696 28.510 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.524 28.061 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.180 23.412 apply_single 127 13.6 0.001 0.001 22.180 23.412 qs_ot_get_p 127 10.4 0.001 0.002 22.448 23.000 ot_diis_step 116 11.6 0.014 0.015 21.875 21.877 multiply_cannon_sync_h2d 29820 15.5 17.735 19.831 17.735 19.831 qs_ot_p2m_diag 82 11.4 0.186 0.214 17.518 17.557 cp_fm_cholesky_invert 11 10.9 16.776 16.789 16.776 16.789 cp_dbcsr_syevd 82 12.4 0.005 0.006 16.291 16.292 make_m2s 4970 13.5 0.085 0.092 14.377 15.973 make_images 4970 14.5 1.161 1.357 14.163 15.755 sum_up_and_integrate 127 10.3 0.002 0.004 13.999 14.027 integrate_v_rspace 127 11.3 0.003 0.004 13.940 13.972 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.423 13.460 calculate_rho_elec 127 8.7 0.086 0.103 13.422 13.459 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.048 13.078 cp_fm_diag_elpa_base 82 14.4 12.793 12.888 13.042 13.067 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 11.164 11.673 init_scf_run 11 5.9 0.000 0.001 11.660 11.663 scf_env_initial_rho_setup 11 6.9 0.014 0.104 11.660 11.663 multiply_cannon_metrocomm4 27335 15.5 0.099 0.113 3.816 11.001 dbcsr_mm_accdrv_process 61726 16.2 5.548 6.083 10.053 10.613 mp_irecv_dv 68888 16.3 3.619 10.596 3.619 10.596 make_images_data 4970 15.5 0.066 0.076 8.466 10.327 hybrid_alltoall_any 5155 16.4 0.345 1.501 7.258 9.821 pw_transfer 1535 11.6 0.084 0.098 8.669 8.733 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.447 8.517 wfi_extrapolate 11 7.9 0.001 0.001 8.421 8.422 density_rs2pw 127 9.7 0.006 0.006 7.163 7.504 fft_wrap_pw1pw2_140 519 13.2 0.926 0.947 7.397 7.486 grid_integrate_task_list 127 12.3 7.103 7.460 7.103 7.460 cp_fm_cholesky_decompose 22 10.9 7.098 7.184 7.098 7.184 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.327 7.073 calculate_dm_sparse 127 9.5 0.001 0.001 6.459 6.614 mp_sum_l 7884 12.9 4.222 6.417 4.222 6.417 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.085 6.093 fft3d_ps 1281 14.7 2.807 2.993 6.023 6.081 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.383 5.452 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.175 5.368 grid_collocate_task_list 127 9.7 4.964 5.340 4.964 5.340 mp_allgather_i34 2485 14.5 1.934 5.126 1.934 5.126 mp_alltoall_d11v 2401 14.1 4.300 4.890 4.300 4.890 potential_pw2rs 127 12.3 0.015 0.017 4.571 4.597 dbcsr_complete_redistribute 393 12.7 0.766 0.856 3.379 4.200 mp_sum_d 4454 12.1 2.727 4.196 2.727 4.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.330000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.545455, yerr=2.934548 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 964.472832E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.099 174.842 174.844 qs_mol_dyn_low 1 2.0 0.003 0.007 174.262 174.281 qs_forces 11 3.9 0.004 0.004 174.158 174.159 qs_energies 11 4.9 0.003 0.010 167.608 167.616 scf_env_do_scf 11 5.9 0.001 0.002 151.817 151.818 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.753 116.754 velocity_verlet 10 3.0 0.001 0.002 111.612 111.615 dbcsr_multiply_generic 2485 12.5 0.181 0.186 81.573 82.659 qs_scf_new_mos 116 7.6 0.001 0.001 79.971 80.324 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.970 80.323 ot_scf_mini 116 9.6 0.004 0.007 75.740 76.151 multiply_cannon 2485 13.5 0.500 0.529 61.925 66.308 multiply_cannon_loop 2485 14.5 1.126 1.198 58.909 61.277 ot_mini 116 10.6 0.001 0.001 42.348 42.758 init_scf_loop 11 6.9 0.003 0.022 34.965 34.966 mp_waitall_1 169034 16.6 24.914 34.791 24.914 34.791 prepare_preconditioner 11 7.9 0.000 0.001 30.929 30.973 make_preconditioner 11 8.9 0.000 0.002 30.928 30.973 rebuild_ks_matrix 127 8.3 0.001 0.001 30.072 30.563 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.026 30.072 30.563 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.489 29.928 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.071 27.525 multiply_cannon_multrec 19880 15.5 12.848 15.941 22.203 25.341 multiply_cannon_metrocomm3 19880 15.5 0.061 0.065 15.531 24.851 qs_ot_get_derivative 116 11.6 0.001 0.002 22.658 23.062 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.745 20.933 apply_single 127 13.6 0.001 0.001 19.745 20.933 ot_diis_step 116 11.6 0.017 0.018 19.582 19.582 qs_ot_get_p 127 10.4 0.001 0.002 18.749 19.202 make_m2s 4970 13.5 0.077 0.081 14.733 15.969 make_images 4970 14.5 1.141 1.229 14.500 15.728 multiply_cannon_sync_h2d 19880 15.5 13.404 15.121 13.404 15.121 qs_ot_p2m_diag 82 11.4 0.262 0.270 14.569 14.576 cp_fm_cholesky_invert 11 10.9 14.364 14.372 14.364 14.372 sum_up_and_integrate 127 10.3 0.002 0.005 13.850 13.875 integrate_v_rspace 127 11.3 0.003 0.003 13.791 13.818 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.519 13.520 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.445 13.479 calculate_rho_elec 127 8.7 0.128 0.142 13.444 13.478 init_scf_run 11 5.9 0.000 0.001 10.710 10.711 scf_env_initial_rho_setup 11 6.9 0.002 