=== 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: aa921764592e607de23a23fbfafae42bb144baa4 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/01 job id: 48144589 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/02 job id: 48144591 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/03 job id: 48144593 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/04 job id: 48144595 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/05 job id: 48144597 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/06 job id: 48144600 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/07 job id: 48144601 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/08 job id: 48144603 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/09 job id: 48144605 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/10 job id: 48144607 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/11 job id: 48144609 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/12 job id: 48144611 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/13 job id: 48144614 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/14 job id: 48144616 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/15 job id: 48144618 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/16 job id: 48144620 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/17 job id: 48144621 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/18 job id: 48144623 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/19 job id: 48144625 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/20 job id: 48144627 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/21 job id: 48144629 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/22 job id: 48144630 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/23 job id: 48144631 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/24 job id: 48144632 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/25 job id: 48144633 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/26 job id: 48144634 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/27 job id: 48144635 --- 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/aa921764592e607de23a23fbfafae42bb144baa4_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 0.043 0.047 135.217 135.218 farming_run 1 2.0 134.572 134.574 135.167 135.171 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.494094E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 115.030 115.030 qs_energies 1 2.0 0.000 0.000 114.814 114.817 mp2_main 1 3.0 0.000 0.000 112.730 112.733 mp2_gpw_main 1 4.0 0.029 0.036 111.741 111.744 mp2_ri_gpw_compute_in 1 5.0 0.179 0.181 92.650 93.011 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.936 55.298 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.166 41.317 46.363 get_2c_integrals 1 6.0 0.008 0.009 37.069 37.534 integrate_v_rspace 273 8.0 0.435 0.451 24.733 29.656 pw_transfer 6555 10.6 0.377 0.404 27.412 28.056 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.020 26.566 grid_integrate_task_list 273 9.0 20.551 25.957 20.551 25.957 fft_wrap_pw1pw2_100 2178 12.4 1.196 1.443 23.544 24.090 compute_2c_integrals 1 7.0 0.002 0.003 19.712 19.712 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.922 19.442 mp2_eri_2c_integrate_gpw 1 9.0 2.347 2.411 18.919 19.439 rpa_ri_compute_en 1 5.0 0.020 0.022 18.971 19.200 cp_fm_cholesky_decompose 12 8.2 17.279 17.753 17.279 17.753 cholesky_decomp 1 7.0 0.000 0.000 16.195 16.662 fft3d_s 5443 13.4 16.157 16.561 16.178 16.582 ao_to_mo_and_store_B_mult_1 272 7.0 10.780 15.323 10.780 15.323 calculate_wavefunction 272 8.0 5.393 5.561 12.489 13.107 rpa_num_int 1 6.0 0.000 0.003 10.635 10.645 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.601 10.611 calc_mat_Q 8 8.0 0.000 0.000 9.470 9.552 contract_S_to_Q 8 9.0 0.000 0.000 8.891 8.973 calc_potential_gpw 544 9.5 0.005 0.006 8.271 8.724 