=== 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: 3c55af5c093d10a5e0ad2f8e06538f92dff62763 ################# 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 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/01 job id: 49214598 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/02 job id: 49214600 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/03 job id: 49214602 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/04 job id: 49214604 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/05 job id: 49214605 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/06 job id: 49214606 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/07 job id: 49214607 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/08 job id: 49214608 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/09 job id: 49214609 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/10 job id: 49214610 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/11 job id: 49214611 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/12 job id: 49214612 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/13 job id: 49214613 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/14 job id: 49214614 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/15 job id: 49214615 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/16 job id: 49214616 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/17 job id: 49214617 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/18 job id: 49214618 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/19 job id: 49214619 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/20 job id: 49214620 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/21 job id: 49214621 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/22 job id: 49214622 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/23 job id: 49214623 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/24 job id: 49214624 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/25 job id: 49214625 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/26 job id: 49214626 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/27 job id: 49214627 --- 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/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 134.829 134.830 farming_run 1 2.0 134.278 134.281 134.796 134.800 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488265E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 115.270 115.270 qs_energies 1 2.0 0.000 0.000 114.997 115.000 mp2_main 1 3.0 0.000 0.000 112.827 112.830 mp2_gpw_main 1 4.0 0.041 0.048 111.773 111.775 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 92.484 92.785 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.687 54.988 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.076 46.089 get_2c_integrals 1 6.0 0.008 0.009 36.753 37.625 integrate_v_rspace 273 8.0 0.442 0.457 24.673 29.540 pw_transfer 6555 10.6 0.376 0.389 26.807 27.366 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.459 26.063 grid_integrate_task_list 273 9.0 20.574 25.916 20.574 25.916 fft_wrap_pw1pw2_100 2178 12.4 1.121 1.315 23.009 23.611 rpa_ri_compute_en 1 5.0 0.029 0.033 19.159 19.409 compute_2c_integrals 1 7.0 0.030 0.036 19.022 19.023 compute_2c_integrals_loop_lm 1 8.0 0.005 0.006 18.545 18.709 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.425 18.539 18.704 cp_fm_cholesky_decompose 12 8.2 17.745 18.627 17.745 18.627 cholesky_decomp 1 7.0 0.000 0.000 16.580 17.457 fft3d_s 5443 13.4 16.148 16.458 16.171 16.481 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.309 10.771 15.309 calculate_wavefunction 272 8.0 5.380 5.522 12.315 13.017 rpa_num_int 1 6.0 0.000 0.006 10.776 10.780 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.700 10.733 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.594 contract_S_to_Q 8 9.0 0.000 0.000 8.910 9.013 calc_potential_gpw 544 9.5 0.005 0.005 8.191 8.662 parallel_gemm_fm 14 9.1 0.000 0.000 8.523 8.617 parallel_gemm_fm_cosma 14 10.1 8.523 8.617 8.523 8.617 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.041 8.249 potential_pw2rs 545 10.0 0.106 0.108 7.490 8.108 create_integ_mat 1 6.0 0.022 0.027 7.977 7.977 collocate_single_gaussian 272 10.0 0.039 0.041 7.304 7.520 array2fm 1 7.0 0.000 0.000 6.802 7.385 pw_scatter_s 2720 13.7 4.363 4.561 4.363 4.561 pw_gather_s 2722 13.2 3.426 3.770 3.426 3.770 array2fm_buffer_send 1 8.0 3.024 3.175 3.024 3.175 pw_poisson_solve 545 10.5 1.117 1.188 2.171 2.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.775454, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.047 424.099 424.106 farming_run 1 2.0 423.096 423.113 424.036 424.043 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.240474E+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 29043579. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 217.387 217.389 qs_energies 1 2.0 0.002 0.005 217.120 217.144 scf_env_do_scf 1 3.0 0.000 0.000 114.949 114.949 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.980 113.987 rebuild_ks_matrix 4 6.0 0.000 0.000 113.979 113.986 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.979 113.986 hfx_ks_matrix 4 8.0 0.001 0.001 113.620 113.625 integrate_four_center 4 9.0 0.153 0.466 113.619 113.624 integrate_four_center_main 4 10.0 0.138 0.691 101.895 105.340 integrate_four_center_bin 266 11.0 101.756 104.849 101.756 104.849 mp2_main 1 3.0 0.012 0.045 101.852 101.872 mp2_gpw_main 1 4.0 0.054 0.075 100.884 100.899 init_scf_loop 1 4.0 0.000 0.000 96.980 96.981 mp2_ri_gpw_compute_in 1 5.0 0.070 0.124 73.669 74.617 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.701 54.642 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 41.475 46.317 integrate_v_rspace 95 8.0 0.398 0.564 27.960 32.636 pw_transfer 2240 10.6 0.146 0.167 29.440 29.796 ao_to_mo_and_store_B_mult_1 91 7.0 10.551 29.130 10.551 29.130 mp2_ri_gpw_compute_en 1 5.0 0.063 0.075 27.026 29.065 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.469 28.833 grid_integrate_task_list 95 9.0 23.377 28.257 23.377 28.257 fft_wrap_pw1pw2_100 730 12.4 1.237 1.425 26.211 26.612 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.899 25.413 25.423 get_2c_integrals 1 6.0 0.002 0.014 19.828 19.909 compute_2c_integrals 1 7.0 0.003 0.004 18.808 18.823 fft3d_s 1823 13.4 18.468 18.798 18.483 18.812 compute_2c_integrals_loop_lm 1 8.0 0.006 0.011 18.494 18.671 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.800 18.488 18.661 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.967 17.967 calculate_wavefunction 91 8.0 2.016 2.046 9.617 9.833 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.603 8.904 9.587 local_gemm 172 8.0 8.347 9.024 8.347 9.024 potential_pw2rs 186 10.0 0.033 0.034 8.394 8.912 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.519 8.302 8.787 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.091 8.403 calc_potential_gpw 182 9.5 0.002 0.002 7.866 8.075 collocate_single_gaussian 91 10.0 0.017 0.025 7.685 7.971 mp_sendrecv_dm3 2068 8.0 6.338 6.814 6.338 6.814 integrate_four_center_load 4 10.0 0.000 0.000 6.783 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.783 6.786 mp_sync 37 10.5 3.714 6.410 3.714 6.410 mp2_ri_gpw_compute_en_ener 172 7.0 6.329 6.406 6.329 6.406 pw_gather_s 912 13.2 4.492 4.845 4.492 4.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.893103, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1500.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.665344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.095 56.202 56.209 qs_mol_dyn_low 1 2.0 0.003 0.004 55.803 55.876 qs_forces 11 3.9 0.002 0.003 55.697 55.698 qs_energies 11 4.9 0.001 0.003 54.143 54.165 scf_env_do_scf 11 5.9 0.001 0.007 47.641 47.641 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 45.339 45.340 dbcsr_multiply_generic 2286 12.5 0.095 0.100 34.621 35.171 qs_scf_new_mos 108 7.5 0.000 0.001 34.870 35.141 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.869 35.141 ot_scf_mini 108 9.5 0.002 0.002 33.146 33.332 multiply_cannon 2286 13.5 0.196 0.212 26.853 28.263 multiply_cannon_loop 2286 14.5 1.826 1.935 26.073 27.553 velocity_verlet 10 3.0 0.002 0.006 27.508 27.509 ot_mini 108 10.5 0.001 0.001 19.981 20.232 qs_ot_get_derivative 108 11.5 0.001 0.001 16.838 17.035 mp_waitall_1 245248 16.5 8.845 15.212 8.845 15.212 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.107 13.137 multiply_cannon_multrec 54864 15.5 3.638 5.674 7.905 11.337 qs_ot_get_p 119 10.4 0.001 0.001 8.436 8.692 rebuild_ks_matrix 119 8.3 0.000 0.000 8.206 8.369 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.025 8.206 8.369 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.226 7.385 mp_sum_l 7287 12.8 5.431 7.264 5.431 7.264 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.038 6.521 multiply_cannon_sync_h2d 54864 15.5 5.126 6.213 5.126 6.213 dbcsr_mm_accdrv_process 76910 16.1 1.836 2.887 4.181 5.986 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.488 5.528 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.283 5.385 init_scf_run 11 5.9 0.000 0.001 5.186 5.187 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.186 5.187 sum_up_and_integrate 119 10.3 0.001 0.002 4.725 4.742 integrate_v_rspace 119 11.3 0.002 0.003 4.714 4.732 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.676 4.677 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.364 4.502 calculate_rho_elec 119 8.7 0.011 0.016 4.363 4.501 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.435 4.435 cp_fm_redistribute_end 50 14.0 2.261 4.400 2.269 4.405 cp_fm_diag_elpa_base 50 14.0 2.128 4.277 2.133 4.288 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.005 3.209 apply_single 119 13.6 0.000 0.000 3.005 3.209 calculate_dm_sparse 119 9.5 0.000 0.001 3.067 3.204 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.900 3.053 jit_kernel_multiply 13 15.8 2.281 3.023 2.281 3.023 calculate_first_density_matrix 1 7.0 0.000 0.000 2.858 2.862 acc_transpose_blocks 54864 15.5 0.226 0.252 2.291 2.849 ot_diis_step 108 11.5 0.006 0.006 2.843 2.846 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.569 2.573 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.482 2.536 density_rs2pw 119 9.7 0.004 0.004 2.370 2.484 init_scf_loop 11 6.9 0.000 0.000 2.274 2.274 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.221 2.271 wfi_extrapolate 11 7.9 0.001 0.001 2.259 2.259 grid_integrate_task_list 119 12.3 2.015 2.127 2.015 2.127 make_m2s 4572 13.5 0.054 0.057 2.022 2.088 mp_sum_d 4135 12.0 1.460 2.057 1.460 2.057 potential_pw2rs 119 12.3 0.004 0.004 2.032 2.052 make_images 4572 14.5 0.134 0.140 1.938 2.003 pw_transfer 1439 11.6 0.051 0.057 1.891 1.972 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.815 1.899 mp_alltoall_d11v 2130 13.8 1.493 1.651 1.493 1.651 fft3d_ps 1201 14.6 0.370 0.480 1.531 1.609 transfer_rs2pw 487 10.6 0.005 0.006 1.465 1.596 mp_waitany 12084 13.8 1.371 1.559 1.371 1.559 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.511 1.536 transfer_pw2rs 487 13.2 0.005 0.006 1.484 1.503 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.415 1.501 acc_transpose_blocks_sync 164592 16.5 1.198 1.443 1.198 1.443 grid_collocate_task_list 119 9.7 1.350 1.429 1.350 1.429 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.828 1.229 prepare_preconditioner 11 7.9 0.000 0.000 1.108 1.144 make_preconditioner 11 8.9 0.000 0.000 1.108 1.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.209000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.452096E+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 793094. 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 0.069 40.966 40.999 qs_mol_dyn_low 1 2.0 0.003 0.004 40.685 40.699 qs_forces 11 3.9 0.002 0.003 40.607 40.608 qs_energies 11 4.9 0.003 0.003 38.893 38.898 scf_env_do_scf 11 5.9 0.000 0.002 33.399 33.400 scf_env_do_scf_inner_loop 108 6.5 0.004 0.013 30.812 30.813 dbcsr_multiply_generic 2286 12.5 0.101 0.103 23.109 23.567 qs_scf_new_mos 108 7.5 0.001 0.001 21.813 22.049 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.813 22.049 ot_scf_mini 108 9.5 0.002 0.003 20.852 21.018 velocity_verlet 10 3.0 0.002 0.003 19.301 19.303 multiply_cannon 2286 13.5 0.225 0.235 17.472 19.294 multiply_cannon_loop 2286 14.5 1.196 1.257 16.235 18.110 ot_mini 108 10.5 0.001 0.001 12.814 13.051 mp_waitall_1 200699 16.5 6.043 11.438 6.043 11.438 qs_ot_get_derivative 108 11.5 0.001 0.001 10.381 10.548 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.346 10.044 multiply_cannon_multrec 27432 15.5 1.833 4.277 6.315 9.207 rebuild_ks_matrix 119 8.3 0.000 0.000 6.996 7.160 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.995 7.159 dbcsr_mm_accdrv_process 47894 16.0 3.496 5.759 4.401 6.572 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.147 6.296 qs_ot_get_p 119 10.4 0.001 0.001 4.998 5.215 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.899 4.821 mp_sum_l 7287 12.8 2.380 4.470 2.380 4.470 init_scf_run 11 5.9 0.000 0.001 4.254 4.255 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.254 4.255 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.119 4.158 apply_single 119 13.6 0.000 0.000 3.119 4.157 sum_up_and_integrate 119 10.3 0.001 0.001 3.871 3.878 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.818 3.867 calculate_rho_elec 119 8.7 0.021 0.024 3.818 3.867 integrate_v_rspace 119 11.3 0.002 0.003 3.857 3.863 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.197 3.218 make_m2s 4572 13.5 0.052 0.054 2.806 3.158 make_images 4572 14.5 0.206 0.245 2.716 3.072 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.784 2.785 calculate_first_density_matrix 1 7.0 0.000 0.000 2.682 2.687 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.493 2.575 init_scf_loop 11 6.9 0.000 0.000 2.563 2.563 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.433 2.433 cp_fm_redistribute_end 50 14.0 1.234 2.408 1.237 2.410 ot_diis_step 108 11.5 0.010 0.011 2.373 2.374 cp_fm_diag_elpa_base 50 14.0 1.140 2.313 1.169 2.354 calculate_dm_sparse 119 9.5 0.000 0.001 2.180 2.262 density_rs2pw 119 9.7 0.004 0.004 2.087 2.200 multiply_cannon_sync_h2d 27432 15.5 1.657 2.184 1.657 2.184 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.074 2.077 acc_transpose_blocks 27432 15.5 0.111 0.117 1.669 2.024 pw_transfer 1439 11.6 0.065 0.068 1.980 2.015 grid_integrate_task_list 119 12.3 1.837 1.964 1.837 1.964 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.888 1.925 jit_kernel_multiply 9 16.2 0.847 1.912 0.847 1.912 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.852 1.897 make_images_data 4572 15.5 0.047 0.054 1.310 1.683 prepare_preconditioner 11 7.9 0.000 0.000 1.646 1.673 make_preconditioner 11 8.9 0.000 0.000 1.646 1.673 potential_pw2rs 119 12.3 0.006 0.006 1.594 1.