0.003 10.710 10.711 make_images_data 4970 15.5 0.064 0.074 8.888 10.469 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.326 10.346 cp_fm_diag_elpa_base 82 14.4 9.920 10.056 10.322 10.343 hybrid_alltoall_any 5155 16.4 0.443 2.028 7.606 10.027 multiply_cannon_metrocomm4 17395 15.5 0.064 0.074 3.491 9.429 mp_irecv_dv 49801 16.2 3.364 9.171 3.364 9.171 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.869 9.162 dbcsr_mm_accdrv_process 41158 16.2 5.695 5.977 8.818 8.984 pw_transfer 1535 11.6 0.084 0.102 8.755 8.845 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.533 8.630 grid_integrate_task_list 127 12.3 7.243 7.804 7.243 7.804 fft_wrap_pw1pw2_140 519 13.2 1.054 1.092 7.567 7.671 wfi_extrapolate 11 7.9 0.001 0.001 7.508 7.509 cp_fm_cholesky_decompose 22 10.9 7.335 7.383 7.335 7.383 cp_fm_upper_to_full 104 14.8 5.737 7.307 5.737 7.307 density_rs2pw 127 9.7 0.005 0.006 6.888 7.224 dbcsr_complete_redistribute 393 12.7 1.160 1.196 4.684 6.443 fft3d_ps 1281 14.7 2.798 3.029 5.866 5.962 calculate_dm_sparse 127 9.5 0.001 0.001 5.831 5.959 grid_collocate_task_list 127 9.7 5.111 5.647 5.111 5.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.482 5.488 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.514 5.277 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.631 5.263 mp_alltoall_d11v 2401 14.1 4.342 4.825 4.342 4.825 mp_allgather_i34 2485 14.5 1.527 4.813 1.527 4.813 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.560 4.705 mp_sum_l 7884 12.9 3.107 4.477 3.107 4.477 potential_pw2rs 127 12.3 0.019 0.021 4.357 4.366 transfer_fm_to_dbcsr 11 9.9 0.019 0.025 2.420 4.175 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.938 3.973 mp_alltoall_i22 712 14.1 1.998 3.876 1.998 3.876 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.771 3.771 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.496 3.535 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.844000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=906.090909, yerr=18.691051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.143439E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496376. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.064 0.080 192.734 192.787 qs_mol_dyn_low 1 2.0 0.003 0.004 192.113 192.128 qs_forces 11 3.9 0.004 0.004 191.991 192.000 qs_energies 11 4.9 0.001 0.002 184.810 184.819 scf_env_do_scf 11 5.9 0.001 0.001 167.835 167.849 velocity_verlet 10 3.0 0.002 0.002 126.873 126.877 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 120.253 120.255 qs_scf_new_mos 116 7.6 0.001 0.001 83.627 83.967 qs_scf_loop_do_ot 116 8.6 0.001 0.001 83.627 83.966 dbcsr_multiply_generic 2485 12.5 0.190 0.195 80.887 81.602 ot_scf_mini 116 9.6 0.003 0.004 79.075 79.420 multiply_cannon 2485 13.5 0.543 0.583 55.703 58.910 multiply_cannon_loop 2485 14.5 1.797 1.914 51.890 53.629 init_scf_loop 11 6.9 0.000 0.000 47.455 47.457 ot_mini 116 10.6 0.001 0.001 43.352 43.697 prepare_preconditioner 11 7.9 0.000 0.000 43.339 43.373 make_preconditioner 11 8.9 0.000 0.000 43.339 43.373 make_full_inverse_cholesky 11 9.9 0.013 0.021 36.886 41.936 multiply_cannon_multrec 29820 15.5 13.334 18.701 26.260 31.346 rebuild_ks_matrix 127 8.3 0.001 0.001 29.522 29.864 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.521 29.864 mp_waitall_1 146592 16.7 17.700 27.107 17.700 27.107 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.616 26.924 qs_ot_get_derivative 116 11.6 0.001 0.002 23.689 24.038 qs_ot_get_p 127 10.4 0.001 0.001 21.815 22.202 make_m2s 4970 13.5 0.092 0.096 20.708 21.886 make_images 4970 14.5 1.942 2.253 20.400 21.575 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.135 19.723 apply_single 127 13.6 0.001 0.001 19.135 19.723 ot_diis_step 116 11.6 0.017 0.018 19.546 19.547 qs_ot_p2m_diag 82 11.4 0.338 0.385 17.518 17.570 cp_fm_upper_to_full 104 14.8 11.642 17.104 11.642 17.104 cp_fm_cholesky_invert 11 10.9 16.714 16.724 16.714 16.724 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.086 16.086 multiply_cannon_metrocomm3 29820 15.5 0.048 0.050 6.582 14.863 sum_up_and_integrate 127 10.3 0.002 0.003 14.159 14.187 integrate_v_rspace 127 11.3 0.003 0.004 14.100 14.130 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.967 14.020 calculate_rho_elec 127 8.7 0.168 0.183 13.966 14.020 dbcsr_complete_redistribute 393 12.7 1.489 1.599 9.343 13.098 dbcsr_mm_accdrv_process 61748 16.2 8.455 9.221 12.499 13.012 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.717 12.733 cp_fm_diag_elpa_base 82 14.4 11.734 12.056 12.709 12.723 make_images_data 4970 15.5 0.065 0.071 11.158 12.643 hybrid_alltoall_any 5155 16.4 0.524 2.181 9.927 11.966 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.872 11.634 multiply_cannon_sync_h2d 29820 15.5 10.382 11.190 10.382 11.190 init_scf_run 11 5.9 0.000 0.001 11.018 11.019 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.018 11.019 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.873 10.134 