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.245 8.580 parallel_gemm_fm 14 9.1 0.000 0.000 8.483 8.573 parallel_gemm_fm_cosma 14 10.1 8.483 8.573 8.483 8.573 potential_pw2rs 545 10.0 0.107 0.109 7.715 8.379 create_integ_mat 1 6.0 0.013 0.027 7.882 7.892 collocate_single_gaussian 272 10.0 0.039 0.042 7.509 7.788 array2fm 1 7.0 0.000 0.000 6.849 7.338 pw_scatter_s 2720 13.7 4.404 4.626 4.404 4.626 pw_gather_s 2722 13.2 3.860 4.202 3.860 4.202 array2fm_buffer_send 1 8.0 2.918 3.145 2.918 3.145 pw_poisson_solve 545 10.5 1.101 1.142 2.200 2.408 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.743870, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2811.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 0.029 0.038 424.088 424.090 farming_run 1 2.0 422.875 422.883 424.043 424.048 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.245450E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 218.321 218.322 qs_energies 1 2.0 0.004 0.031 218.078 218.084 scf_env_do_scf 1 3.0 0.000 0.000 115.394 115.395 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.471 114.478 rebuild_ks_matrix 4 6.0 0.000 0.000 114.470 114.477 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.470 114.477 hfx_ks_matrix 4 8.0 0.001 0.001 114.093 114.098 integrate_four_center 4 9.0 0.152 0.465 114.092 114.097 integrate_four_center_main 4 10.0 0.103 0.648 101.953 105.886 integrate_four_center_bin 268 11.0 101.849 105.718 101.849 105.718 mp2_main 1 3.0 0.003 0.024 102.384 102.394 mp2_gpw_main 1 4.0 0.055 0.128 101.502 101.511 init_scf_loop 1 4.0 0.000 0.000 96.712 96.713 mp2_ri_gpw_compute_in 1 5.0 0.072 0.131 74.643 75.807 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.124 55.279 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.167 41.805 46.885 integrate_v_rspace 95 8.0 0.393 0.564 28.173 33.048 pw_transfer 2240 10.6 0.146 0.191 29.922 30.344 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.937 29.384 ao_to_mo_and_store_B_mult_1 91 7.0 10.629 28.570 10.629 28.570 grid_integrate_task_list 95 9.0 23.468 28.549 23.468 28.549 mp2_ri_gpw_compute_en 1 5.0 0.057 0.101 26.668 28.431 fft_wrap_pw1pw2_100 730 12.4 1.271 1.517 26.651 27.147 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.903 24.795 24.805 get_2c_integrals 1 6.0 0.002 0.017 20.424 20.463 compute_2c_integrals 1 7.0 0.011 0.047 19.402 19.420 compute_2c_integrals_loop_lm 1 8.0 0.003 0.016 18.833 19.270 mp2_eri_2c_integrate_gpw 1 9.0 1.726 1.888 18.830 19.269 fft3d_s 1823 13.4 18.419 18.811 18.433 18.824 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.680 18.680 calculate_wavefunction 91 8.0 2.004 2.029 9.736 9.979 potential_pw2rs 186 10.0 0.033 0.036 8.629 9.218 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.591 8.742 9.212 local_gemm 172 8.0 8.185 8.650 8.185 8.650 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.237 8.574 calc_potential_gpw 182 9.5 0.002 0.003 7.919 8.224 collocate_single_gaussian 91 10.0 0.018 0.053 7.868 8.117 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.518 7.819 8.079 mp_sync 37 10.5 4.103 7.220 4.103 7.220 integrate_four_center_load 4 10.0 0.000 0.000 6.768 6.774 hfx_load_balance 1 11.0 0.000 0.000 6.768 6.773 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.407 6.348 6.407 mp_sendrecv_dm3 2068 8.0 5.850 6.110 5.850 6.110 pw_gather_s 912 13.2 4.914 5.344 4.914 5.344 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.504918, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1506.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 451.997696E+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 525137. 