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.539 1.592 fft3d_ps 1201 14.6 0.527 0.582 1.524 1.557 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.160 1.556 mp_alltoall_d11v 2130 13.8 1.340 1.517 1.340 1.517 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.470 1.505 wfi_extrapolate 11 7.9 0.001 0.001 1.505 1.505 grid_collocate_task_list 119 9.7 1.286 1.353 1.286 1.353 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.296 1.336 transfer_rs2pw 487 10.6 0.005 0.005 1.200 1.271 mp_allgather_i34 2286 14.5 0.642 1.081 0.642 1.081 mp_sum_d 4135 12.0 0.595 1.032 0.595 1.032 transfer_pw2rs 487 13.2 0.004 0.005 1.001 1.005 acc_transpose_blocks_kernels 27432 16.5 0.188 0.279 0.717 0.954 qs_energies_init_hamiltonians 11 5.9 0.001 0.007 0.947 0.948 acc_transpose_blocks_sync 82296 16.5 0.815 0.940 0.815 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.922 make_images_sizes 4572 15.5 0.005 0.005 0.628 0.895 mp_alltoall_i44 4572 16.5 0.623 0.890 0.623 0.890 mp_alltoall_z22v 1201 16.6 0.777 0.844 0.777 0.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.999000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 520.744960E+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 0.021 0.041 35.027 35.031 qs_mol_dyn_low 1 2.0 0.003 0.004 34.683 34.696 qs_forces 11 3.9 0.003 0.004 34.603 34.606 qs_energies 11 4.9 0.002 0.002 33.020 33.027 scf_env_do_scf 11 5.9 0.001 0.004 27.879 27.879 scf_env_do_scf_inner_loop 108 6.5 0.005 0.012 25.151 25.151 dbcsr_multiply_generic 2286 12.5 0.096 0.100 18.363 18.527 qs_scf_new_mos 108 7.5 0.001 0.001 17.001 17.031 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.000 17.030 velocity_verlet 10 3.0 0.003 0.005 16.285 16.286 ot_scf_mini 108 9.5 0.002 0.002 16.200 16.236 multiply_cannon 2286 13.5 0.197 0.199 14.580 15.486 multiply_cannon_loop 2286 14.5 0.864 0.904 13.724 14.665 ot_mini 108 10.5 0.001 0.001 10.025 10.067 qs_ot_get_derivative 108 11.5 0.001 0.001 8.439 8.468 multiply_cannon_multrec 18288 15.5 1.938 3.032 7.319 7.668 rebuild_ks_matrix 119 8.3 0.000 0.000 6.291 6.352 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.290 6.351 dbcsr_mm_accdrv_process 38222 16.0 5.229 6.148 5.287 6.214 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.547 5.606 mp_waitall_1 158411 16.6 3.286 4.392 3.286 4.392 init_scf_run 11 5.9 0.000 0.001 3.883 3.883 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.882 3.882 qs_ot_get_p 119 10.4 0.001 0.001 3.819 3.862 sum_up_and_integrate 119 10.3 0.001 0.001 3.780 3.786 integrate_v_rspace 119 11.3 0.003 0.003 3.768 3.775 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.174 3.737 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.499 3.513 calculate_rho_elec 119 8.7 0.031 0.031 3.498 3.513 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.813 2.977 init_scf_loop 11 6.9 0.000 0.000 2.703 2.704 calculate_first_density_matrix 1 7.0 0.000 0.000 2.583 2.584 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.534 2.544 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.025 2.415 apply_single 119 13.6 0.000 0.000 2.025 2.415 make_m2s 4572 13.5 0.045 0.046 2.087 2.213 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.202 2.203 make_images 4572 14.5 0.193 0.204 2.001 2.126 density_rs2pw 119 9.7 0.004 0.004 2.028 2.125 calculate_dm_sparse 119 9.5 0.000 0.001 2.010 2.027 pw_transfer 1439 11.6 0.065 0.069 1.999 2.010 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.980 1.995 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.907 1.920 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.903 1.913 cp_fm_diag_elpa_base 50 14.0 1.877 1.892 1.901 1.912 grid_integrate_task_list 119 12.3 1.806 1.906 1.806 1.906 prepare_preconditioner 11 7.9 0.000 0.000 1.886 1.889 make_preconditioner 11 8.9 0.000 0.000 1.886 1.889 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.845 1.847 acc_transpose_blocks 18288 15.5 0.078 0.080 1.738 1.823 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.727 1.815 mp_sum_l 7287 12.8 1.308 1.792 1.308 1.792 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.679 1.695 ot_diis_step 108 11.5 0.011 0.011 1.550 1.551 potential_pw2rs 119 12.3 0.007 0.009 1.523 1.527 fft_wrap_pw1pw2_140 487 13.2 0.211 0.218 1.491 1.501 fft3d_ps 1201 14.6 0.529 0.547 1.476 1.490 grid_collocate_task_list 119 9.7 1.234 1.330 1.234 1.330 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.310 1.320 wfi_extrapolate 11 7.9 0.001 0.001 1.247 1.248 transfer_rs2pw 487 10.6 0.005 0.005 1.107 1.208 multiply_cannon_sync_h2d 18288 15.5 0.998 1.137 0.998 1.137 make_images_data 4572 15.5 0.047 0.052 0.923 1.085 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.031 1.055 hybrid_alltoall_any 4725 16.4 0.057 0.116 0.812 1.013 mp_alltoall_d11v 2130 13.8 0.844 0.982 0.844 0.982 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.957 0.960 transfer_pw2rs 487 13.2 0.004 0.004 0.940 0.944 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.891 0.905 mp_alltoall_z22v 1201 16.6 0.794 0.868 0.794 0.868 cp_fm_cholesky_invert 11 10.9 0.862 0.867 0.862 0.867 acc_transpose_blocks_sync 54864 16.5 0.750 0.835 0.750 0.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.822 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.723 mp_waitany 9880 13.7 0.566 0.704 0.566 0.704 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.031000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=0.715819 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 559.542272E+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 0.018 0.040 38.785 38.788 qs_mol_dyn_low 1 2.0 0.003 0.004 38.589 38.607 qs_forces 11 3.9 0.003 0.003 38.118 38.122 qs_energies 11 4.9 0.002 0.003 36.376 36.384 scf_env_do_scf 11 5.9 0.001 0.003 30.942 30.943 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.331 27.331 dbcsr_multiply_generic 2286 12.5 0.100 0.104 20.681 20.837 velocity_verlet 10 3.0 0.002 0.002 19.793 19.811 qs_scf_new_mos 108 7.5 0.001 0.001 18.891 18.945 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.890 18.944 ot_scf_mini 108 9.5 0.002 0.003 17.817 17.875 multiply_cannon 2286 13.5 0.220 0.226 16.445 16.754 multiply_cannon_loop 2286 14.5 1.512 1.603 15.432 15.808 ot_mini 108 10.5 0.001 0.001 10.950 11.002 multiply_cannon_multrec 27432 15.5 2.451 3.125 9.030 9.328 qs_ot_get_derivative 108 11.5 0.001 0.001 9.002 9.057 dbcsr_mm_accdrv_process 47916 15.9 6.013 7.707 6.476 7.828 rebuild_ks_matrix 119 8.3 0.000 0.000 6.565 6.612 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.565 6.611 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.832 5.873 init_scf_run 11 5.9 0.000 0.001 4.047 4.047 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.047 4.047 qs_ot_get_p 119 10.4 0.001 0.001 3.913 3.986 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.293 3.688 sum_up_and_integrate 119 10.3 0.001 0.001 3.676 3.686 integrate_v_rspace 119 11.3 0.003 0.003 3.665 3.674 init_scf_loop 11 6.9 0.000 0.000 3.584 3.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.534 3.582 calculate_rho_elec 119 8.7 0.040 0.046 3.534 3.581 mp_waitall_1 137007 16.6 2.278 2.852 2.278 2.852 prepare_preconditioner 11 7.9 0.000 0.000 2.719 2.724 make_preconditioner 11 8.9 0.000 0.000 2.719 2.724 make_m2s 4572 13.5 0.055 0.057 2.589 2.672 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.310 2.639 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.541 2.564 make_images 4572 14.5 0.273 0.336 2.479 2.561 calculate_first_density_matrix 1 7.0 0.000 0.000 2.536 2.538 acc_transpose_blocks 27432 15.5 0.116 0.120 2.396 2.529 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.375 2.386 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.163 2.281 apply_single 119 13.6 0.000 0.000 2.163 2.281 calculate_dm_sparse 119 9.5 0.000 0.000 2.213 2.255 pw_transfer 1439 11.6 0.065 0.069 2.084 2.120 density_rs2pw 119 9.7 0.004 0.004 1.944 2.036 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.992 2.031 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.955 1.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.938 1.939 grid_integrate_task_list 119 12.3 1.837 1.904 1.837 1.904 ot_diis_step 108 11.5 0.012 0.012 1.898 1.899 jit_kernel_multiply 10 15.9 0.397 1.838 0.397 1.838 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.707 1.720 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.652 1.660 cp_fm_diag_elpa_base 50 14.0 1.615 1.632 1.649 1.658 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.598 1.639 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.950 1.552 fft3d_ps 1201 14.6 0.562 0.608 1.511 1.538 acc_transpose_blocks_sync 82296 16.5 1.382 1.505 1.382 1.505 mp_sum_l 7287 12.8 1.221 1.502 1.221 1.502 potential_pw2rs 119 12.3 0.009 0.009 1.454 1.459 wfi_extrapolate 11 7.9 0.001 0.001 1.451 1.451 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.364 1.376 grid_collocate_task_list 119 9.7 1.247 1.339 1.247 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.278 1.288 cp_fm_upper_to_full 72 14.2 0.825 1.175 0.825 1.175 dbcsr_complete_redistribute 329 12.2 0.127 0.168 0.879 1.150 make_images_data 4572 15.5 0.047 0.051 1.021 1.133 transfer_rs2pw 487 10.6 0.004 0.005 1.016 1.123 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.101 1.104 hybrid_alltoall_any 4725 16.4 0.065 0.153 0.881 1.041 mp_alltoall_d11v 2130 13.8 0.888 0.996 0.888 0.996 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.657 0.923 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.869 0.908 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.879 mp_alltoall_z22v 1201 16.6 0.831 0.857 0.831 0.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.832 cp_fm_cholesky_invert 11 10.9 0.824 0.827 0.824 0.827 transfer_pw2rs 487 13.2 0.004 0.004 0.818 0.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.788000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.818182, yerr=3.009626 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 617.324544E+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 0.018 0.039 31.314 31.317 qs_mol_dyn_low 1 2.0 0.003 0.003 31.006 31.027 qs_forces 11 3.9 0.003 0.005 30.877 30.880 qs_energies 11 4.9 0.003 0.004 29.124 29.129 scf_env_do_scf 11 5.9 0.002 0.021 23.799 23.799 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 21.002 21.003 velocity_verlet 10 3.0 0.060 0.067 16.013 16.016 dbcsr_multiply_generic 2286 12.5 0.098 0.102 14.230 14.333 qs_scf_new_mos 108 7.5 0.001 0.001 12.867 12.904 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.866 12.904 ot_scf_mini 108 9.5 0.002 0.002 12.102 12.135 multiply_cannon 2286 13.5 0.223 0.228 11.161 11.718 multiply_cannon_loop 2286 14.5 0.641 0.665 10.149 10.365 ot_mini 108 10.5 0.001 0.001 6.892 6.929 multiply_cannon_multrec 9144 15.5 1.726 1.953 6.230 6.466 rebuild_ks_matrix 119 8.3 0.000 0.000 6.105 6.134 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.105 6.134 qs_ot_get_derivative 108 11.5 0.001 0.001 5.497 5.526 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.439 5.462 dbcsr_mm_accdrv_process 12550 15.8 3.595 4.437 4.393 4.480 init_scf_run 11 5.9 0.000 0.001 3.771 3.771 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.770 3.771 sum_up_and_integrate 119 10.3 0.001 0.001 3.626 3.632 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.624 3.628 calculate_rho_elec 119 8.7 0.060 0.061 3.623 3.628 integrate_v_rspace 119 11.3 0.003 0.003 3.616 3.622 qs_ot_get_p 119 10.4 0.001 0.001 3.200 3.232 init_scf_loop 11 6.9 0.000 0.000 2.753 2.755 mp_waitall_1 115863 16.7 1.891 2.503 1.891 2.503 calculate_first_density_matrix 1 7.0 0.000 0.000 2.449 2.450 make_m2s 4572 13.5 0.035 0.036 2.137 2.273 pw_transfer 1439 11.6 0.065 0.068 2.181 2.192 make_images 4572 14.5 0.269 0.300 2.047 2.182 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.088 2.101 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.070 2.074 density_rs2pw 119 9.7 0.003 0.004 1.989 2.065 jit_kernel_multiply 10 15.8 0.759 1.995 0.759 1.995 prepare_preconditioner 11 7.9 0.000 0.000 1.946 1.951 make_preconditioner 11 8.9 0.000 0.000 1.946 1.951 calculate_dm_sparse 119 9.5 0.000 0.000 1.922 1.944 grid_integrate_task_list 119 12.3 1.870 1.943 1.870 1.943 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.915 1.937 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.824 1.855 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.808 1.809 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.795 1.797 fft_wrap_pw1pw2_140 487 13.2 0.323 0.332 1.691 1.