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.431 10.110 pw_transfer 1535 11.6 0.085 0.101 9.362 9.435 mp_alltoall_i22 712 14.1 5.660 9.352 5.660 9.352 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.137 9.213 fft_wrap_pw1pw2_140 519 13.2 1.240 1.287 8.066 8.153 wfi_extrapolate 11 7.9 0.001 0.001 7.900 7.901 cp_fm_cholesky_decompose 22 10.9 7.801 7.895 7.801 7.895 grid_integrate_task_list 127 12.3 7.508 7.879 7.508 7.879 density_rs2pw 127 9.7 0.005 0.005 7.021 7.401 multiply_cannon_metrocomm4 24850 15.5 0.079 0.089 2.829 7.393 mp_irecv_dv 75445 16.2 2.679 7.122 2.679 7.122 calculate_dm_sparse 127 9.5 0.001 0.001 6.238 6.335 fft3d_ps 1281 14.7 2.964 3.017 6.207 6.265 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.547 5.584 mp_alltoall_d11v 2401 14.1 5.102 5.568 5.102 5.568 grid_collocate_task_list 127 9.7 5.241 5.516 5.241 5.516 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.608 4.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.406 4.530 potential_pw2rs 127 12.3 0.023 0.024 4.473 4.488 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.442 4.443 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.280 4.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.787000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.727273, yerr=13.280819 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.524015E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.035 176.217 176.218 qs_mol_dyn_low 1 2.0 0.003 0.003 175.811 175.827 qs_forces 11 3.9 0.004 0.004 175.691 175.695 qs_energies 11 4.9 0.058 0.064 168.268 168.276 scf_env_do_scf 11 5.9 0.001 0.001 150.713 150.720 velocity_verlet 10 3.0 0.002 0.002 115.137 115.143 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.116 114.117 qs_scf_new_mos 117 7.6 0.001 0.001 76.328 76.490 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.328 76.490 dbcsr_multiply_generic 2507 12.6 0.183 0.190 73.673 74.105 ot_scf_mini 117 9.6 0.003 0.004 71.863 72.011 multiply_cannon 2507 13.6 0.558 0.583 54.230 57.432 multiply_cannon_loop 2507 14.6 0.813 0.838 51.108 51.951 ot_mini 117 10.6 0.001 0.001 38.645 38.803 init_scf_loop 11 6.9 0.000 0.000 36.446 36.448 prepare_preconditioner 11 7.9 0.000 0.000 32.521 32.548 make_preconditioner 11 8.9 0.000 0.000 32.521 32.548 mp_waitall_1 125778 16.7 25.986 32.415 25.986 32.415 make_full_inverse_cholesky 11 9.9 0.014 0.026 30.450 30.715 rebuild_ks_matrix 128 8.3 0.001 0.001 29.664 29.755 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 29.663 29.754 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.084 27.168 multiply_cannon_multrec 10028 15.6 10.338 14.361 17.952 22.352 qs_ot_get_derivative 117 11.6 0.002 0.002 21.250 21.393 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 13.233 20.640 qs_ot_get_p 128 10.4 0.001 0.001 20.233 20.387 cp_fm_cholesky_invert 11 10.9 18.675 18.681 18.675 18.681 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.423 17.573 apply_single 128 13.6 0.001 0.001 17.423 17.572 ot_diis_step 117 11.6 0.020 0.020 17.316 17.316 make_m2s 5014 13.6 0.064 0.068 15.399 16.300 qs_ot_p2m_diag 83 11.4 0.495 0.501 16.267 16.283 make_images 5014 14.6 2.164 2.592 15.092 15.992 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.048 15.049 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.563 14.604 calculate_rho_elec 128 8.7 0.253 0.264 14.563 14.603 sum_up_and_integrate 128 10.3 0.002 0.002 14.435 14.476 integrate_v_rspace 128 11.3 0.004 0.004 14.375 14.418 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.958 11.968 cp_fm_diag_elpa_base 83 14.4 11.708 11.791 11.953 11.964 multiply_cannon_sync_h2d 10028 15.6 10.681 11.329 10.681 11.329 init_scf_run 11 5.9 0.000 0.001 10.611 10.612 scf_env_initial_rho_setup 11 6.9 0.011 0.017 10.611 10.611 make_images_data 5014 15.6 0.055 0.066 8.934 10.291 pw_transfer 1547 11.6 0.085 0.094 10.149 10.182 hybrid_alltoall_any 5200 16.5 0.845 3.779 8.701 10.058 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.927 9.966 fft_wrap_pw1pw2_140 523 13.2 1.672 1.707 8.705 8.743 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.405 8.511 cp_fm_cholesky_decompose 22 10.9 8.194 8.314 8.194 8.314 grid_integrate_task_list 128 12.3 7.789 8.133 7.789 8.133 dbcsr_mm_accdrv_process 20762 16.1 3.275 4.514 7.269 7.970 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.496 7.759 wfi_extrapolate 11 7.9 0.001 0.001 7.466 7.466 density_rs2pw 128 9.7 0.005 0.005 7.203 7.461 fft3d_ps 1291 14.7 3.140 3.249 6.383 6.418 calculate_dm_sparse 128 9.5 0.001 0.001 6.231 6.303 grid_collocate_task_list 128 9.7 5.572 5.809 5.572 5.809 dbcsr_complete_redistribute 395 12.7 2.067 2.110 5.276 5.717 multiply_cannon_metrocomm4 7521 15.6 0.025 0.029 1.849 5.431 mp_irecv_dv 28860 15.9 1.811 5.352 1.811 5.352 qs_energies_init_hamiltonians 11 5.9 0.007 0.009 5.322 5.322 mp_alltoall_d11v 2415 14.1 4.801 5.180 4.801 5.180 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.165 5.173 mp_allgather_i34 2507 14.6 1.316 5.018 1.316 5.018 potential_pw2rs 128 12.3 0.026 0.026 4.554 4.