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 0.023 0.067 55.005 55.010 qs_mol_dyn_low 1 2.0 0.004 0.008 54.636 54.643 qs_forces 11 3.9 0.004 0.015 54.571 54.572 qs_energies 11 4.9 0.002 0.007 53.039 53.069 scf_env_do_scf 11 5.9 0.000 0.001 46.496 46.497 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 44.423 44.424 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.157 34.600 qs_scf_new_mos 108 7.5 0.000 0.001 34.248 34.532 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.248 34.531 ot_scf_mini 108 9.5 0.002 0.002 32.558 32.785 multiply_cannon 2286 13.5 0.191 0.203 26.849 28.314 multiply_cannon_loop 2286 14.5 1.824 1.925 26.191 27.678 velocity_verlet 10 3.0 0.001 0.002 26.485 26.486 ot_mini 108 10.5 0.001 0.001 19.698 19.943 qs_ot_get_derivative 108 11.5 0.001 0.001 16.691 16.876 mp_waitall_1 245248 16.5 8.496 14.911 8.496 14.911 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.152 13.160 multiply_cannon_multrec 54864 15.5 3.645 5.736 7.955 11.573 qs_ot_get_p 119 10.4 0.001 0.001 8.174 8.479 rebuild_ks_matrix 119 8.3 0.000 0.000 7.905 8.069 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.905 8.068 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.947 7.101 mp_sum_l 7287 12.8 5.448 7.101 5.448 7.101 multiply_cannon_sync_h2d 54864 15.5 5.159 6.581 5.159 6.581 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.710 6.136 dbcsr_mm_accdrv_process 76910 16.1 1.834 2.870 4.223 6.067 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.328 5.387 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.071 5.194 init_scf_run 11 5.9 0.000 0.001 5.161 5.161 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.161 5.161 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.441 4.559 calculate_rho_elec 119 8.7 0.033 0.189 4.440 4.559 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.500 4.500 sum_up_and_integrate 119 10.3 0.012 0.014 4.477 4.485 integrate_v_rspace 119 11.3 0.002 0.003 4.465 4.474 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.302 4.302 cp_fm_redistribute_end 50 14.0 2.195 4.278 2.202 4.281 cp_fm_diag_elpa_base 50 14.0 2.074 4.171 2.078 4.180 calculate_dm_sparse 119 9.5 0.000 0.001 3.234 3.422 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.950 3.151 apply_single 119 13.6 0.000 0.000 2.950 3.151 jit_kernel_multiply 13 15.8 2.328 3.126 2.328 3.126 acc_transpose_blocks 54864 15.5 0.228 0.249 2.469 2.973 rs_pw_transfer 974 11.9 0.012 0.013 2.756 2.857 calculate_first_density_matrix 1 7.0 0.000 0.002 2.745 2.763 ot_diis_step 108 11.5 0.006 0.006 2.745 2.745 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.711 2.707 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.498 2.559 density_rs2pw 119 9.7 0.004 0.005 2.183 2.306 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.193 2.195 wfi_extrapolate 11 7.9 0.001 0.001 2.172 2.172 grid_integrate_task_list 119 12.3 2.001 2.107 2.001 2.107 init_scf_loop 11 6.9 0.001 0.005 2.057 2.057 mp_sum_d 4135 12.0 1.405 1.977 1.405 1.977 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.887 1.948 potential_pw2rs 119 12.3 0.004 0.004 1.841 1.852 pw_transfer 1439 11.6 0.052 0.057 1.706 1.777 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.716 1.761 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.630 1.703 make_m2s 4572 13.5 0.053 0.055 1.589 1.631 mp_alltoall_d11v 2130 13.8 1.450 1.616 1.450 1.616 make_images 4572 14.5 0.133 0.138 1.508 1.549 mp_waitany 12084 13.8 1.282 1.453 1.282 1.453 acc_transpose_blocks_sync 164592 16.5 1.199 1.429 1.199 1.429 grid_collocate_task_list 119 9.7 1.359 1.426 1.359 1.426 fft3d_ps 1201 14.6 0.368 0.472 1.269 1.334 fft_wrap_pw1pw2_140 487 13.2 0.188 0.210 1.260 1.333 acc_transpose_blocks_kernels 54864 16.5 0.248 0.382 0.995 1.265 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.761 1.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.010000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.909091, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 488.927232E+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 1085756. 