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.611 1.619 acc_transpose_blocks 9144 15.5 0.040 0.041 1.563 1.605 fft3d_ps 1201 14.6 0.562 0.572 1.502 1.515 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.492 1.500 cp_fm_diag_elpa_base 50 14.0 1.461 1.479 1.490 1.498 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.484 1.498 potential_pw2rs 119 12.3 0.010 0.011 1.401 1.405 grid_collocate_task_list 119 9.7 1.298 1.403 1.298 1.403 ot_diis_step 108 11.5 0.012 0.013 1.371 1.372 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.296 1.324 apply_single 119 13.6 0.000 0.000 1.295 1.324 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.267 wfi_extrapolate 11 7.9 0.001 0.001 1.261 1.262 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 1.240 1.243 make_images_data 4572 15.5 0.042 0.045 0.972 1.168 hybrid_alltoall_any 4725 16.4 0.064 0.175 0.933 1.167 transfer_rs2pw 487 10.6 0.004 0.004 0.980 1.038 cp_fm_cholesky_invert 11 10.9 1.021 1.023 1.021 1.023 mp_alltoall_d11v 2130 13.8 0.884 1.003 0.884 1.003 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.931 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.883 0.888 mp_alltoall_z22v 1201 16.6 0.814 0.850 0.814 0.850 mp_allgather_i34 2286 14.5 0.321 0.843 0.321 0.843 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.383 0.825 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.786 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.796 acc_transpose_blocks_sync 27432 16.5 0.721 0.762 0.721 0.762 transfer_pw2rs 487 13.2 0.003 0.004 0.742 0.747 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.685 0.732 jit_kernel_transpose 5 15.6 0.668 0.680 0.668 0.680 mp_sum_l 7287 12.8 0.483 0.665 0.483 0.665 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.317000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.272727, yerr=3.910148 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 773.541888E+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 0.024 0.044 89.798 89.799 qs_mol_dyn_low 1 2.0 0.003 0.004 89.407 89.438 qs_forces 11 3.9 0.002 0.002 89.306 89.307 qs_energies 11 4.9 0.002 0.004 86.449 86.469 scf_env_do_scf 11 5.9 0.001 0.004 77.602 77.602 scf_env_do_scf_inner_loop 108 6.5 0.004 0.011 59.266 59.271 velocity_verlet 10 3.0 0.011 0.014 51.259 51.268 qs_scf_new_mos 108 7.5 0.001 0.001 39.793 39.877 qs_scf_loop_do_ot 108 8.5 0.001 0.001 39.792 39.876 ot_scf_mini 108 9.5 0.002 0.002 37.685 37.805 dbcsr_multiply_generic 2286 12.5 0.105 0.106 36.563 37.251 multiply_cannon 2286 13.5 0.320 0.338 23.970 26.894 multiply_cannon_loop 2286 14.5 0.877 0.896 21.027 22.874 ot_mini 108 10.5 0.001 0.001 21.157 21.326 init_scf_loop 11 6.9 0.000 0.000 18.253 18.263 mp_waitall_1 94719 16.7 15.426 17.720 15.426 17.720 qs_ot_get_derivative 108 11.5 0.001 0.001 17.543 17.673 prepare_preconditioner 11 7.9 0.000 0.000 16.469 16.487 make_preconditioner 11 8.9 0.000 0.000 16.469 16.487 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.624 16.197 rebuild_ks_matrix 119 8.3 0.000 0.000 13.606 13.781 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 13.606 13.781 qs_ks_update_qs_env 119 7.6 0.001 0.001 12.243 12.403 qs_ot_get_p 119 10.4 0.001 0.001 10.773 10.949 make_m2s 4572 13.5 0.038 0.039 8.698 10.026 make_images 4572 14.5 0.360 0.386 8.578 9.908 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 7.938 9.901 qs_rho_update_rho_low 119 7.7 0.001 0.001 9.194 9.232 calculate_rho_elec 119 8.7 0.118 0.121 9.194 9.232 multiply_cannon_multrec 9144 15.5 3.566 4.928 9.070 9.182 sum_up_and_integrate 119 10.3 0.001 0.001 8.180 8.191 integrate_v_rspace 119 11.3 0.003 0.003 8.170 8.181 cp_fm_cholesky_invert 11 10.9 7.584 7.602 7.584 7.602 make_images_data 4572 15.5 0.046 0.049 5.323 7.314 pw_transfer 1439 11.6 0.069 0.070 6.924 6.970 qs_ot_p2m_diag 50 11.0 0.044 0.044 6.886 6.906 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 6.825 6.871 dbcsr_mm_accdrv_process 12550 15.8 4.577 6.471 5.358 6.738 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.129 6.554 hybrid_alltoall_any 4725 16.4 0.090 0.150 4.650 6.552 density_rs2pw 119 9.7 0.003 0.003 6.468 6.479 init_scf_run 11 5.9 0.000 0.001 6.184 6.184 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.183 6.184 fft3d_ps 1201 14.6 0.622 0.632 5.919 5.963 cp_dbcsr_syevd 50 12.0 0.003 0.003 5.928 5.930 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.422 5.491 mp_alltoall_z22v 1201 16.6 5.145 5.196 5.145 5.196 cp_fm_upper_to_full 72 14.2 3.489 5.161 3.489 5.161 fft_wrap_pw1pw2_140 487 13.2 0.577 0.582 5.100 5.147 potential_pw2rs 119 12.3 0.014 0.014 4.971 5.009 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.835 4.837 cp_fm_diag_elpa_base 50 14.0 4.614 4.732 4.824 4.825 dbcsr_complete_redistribute 329 12.2 0.267 0.273 3.599 4.666 mp_sum_l 7287 12.8 3.398 4.449 3.398 4.449 make_images_sizes 4572 15.5 0.005 0.005 2.239 4.126 mp_alltoall_i44 4572 16.5 2.235 4.121 2.235 4.121 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.490 3.879 apply_single 119 13.6 0.000 0.000 3.489 3.879 mp_allgather_i34 2286 14.5 1.862 3.817 1.862 3.817 mp_alltoall_d11v 2130 13.8 3.716 3.794 3.716 3.794 copy_fm_to_dbcsr 176 11.2 0.001 0.001 2.545 3.597 calculate_dm_sparse 119 9.5 0.000 0.000 3.412 3.499 wfi_extrapolate 11 7.9 0.001 0.001 3.471 3.474 ot_diis_step 108 11.5 0.016 0.016 3.457 3.465 cp_fm_cholesky_decompose 22 10.9 3.299 3.311 3.299 3.311 transfer_rs2pw 487 10.6 0.005 0.005 3.120 3.173 mp_alltoall_i22 627 13.8 1.973 3.014 1.973 3.014 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.926 3.002 transfer_pw2rs 487 13.2 0.003 0.004 2.877 2.918 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.839 2.889 yz_to_x 606 15.1 0.057 0.059 2.789 2.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.739 2.743 calculate_first_density_matrix 1 7.0 0.000 0.000 2.576 2.581 x_to_yz 595 16.2 0.082 0.083 2.495 2.503 grid_integrate_task_list 119 12.3 2.170 2.182 2.170 2.182 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.054 2.091 mp_sendrecv_dv 2499 12.7 2.002 2.050 2.002 2.050 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.994 1.995 acc_transpose_blocks 9144 15.5 0.043 0.043 1.845 1.895 copy_dbcsr_to_fm 153 11.3 0.002 0.002 1.801 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=89.799000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=725.090909, yerr=13.865946 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.665792E+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 0.019 0.043 85.691 85.692 qs_mol_dyn_low 1 2.0 0.004 0.009 85.209 85.265 qs_forces 11 3.9 0.004 0.005 84.601 84.602 qs_energies 11 4.9 0.002 0.003 81.681 81.700 scf_env_do_scf 11 5.9 0.000 0.001 72.436 72.438 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.619 66.620 dbcsr_multiply_generic 2055 12.4 0.108 0.112 52.611 52.808 qs_scf_new_mos 99 7.5 0.000 0.001 49.079 49.221 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.078 49.220 ot_scf_mini 99 9.5 0.002 0.002 46.703 46.821 velocity_verlet 10 3.0 0.005 0.018 44.801 44.807 multiply_cannon 2055 13.4 0.187 0.191 43.417 44.407 multiply_cannon_loop 2055 14.4 1.790 1.834 42.355 43.391 ot_mini 99 10.5 0.001 0.001 27.824 27.918 qs_ot_get_derivative 99 11.5 0.001 0.001 20.980 21.074 multiply_cannon_multrec 49320 15.4 11.371 12.073 17.506 18.272 rebuild_ks_matrix 110 8.3 0.000 0.000 14.813 14.970 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.813 14.970 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.028 13.174 mp_waitall_1 220248 16.4 11.666 12.567 11.666 12.567 qs_ot_get_p 110 10.4 0.001 0.001 10.275 10.411 multiply_cannon_sync_h2d 49320 15.4 9.551 10.214 9.551 10.214 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.258 8.943 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.431 8.270 apply_single 110 13.6 0.000 0.001 7.431 8.270 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.837 8.254 sum_up_and_integrate 110 10.3 0.002 0.003 7.205 7.219 integrate_v_rspace 110 11.3 0.003 0.003 7.179 7.199 init_scf_run 11 5.9 0.000 0.001 7.150 7.150 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.150 7.150 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.965 7.002 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.577 6.685 calculate_rho_elec 110 8.6 0.020 0.025 6.577 6.685 ot_diis_step 99 11.5 0.006 0.006 6.636 6.637 dbcsr_mm_accdrv_process 87628 16.1 3.045 3.115 6.005 6.625 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.996 5.996 init_scf_loop 11 6.9 0.000 0.000 5.762 5.763 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.571 5.623 mp_sum_l 6594 12.7 4.468 5.437 4.468 5.437 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.359 5.382 cp_fm_diag_elpa_base 48 14.0 5.344 5.366 5.358 5.381 make_m2s 4110 13.4 0.062 0.067 4.273 4.400 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.778 4.316 make_images 4110 14.4 0.177 0.190 4.175 4.306 wfi_extrapolate 11 7.9 0.001 0.001 4.189 4.189 calculate_dm_sparse 110 9.5 0.001 0.001 3.797 3.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.737 3.741 density_rs2pw 110 9.6 0.004 0.005 3.563 3.738 prepare_preconditioner 11 7.9 0.000 0.000 3.527 3.547 make_preconditioner 11 8.9 0.000 0.000 3.527 3.547 pw_transfer 1331 11.6 0.055 0.068 3.358 3.440 grid_integrate_task_list 110 12.3 3.238 3.418 3.238 3.418 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.327 3.365 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.315 3.354 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.269 3.354 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.265 3.324 jit_kernel_multiply 13 15.9 2.683 3.089 2.683 3.089 calculate_first_density_matrix 1 7.0 0.000 0.000 2.862 2.867 fft_wrap_pw1pw2_140 451 13.1 0.389 0.427 2.735 2.840 mp_alltoall_d11v 2046 13.8 2.234 2.821 2.234 2.821 potential_pw2rs 110 12.3 0.005 0.006 2.771 2.793 acc_transpose_blocks 49320 15.4 0.210 0.216 2.589 2.647 fft3d_ps 1111 14.6 0.798 0.896 2.564 2.625 grid_collocate_task_list 110 9.6 2.161 2.266 2.161 2.266 mp_waitany 14300 13.8 1.894 2.167 1.894 2.167 transfer_rs2pw 451 10.6 0.005 0.006 1.981 2.127 make_images_data 4110 15.4 0.043 0.047 1.892 2.032 mp_sum_d 3889 11.9 1.427 1.974 1.427 1.974 cp_fm_cholesky_invert 11 10.9 1.967 1.973 1.967 1.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.937 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.648 1.925 transfer_pw2rs 451 13.1 0.006 0.007 1.872 1.884 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.704 1.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.692000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.699862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 592.113664E+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 3061714. 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 0.028 0.068 73.792 73.793 qs_mol_dyn_low 1 2.0 0.003 0.004 73.304 73.370 qs_forces 11 3.9 0.004 0.008 73.196 73.198 qs_energies 11 4.9 0.006 0.019 69.812 69.819 scf_env_do_scf 11 5.9 0.001 0.001 60.558 60.562 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 52.252 52.253 dbcsr_multiply_generic 2055 12.4 0.121 0.134 39.396 39.750 velocity_verlet 10 3.0 0.005 0.011 37.894 37.896 qs_scf_new_mos 99 7.5 0.001 0.001 35.350 35.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.350 35.446 ot_scf_mini 99 9.5 0.003 0.003 33.577 33.653 multiply_cannon 2055 13.4 0.226 0.244 31.979 33.094 multiply_cannon_loop 2055 14.4 1.165 1.203 30.596 31.591 ot_mini 99 10.5 0.001 0.001 19.423 19.524 multiply_cannon_multrec 24660 15.4 6.862 8.271 13.986 15.550 rebuild_ks_matrix 110 8.3 0.000 0.001 13.966 14.051 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.022 13.966 14.051 qs_ot_get_derivative 99 11.5 0.001 0.001 13.496 13.576 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.301 12.376 mp_waitall_1 176588 16.5 8.534 11.389 8.534 11.389 multiply_cannon_metrocomm3 24660 15.4 0.072 0.075 5.698 8.679 init_scf_loop 11 6.9 0.001 0.009 8.266 8.266 qs_ot_get_p 110 10.4 0.001 0.001 7.279 7.393 dbcsr_mm_accdrv_process 52282 16.1 5.552 6.303 6.958 7.279 multiply_cannon_sync_h2d 24660 15.4 6.305 7.232 6.305 7.232 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.529 7.020 apply_single 110 13.6 0.000 0.001 6.528 7.020 sum_up_and_integrate 110 10.3 0.001 0.003 6.698 6.711 integrate_v_rspace 110 11.3 0.003 0.004 6.671 6.682 init_scf_run 11 5.9 0.000 0.001 6.538 6.539 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.538 6.539 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.210 6.238 calculate_rho_elec 110 8.6 0.039 0.047 6.209 6.237 prepare_preconditioner 11 7.9 0.000 0.001 6.166 6.185 make_preconditioner 11 8.9 0.000 0.001 6.166 6.185 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.708 5.870 ot_diis_step 99 11.5 0.010 0.010 5.855 5.856 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.002 5.707 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.191 5.220 make_m2s 4110 13.4 0.058 0.061 4.671 5.070 make_images 4110 14.4 0.407 0.468 4.561 4.959 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.663 4.664 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.013 4.028 cp_fm_diag_elpa_base 48 14.0 3.964 3.980 4.010 4.025 pw_transfer 1331 11.6 0.066 0.073 3.834 3.981 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.727 3.876 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.659 3.701 wfi_extrapolate 11 7.9 0.001 0.001 3.663 3.663 density_rs2pw 110 9.6 0.004 0.005 3.369 3.