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.067 4.134 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.588 3.864 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.738 3.794 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.735 3.769 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.539 3.714 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.366 3.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.218000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1438.545455, yerr=12.550724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.185119E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3967 59639. MP_Allreduce 10995 1506. MP_Sync 86 MP_Alltoall 1700 36954412. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.041 308.794 308.795 qs_mol_dyn_low 1 2.0 0.003 0.004 308.255 308.269 qs_forces 11 3.9 0.004 0.004 308.149 308.151 qs_energies 11 4.9 0.054 0.056 298.960 298.965 scf_env_do_scf 11 5.9 0.001 0.001 275.507 275.511 velocity_verlet 10 3.0 0.002 0.002 221.252 221.261 scf_env_do_scf_inner_loop 116 6.6 0.005 0.010 145.576 145.579 init_scf_loop 11 6.9 0.000 0.000 129.640 129.641 prepare_preconditioner 11 7.9 0.000 0.000 124.510 124.539 make_preconditioner 11 8.9 0.000 0.000 124.510 124.539 make_full_inverse_cholesky 11 9.9 0.038 0.039 100.828 121.584 qs_scf_new_mos 116 7.6 0.001 0.001 98.340 98.478 qs_scf_loop_do_ot 116 8.6 0.001 0.001 98.339 98.477 ot_scf_mini 116 9.6 0.004 0.004 93.392 93.503 dbcsr_multiply_generic 2485 12.5 0.213 0.223 85.188 85.841 cp_fm_upper_to_full 104 14.8 52.548 75.142 52.548 75.142 multiply_cannon 2485 13.5 0.678 0.727 60.431 61.687 multiply_cannon_loop 2485 14.5 1.044 1.065 56.093 57.759 ot_mini 116 10.6 0.001 0.001 46.649 46.785 dbcsr_complete_redistribute 393 12.7 3.965 4.004 29.991 42.773 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.300 39.051 cp_fm_cholesky_invert 11 10.9 37.677 37.684 37.677 37.684 rebuild_ks_matrix 127 8.3 0.001 0.001 36.428 36.517 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 36.428 36.516 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.640 36.188 mp_waitall_1 102768 16.8 30.468 35.959 30.468 35.959 mp_alltoall_i22 712 14.1 21.355 34.241 21.355 34.241 qs_ks_update_qs_env 127 7.6 0.001 0.001 33.713 33.801 qs_ot_get_p 127 10.4 0.001 0.001 31.152 31.277 qs_ot_p2m_diag 82 11.4 0.868 0.872 26.567 26.594 qs_ot_get_derivative 116 11.6 0.002 0.002 26.318 26.429 cp_dbcsr_syevd 82 12.4 0.006 0.006 24.713 24.717 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 20.115 21.823 make_m2s 4970 13.5 0.073 0.076 19.860 21.500 cp_fm_diag_elpa 82 13.4 0.000 0.000 21.028 21.029 cp_fm_diag_elpa_base 82 14.4 16.637 18.278 21.023 21.024 make_images 4970 14.5 3.042 3.245 19.382 21.022 ot_diis_step 116 11.6 0.023 0.032 20.272 20.272 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.577 19.804 apply_single 127 13.6 0.001 0.001 19.577 19.803 qs_rho_update_rho_low 127 7.7 0.001 0.001 18.531 18.552 calculate_rho_elec 127 8.7 0.473 0.474 18.530 18.551 multiply_cannon_multrec 9940 15.5 10.297 12.121 18.109 18.169 sum_up_and_integrate 127 10.3 0.002 0.002 17.599 17.688 integrate_v_rspace 127 11.3 0.004 0.004 17.537 17.625 pw_transfer 1535 11.6 0.094 0.095 14.358 14.371 multiply_cannon_sync_h2d 9940 15.5 14.188 14.214 14.188 14.214 hybrid_alltoall_any 5155 16.4 1.301 3.033 11.375 14.138 make_images_data 4970 15.5 0.063 0.069 11.309 14.136 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 14.121 14.134 init_scf_run 11 5.9 0.000 0.001 12.864 12.865 scf_env_initial_rho_setup 11 6.9 0.002 0.002 12.864 12.864 fft_wrap_pw1pw2_140 519 13.2 3.206 3.254 12.426 12.433 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.377 10.465 cp_fm_cholesky_decompose 22 10.9 10.424 10.437 10.424 10.437 wfi_extrapolate 11 7.9 0.001 0.001 9.668 9.668 dbcsr_mm_accdrv_process 20590 16.0 4.340 6.220 7.561 9.473 density_rs2pw 127 9.7 0.005 0.005 9.312 9.360 grid_integrate_task_list 127 12.3 8.520 8.740 8.520 8.740 fft3d_ps 1281 14.7 3.951 3.970 8.616 8.662 qs_energies_init_hamiltonians 11 5.9 0.013 0.024 8.258 8.259 mp_alltoall_d11v 2401 14.1 8.058 8.185 8.058 8.185 calculate_dm_sparse 127 9.5 0.001 0.001 6.862 6.921 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.623 6.661 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.420 6.496 grid_collocate_task_list 127 9.7 6.382 6.473 6.382 6.473 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=308.795000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2893.000000, yerr=155.472886 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 1.260364E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.069 85.193 85.194 qs_energies 1 2.0 0.000 0.001 84.648 84.654 ls_scf 1 3.0 0.000 0.003 83.731 83.736 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.583 72.719 multiply_cannon 111 7.7 0.017 0.020 55.958 56.861 multiply_cannon_loop 111 