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 0.043 0.088 40.649 40.651 qs_mol_dyn_low 1 2.0 0.003 0.006 39.768 39.917 qs_forces 11 3.9 0.002 0.002 39.700 39.704 qs_energies 11 4.9 0.001 0.001 37.983 37.990 scf_env_do_scf 11 5.9 0.000 0.001 32.398 32.398 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.917 29.918 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.614 22.970 qs_scf_new_mos 108 7.5 0.001 0.001 20.749 20.980 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.749 20.979 ot_scf_mini 108 9.5 0.002 0.003 19.829 19.994 multiply_cannon 2286 13.5 0.209 0.219 17.584 19.072 velocity_verlet 10 3.0 0.001 0.001 18.588 18.589 multiply_cannon_loop 2286 14.5 1.189 1.244 16.420 17.855 ot_mini 108 10.5 0.001 0.001 12.263 12.495 mp_waitall_1 200699 16.5 5.852 11.134 5.852 11.134 qs_ot_get_derivative 108 11.5 0.001 0.001 9.905 10.071 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.341 9.812 multiply_cannon_multrec 27432 15.5 1.832 4.284 6.449 9.597 rebuild_ks_matrix 119 8.3 0.000 0.000 7.285 7.420 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.285 7.420 dbcsr_mm_accdrv_process 47894 16.0 3.707 6.017 4.536 6.890 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.421 6.543 qs_ot_get_p 119 10.4 0.001 0.001 4.652 4.873 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.874 4.707 init_scf_run 11 5.9 0.000 0.001 4.391 4.391 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.391 4.391 sum_up_and_integrate 119 10.3 0.025 0.028 4.192 4.197 integrate_v_rspace 119 11.3 0.002 0.002 4.167 4.175 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.985 4.172 apply_single 119 13.6 0.000 0.000 2.985 4.172 mp_sum_l 7287 12.8 2.000 4.072 2.000 4.072 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.775 3.812 calculate_rho_elec 119 8.7 0.021 0.024 3.775 3.812 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.021 3.039 rs_pw_transfer 974 11.9 0.010 0.012 2.515 2.962 make_m2s 4572 13.5 0.052 0.054 2.593 2.936 calculate_first_density_matrix 1 7.0 0.000 0.000 2.875 2.883 make_images 4572 14.5 0.206 0.246 2.506 2.850 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.624 2.624 jit_kernel_multiply 10 16.4 0.772 2.549 0.772 2.549 density_rs2pw 119 9.7 0.004 0.004 2.033 2.502 init_scf_loop 11 6.9 0.000 0.000 2.461 2.462 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.330 2.416 calculate_dm_sparse 119 9.5 0.000 0.001 2.316 2.395 ot_diis_step 108 11.5 0.011 0.011 2.312 2.312 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.307 2.307 cp_fm_redistribute_end 50 14.0 1.168 2.279 1.172 2.282 cp_fm_diag_elpa_base 50 14.0 1.077 2.191 1.106 2.226 multiply_cannon_sync_h2d 27432 15.5 1.686 2.206 1.686 2.206 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.147 2.149 acc_transpose_blocks 27432 15.5 0.114 0.120 1.726 2.133 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.948 1.988 pw_transfer 1439 11.6 0.066 0.070 1.926 1.955 grid_integrate_task_list 119 12.3 1.839 1.932 1.839 1.932 potential_pw2rs 119 12.3 0.006 0.006 1.914 1.922 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.833 1.864 make_images_data 4572 15.5 0.047 0.054 1.232 1.578 prepare_preconditioner 11 7.9 0.000 0.000 1.518 1.545 make_preconditioner 11 8.9 0.000 0.000 1.518 1.545 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.504 1.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.420 1.476 fft_wrap_pw1pw2_140 487 13.2 0.204 0.213 1.440 1.471 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.088 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.450 1.450 grid_collocate_task_list 119 9.7 1.282 1.419 1.282 1.419 fft3d_ps 1201 14.6 0.519 0.574 1.383 1.408 mp_alltoall_d11v 2130 13.8 1.172 1.311 1.172 1.311 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.188 1.233 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.768 1.047 mp_sum_d 4135 12.0 0.541 1.016 0.541 1.016 mp_waitany 5720 13.7 0.517 1.003 0.517 1.003 mp_allgather_i34 2286 14.5 0.584 0.996 0.584 0.996 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.545 0.989 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.937 0.941 acc_transpose_blocks_sync 82296 16.5 0.817 0.941 0.817 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.920 0.