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.382 3.383 grid_integrate_task_list 110 12.3 3.159 3.353 3.159 3.353 fft_wrap_pw1pw2_140 451 13.1 0.459 0.474 3.120 3.267 calculate_dm_sparse 110 9.5 0.001 0.001 3.132 3.175 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.969 2.993 fft3d_ps 1111 14.6 1.119 1.349 2.825 2.983 make_images_data 4110 15.4 0.049 0.054 2.547 2.979 cp_fm_cholesky_invert 11 10.9 2.941 2.948 2.941 2.948 hybrid_alltoall_any 4261 16.3 0.106 0.444 2.236 2.932 mp_sum_l 6594 12.7 2.144 2.793 2.144 2.793 calculate_first_density_matrix 1 7.0 0.001 0.010 2.776 2.778 potential_pw2rs 110 12.3 0.008 0.009 2.504 2.518 grid_collocate_task_list 110 9.6 2.172 2.304 2.172 2.304 mp_alltoall_d11v 2046 13.8 2.013 2.231 2.013 2.231 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.126 2.151 acc_transpose_blocks 24660 15.4 0.112 0.117 2.008 2.036 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.995 1.998 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.832 1.843 transfer_rs2pw 451 10.6 0.006 0.007 1.613 1.834 jit_kernel_multiply 9 16.3 1.046 1.833 1.046 1.833 mp_allgather_i34 2055 14.4 0.688 1.829 0.688 1.829 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.779 1.814 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.606 1.619 dbcsr_complete_redistribute 325 12.2 0.255 0.333 1.305 1.595 mp_irecv_dv 57340 16.2 0.651 1.588 0.651 1.588 mp_waitany 10164 13.8 1.346 1.553 1.346 1.553 cp_fm_cholesky_decompose 22 10.9 1.522 1.529 1.522 1.529 mp_alltoall_z22v 1111 16.6 1.441 1.514 1.441 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.489 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.793000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.545455, yerr=6.401188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 667.041792E+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 0.028 0.056 63.637 63.648 qs_mol_dyn_low 1 2.0 0.003 0.004 63.330 63.349 qs_forces 11 3.9 0.003 0.005 63.202 63.204 qs_energies 11 4.9 0.004 0.007 59.954 59.960 scf_env_do_scf 11 5.9 0.000 0.001 51.705 51.705 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 42.626 42.626 velocity_verlet 10 3.0 0.006 0.013 34.333 34.336 dbcsr_multiply_generic 2055 12.4 0.122 0.144 30.326 30.692 qs_scf_new_mos 99 7.5 0.001 0.001 27.302 27.414 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.302 27.413 ot_scf_mini 99 9.5 0.002 0.002 25.969 26.110 multiply_cannon 2055 13.4 0.220 0.239 23.244 24.514 multiply_cannon_loop 2055 14.4 0.817 0.853 21.932 23.336 ot_mini 99 10.5 0.001 0.001 14.726 14.879 rebuild_ks_matrix 110 8.3 0.000 0.000 12.543 12.737 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.543 12.737 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.061 11.239 multiply_cannon_multrec 16440 15.4 3.570 4.574 9.906 10.882 mp_waitall_1 139946 16.5 7.851 10.753 7.851 10.753 qs_ot_get_derivative 99 11.5 0.001 0.001 10.106 10.247 init_scf_loop 11 6.9 0.000 0.000 9.039 9.040 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.828 7.909 prepare_preconditioner 11 7.9 0.000 0.000 7.229 7.244 make_preconditioner 11 8.9 0.000 0.000 7.229 7.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.531 6.911 sum_up_and_integrate 110 10.3 0.001 0.002 6.512 6.528 integrate_v_rspace 110 11.3 0.003 0.003 6.486 6.503 dbcsr_mm_accdrv_process 34862 16.1 5.362 5.884 6.184 6.321 qs_ot_get_p 110 10.4 0.001 0.001 6.143 6.315 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.130 6.150 calculate_rho_elec 110 8.6 0.058 0.059 6.129 6.150 init_scf_run 11 5.9 0.000 0.001 5.777 5.777 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.777 5.777 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.063 5.557 apply_single 110 13.6 0.000 0.000 5.063 5.556 make_m2s 4110 13.4 0.050 0.053 4.638 5.016 make_images 4110 14.4 0.398 0.523 4.522 4.901 ot_diis_step 99 11.5 0.011 0.011 4.566 4.567 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.305 4.313 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.386 4.049 multiply_cannon_sync_h2d 16440 15.4 3.195 3.908 3.195 3.908 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.903 3.904 pw_transfer 1331 11.6 0.065 0.072 3.880 3.890 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.773 3.786 density_rs2pw 110 9.6 0.004 0.004 3.282 3.469 grid_integrate_task_list 110 12.3 3.186 3.387 3.186 3.387 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.304 3.314 cp_fm_diag_elpa_base 48 14.0 3.233 3.268 3.302 3.312 fft_wrap_pw1pw2_140 451 13.1 0.578 0.584 3.205 3.216 wfi_extrapolate 11 7.9 0.001 0.001 3.156 3.156 make_images_data 4110 15.4 0.045 0.049 2.635 3.095 hybrid_alltoall_any 4261 16.3 0.108 0.383 2.331 2.973 cp_fm_cholesky_invert 11 10.9 2.950 2.957 2.950 2.957 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.917 2.918 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.761 2.839 fft3d_ps 1111 14.6 1.097 1.102 2.717 2.728 calculate_dm_sparse 110 9.5 0.001 0.001 2.658 2.690 mp_sum_l 6594 12.7 1.805 2.617 1.805 2.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.524 2.581 calculate_first_density_matrix 1 7.0 0.000 0.000 2.526 2.527 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.896 2.493 grid_collocate_task_list 110 9.6 2.221 2.383 2.221 2.383 mp_irecv_dv 48980 15.7 0.821 2.356 0.821 2.356 potential_pw2rs 110 12.3 0.011 0.012 2.291 2.297 mp_alltoall_d11v 2046 13.8 2.000 2.265 2.000 2.265 dbcsr_complete_redistribute 325 12.2 0.328 0.365 1.550 2.026 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.997 2.000 cp_fm_upper_to_full 70 14.2 1.433 1.817 1.433 1.817 mp_allgather_i34 2055 14.4 0.618 1.791 0.618 1.791 cp_fm_cholesky_decompose 22 10.9 1.704 1.723 1.704 1.723 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.681 1.698 acc_transpose_blocks 16440 15.4 0.075 0.077 1.599 1.664 transfer_rs2pw 451 10.6 0.005 0.006 1.420 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.538 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.375 1.511 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.038 1.502 mp_waitany 17072 13.8 1.214 1.466 1.214 1.466 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.419 1.430 mp_alltoall_z22v 1111 16.6 1.406 1.427 1.406 1.427 transfer_pw2rs 451 13.1 0.005 0.005 1.279 1.284 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.648000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.000000, yerr=9.816498 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 731.865088E+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 0.100 0.127 121.667 121.674 qs_mol_dyn_low 1 2.0 0.003 0.004 121.223 121.259 qs_forces 11 3.9 0.003 0.005 121.115 121.119 qs_energies 11 4.9 0.004 0.007 116.764 116.789 scf_env_do_scf 11 5.9 0.000 0.001 104.636 104.639 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 78.227 78.230 velocity_verlet 10 3.0 0.009 0.011 68.284 68.288 qs_scf_new_mos 99 7.5 0.001 0.001 53.972 54.120 qs_scf_loop_do_ot 99 8.5 0.001 0.001 53.971 54.120 ot_scf_mini 99 9.5 0.003 0.003 51.313 51.437 dbcsr_multiply_generic 2055 12.4 0.118 0.124 46.455 48.190 multiply_cannon 2055 13.4 0.251 0.274 29.792 33.729 multiply_cannon_loop 2055 14.4 1.407 1.482 26.116 31.077 ot_mini 99 10.5 0.001 0.001 26.670 26.855 init_scf_loop 11 6.9 0.000 0.000 26.311 26.314 prepare_preconditioner 11 7.9 0.000 0.000 23.860 23.882 make_preconditioner 11 8.9 0.000 0.000 23.860 23.881 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.741 23.414 mp_waitall_1 121746 16.5 14.064 22.994 14.064 22.994 qs_ot_get_derivative 99 11.5 0.001 0.001 20.974 21.094 rebuild_ks_matrix 110 8.3 0.000 0.000 17.698 18.046 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 17.697 18.046 qs_ot_get_p 110 10.4 0.001 0.001 17.215 17.407 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.672 15.978 multiply_cannon_multrec 24660 15.4 3.866 6.727 12.993 14.175 make_m2s 4110 13.4 0.061 0.063 11.295 12.700 qs_ot_p2m_diag 48 11.0 0.055 0.064 12.582 12.653 make_images 4110 14.4 0.578 0.704 11.148 12.553 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 3.516 12.004 cp_dbcsr_syevd 48 12.0 0.003 0.003 11.480 11.484 cp_fm_cholesky_invert 11 10.9 11.430 11.471 11.430 11.471 qs_rho_update_rho_low 110 7.6 0.001 0.001 11.073 11.161 calculate_rho_elec 110 8.6 0.078 0.081 11.072 11.160 sum_up_and_integrate 110 10.3 0.001 0.002 10.866 10.881 integrate_v_rspace 110 11.3 0.003 0.003 10.840 10.854 dbcsr_mm_accdrv_process 52304 16.0 7.776 9.147 8.977 9.983 cp_fm_diag_elpa 48 13.0 0.000 0.000 9.715 9.727 cp_fm_diag_elpa_base 48 14.0 9.434 9.516 9.711 9.723 init_scf_run 11 5.9 0.000 0.001 8.647 8.648 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.647 8.648 make_images_data 4110 15.4 0.048 0.052 6.298 8.579 density_rs2pw 110 9.6 0.004 0.004 7.318 7.473 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.204 7.263 pw_transfer 1331 11.6 0.066 0.076 6.913 7.022 hybrid_alltoall_any 4261 16.3 0.122 0.454 5.184 6.964 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 6.805 6.918 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.591 6.658 mp_sum_l 6594 12.7 4.699 6.333 4.699 6.333 cp_fm_cholesky_decompose 22 10.9 6.001 6.051 6.001 6.051 potential_pw2rs 110 12.3 0.012 0.013 5.930 5.989 fft3d_ps 1111 14.6 1.103 1.132 5.695 5.784 cp_fm_upper_to_full 70 14.2 4.003 5.757 4.003 5.757 dbcsr_complete_redistribute 325 12.2 0.409 0.467 4.253 5.577 wfi_extrapolate 11 7.9 0.001 0.001 5.521 5.522 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.769 5.444 apply_single 110 13.6 0.000 0.000 4.769 5.443 ot_diis_step 99 11.5 0.011 0.012 5.348 5.349 mp_alltoall_d11v 2046 13.8 4.774 5.016 4.774 5.016 fft_wrap_pw1pw2_140 451 13.1 0.602 0.623 4.873 4.986 make_images_sizes 4110 15.4 0.005 0.005 2.688 4.637 mp_alltoall_i44 4110 16.4 2.683 4.632 2.683 4.632 mp_alltoall_z22v 1111 16.6 4.369 4.490 4.369 4.490 copy_fm_to_dbcsr 174 11.2 0.001 0.001 3.065 4.378 calculate_dm_sparse 110 9.5 0.001 0.001 4.181 4.257 multiply_cannon_metrocomm1 24660 15.4 0.036 0.036 2.164 4.216 transfer_rs2pw 451 10.6 0.005 0.005 4.025 4.187 mp_allgather_i34 2055 14.4 2.743 3.937 2.743 3.937 transfer_pw2rs 451 13.1 0.005 0.005 3.752 3.811 mp_alltoall_i22 605 13.7 2.369 3.750 2.369 3.750 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.665 3.669 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.475 3.621 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.110 3.429 grid_integrate_task_list 110 12.3 3.224 3.381 3.224 3.381 calculate_first_density_matrix 1 7.0 0.000 0.000 2.942 2.945 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.759 2.880 mp_waitany 13376 13.8 2.401 2.710 2.401 2.710 acc_transpose_blocks 24660 15.4 0.111 0.113 2.464 2.650 mp_sum_d 3887 11.9 1.902 2.551 1.902 2.551 yz_to_x 561 15.1 0.093 0.100 2.432 2.530 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.436 2.439 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=121.674000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.090909, yerr=7.844564 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 852.148224E+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 0.184 0.267 62.757 62.759 qs_mol_dyn_low 1 2.0 0.003 0.003 62.111 62.127 qs_forces 11 3.9 0.003 0.003 61.967 61.970 qs_energies 11 4.9 0.003 0.009 58.191 58.198 scf_env_do_scf 11 5.9 0.000 0.001 49.219 49.219 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.313 40.314 velocity_verlet 10 3.0 0.006 0.012 34.696 34.703 dbcsr_multiply_generic 2055 12.4 0.106 0.111 25.855 26.121 qs_scf_new_mos 99 7.5 0.001 0.001 24.560 24.619 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.559 24.618 ot_scf_mini 99 9.5 0.002 0.002 23.147 23.184 multiply_cannon 2055 13.4 0.256 0.282 18.771 20.463 multiply_cannon_loop 2055 14.4 0.602 0.631 17.191 17.738 rebuild_ks_matrix 110 8.3 0.000 0.000 12.440 12.