8.7 0.230 0.266 52.539 53.583 ls_scf_main 1 4.0 0.000 0.003 52.279 52.280 density_matrix_trs4 2 5.0 0.002 0.004 46.699 46.750 ls_scf_init_scf 1 4.0 0.000 0.003 28.429 28.433 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.284 27.340 mp_waitall_1 11031 10.9 22.496 25.619 22.496 25.619 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.179 25.192 multiply_cannon_multrec 2664 9.7 8.137 8.885 15.601 17.350 multiply_cannon_sync_h2d 2664 9.7 13.548 15.675 13.548 15.675 make_m2s 222 7.7 0.009 0.011 13.105 13.514 make_images 222 8.7 0.099 0.108 13.082 13.493 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.684 11.887 make_images_data 222 9.7 0.004 0.005 7.675 8.263 dbcsr_mm_accdrv_process 4760 10.4 0.604 0.731 7.083 8.140 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.461 8.098 hybrid_alltoall_any 227 10.6 0.216 1.846 6.596 8.015 dbcsr_mm_accdrv_process_sort 4760 11.4 6.277 7.219 6.277 7.219 calculate_norms 4752 9.8 5.507 6.157 5.507 6.157 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.187 mp_sum_l 887 5.1 3.015 4.471 3.015 4.471 make_images_sizes 222 9.7 0.000 0.000 0.711 3.689 mp_alltoall_i44 222 10.7 0.710 3.689 0.710 3.689 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.060 3.643 mp_irecv_dv 6231 10.9 2.043 3.613 2.043 3.613 arnoldi_extremal 4 6.8 0.000 0.000 3.306 3.327 arnoldi_normal_ev 4 7.8 0.001 0.003 3.306 3.327 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.220 3.271 build_subspace 16 8.4 0.009 0.013 3.209 3.211 ls_scf_post 1 4.0 0.000 0.001 3.023 3.031 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.886 dbcsr_special_finalize 555 9.7 0.005 0.006 2.366 2.761 dbcsr_merge_single_wm 555 10.7 0.468 0.597 2.358 2.752 make_images_pack 222 9.7 2.205 2.623 2.207 2.624 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.322 2.569 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.457 2.065 2.459 dbcsr_sort_data 658 11.4 2.146 2.459 2.146 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.335 2.413 buffer_matrices_ensure_size 222 8.7 1.779 2.108 1.779 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.785 1.791 rebuild_ks_matrix 3 7.3 0.000 0.000 1.776 1.782 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.008 1.776 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.194000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 2.131341E+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 0.043 0.052 90.676 90.678 qs_energies 1 2.0 0.000 0.001 90.117 90.139 ls_scf 1 3.0 0.000 0.004 88.724 88.749 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.781 75.082 multiply_cannon 111 7.7 0.028 0.065 52.876 57.612 ls_scf_main 1 4.0 0.000 0.004 55.182 55.189 multiply_cannon_loop 111 8.7 0.136 0.146 50.026 53.530 density_matrix_trs4 2 5.0 0.002 0.004 49.470 49.669 mp_waitall_1 9105 10.9 20.941 30.396 20.941 30.396 ls_scf_init_scf 1 4.0 0.000 0.003 29.856 29.861 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.664 28.756 multiply_cannon_multrec 1332 9.7 13.341 17.508 22.684 28.011 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.267 26.278 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.508 21.021 make_m2s 222 7.7 0.006 0.007 15.250 15.861 make_images 222 8.7 1.367 1.701 15.219 15.831 dbcsr_mm_accdrv_process 4041 10.4 0.344 0.536 8.940 10.462 dbcsr_mm_accdrv_process_sort 4041 11.4 8.457 9.926 8.457 9.926 make_images_data 222 9.7 0.004 0.005 8.888 9.854 hybrid_alltoall_any 227 10.6 0.544 2.552 8.268 9.606 mp_sum_l 887 5.1 5.181 8.479 5.181 8.479 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.182 7.835 mp_irecv_dv 3311 11.0 3.162 7.785 3.162 7.785 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.946 6.914 calculate_norms 2376 9.8 6.067 6.710 6.067 6.710 multiply_cannon_sync_h2d 1332 9.7 4.902 6.468 4.902 6.468 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.946 5.166 arnoldi_extremal 4 6.8 0.000 0.000 4.718 4.734 arnoldi_normal_ev 4 7.8 0.001 0.005 4.718 4.734 build_subspace 16 8.4 0.014 0.021 4.469 4.472 ls_scf_post 1 4.0 0.000 0.002 3.686 3.709 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.506 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.181 3.409 dbcsr_matrix_vector_mult_local 304 10.0 2.766 3.243 2.768 3.245 mp_allgather_i34 111 8.7 0.840 3.114 0.840 3.114 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.611 2.701 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.187 2.693 dbcsr_data_new 4174 10.1 2.117 2.429 2.117 2.429 make_images_pack 222 9.7 1.816 2.127 1.819 2.129 dbcsr_sort_data 436 11.2 1.850 2.072 1.850 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.858 1.863 rebuild_ks_matrix 3 7.3 0.000 0.000 1.845 1.850 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.008 1.845 1.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.678000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1771.