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.651000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=2.135880 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 519.659520E+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 1607886. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.026 33.390 33.390 qs_mol_dyn_low 1 2.0 0.003 0.003 33.175 33.183 qs_forces 11 3.9 0.002 0.002 33.115 33.116 qs_energies 11 4.9 0.001 0.001 31.505 31.507 scf_env_do_scf 11 5.9 0.000 0.001 26.855 26.855 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.222 24.223 dbcsr_multiply_generic 2286 12.5 0.095 0.097 17.379 17.503 qs_scf_new_mos 108 7.5 0.001 0.001 15.956 15.974 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.956 15.974 velocity_verlet 10 3.0 0.001 0.002 15.619 15.621 ot_scf_mini 108 9.5 0.002 0.002 15.192 15.206 multiply_cannon 2286 13.5 0.193 0.198 13.972 14.860 multiply_cannon_loop 2286 14.5 0.857 0.900 13.206 14.113 ot_mini 108 10.5 0.001 0.001 9.496 9.520 qs_ot_get_derivative 108 11.5 0.001 0.001 7.953 7.968 multiply_cannon_multrec 18288 15.5 2.021 3.113 7.320 7.575 rebuild_ks_matrix 119 8.3 0.000 0.000 6.597 6.622 qs_ks_build_kohn_sham_matrix 119 9.3 0.022 0.034 6.597 6.622 dbcsr_mm_accdrv_process 38222 16.0 5.095 6.211 5.206 6.273 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.843 5.866 sum_up_and_integrate 119 10.3 0.032 0.032 4.010 4.032 integrate_v_rspace 119 11.3 0.002 0.003 3.978 4.003 mp_waitall_1 158411 16.6 2.615 3.741 2.615 3.741 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.900 3.536 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.488 3.498 calculate_rho_elec 119 8.7 0.031 0.032 3.488 3.497 init_scf_run 11 5.9 0.000 0.001 3.480 3.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.480 3.480 qs_ot_get_p 119 10.4 0.001 0.001 3.444 3.470 init_scf_loop 11 6.9 0.000 0.000 2.616 2.616 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.504 2.522 rs_pw_transfer 974 11.9 0.009 0.010 2.254 2.517 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.074 2.456 apply_single 119 13.6 0.000 0.000 2.074 2.456 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.287 2.297 calculate_first_density_matrix 1 7.0 0.000 0.000 2.236 2.237 density_rs2pw 119 9.7 0.004 0.004 1.962 2.231 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.972 1.972 pw_transfer 1439 11.6 0.065 0.069 1.935 1.962 make_m2s 4572 13.5 0.044 0.045 1.826 1.955 grid_integrate_task_list 119 12.3 1.807 1.894 1.807 1.894 calculate_dm_sparse 119 9.5 0.000 0.000 1.863 1.874 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.842 1.873 make_images 4572 14.5 0.193 0.205 1.742 1.868 potential_pw2rs 119 12.3 0.007 0.009 1.760 1.785 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.745 1.753 acc_transpose_blocks 18288 15.5 0.078 0.080 1.625 1.745 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.718 1.729 cp_fm_diag_elpa_base 50 14.0 1.695 1.711 1.716 1.727 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.715 1.717 mp_sum_l 7287 12.8 1.193 1.699 1.193 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.645 1.657 make_preconditioner 11 8.9 0.000 0.000 1.645 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.502 1.