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.439 12.540 ot_mini 99 10.5 0.001 0.001 12.465 12.495 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.100 11.190 mp_waitall_1 103326 16.6 6.922 9.647 6.922 9.647 init_scf_loop 11 6.9 0.000 0.000 8.856 8.857 multiply_cannon_multrec 8220 15.4 3.194 4.353 7.664 8.627 qs_ot_get_derivative 99 11.5 0.001 0.001 8.325 8.386 prepare_preconditioner 11 7.9 0.000 0.000 7.118 7.125 make_preconditioner 11 8.9 0.000 0.000 7.118 7.125 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.676 6.761 sum_up_and_integrate 110 10.3 0.001 0.002 6.707 6.724 integrate_v_rspace 110 11.3 0.003 0.003 6.681 6.698 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.506 6.517 calculate_rho_elec 110 8.6 0.114 0.114 6.505 6.516 qs_ot_get_p 110 10.4 0.001 0.001 6.374 6.432 init_scf_run 11 5.9 0.000 0.001 5.657 5.657 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.656 5.656 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.121 5.275 dbcsr_mm_accdrv_process 17442 15.9 3.185 4.204 4.329 5.261 make_m2s 4110 13.4 0.039 0.040 4.834 5.213 make_images 4110 14.4 0.650 0.714 4.703 5.081 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.500 4.504 pw_transfer 1331 11.6 0.066 0.071 4.421 4.445 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.314 4.341 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.097 4.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.948 4.087 apply_single 110 13.6 0.000 0.000 3.948 4.087 ot_diis_step 99 11.5 0.012 0.012 4.056 4.057 density_rs2pw 110 9.6 0.004 0.004 3.392 3.625 fft_wrap_pw1pw2_140 451 13.1 0.777 0.786 3.590 3.618 cp_fm_cholesky_invert 11 10.9 3.613 3.617 3.613 3.617 grid_integrate_task_list 110 12.3 3.377 3.517 3.377 3.517 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.482 3.490 cp_fm_diag_elpa_base 48 14.0 3.423 3.452 3.480 3.488 wfi_extrapolate 11 7.9 0.001 0.001 3.074 3.074 fft3d_ps 1111 14.6 1.139 1.165 2.980 3.001 make_images_data 4110 15.4 0.041 0.047 2.481 2.988 hybrid_alltoall_any 4261 16.3 0.199 0.843 2.377 2.824 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.792 2.797 calculate_dm_sparse 110 9.5 0.001 0.001 2.719 2.762 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.743 2.745 multiply_cannon_sync_h2d 8220 15.4 2.356 2.502 2.356 2.502 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.467 2.491 calculate_first_density_matrix 1 7.0 0.000 0.000 2.475 2.476 grid_collocate_task_list 110 9.6 2.363 2.465 2.363 2.465 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.417 2.453 potential_pw2rs 110 12.3 0.015 0.015 2.343 2.349 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.267 2.317 mp_alltoall_d11v 2046 13.8 2.093 2.293 2.093 2.293 cp_fm_cholesky_decompose 22 10.9 2.006 2.026 2.006 2.026 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 1.377 2.013 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.998 mp_allgather_i34 2055 14.4 0.727 1.939 0.727 1.939 dbcsr_complete_redistribute 325 12.2 0.584 0.628 1.682 1.794 mp_sum_l 6594 12.7 1.505 1.751 1.505 1.751 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.723 1.731 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.566 1.682 mp_alltoall_z22v 1111 16.6 1.606 1.635 1.606 1.635 transfer_rs2pw 451 10.6 0.005 0.005 1.361 1.619 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.549 mp_waitany 9240 13.8 1.245 1.523 1.245 1.523 acc_transpose_blocks 8220 15.4 0.038 0.039 1.385 1.422 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.329 1.374 qs_create_task_list 11 7.9 0.001 0.002 1.254 1.353 generate_qs_task_list 11 8.9 0.375 0.443 1.253 1.352 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.244 1.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=62.759000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.545455, yerr=9.948628 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 1.416180E+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 0.030 0.053 154.811 154.817 qs_mol_dyn_low 1 2.0 0.003 0.003 154.287 154.300 qs_forces 11 3.9 0.003 0.004 154.161 154.168 qs_energies 11 4.9 0.003 0.005 148.824 148.837 scf_env_do_scf 11 5.9 0.001 0.001 134.507 134.507 velocity_verlet 10 3.0 0.002 0.003 96.169 96.178 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 84.337 84.340 qs_scf_new_mos 99 7.5 0.001 0.001 55.138 55.300 qs_scf_loop_do_ot 99 8.5 0.001 0.001 55.138 55.300 ot_scf_mini 99 9.5 0.002 0.002 52.340 52.507 init_scf_loop 11 6.9 0.000 0.000 50.022 50.030 dbcsr_multiply_generic 2055 12.4 0.135 0.140 49.192 49.563 prepare_preconditioner 11 7.9 0.000 0.000 46.888 46.938 make_preconditioner 11 8.9 0.000 0.000 46.888 46.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 40.034 45.961 multiply_cannon 2055 13.4 0.356 0.391 33.261 35.749 multiply_cannon_loop 2055 14.4 0.838 0.855 29.689 30.632 ot_mini 99 10.5 0.001 0.001 26.716 27.060 mp_waitall_1 84994 16.7 22.954 26.533 22.954 26.533 rebuild_ks_matrix 110 8.3 0.000 0.001 21.504 21.749 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 21.503 21.748 qs_ot_get_derivative 99 11.5 0.001 0.001 20.137 20.317 cp_fm_upper_to_full 70 14.2 13.310 19.635 13.310 19.635 qs_ks_update_qs_env 110 7.6 0.001 0.001 19.298 19.511 cp_fm_cholesky_invert 11 10.9 18.611 18.627 18.611 18.627 qs_ot_get_p 110 10.4 0.001 0.001 17.024 17.329 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 13.219 14.321 make_m2s 4110 13.4 0.044 0.044 12.017 13.861 make_images 4110 14.4 0.892 0.973 11.822 13.666 dbcsr_complete_redistribute 325 12.2 1.023 1.044 9.886 13.567 qs_rho_update_rho_low 110 7.6 0.001 0.001 13.423 13.500 calculate_rho_elec 110 8.6 0.225 0.225 13.423 13.500 qs_ot_p2m_diag 48 11.0 0.151 0.156 12.635 12.660 sum_up_and_integrate 110 10.3 0.002 0.002 11.977 12.001 integrate_v_rspace 110 11.3 0.003 0.004 11.949 11.973 copy_fm_to_dbcsr 174 11.2 0.001 0.001 8.058 11.736 cp_dbcsr_syevd 48 12.0 0.003 0.003 11.561 11.565 pw_transfer 1331 11.6 0.075 0.076 10.570 10.649 make_images_data 4110 15.4 0.044 0.048 7.643 10.563 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 10.452 10.532 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.838 10.479 multiply_cannon_multrec 8220 15.4 4.177 4.332 10.062 10.187 mp_alltoall_i22 605 13.7 6.380 10.089 6.380 10.089 cp_fm_diag_elpa 48 13.0 0.000 0.000 9.874 9.875 cp_fm_diag_elpa_base 48 14.0 9.168 9.529 9.867 9.867 hybrid_alltoall_any 4261 16.3 0.262 0.563 6.817 9.302 init_scf_run 11 5.9 0.000 0.001 9.250 9.252 scf_env_initial_rho_setup 11 6.9 0.000 0.000 9.250 9.252 density_rs2pw 110 9.6 0.004 0.004 8.686 8.717 fft3d_ps 1111 14.6 1.313 1.319 8.505 8.587 fft_wrap_pw1pw2_140 451 13.1 1.278 1.284 8.300 8.378 cp_fm_cholesky_decompose 22 10.9 7.537 7.550 7.537 7.550 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.587 7.100 apply_single 110 13.6 0.000 0.000 6.587 7.100 mp_alltoall_z22v 1111 16.6 6.931 7.014 6.931 7.014 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.248 6.738 ot_diis_step 99 11.5 0.016 0.016 6.465 6.468 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.343 6.451 wfi_extrapolate 11 7.9 0.001 0.001 6.333 6.335 potential_pw2rs 110 12.3 0.021 0.021 6.268 6.315 mp_alltoall_d11v 2046 13.8 5.947 6.016 5.947 6.016 dbcsr_mm_accdrv_process 11614 15.7 3.994 4.185 5.722 5.934 mp_sum_l 6594 12.7 2.999 5.149 2.999 5.149 mp_allgather_i34 2055 14.4 2.283 5.112 2.283 5.112 make_images_sizes 4110 15.4 0.004 0.005 2.186 4.518 mp_alltoall_i44 4110 16.4 2.182 4.514 2.182 4.514 calculate_dm_sparse 110 9.5 0.001 0.001 4.429 4.470 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.042 4.048 grid_integrate_task_list 110 12.3 3.908 3.963 3.908 3.963 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.929 3.935 yz_to_x 561 15.1 0.110 0.112 3.700 3.780 transfer_rs2pw 451 10.6 0.006 0.006 3.489 3.566 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.432 3.483 x_to_yz 550 16.1 0.137 0.138 3.478 3.483 transfer_pw2rs 451 13.1 0.004 0.004 3.243 3.303 multiply_cannon_sync_h2d 8220 15.4 3.151 3.178 3.151 3.178 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.988 3.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=154.817000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1267.272727, yerr=66.191308 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.934976E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2529110. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.078 213.729 213.732 qs_mol_dyn_low 1 2.0 0.003 0.005 212.896 212.937 qs_forces 11 3.9 0.015 0.036 212.772 212.778 qs_energies 11 4.9 0.006 0.020 206.940 206.955 scf_env_do_scf 11 5.9 0.001 0.001 189.628 189.632 scf_env_do_scf_inner_loop 117 6.6 0.004 0.016 167.796 167.797 dbcsr_multiply_generic 2507 12.6 0.182 0.189 127.353 128.121 velocity_verlet 10 3.0 0.005 0.012 127.829 127.830 qs_scf_new_mos 117 7.6 0.001 0.001 127.251 127.501 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.250 127.500 ot_scf_mini 117 9.6 0.003 0.003 120.468 120.725 multiply_cannon 2507 13.6 0.251 0.283 102.233 104.186 multiply_cannon_loop 2507 14.6 2.405 2.488 99.870 101.913 ot_mini 117 10.6 0.001 0.001 67.215 67.468 multiply_cannon_multrec 60168 15.6 31.713 33.602 41.543 43.497 qs_ot_get_derivative 117 11.6 0.001 0.001 42.325 42.588 rebuild_ks_matrix 128 8.3 0.001 0.001 34.211 34.613 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 34.210 34.612 mp_waitall_1 267128 16.5 29.942 33.487 29.942 33.487 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.703 31.057 qs_ot_get_p 128 10.4 0.001 0.001 30.795 31.032 multiply_cannon_sync_h2d 60168 15.6 26.256 28.295 26.256 28.295 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.357 25.004 apply_single 128 13.6 0.001 0.001 24.357 25.004 ot_diis_step 117 11.6 0.007 0.008 24.620 24.622 qs_ot_p2m_diag 83 11.4 0.091 0.123 24.060 24.167 init_scf_loop 11 6.9 0.000 0.001 21.746 21.747 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.023 21.037 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.288 20.431 multiply_cannon_metrocomm3 60168 15.6 0.118 0.124 16.087 18.521 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.945 17.983 cp_fm_diag_elpa_base 83 14.4 17.870 17.909 17.941 17.978 prepare_preconditioner 11 7.9 0.000 0.000 17.070 17.117 make_preconditioner 11 8.9 0.000 0.000 17.070 17.116 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.285 16.448 make_m2s 5014 13.6 0.105 0.113 14.615 14.985 make_images 5014 14.6 0.398 0.416 14.430 14.812 sum_up_and_integrate 128 10.3 0.002 0.004 14.505 14.521 integrate_v_rspace 128 11.3 0.005 0.009 14.446 14.464 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.843 14.032 calculate_rho_elec 128 8.7 0.045 0.064 13.842 14.031 init_scf_run 11 5.9 0.000 0.001 12.941 12.943 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.941 12.943 mp_sum_l 7950 12.9 9.252 10.373 9.252 10.373 dbcsr_mm_accdrv_process 124484 16.2 4.731 4.879 9.394 9.922 cp_fm_cholesky_invert 11 10.9 9.779 9.787 9.779 9.787 multiply_cannon_metrocomm1 60168 15.6 0.092 0.099 6.764 9.388 wfi_extrapolate 11 7.9 0.001 0.001 9.328 9.328 calculate_dm_sparse 128 9.5 0.001 0.001 8.746 8.857 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.435 8.595 pw_transfer 1547 11.6 0.074 0.089 8.193 8.386 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.275 8.353 make_images_data 5014 15.6 0.067 0.074 7.158 8.205 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.990 8.185 density_rs2pw 128 9.7 0.006 0.007 7.365 7.892 hybrid_alltoall_any 5200 16.5 0.294 2.255 6.228 7.570 grid_integrate_task_list 128 12.3 7.079 7.455 7.079 7.455 fft_wrap_pw1pw2_140 523 13.2 1.123 1.169 6.882 7.072 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.866 6.879 fft3d_ps 1291 14.7 2.192 2.849 5.999 6.350 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.898 5.990 mp_alltoall_d11v 2415 14.1 4.745 5.873 4.745 5.873 grid_collocate_task_list 128 9.7 4.841 5.268 4.841 5.268 cp_fm_cholesky_decompose 22 10.9 5.018 5.032 5.018 5.032 potential_pw2rs 128 12.3 0.009 0.010 4.889 4.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=213.732000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.545455, yerr=6.624522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 837.869568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 4883162. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.066 0.117 191.171 191.177 qs_mol_dyn_low 1 2.0 0.003 0.004 190.660 190.680 qs_forces 11 3.9 0.005 0.006 190.545 190.551 qs_energies 11 4.9 0.004 0.011 183.723 183.743 scf_env_do_scf 11 5.9 0.001 0.013 167.303 167.313 scf_env_do_scf_inner_loop 117 6.6 0.008 0.024 133.699 133.705 velocity_verlet 10 3.0 0.001 0.002 119.675 119.677 dbcsr_multiply_generic 2507 12.6 0.190 0.195 97.849 99.155 qs_scf_new_mos 117 7.6 0.001 0.001 95.101 95.676 qs_scf_loop_do_ot 117 8.6 0.001 0.002 95.100 95.675 ot_scf_mini 117 9.6 0.004 0.005 90.272 90.920 multiply_cannon 2507 13.6 0.505 0.567 77.357 82.366 multiply_cannon_loop 2507 14.6 1.573 1.639 73.917 76.815 ot_mini 117 10.6 0.001 0.001 50.667 51.227 mp_waitall_1 214728 16.6 24.632 38.977 24.632 38.977 multiply_cannon_multrec 30084 15.6 20.720 25.628 31.309 36.590 init_scf_loop 11 6.9 0.001 0.005 33.503 33.528 rebuild_ks_matrix 128 8.3 0.001 0.001 32.369 33.107 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 32.368 33.107 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.057 29.728 qs_ot_get_derivative 117 11.6 0.001 0.002 28.401 29.034 prepare_preconditioner 11 7.9 0.000 0.000 28.646 28.724 make_preconditioner 11 8.9 0.000 0.002 28.646 28.724 multiply_cannon_metrocomm3 30084 15.6 0.096 0.103 16.136 28.680 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.336 27.902 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.348 23.562 apply_single 128 13.6 0.001 0.001 22.348 23.562 qs_ot_get_p 128 10.4 0.036 0.279 22.085 22.752 ot_diis_step 117 11.6 0.014 0.015 22.079 22.081 multiply_cannon_sync_h2d 30084 15.6 17.747 20.443 17.747 20.443 cp_fm_cholesky_invert 11 10.9 16.763 16.776 16.763 16.776 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.638 16.676 make_m2s 5014 13.6 0.091 0.098 14.355 16.088 make_images 5014 14.6 1.173 1.376 14.142 15.878 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.424 15.424 sum_up_and_integrate 128 10.3 0.002 0.004 14.069 14.107 integrate_v_rspace 128 11.3 0.003 0.004 14.008 14.051 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.390 13.420 calculate_rho_elec 128 8.7 0.087 0.104 13.389 13.419 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.242 12.277 cp_fm_diag_elpa_base 83 14.4 11.983 12.084 12.236 12.267 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.283 11.723 init_scf_run 11 5.9 0.000 0.001 11.546 11.548 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.546 11.547 multiply_cannon_metrocomm4 27577 15.6 0.105 0.119 3.911 10.822 dbcsr_mm_accdrv_process 62242 16.2 5.536 6.541 10.051 10.612 make_images_data 5014 15.6 0.067 0.077 8.457 10.526 mp_irecv_dv 69486 16.3 3.705 10.427 3.705 10.427 hybrid_alltoall_any 5200 16.5 0.353 1.573 7.220 9.738 pw_transfer 1547 11.6 0.085 0.099 8.672 8.737 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.449 8.518 wfi_extrapolate 11 7.9 0.001 0.001 8.356 8.356 density_rs2pw 128 9.7 0.006 0.007 7.084 7.612 fft_wrap_pw1pw2_140 523 13.2 1.213 1.242 7.483 7.564 grid_integrate_task_list 128 12.3 7.164 7.512 7.164 7.512 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.306 7.057 cp_fm_cholesky_decompose 22 10.9 6.935 7.032 6.935 7.032 calculate_dm_sparse 128 9.5 0.001 0.001 6.495 6.643 mp_sum_l 7950 12.9 4.355 6.546 4.355 6.