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 2.926416E+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 0.046 93.094 93.095 qs_energies 1 2.0 0.000 0.000 92.567 92.572 ls_scf 1 3.0 0.000 0.000 91.142 91.145 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.664 75.969 multiply_cannon 111 7.7 0.039 0.104 52.294 57.015 ls_scf_main 1 4.0 0.000 0.000 56.843 56.846 multiply_cannon_loop 111 8.7 0.116 0.129 49.475 53.223 density_matrix_trs4 2 5.0 0.002 0.003 50.948 51.116 mp_waitall_1 7281 11.0 23.814 33.280 23.814 33.280 ls_scf_init_scf 1 4.0 0.000 0.000 30.635 30.639 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.434 29.522 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.076 27.094 multiply_cannon_multrec 888 9.7 12.664 15.241 21.315 24.462 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.325 22.765 make_m2s 222 7.7 0.006 0.007 16.655 17.412 make_images 222 8.7 1.589 1.854 16.617 17.374 make_images_data 222 9.7 0.004 0.004 9.649 10.841 hybrid_alltoall_any 227 10.6 0.639 2.937 9.151 10.500 dbcsr_mm_accdrv_process 3754 10.4 0.301 0.496 8.167 9.405 mp_sum_l 887 5.1 5.005 8.919 5.005 8.919 dbcsr_mm_accdrv_process_sort 3754 11.4 7.730 8.909 7.730 8.909 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.483 8.499 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.481 7.139 multiply_cannon_sync_h2d 888 9.7 6.076 7.116 6.076 7.116 mp_irecv_dv 2335 11.1 2.464 7.079 2.464 7.079 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.918 6.916 arnoldi_extremal 4 6.8 0.000 0.000 5.278 5.299 arnoldi_normal_ev 4 7.8 0.001 0.005 5.278 5.299 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.729 4.990 build_subspace 16 8.4 0.014 0.020 4.972 4.977 calculate_norms 1584 9.8 4.332 4.738 4.332 4.738 mp_allgather_i34 111 8.7 0.924 3.865 0.924 3.865 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.495 3.833 ls_scf_post 1 4.0 0.000 0.000 3.664 3.669 dbcsr_matrix_vector_mult_local 304 10.0 3.043 3.630 3.046 3.632 ls_scf_store_result 1 5.0 0.000 0.000 3.397 3.488 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.855 2.966 dbcsr_data_new 4116 9.9 2.116 2.458 2.116 2.458 make_images_sizes 222 9.7 0.000 0.000 1.152 2.282 mp_alltoall_i44 222 10.7 1.152 2.282 1.152 2.282 dbcsr_sort_data 325 11.1 1.890 2.116 1.890 2.116 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.894 1.896 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.876 1.878 make_images_pack 222 9.7 1.623 1.863 1.626 1.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.095000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2245.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 3.341271E+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 0.027 0.045 100.098 100.099 qs_energies 1 2.0 0.000 0.000 99.474 99.480 ls_scf 1 3.0 0.000 0.000 97.812 97.819 dbcsr_multiply_generic 111 6.7 0.016 0.017 81.597 81.843 multiply_cannon 111 7.7 0.046 0.091 55.008 59.714 ls_scf_main 1 4.0 0.000 0.000 58.568 58.569 multiply_cannon_loop 111 8.7 0.155 0.171 49.904 52.589 density_matrix_trs4 2 5.0 0.002 0.003 52.434 52.560 ls_scf_init_scf 1 4.0 0.000 0.000 36.002 36.004 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.780 34.863 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.300 32.311 mp_waitall_1 6369 11.0 23.505 32.138 23.505 32.138 multiply_cannon_multrec 1332 9.7 14.164 17.199 22.138 25.229 make_m2s 222 7.7 0.006 0.008 21.000 22.519 make_images 222 8.7 3.149 3.602 20.950 22.471 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.854 19.033 make_images_data 222 9.7 0.004 0.004 11.696 13.278 hybrid_alltoall_any 227 10.6 0.798 3.785 11.070 12.962 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.478 7.610 9.172 dbcsr_mm_accdrv_process_sort 3641 11.4 7.168 8.674 7.168 8.674 mp_sum_l 887 5.1 4.013 7.296 4.013 7.296 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.786 6.747 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.070 6.157 mp_irecv_dv 3229 10.9 2.045 6.078 2.045 6.078 multiply_cannon_sync_h2d 1332 9.7 5.437 6.070 5.437 6.070 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.139 5.817 arnoldi_extremal 4 6.8 0.000 0.000 5.216 5.226 arnoldi_normal_ev 4 7.8 0.001 0.005 5.216 5.226 build_subspace 16 8.4 0.014 0.021 4.883 4.891 mp_allgather_i34 111 8.7 2.219 4.815 2.219 4.815 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.564 4.777 calculate_norms 2376 9.8 4.175 4.495 4.175 4.495 compute_matrix_preconditioner 1 6.0 0.001 0.001 4.202 4.209 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.598 3.894 acc_transpose_blocks 1332 9.7 0.008 0.008 2.596 3.758 dbcsr_matrix_vector_mult_local 304 10.0 3.215 3.715 3.217 3.717 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 2.481 3.639 jit_kernel_transpose 1 13.0 2.463 3.621 2.463 3.621 dbcsr_sort_data 658 11.4 3.091 3.489 3.091 3.489 dbcsr_special_finalize 555 9.7 0.006 0.007 2.844 3.249 ls_scf_post 1 4.0 0.000 0.000 3.241 3.247 dbcsr_merge_single_wm 555 10.7 0.542 0.672 2.836 3.241 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.012 3.056 ls_scf_store_result 1 5.0 0.000 0.000 2.987 3.042 dbcsr_data_release 10477 10.7 1.584 2.414 1.584 2.