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.122 6.130 fft3d_ps 1291 14.7 2.843 3.009 6.012 6.081 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.425 5.509 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.238 5.433 grid_collocate_task_list 128 9.7 5.003 5.389 5.003 5.389 mp_allgather_i34 2507 14.6 1.957 5.199 1.957 5.199 mp_alltoall_d11v 2415 14.1 4.187 4.942 4.187 4.942 potential_pw2rs 128 12.3 0.016 0.018 4.568 4.596 dbcsr_complete_redistribute 395 12.7 0.768 0.859 3.175 4.019 mp_sum_d 4465 12.1 2.781 4.016 2.781 4.016 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.177000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.181818, yerr=0.385695 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 941.916160E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.109 182.175 182.176 qs_mol_dyn_low 1 2.0 0.003 0.004 181.420 181.441 qs_forces 11 3.9 0.008 0.043 181.306 181.309 qs_energies 11 4.9 0.002 0.006 174.706 174.724 scf_env_do_scf 11 5.9 0.001 0.001 158.557 158.561 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.570 121.571 velocity_verlet 10 3.0 0.002 0.002 116.524 116.526 dbcsr_multiply_generic 2507 12.6 0.181 0.186 84.135 85.339 qs_scf_new_mos 117 7.6 0.001 0.001 83.627 83.979 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.626 83.978 ot_scf_mini 117 9.6 0.003 0.004 79.244 79.673 multiply_cannon 2507 13.6 0.511 0.526 63.757 67.874 multiply_cannon_loop 2507 14.6 1.129 1.188 60.383 63.152 ot_mini 117 10.6 0.001 0.001 43.996 44.441 init_scf_loop 11 6.9 0.000 0.000 36.881 36.882 mp_waitall_1 170520 16.6 26.130 36.123 26.130 36.123 prepare_preconditioner 11 7.9 0.000 0.000 32.810 32.864 make_preconditioner 11 8.9 0.000 0.000 32.810 32.864 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.376 31.862 rebuild_ks_matrix 128 8.3 0.001 0.001 30.794 31.282 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.793 31.281 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.770 28.217 multiply_cannon_metrocomm3 20056 15.6 0.062 0.066 16.407 26.173 multiply_cannon_multrec 20056 15.6 12.916 16.399 22.392 25.879 qs_ot_get_derivative 117 11.6 0.002 0.002 23.989 24.424 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.049 21.158 apply_single 128 13.6 0.001 0.001 20.048 21.157 qs_ot_get_p 128 10.4 0.001 0.001 20.347 20.844 ot_diis_step 117 11.6 0.018 0.020 19.867 19.868 make_m2s 5014 13.6 0.081 0.086 15.022 16.377 make_images 5014 14.6 1.153 1.254 14.788 16.143 qs_ot_p2m_diag 83 11.4 0.266 0.272 15.949 15.969 cp_fm_cholesky_invert 11 10.9 15.490 15.499 15.490 15.499 multiply_cannon_sync_h2d 20056 15.6 13.512 15.304 13.512 15.304 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.786 14.788 sum_up_and_integrate 128 10.3 0.002 0.003 14.373 14.397 integrate_v_rspace 128 11.3 0.003 0.005 14.313 14.340 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.967 14.030 calculate_rho_elec 128 8.7 0.130 0.145 13.966 14.029 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.459 11.478 cp_fm_diag_elpa_base 83 14.4 11.020 11.203 11.456 11.475 init_scf_run 11 5.9 0.000 0.001 11.024 11.024 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.024 11.024 make_images_data 5014 15.6 0.063 0.072 8.939 10.781 hybrid_alltoall_any 5200 16.5 0.449 2.036 7.663 9.932 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.546 9.866 multiply_cannon_metrocomm4 17549 15.6 0.066 0.076 3.529 9.579 mp_irecv_dv 50230 16.2 3.400 9.318 3.400 9.318 pw_transfer 1547 11.6 0.085 0.102 9.060 9.169 dbcsr_mm_accdrv_process 41502 16.2 5.834 6.118 8.952 9.080 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.836 8.953 cp_fm_cholesky_decompose 22 10.9 7.954 7.978 7.954 7.978 fft_wrap_pw1pw2_140 523 13.2 1.299 1.330 7.738 7.864 grid_integrate_task_list 128 12.3 7.301 7.768 7.301 7.768 wfi_extrapolate 11 7.9 0.001 0.001 7.697 7.697 density_rs2pw 128 9.7 0.006 0.006 7.206 7.599 cp_fm_upper_to_full 105 14.8 5.880 7.438 5.880 7.438 dbcsr_complete_redistribute 395 12.7 1.175 1.210 4.821 6.673 fft3d_ps 1291 14.7 2.768 2.995 6.217 6.303 calculate_dm_sparse 128 9.5 0.001 0.001 6.098 6.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.682 5.688 grid_collocate_task_list 128 9.7 5.153 5.549 5.153 5.549 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.881 5.526 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.588 5.425 mp_sum_l 7950 12.9 3.544 5.201 3.544 5.201 mp_allgather_i34 2507 14.6 1.855 5.152 1.855 5.152 mp_alltoall_d11v 2415 14.1 4.647 5.138 4.647 5.138 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.716 4.874 potential_pw2rs 128 12.3 0.020 0.022 4.722 4.738 transfer_fm_to_dbcsr 11 9.9 0.020 0.024 2.416 4.233 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.100 4.130 mp_alltoall_i22 716 14.1 2.024 3.996 2.024 3.996 qs_energies_init_hamiltonians 11 5.9 0.019 0.033 3.797 3.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.176000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.818182, yerr=8.054566 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.130017E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.056 196.847 196.850 qs_mol_dyn_low 1 2.0 0.003 0.004 196.257 196.280 qs_forces 11 3.9 0.005 0.006 196.116 196.126 qs_energies 11 4.9 0.025 0.136 188.837 188.953 scf_env_do_scf 11 5.9 0.001 0.005 171.265 171.276 velocity_verlet 10 3.0 0.002 0.002 129.855 129.858 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 122.351 122.352 qs_scf_new_mos 117 7.6 0.001 0.001 84.973 85.316 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.972 85.315 dbcsr_multiply_generic 2507 12.6 0.188 0.193 83.420 84.147 ot_scf_mini 117 9.6 0.003 0.004 80.312 80.672 multiply_cannon 2507 13.6 0.614 0.761 57.023 60.356 multiply_cannon_loop 2507 14.6 1.827 1.926 53.060 54.808 init_scf_loop 11 6.9 0.001 0.003 48.779 48.780 ot_mini 117 10.6 0.001 0.001 44.730 45.088 prepare_preconditioner 11 7.9 0.000 0.000 44.273 44.309 make_preconditioner 11 8.9 0.000 0.000 44.273 44.309 make_full_inverse_cholesky 11 9.9 0.012 0.022 37.880 42.897 multiply_cannon_multrec 30084 15.6 13.237 18.742 26.201 31.408 rebuild_ks_matrix 128 8.3 0.001 0.001 30.000 30.316 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.000 30.316 mp_waitall_1 147882 16.7 19.070 29.744 19.070 29.744 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.069 27.369 qs_ot_get_derivative 117 11.6 0.002 0.002 24.743 25.102 make_m2s 5014 13.6 0.097 0.102 21.238 22.359 make_images 5014 14.6 1.960 2.247 20.927 22.045 qs_ot_get_p 128 10.4 0.001 0.001 21.339 21.713 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.381 19.874 apply_single 128 13.6 0.001 0.001 19.380 19.874 ot_diis_step 117 11.6 0.017 0.020 19.859 19.862 cp_fm_cholesky_invert 11 10.9 17.596 17.606 17.596 17.606 cp_fm_upper_to_full 105 14.8 11.572 16.997 11.572 16.997 qs_ot_p2m_diag 83 11.4 0.343 0.391 16.866 16.934 multiply_cannon_metrocomm3 30084 15.6 0.050 0.052 7.365 16.382 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.384 15.386 sum_up_and_integrate 128 10.3 0.002 0.003 14.447 14.479 integrate_v_rspace 128 11.3 0.003 0.004 14.387 14.420 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.265 14.303 calculate_rho_elec 128 8.7 0.173 0.189 14.265 14.303 make_images_data 5014 15.6 0.066 0.073 11.493 13.454 dbcsr_mm_accdrv_process 62264 16.2 8.496 9.297 12.538 13.022 dbcsr_complete_redistribute 395 12.7 1.503 1.617 9.190 12.949 hybrid_alltoall_any 5200 16.5 0.532 2.225 10.273 12.587 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.157 12.169 cp_fm_diag_elpa_base 83 14.4 11.159 11.501 12.151 12.162 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.727 11.482 init_scf_run 11 5.9 0.000 0.001 11.391 11.395 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.391 11.394 multiply_cannon_sync_h2d 30084 15.6 10.403 11.170 10.403 11.170 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.421 10.683 transfer_fm_to_dbcsr 11 9.9 0.002 0.012 6.371 10.002 pw_transfer 1547 11.6 0.086 0.101 9.414 9.512 mp_alltoall_i22 716 14.1 5.673 9.439 5.673 9.439 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.188 9.290 fft_wrap_pw1pw2_140 523 13.2 1.446 1.486 8.079 8.195 cp_fm_cholesky_decompose 22 10.9 7.963 8.059 7.963 8.059 wfi_extrapolate 11 7.9 0.001 0.001 7.943 7.944 grid_integrate_task_list 128 12.3 7.531 7.909 7.531 7.909 multiply_cannon_metrocomm4 25070 15.6 0.084 0.097 2.919 7.685 density_rs2pw 128 9.7 0.006 0.006 7.227 7.595 mp_irecv_dv 76098 16.2 2.763 7.396 2.763 7.396 fft3d_ps 1291 14.7 2.855 2.929 6.366 6.431 calculate_dm_sparse 128 9.5 0.001 0.001 6.363 6.424 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.799 5.843 mp_alltoall_d11v 2415 14.1 5.238 5.694 5.238 5.694 grid_collocate_task_list 128 9.7 5.283 5.633 5.283 5.633 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.702 4.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.651 4.757 potential_pw2rs 128 12.3 0.023 0.023 4.576 4.595 qs_energies_init_hamiltonians 11 5.9 0.019 0.035 4.575 4.575 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.460 4.511 mp_sum_l 7950 12.9 3.245 4.380 3.245 4.380 mp_allgather_i34 2507 14.6 1.985 4.031 1.985 4.031 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=196.850000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.272727, yerr=13.618777 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.550586E+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 5.532237E+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 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838186. 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 0.056 0.068 178.207 178.214 qs_mol_dyn_low 1 2.0 0.003 0.004 177.659 177.687 qs_forces 11 3.9 0.005 0.007 177.551 177.555 qs_energies 11 4.9 0.007 0.014 170.029 170.039 scf_env_do_scf 11 5.9 0.001 0.006 152.159 152.172 velocity_verlet 10 3.0 0.007 0.008 116.710 116.714 scf_env_do_scf_inner_loop 117 6.6 0.009 0.100 114.176 114.177 qs_scf_new_mos 117 7.6 0.001 0.001 76.231 76.349 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.230 76.349 dbcsr_multiply_generic 2507 12.6 0.182 0.193 74.925 75.379 ot_scf_mini 117 9.6 0.003 0.004 71.673 71.802 multiply_cannon 2507 13.6 0.572 0.615 54.891 58.415 multiply_cannon_loop 2507 14.6 0.810 0.849 51.622 52.714 ot_mini 117 10.6 0.001 0.001 39.301 39.452 init_scf_loop 11 6.9 0.000 0.000 37.823 37.825 prepare_preconditioner 11 7.9 0.000 0.000 33.857 33.892 make_preconditioner 11 8.9 0.000 0.000 33.857 33.892 mp_waitall_1 125778 16.7 26.800 33.836 26.800 33.836 make_full_inverse_cholesky 11 9.9 0.015 0.026 31.719 32.023 rebuild_ks_matrix 128 8.3 0.001 0.001 29.730 29.890 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.729 29.889 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.124 27.264 multiply_cannon_multrec 10028 15.6 10.341 14.676 18.094 22.874 qs_ot_get_derivative 117 11.6 0.001 0.002 21.819 21.961 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 13.393 21.292 cp_fm_cholesky_invert 11 10.9 19.771 19.777 19.771 19.777 qs_ot_get_p 128 10.4 0.001 0.001 19.206 19.367 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.505 17.630 apply_single 128 13.6 0.001 0.001 17.505 17.630 ot_diis_step 117 11.6 0.020 0.021 17.392 17.394 make_m2s 5014 13.6 0.067 0.071 15.806 16.769 make_images 5014 14.6 2.170 2.602 15.499 16.462 qs_ot_p2m_diag 83 11.4 0.496 0.502 15.121 15.141 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.554 14.582 calculate_rho_elec 128 8.7 0.256 0.267 14.553 14.581 sum_up_and_integrate 128 10.3 0.002 0.002 14.532 14.579 integrate_v_rspace 128 11.3 0.004 0.004 14.472 14.521 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.879 13.880 multiply_cannon_sync_h2d 10028 15.6 10.646 11.207 10.646 11.207 init_scf_run 11 5.9 0.000 0.001 10.931 10.932 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.931 10.932 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.868 10.880 cp_fm_diag_elpa_base 83 14.4 10.623 10.708 10.864 10.877 make_images_data 5014 15.6 0.055 0.064 9.277 10.747 hybrid_alltoall_any 5200 16.5 0.840 3.772 8.926 10.350 pw_transfer 1547 11.6 0.085 0.094 9.801 9.837 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.579 9.620 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.621 8.712 cp_fm_cholesky_decompose 22 10.9 8.366 8.465 8.366 8.465 fft_wrap_pw1pw2_140 523 13.2 1.797 1.848 8.360 8.403 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.800 8.196 dbcsr_mm_accdrv_process 20762 16.1 3.420 4.646 7.404 8.116 grid_integrate_task_list 128 12.3 7.798 8.094 7.798 8.094 wfi_extrapolate 11 7.9 0.001 0.001 7.613 7.613 density_rs2pw 128 9.7 0.005 0.006 7.103 7.270 calculate_dm_sparse 128 9.5 0.001 0.001 6.345 6.433 fft3d_ps 1291 14.7 2.770 2.854 6.317 6.349 dbcsr_complete_redistribute 395 12.7 2.104 2.141 5.441 5.843 multiply_cannon_metrocomm4 7521 15.6 0.026 0.029 1.855 5.784 grid_collocate_task_list 128 9.7 5.575 5.775 5.575 5.775 mp_irecv_dv 28860 15.9 1.814 5.685 1.814 5.685 mp_alltoall_d11v 2415 14.1 5.053 5.517 5.053 5.517 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.362 5.388 mp_allgather_i34 2507 14.6 1.436 5.333 1.436 5.333 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.298 5.302 potential_pw2rs 128 12.3 0.026 0.026 4.548 4.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.199 4.255 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.925 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.825 3.912 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.526 3.892 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.834 3.886 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.556 3.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.214000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1452.636364, yerr=44.344258 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.095532E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.063 413.523 413.533 qs_mol_dyn_low 1 2.0 0.003 0.003 412.735 412.760 qs_forces 11 3.9 0.004 0.005 412.571 412.577 qs_energies 11 4.9 0.008 0.013 402.363 402.389 scf_env_do_scf 11 5.9 0.006 0.045 374.373 374.379 velocity_verlet 10 3.0 0.032 0.036 292.704 292.717 scf_env_do_scf_inner_loop 118 6.6 0.013 0.021 208.742 208.744 init_scf_loop 11 6.9 0.000 0.000 165.210 165.212 prepare_preconditioner 11 7.9 0.000 0.000 159.059 159.190 make_preconditioner 11 8.9 0.000 0.000 159.059 159.190 make_full_inverse_cholesky 11 9.9 0.038 0.039 134.203 155.221 qs_scf_new_mos 118 7.6 0.001 0.001 148.959 149.772 qs_scf_loop_do_ot 118 8.6 0.001 0.001 148.958 149.771 ot_scf_mini 118 9.6 0.004 0.004 142.719 143.386 dbcsr_multiply_generic 2529 12.6 0.214 0.224 111.818 116.788 multiply_cannon 2529 13.6 0.766 0.836 72.578 81.178 multiply_cannon_loop 2529 14.6 1.072 1.108 64.987 77.569 cp_fm_upper_to_full 106 14.8 53.059 75.861 53.059 75.861 ot_mini 118 10.6 0.001 0.001 66.808 67.772 mp_waitall_1 104580 16.8 44.872 67.095 44.872 67.095 cp_fm_cholesky_invert 11 10.9 61.402 61.429 61.402 61.429 qs_ot_get_p 129 10.4 0.001 0.001 56.874 57.880 qs_ot_p2m_diag 84 11.4 0.891 0.896 48.592 48.793 dbcsr_complete_redistribute 397 12.7 3.975 4.060 32.870 45.974 rebuild_ks_matrix 129 8.3 0.001 0.001 44.462 45.921 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 44.461 45.921 cp_dbcsr_syevd 84 12.4 0.006 0.006 45.706 45.720 qs_ks_update_qs_env 129 7.6 0.001 0.001 41.017 42.334 qs_ot_get_derivative 118 11.6 0.002 0.002 41.064 41.726 copy_fm_to_dbcsr 210 11.7 0.001 0.001 28.034 41.118 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 28.120 40.709 cp_fm_diag_elpa 84 13.4 0.000 0.000 39.600 39.601 cp_fm_diag_elpa_base 84 14.4 35.186 36.785 39.591 39.592 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.813 37.869 mp_alltoall_i22 720 14.1 22.583 35.742 22.583 35.742 make_m2s 5058 13.6 0.079 0.081 27.541 32.244 make_images 5058 14.6 3.111 3.332 27.050 31.753 ot_diis_step 118 11.6 0.025 0.032 25.352 25.354 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.578 25.142 apply_single 129 13.6 0.001 0.001 22.578 25.141 qs_rho_update_rho_low 129 7.7 0.001 0.001 24.869 25.060 calculate_rho_elec 129 8.7 0.484 0.484 24.869 25.059 make_images_data 5058 15.6 0.064 0.070 16.003 24.332 sum_up_and_integrate 129 10.3 0.002 0.002 22.839 22.946 integrate_v_rspace 129 11.3 0.004 0.004 22.776 22.883 hybrid_alltoall_any 5245 16.5 1.318 3.073 14.058 22.164 pw_transfer 1559 11.6 0.096 0.097 19.693 19.846 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 19.452 19.604 cp_fm_cholesky_decompose 22 10.9 19.323 19.346 19.323 19.346 multiply_cannon_multrec 10116 15.6 10.498 12.378 18.504 18.737 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 17.708 18.219 init_scf_run 11 5.9 0.000 0.001 16.784 16.785 scf_env_initial_rho_setup 11 6.9 0.000 0.000 16.784 16.784 fft_wrap_pw1pw2_140 527 13.2 3.048 3.090 16.447 16.597 mp_sum_l 8016 12.9 8.961 15.135 8.961 15.135 fft3d_ps 1301 14.7 2.892 2.911 14.684 14.831 multiply_cannon_sync_h2d 10116 15.6 14.447 14.474 14.447 14.474 density_rs2pw 129 9.7 0.005 0.005 14.160 14.231 mp_alltoall_d11v 2429 14.1 13.112 13.534 13.112 13.534 wfi_extrapolate 11 7.9 0.001 0.001 13.001 13.002 mp_alltoall_z22v 1301 16.7 11.135 11.278 11.135 11.278 mp_allgather_i34 2529 14.6 4.969 10.709 4.969 10.709 potential_pw2rs 129 12.3 0.041 0.041 9.632 9.738 dbcsr_mm_accdrv_process 20934 16.1 4.321 6.439 7.740 9.699 dbcsr_dot_sd 1331 12.0 3.736 3.766 6.757 8.983 grid_integrate_task_list 129 12.3 8.663 8.839 8.663 8.839 copy_dbcsr_to_fm 187 11.8 0.004 0.004 8.282 8.653 calculate_dm_sparse 129 9.5 0.001 0.001 8.307 8.460 qs_energies_init_hamiltonians 11 5.9 0.015 0.025 8.404 8.408 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=413.533000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2778.909091, yerr=163.965326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 1.260065E+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 10878587. 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 0.082 0.182 88.973 88.977 qs_energies 1 2.0 0.000 0.000 87.828 87.836 ls_scf 1 3.0 0.000 0.000 86.900 86.909 dbcsr_multiply_generic 111 6.7 0.014 0.015 74.155 74.405 multiply_cannon 111 7.7 0.018 0.021 56.361 58.031 multiply_cannon_loop 111 8.7 0.226 0.244 52.851 54.638 ls_scf_main 1 4.0 0.000 0.000 54.188 54.188 density_matrix_trs4 2 5.0 0.002 0.003 48.087 48.191 ls_scf_init_scf 1 4.0 0.000 0.000 29.537 29.538 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.200 28.268 mp_waitall_1 11031 10.9 24.170 27.683 24.170 27.683 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.017 26.046 multiply_cannon_multrec 2664 9.7 8.094 8.857 15.456 17.207 multiply_cannon_sync_h2d 2664 9.7 12.944 14.600 12.944 14.600 make_m2s 222 7.7 0.009 0.011 13.717 14.296 make_images 222 8.7 0.099 0.109 13.695 14.274 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 10.615 13.570 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.655 10.266 make_images_data 222 9.7 0.004 0.005 8.152 8.755 hybrid_alltoall_any 227 10.6 0.216 1.834 6.972 8.250 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.700 6.983 7.932 dbcsr_mm_accdrv_process_sort 4760 11.4 6.197 7.068 6.197 7.068 calculate_norms 4752 9.8 5.436 6.119 5.436 6.119 mp_sum_l 887 5.1 3.724 5.738 3.724 5.738 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.304 5.549 arnoldi_extremal 4 6.8 0.000 0.000 4.290 4.325 arnoldi_normal_ev 4 7.8 0.001 0.002 4.290 4.325 build_subspace 16 8.4 0.009 0.012 4.172 4.175 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.572 4.116 make_images_sizes 222 9.7 0.000 0.000 0.904 3.867 mp_alltoall_i44 222 10.7 0.903 3.866 0.903 3.866 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.077 3.483 mp_irecv_dv 6231 10.9 2.061 3.453 2.061 3.453 ls_scf_post 1 4.0 0.000 0.000 3.175 3.183 ls_scf_store_result 1 5.0 0.000 0.000 2.988 3.027 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.635 2.861 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.671 2.802 dbcsr_special_finalize 555 9.7 0.005 0.006 2.340 2.745 dbcsr_merge_single_wm 555 10.7 0.453 0.581 2.331 2.735 make_images_pack 222 9.7 2.174 2.592 2.176 2.593 dbcsr_sort_data 658 11.4 2.136 2.444 2.136 2.444 dbcsr_matrix_vector_mult_local 304 10.0 2.008 2.405 2.010 2.407 buffer_matrices_ensure_size 222 8.7 1.742 2.055 1.742 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.004 2.005 rebuild_ks_matrix 3 7.3 0.000 0.000 1.994 1.995 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.019 1.994 1.995 mp_sum_dv 2129 10.8 1.506 1.845 1.506 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=88.977000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 2.135982E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.045 92.525 92.526 qs_energies 1 2.0 0.000 0.000 91.964 91.968 ls_scf 1 3.0 0.000 0.000 90.608 90.613 dbcsr_multiply_generic 111 6.7 0.015 0.018 74.885 75.205 multiply_cannon 111 7.7 0.030 0.044 53.097 57.542 ls_scf_main 1 4.0 0.000 0.000 55.585 55.590 multiply_cannon_loop 111 8.7 0.135 0.147 50.336 53.796 density_matrix_trs4 2 5.0 0.002 0.003 49.553 49.730 ls_scf_init_scf 1 4.0 0.000 0.000 31.219 31.220 mp_waitall_1 9105 10.9 21.338 30.200 21.338 30.200 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.920 30.010 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.482 27.493 multiply_cannon_multrec 1332 9.7 13.292 16.636 22.592 27.054 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.736 20.616 make_m2s 222 7.7 0.006 0.008 15.194 15.659 make_images 222 8.7 1.368 1.691 15.163 15.628 dbcsr_mm_accdrv_process 4041 10.4 0.344 0.564 8.899 10.484 dbcsr_mm_accdrv_process_sort 4041 11.4 8.404 9.940 8.404 9.940 hybrid_alltoall_any 227 10.6 0.542 2.555 8.049 9.521 make_images_data 222 9.7 0.004 0.005 8.750 9.487 mp_sum_l 887 5.1 5.148 8.676 5.148 8.676 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.208 7.567 mp_irecv_dv 3311 11.0 3.188 7.517 3.188 7.517 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.869 6.962 calculate_norms 2376 9.8 6.058 6.716 6.058 6.716 multiply_cannon_sync_h2d 1332 9.7 4.835 6.231 4.835 6.231 arnoldi_extremal 4 6.8 0.000 0.000 6.002 6.020 arnoldi_normal_ev 4 7.8 0.001 0.005 6.002 6.020 build_subspace 16 8.4 0.014 0.020 5.712 5.719 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.246 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.660 3.850 ls_scf_post 1 4.0 0.000 0.000 3.804 3.808 ls_scf_store_result 1 5.0 0.000 0.000 3.475 3.624 dbcsr_matrix_vector_mult_local 304 10.0 2.727 3.214 2.729 3.215 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.378 3.146 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.885 2.988 mp_allgather_i34 111 8.7 0.742 2.857 0.742 2.857 dbcsr_data_new 4174 10.1 2.111 2.398 2.111 2.398 mp_sum_dv 2129 10.8 1.955 2.281 1.955 2.281 make_images_pack 222 9.7 1.813 2.130 1.815 2.132 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.070 2.073 rebuild_ks_matrix 3 7.3 0.000 0.000 2.057 2.060 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 2.057 2.060 dbcsr_sort_data 436 11.2 1.771 1.996 1.771 1.996 make_images_sizes 222 9.7 0.000 0.000 0.764 1.961 mp_alltoall_i44 222 10.7 0.764 1.961 0.764 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.526000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 2.869850E+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 0.075 0.137 96.387 96.392 qs_energies 1 2.0 0.001 0.004 95.272 95.301 ls_scf 1 3.0 0.000 0.001 93.844 93.869 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.091 76.477 ls_scf_main 1 4.0 0.000 0.006 58.243 58.247 multiply_cannon 111 7.7 0.035 0.080 52.460 56.992 multiply_cannon_loop 111 8.7 0.117 0.130 49.608 53.175 density_matrix_trs4 2 5.0 0.004 0.046 51.951 52.163 mp_waitall_1 7281 11.0 24.307 33.895 24.307 33.895 ls_scf_init_scf 1 4.0 0.000 0.003 31.865 31.868 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.393 30.470 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.961 27.975 multiply_cannon_multrec 888 9.7 12.679 15.272 21.254 24.416 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.185 22.977 make_m2s 222 7.7 0.006 0.007 16.965 17.756 make_images 222 8.7 1.581 1.852 16.927 17.718 make_images_data 222 9.7 0.004 0.004 9.900 11.279 hybrid_alltoall_any 227 10.6 0.641 2.957 9.240 10.883 dbcsr_mm_accdrv_process 3754 10.4 0.330 0.500 8.098 9.347 mp_sum_l 887 5.1 5.009 8.959 5.009 8.959 dbcsr_mm_accdrv_process_sort 3754 11.4 7.662 8.847 7.662 8.847 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.823 7.248 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.485 7.233 mp_irecv_dv 2335 11.1 2.470 7.184 2.470 7.184 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.772 6.996 multiply_cannon_sync_h2d 888 9.7 6.033 6.910 6.033 6.910 arnoldi_extremal 4 6.8 0.000 0.000 6.646 6.668 arnoldi_normal_ev 4 7.8 0.001 0.005 6.646 6.667 build_subspace 16 8.4 0.014 0.020 6.304 6.310 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.849 5.093 calculate_norms 1584 9.8 4.329 4.729 4.329 4.729 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.065 4.421 mp_allgather_i34 111 8.7 0.955 3.929 0.955 3.929 ls_scf_post 1 4.0 0.002 0.014 3.735 3.762 dbcsr_matrix_vector_mult_local 304 10.0 3.012 3.588 3.014 3.590 ls_scf_store_result 1 5.0 0.000 0.000 3.429 3.523 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.148 3.246 make_images_sizes 222 9.7 0.000 0.000 1.327 3.025 mp_alltoall_i44 222 10.7 1.326 3.025 1.326 3.025 mp_sum_dv 2129 10.8 2.296 2.567 2.296 2.567 dbcsr_data_new 4116 9.9 2.109 2.456 2.109 2.456 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.128 2.130 rebuild_ks_matrix 3 7.3 0.000 0.000 2.109 2.112 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.109 2.111 dbcsr_sort_data 325 11.1 1.801 2.064 1.801 2.064 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.392000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2230.