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.099000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2758.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 4.743344E+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 0.036 0.055 93.047 93.048 qs_energies 1 2.0 0.000 0.000 92.178 92.182 ls_scf 1 3.0 0.000 0.000 90.156 90.160 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.595 71.893 ls_scf_main 1 4.0 0.000 0.000 57.266 57.266 multiply_cannon 111 7.7 0.084 0.129 52.967 55.931 multiply_cannon_loop 111 8.7 0.088 0.094 50.359 52.141 density_matrix_trs4 2 5.0 0.002 0.003 50.284 50.377 ls_scf_init_scf 1 4.0 0.000 0.000 29.461 29.462 mp_waitall_1 5436 11.0 25.025 29.070 25.025 29.070 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.191 28.237 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.095 26.112 multiply_cannon_multrec 444 9.7 13.733 16.225 20.872 22.112 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.044 15.777 make_m2s 222 7.7 0.005 0.005 13.834 14.799 make_images 222 8.7 2.040 2.474 13.766 14.730 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.227 14.306 hybrid_alltoall_any 227 10.6 0.801 3.815 8.330 10.077 make_images_data 222 9.7 0.003 0.004 8.570 9.857 multiply_cannon_sync_h2d 444 9.7 6.723 8.072 6.723 8.072 dbcsr_mm_accdrv_process 3003 10.4 0.366 0.406 6.837 7.956 dbcsr_mm_accdrv_process_sort 3003 11.4 6.471 7.551 6.471 7.551 arnoldi_extremal 4 6.8 0.000 0.000 5.914 5.932 arnoldi_normal_ev 4 7.8 0.002 0.005 5.914 5.932 build_subspace 16 8.4 0.015 0.020 5.515 5.521 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.416 4.657 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.246 4.456 mp_sum_l 887 5.1 2.793 4.256 2.793 4.256 dbcsr_matrix_vector_mult_local 304 10.0 3.737 4.222 3.739 4.225 mp_allgather_i34 111 8.7 1.193 3.837 1.193 3.837 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.543 3.802 mp_irecv_dv 1241 11.2 1.533 3.757 1.533 3.757 calculate_norms 792 9.8 3.607 3.731 3.607 3.731 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.551 3.633 make_images_sizes 222 9.7 0.000 0.000 0.884 3.446 mp_alltoall_i44 222 10.7 0.883 3.445 0.883 3.445 ls_scf_post 1 4.0 0.000 0.000 3.429 3.433 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.983 3.300 ls_scf_store_result 1 5.0 0.000 0.000 3.214 3.256 dbcsr_finalize 304 7.8 0.062 0.077 2.191 2.260 dbcsr_data_new 4608 9.7 1.773 2.253 1.773 2.253 dbcsr_merge_all 275 8.9 0.478 0.524 2.049 2.105 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.034 2.035 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.006 2.006 rebuild_ks_matrix 3 7.3 0.000 0.000 2.002 2.003 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.002 2.003 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.048000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 8.835781E+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 0.072 0.090 105.280 105.281 qs_energies 1 2.0 0.000 0.000 103.942 103.949 ls_scf 1 3.0 0.000 0.000 100.899 100.905 dbcsr_multiply_generic 111 6.7 0.023 0.027 74.441 74.578 ls_scf_main 1 4.0 0.000 0.000 63.556 63.556 density_matrix_trs4 2 5.0 0.002 0.003 54.521 54.575 multiply_cannon 111 7.7 0.119 0.213 48.399 50.494 multiply_cannon_loop 111 8.7 0.098 0.099 45.387 46.068 ls_scf_init_scf 1 4.0 0.000 0.000 33.555 33.556 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.988 32.015 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.295 29.304 mp_waitall_1 4527 11.1 21.523 25.186 21.523 25.186 make_m2s 222 7.7 0.005 0.005 22.373 23.489 make_images 222 8.7 3.573 3.869 22.266 23.379 multiply_cannon_multrec 444 9.7 17.833 18.649 22.502 23.359 hybrid_alltoall_any 227 10.6 1.654 3.614 12.583 15.374 make_images_data 222 9.7 0.004 0.004 12.765 14.906 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.161 10.461 multiply_cannon_sync_h2d 444 9.7 8.786 8.841 8.786 8.841 arnoldi_extremal 4 6.8 0.000 0.000 7.464 7.471 arnoldi_normal_ev 4 7.8 0.003 0.009 7.464 7.471 build_subspace 16 8.4 0.026 0.036 6.907 6.917 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.537 5.687 dbcsr_matrix_vector_mult_local 304 10.0 5.095 5.409 5.097 5.412 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.156 5.251 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.880 5.136 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.357 4.495 4.626 dbcsr_mm_accdrv_process_sort 1814 11.4 4.159 4.292 4.159 4.292 ls_scf_post 1 4.0 0.000 0.000 3.788 3.796 make_images_sizes 222 9.7 0.000 0.000 1.473 3.596 mp_alltoall_i44 222 10.7 1.473 3.596 1.473 3.596 mp_allgather_i34 111 8.7 1.125 3.577 1.125 3.577 ls_scf_store_result 1 5.0 0.000 0.000 3.512 3.548 calculate_norms 792 9.8 3.235 3.299 3.235 3.299 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.160 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 3.013 3.013 dbcsr_merge_all 275 8.9 0.885 0.912 2.855 2.939 dbcsr_complete_redistribute 5 7.6 1.431 1.468 2.731 2.851 dbcsr_sort_data 325 11.1 2.439 2.500 2.439 2.500 matrix_ls_to_qs 2 6.0 0.000 0.000 2.380 2.500 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.467 2.469 rebuild_ks_matrix 3 7.3 0.000 0.000 2.400 2.402 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.400 2.402 dbcsr_new_transposed 4 7.5 0.242 0.252 2.295 2.314 dbcsr_data_new 6591 9.6 1.866 2.283 1.866 2.283 dbcsr_frobenius_norm 74 6.6 2.055 2.132 2.203 2.239 dbcsr_data_release 12724 10.6 1.976 2.223 1.976 2.223 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.202 dbcsr_add_anytype 103 7.2 0.858 0.891 2.124 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.281000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7040.