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 3.383693E+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 0.085 0.100 104.696 104.697 qs_energies 1 2.0 0.000 0.000 103.787 103.792 ls_scf 1 3.0 0.000 0.000 102.097 102.102 dbcsr_multiply_generic 111 6.7 0.017 0.018 81.464 81.780 ls_scf_main 1 4.0 0.000 0.000 63.152 63.154 multiply_cannon 111 7.7 0.039 0.070 53.208 57.567 density_matrix_trs4 2 5.0 0.002 0.003 56.283 56.448 multiply_cannon_loop 111 8.7 0.153 0.168 48.068 52.283 ls_scf_init_scf 1 4.0 0.000 0.000 35.509 35.511 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.041 34.123 mp_waitall_1 6369 11.0 25.126 32.108 25.126 32.108 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.499 31.513 multiply_cannon_multrec 1332 9.7 14.006 17.159 21.906 25.075 make_m2s 222 7.7 0.007 0.008 21.985 23.416 make_images 222 8.7 3.141 3.592 21.934 23.368 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 10.400 18.800 make_images_data 222 9.7 0.004 0.004 12.480 14.075 hybrid_alltoall_any 227 10.6 0.797 3.755 11.774 13.490 dbcsr_mm_accdrv_process 3641 10.4 0.289 0.498 7.542 9.033 dbcsr_mm_accdrv_process_sort 3641 11.4 7.099 8.540 7.099 8.540 arnoldi_extremal 4 6.8 0.000 0.000 8.435 8.450 arnoldi_normal_ev 4 7.8 0.001 0.004 8.434 8.450 mp_sum_l 887 5.1 4.717 8.132 4.717 8.132 build_subspace 16 8.4 0.013 0.020 8.041 8.052 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.529 6.370 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.073 6.123 mp_irecv_dv 3229 10.9 2.050 6.042 2.050 6.042 multiply_cannon_sync_h2d 1332 9.7 5.367 6.039 5.367 6.039 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 3.039 5.703 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.871 5.534 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.921 5.122 mp_allgather_i34 111 8.7 2.212 4.500 2.212 4.500 calculate_norms 2376 9.8 4.161 4.475 4.161 4.475 mp_sum_dv 2129 10.8 3.661 4.136 3.661 4.136 dbcsr_matrix_vector_mult_local 304 10.0 3.192 3.694 3.193 3.695 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.398 3.441 ls_scf_post 1 4.0 0.000 0.000 3.436 3.441 dbcsr_sort_data 658 11.4 2.987 3.341 2.987 3.341 ls_scf_store_result 1 5.0 0.000 0.000 3.154 3.223 dbcsr_special_finalize 555 9.7 0.006 0.007 2.710 3.129 dbcsr_merge_single_wm 555 10.7 0.527 0.662 2.702 3.121 dbcsr_data_release 10477 10.7 1.594 2.400 1.594 2.400 dbcsr_sym_m_v_mult 184 10.2 0.006 0.011 2.097 2.255 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.220 2.222 rebuild_ks_matrix 3 7.3 0.000 0.000 2.193 2.194 qs_ks_build_kohn_sham_matrix 3 8.3 0.026 0.033 2.193 2.194 make_images_sizes 222 9.7 0.000 0.000 0.955 2.130 mp_alltoall_i44 222 10.7 0.955 2.130 0.955 2.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=104.697000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 4.758974E+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 0.059 0.091 98.499 98.499 qs_energies 1 2.0 0.000 0.000 97.625 97.629 ls_scf 1 3.0 0.000 0.000 95.631 95.635 dbcsr_multiply_generic 111 6.7 0.017 0.018 74.257 74.487 ls_scf_main 1 4.0 0.000 0.000 60.353 60.353 multiply_cannon 111 7.7 0.084 0.224 54.122 58.004 multiply_cannon_loop 111 8.7 0.089 0.102 51.424 52.772 density_matrix_trs4 2 5.0 0.002 0.003 52.506 52.569 mp_waitall_1 5436 11.0 26.946 31.912 26.946 31.912 ls_scf_init_scf 1 4.0 0.000 0.000 31.634 31.637 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.175 30.215 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.981 27.988 multiply_cannon_multrec 444 9.7 13.557 16.369 20.662 22.270 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 13.519 17.445 make_m2s 222 7.7 0.005 0.005 15.198 16.103 make_images 222 8.7 2.039 2.475 15.129 16.035 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.028 11.177 make_images_data 222 9.7 0.003 0.004 9.173 11.124 hybrid_alltoall_any 227 10.6 0.799 3.815 8.893 10.871 multiply_cannon_sync_h2d 444 9.7 6.598 7.983 6.598 7.983 dbcsr_mm_accdrv_process 3003 10.4 0.347 0.448 6.798 7.963 dbcsr_mm_accdrv_process_sort 3003 11.4 6.422 7.558 6.422 7.558 arnoldi_extremal 4 6.8 0.000 0.000 7.472 7.482 arnoldi_normal_ev 4 7.8 0.001 0.004 7.472 7.482 build_subspace 16 8.4 0.014 0.019 7.061 7.068 mp_sum_l 887 5.1 2.977 5.626 2.977 5.626 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.785 4.963 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.789 4.913 make_images_sizes 222 9.7 0.000 0.000 1.641 4.893 mp_alltoall_i44 222 10.7 1.640 4.892 1.640 4.892 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.091 4.478 dbcsr_matrix_vector_mult_local 304 10.0 3.717 4.190 3.719 4.193 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.042 4.100 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.659 4.026 mp_irecv_dv 1241 11.2 1.642 3.987 1.642 3.987 mp_allgather_i34 111 8.7 1.260 3.837 1.260 3.837 calculate_norms 792 9.8 3.585 3.691 3.585 3.691 ls_scf_post 1 4.0 0.000 0.000 3.644 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.454 mp_sum_dv 2129 10.8 2.288 2.418 2.288 2.418 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.324 2.325 rebuild_ks_matrix 3 7.3 0.000 0.000 2.289 2.289 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 2.288 2.289 dbcsr_finalize 304 7.8 0.062 0.077 2.191 2.272 dbcsr_data_new 4608 9.7 1.790 2.239 1.790 2.239 dbcsr_merge_all 275 8.9 0.478 0.517 2.049 2.113 mp_alltoall_d11v 48 9.2 2.003 2.079 2.003 2.079 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 1.977 1.977 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.499000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3756.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 8.813568E+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 0.103 0.131 117.748 117.749 qs_energies 1 2.0 0.000 0.000 116.138 116.141 ls_scf 1 3.0 0.000 0.000 113.129 113.132 dbcsr_multiply_generic 111 6.7 0.023 0.027 82.946 83.251 ls_scf_main 1 4.0 0.000 0.000 71.268 71.268 density_matrix_trs4 2 5.0 0.002 0.003 60.950 61.118 multiply_cannon 111 7.7 0.145 0.201 51.355 53.290 multiply_cannon_loop 111 8.7 0.099 0.103 48.023 51.368 ls_scf_init_scf 1 4.0 0.000 0.000 37.677 37.678 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.836 35.893 mp_waitall_1 4527 11.1 26.213 33.448 26.213 33.448 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.955 32.963 make_m2s 222 7.7 0.005 0.005 25.417 26.818 make_images 222 8.7 3.576 3.877 25.309 26.710 multiply_cannon_multrec 444 9.7 17.861 18.608 22.498 23.274 hybrid_alltoall_any 227 10.6 1.655 3.614 14.071 18.550 make_images_data 222 9.7 0.003 0.004 14.798 18.240 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 12.786 15.972 arnoldi_extremal 4 6.8 0.000 0.000 9.972 9.983 arnoldi_normal_ev 4 7.8 0.002 0.008 9.971 9.983 build_subspace 16 8.4 0.024 0.034 9.378 9.388 multiply_cannon_sync_h2d 444 9.7 8.792 8.851 8.792 8.851 dbcsr_matrix_vector_mult 304 9.0 0.015 0.030 6.655 7.142 mp_sum_l 887 5.1 3.833 6.607 3.833 6.607 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.675 6.034 make_images_sizes 222 9.7 0.000 0.000 2.490 5.887 mp_alltoall_i44 222 10.7 2.490 5.887 2.490 5.887 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.723 5.831 dbcsr_matrix_vector_mult_local 304 10.0 5.119 5.437 5.121 5.439 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.606 4.643 dbcsr_mm_accdrv_process 1814 10.4 0.293 0.358 4.459 4.579 mp_allgather_i34 111 8.7 1.433 4.322 1.433 4.322 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.249 4.120 4.249 ls_scf_post 1 4.0 0.000 0.000 4.184 4.187 ls_scf_store_result 1 5.0 0.000 0.000 3.844 3.913 calculate_norms 792 9.8 3.231 3.279 3.231 3.279 dbcsr_complete_redistribute 5 7.6 1.427 1.464 3.063 3.203 mp_sum_dv 2129 10.8 2.858 3.199 2.858 3.199 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.179 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.978 2.978 dbcsr_merge_all 275 8.9 0.885 0.920 2.861 2.955 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.775 2.777 rebuild_ks_matrix 3 7.3 0.000 0.000 2.709 2.711 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.008 2.709 2.711 matrix_ls_to_qs 2 6.0 0.000 0.000 2.568 2.696 dbcsr_sort_data 325 11.1 2.441 2.504 2.441 2.504 mp_alltoall_d11v 48 9.2 2.340 2.413 2.340 2.413 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=117.749000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7017.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3c55af5c093d10a5e0ad2f8e06538f92dff62763_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 593.252352E+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 4808541. 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 0.086 0.107 556.634 556.637 qs_mol_dyn_low 1 2.0 0.004 0.034 554.801 554.898 qs_forces 5 3.8 0.006 0.057 554.421 554.427 qs_energies 5 4.8 0.004 0.060 548.848 548.946 scf_env_do_scf 5 5.8 0.000 0.006 527.184 527.187 scf_env_do_scf_inner_loop 105 6.6 0.004 0.049 444.395 444.401 qs_scf_new_mos 105 7.6 0.000 0.001 348.673 350.067 qs_scf_loop_do_ot 105 8.6 0.001 0.001 348.673 350.067 ot_scf_mini 105 9.6 0.006 0.165 334.045 335.028 qs_ot_get_p 112 10.4 0.001 0.001 206.695 207.702 velocity_verlet 4 3.0 0.004 0.049 203.235 203.243 dbcsr_multiply_generic 1445 12.2 0.127 0.133 183.234 191.156 qs_ot_p2m_diag 40 11.0 0.020 0.030 183.978 184.330 cp_dbcsr_syevd 40 12.0 0.002 0.002 178.683 178.698 cp_fm_syevd 40 13.0 0.000 0.000 168.247 168.457 multiply_cannon 1445 13.2 0.283 0.309 141.403 163.972 cp_fm_redistribute_end 40 14.0 79.784 159.196 79.816 159.218 cp_fm_syevd_base 40 14.0 79.320 158.762 79.320 158.762 multiply_cannon_loop 1445 14.2 2.873 3.028 135.089 157.758 mp_waitall_1 488190 16.1 80.405 114.997 80.405 114.997 multiply_cannon_metrocomm3 69360 15.2 0.198 0.208 45.057 92.072 ot_mini 105 10.6 0.001 0.016 87.376 87.856 init_scf_loop 7 6.6 0.000 0.009 82.635 82.640 prepare_preconditioner 7 7.6 0.000 0.000 74.003 74.083 make_preconditioner 7 8.6 0.000 0.018 74.003 74.083 qs_ot_get_derivative 55 11.6 0.001 0.022 64.107 64.449 multiply_cannon_metrocomm1 69360 15.2 0.094 0.103 14.242 63.614 make_full_inverse_cholesky 7 9.6 0.000 0.000 59.447 59.565 qs_rho_update_rho_low 110 7.6 0.001 0.008 54.432 55.289 calculate_rho_elec 110 8.6 0.029 0.033 54.432 55.288 rebuild_ks_matrix 110 8.4 0.000 0.000 52.015 53.531 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.034 52.015 53.531 qs_ks_update_qs_env 112 7.6 0.001 0.001 47.790 49.171 density_rs2pw 110 9.6 0.005 0.006 44.721 45.407 multiply_cannon_multrec 69360 15.2 26.231 33.362 36.567 43.246 cp_fm_cholesky_invert 7 10.6 40.490 40.561 40.490 40.561 transfer_rs2pw 445 10.6 0.007 0.008 37.990 38.674 mp_sum_l 4764 12.2 23.808 33.736 23.808 33.736 mp_sendrecv_dv 168740 12.6 31.182 31.598 31.182 31.598 multiply_cannon_sync_h2d 69360 15.2 24.256 30.895 24.256 30.895 transfer_rs2pw_30 110 11.6 1.246 1.310 26.952 30.707 sum_up_and_integrate 60 10.3 0.001 0.002 26.422 26.445 integrate_v_rspace 60 11.3 0.002 0.002 26.405 26.428 qs_ot_get_orbitals 105 10.6 0.001 0.001 22.999 24.357 apply_preconditioner_dbcsr 62 12.6 0.000 0.009 23.714 24.040 apply_single 62 13.6 0.000 0.000 23.714 24.040 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 23.740 24.011 ot_new_cg_direction 55 11.6 0.001 0.005 21.844 21.849 pw_transfer 1645 12.4 0.078 0.099 20.149 21.003 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 20.011 20.869 potential_pw2rs 60 12.3 0.003 0.003 20.411 20.623 make_m2s 2890 13.2 0.077 0.083 17.729 18.906 make_images 2890 14.2 0.240 0.259 17.623 18.803 cp_fm_cholesky_decompose 14 10.2 18.364 18.377 18.364 18.377 transfer_pw2rs 245 13.2 0.003 0.004 18.064 18.288 init_scf_run 5 5.8 0.000 0.001 17.644 17.646 scf_env_initial_rho_setup 5 6.8 0.000 0.012 17.644 17.646 check_diag 80 13.5 8.594 8.891 17.103 17.335 calculate_dm_sparse 110 9.5 0.000 0.001 16.509 17.237 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 16.763 16.927 fft_wrap_pw1pw2_240 915 15.0 1.096 1.211 15.620 16.433 fft3d_pb 915 16.0 2.331 2.546 13.906 14.906 qs_vxc_create 110 10.4 0.003 0.015 14.029 14.130 transfer_pw2rs_30 60 14.3 0.278 0.309 12.308 14.058 make_full_single_inverse 7 9.6 0.001 0.002 13.684 13.763 mp_alltoall_z22v 2340 17.7 11.480 13.743 11.480 13.743 mp_alltoall_d11v 1300 13.8 12.158 13.646 12.158 13.646 make_images_sizes 2890 15.2 0.004 0.004 10.820 12.758 mp_alltoall_i44 2890 16.2 10.816 12.753 10.816 12.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=556.637000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 3c55af5c093d10a5e0ad2f8e06538f92dff62763 Summary: empty Status: OK