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/96ef312a72745dc8e373ed79515b61e51a8022ea_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 594.325504E+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 4806034. 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 0.032 0.063 254.246 254.248 qs_mol_dyn_low 1 2.0 0.004 0.014 253.034 253.141 qs_forces 5 3.8 0.005 0.012 252.945 252.947 qs_energies 5 4.8 0.004 0.053 249.632 249.653 scf_env_do_scf 5 5.8 0.000 0.002 234.918 234.921 scf_env_do_scf_inner_loop 105 6.6 0.003 0.018 205.512 205.525 qs_scf_new_mos 105 7.6 0.000 0.001 160.337 160.533 qs_scf_loop_do_ot 105 8.6 0.001 0.001 160.337 160.532 ot_scf_mini 105 9.6 0.003 0.005 150.367 150.516 dbcsr_multiply_generic 1445 12.2 0.127 0.135 134.395 134.789 multiply_cannon 1445 13.2 0.275 0.286 114.088 116.104 multiply_cannon_loop 1445 14.2 2.879 3.026 112.181 113.465 velocity_verlet 4 3.0 0.001 0.002 111.924 111.925 ot_mini 105 10.6 0.001 0.003 61.009 61.131 qs_ot_get_p 112 10.4 0.001 0.002 58.986 59.235 qs_ot_p2m_diag 40 11.0 0.020 0.030 47.364 47.455 multiply_cannon_multrec 69360 15.2 29.561 34.394 39.552 44.920 cp_dbcsr_syevd 40 12.0 0.002 0.007 44.013 44.014 mp_waitall_1 488190 16.1 35.562 42.128 35.562 42.128 qs_ot_get_derivative 55 11.6 0.001 0.002 39.200 39.339 cp_fm_syevd 40 13.0 0.000 0.002 38.460 38.593 cp_fm_redistribute_end 40 14.0 16.665 33.261 16.671 33.263 cp_fm_syevd_base 40 14.0 16.581 33.176 16.581 33.176 multiply_cannon_sync_h2d 69360 15.2 28.914 32.645 28.914 32.645 multiply_cannon_metrocomm3 69360 15.2 0.199 0.212 25.666 32.574 rebuild_ks_matrix 110 8.4 0.000 0.000 32.139 32.316 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.048 32.139 32.316 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.556 29.713 init_scf_loop 7 6.6 0.002 0.063 29.369 29.372 prepare_preconditioner 7 7.6 0.000 0.000 23.905 23.938 make_preconditioner 7 8.6 0.001 0.007 23.904 23.938 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.040 23.267 apply_single 62 13.6 0.000 0.000 23.039 23.267 ot_new_cg_direction 55 11.6 0.001 0.003 21.066 21.067 qs_rho_update_rho_low 110 7.6 0.001 0.001 19.603 20.004 calculate_rho_elec 110 8.6 0.030 0.033 19.602 20.004 make_full_inverse_cholesky 7 9.6 0.000 0.001 16.430 16.502 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.070 15.208 pw_transfer 1645 12.4 0.078 0.103 14.624 14.888 fft_wrap_pw1pw2 1425 13.5 0.011 0.015 14.487 14.744 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.114 14.209 density_rs2pw 110 9.6 0.005 0.010 13.398 13.837 mp_sum_l 4764 12.2 12.465 13.170 12.465 13.170 fft_wrap_pw1pw2_240 915 15.0 0.809 0.889 12.880 13.119 qs_vxc_create 110 10.4 0.002 0.007 12.137 12.180 init_scf_run 5 5.8 0.000 0.001 11.917 11.918 scf_env_initial_rho_setup 5 6.8 0.002 0.004 11.917 11.918 calculate_dm_sparse 110 9.5 0.000 0.001 11.533 11.731 fft3d_pb 915 16.0 2.349 2.571 11.145 11.399 cp_fm_cholesky_invert 7 10.6 10.995 11.003 10.995 11.003 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.640 10.726 dbcsr_mm_accdrv_process 154766 15.8 6.370 6.569 9.858 10.658 check_diag 80 13.5 8.683 8.935 10.121 10.257 sum_up_and_integrate 60 10.3 0.001 0.005 9.735 9.768 integrate_v_rspace 60 11.3 0.001 0.002 9.717 9.751 transfer_rs2pw 445 10.6 0.007 0.008 9.320 9.747 mp_alltoall_z22v 2340 17.7 8.265 8.598 8.265 8.598 xc_rho_set_and_dset_create 110 12.4 0.076 0.099 8.325 8.584 xc_vxc_pw_create 60 11.3 0.039 0.052 8.212 8.254 make_m2s 2890 13.2 0.078 0.086 7.429 8.012 acc_transpose_blocks 69360 15.2 0.364 0.399 7.459 7.945 make_images 2890 14.2 0.239 0.259 7.321 7.905 xc_pw_derive 510 13.4 0.005 0.006 7.757 7.829 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.328 7.347 make_full_single_inverse 7 9.6 0.001 0.004 7.119 7.153 multiply_cannon_metrocomm1 69360 15.2 0.097 0.103 4.611 7.126 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.853 6.918 calculate_first_density_matrix 1 7.0 0.000 0.006 6.818 6.831 potential_pw2rs 60 12.3 0.002 0.003 5.662 5.713 mp_alltoall_d11v 1300 13.8 4.481 5.416 4.481 5.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=254.248000, 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: 96ef312a72745dc8e373ed79515b61e51a8022ea Summary: empty Status: OK