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.532 1.544 ot_diis_step 108 11.5 0.011 0.011 1.525 1.526 fft_wrap_pw1pw2_140 487 13.2 0.253 0.260 1.480 1.510 grid_collocate_task_list 119 9.7 1.247 1.405 1.247 1.405 fft3d_ps 1201 14.6 0.528 0.547 1.331 1.360 multiply_cannon_sync_h2d 18288 15.5 1.023 1.198 1.023 1.198 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.194 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.191 1.191 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.975 0.994 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.941 0.942 make_images_data 4572 15.5 0.047 0.051 0.784 0.936 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.672 0.871 acc_transpose_blocks_sync 54864 16.5 0.741 0.862 0.741 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.829 mp_alltoall_d11v 2130 13.8 0.662 0.799 0.662 0.799 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.785 0.794 mp_waitany 9880 13.7 0.525 0.789 0.525 0.789 jit_kernel_multiply 4 15.2 0.054 0.785 0.054 0.785 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.133 0.518 0.777 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.763 mp_alltoall_z22v 1201 16.6 0.657 0.729 0.657 0.729 cp_fm_cholesky_invert 11 10.9 0.690 0.694 0.690 0.694 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.390000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.434560E+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 0.016 0.034 37.223 37.223 qs_mol_dyn_low 1 2.0 0.003 0.003 37.028 37.036 qs_forces 11 3.9 0.002 0.002 36.969 36.970 qs_energies 11 4.9 0.002 0.005 35.234 35.241 scf_env_do_scf 11 5.9 0.001 0.002 29.455 29.456 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.037 26.037 dbcsr_multiply_generic 2286 12.5 0.098 0.105 19.995 20.103 velocity_verlet 10 3.0 0.002 0.002 18.496 18.498 qs_scf_new_mos 108 7.5 0.001 0.001 17.599 17.653 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.598 17.652 multiply_cannon 2286 13.5 0.219 0.250 16.460 17.296 ot_scf_mini 108 9.5 0.002 0.003 16.577 16.630 multiply_cannon_loop 2286 14.5 1.524 1.601 15.526 16.194 ot_mini 108 10.5 0.001 0.001 10.200 10.265 multiply_cannon_multrec 27432 15.5 2.455 3.104 9.274 9.615 qs_ot_get_derivative 108 11.5 0.001 0.001 8.346 8.401 dbcsr_mm_accdrv_process 47916 15.9 6.174 7.995 6.715 8.104 rebuild_ks_matrix 119 8.3 0.000 0.000 6.594 6.645 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.016 6.594 6.644 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.841 5.884 init_scf_run 11 5.9 0.000 0.001 3.995 3.995 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.995 3.995 sum_up_and_integrate 119 10.3 0.036 0.038 3.770 3.777 integrate_v_rspace 119 11.3 0.003 0.003 3.734 3.742 qs_ot_get_p 119 10.4 0.001 0.001 3.546 3.620 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.518 3.547 calculate_rho_elec 119 8.7 0.040 0.046 3.517 3.547 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.170 3.501 init_scf_loop 11 6.9 0.001 0.002 3.399 3.399 calculate_first_density_matrix 1 7.0 0.000 0.002 2.604 2.607 rs_pw_transfer 974 11.9 0.009 0.009 2.326 2.581 acc_transpose_blocks 27432 15.5 0.117 0.121 2.386 2.565 prepare_preconditioner 11 7.9 0.000 0.000 2.546 2.556 make_preconditioner 11 8.9 0.008 0.065 2.546 2.556 calculate_dm_sparse 119 9.5 0.000 0.000 2.468 2.522 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.150 2.481 mp_waitall_1 137007 16.6 1.817 2.390 1.817 2.390 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.309 2.337 make_m2s 4572 13.5 0.054 0.056 2.209 2.337 make_images 4572 14.5 0.272 0.334 2.102 2.228 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.027 2.212 apply_single 119 13.6 0.000 0.000 2.027 2.212 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.171 2.183 density_rs2pw 119 9.7 0.004 0.004 1.904 2.093 pw_transfer 1439 11.6 0.065 0.070 1.975 2.007 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.882 1.919 grid_integrate_task_list 119 12.3 1.811 1.897 1.811 1.897 jit_kernel_multiply 9 15.9 0.476 1.843 0.476 1.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.841 1.842 ot_diis_step 108 11.5 0.012 0.012 1.814 1.814 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.772 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.660 1.673 fft_wrap_pw1pw2_140 487 13.2 0.289 0.305 1.562 1.600 acc_transpose_blocks_sync 82296 16.5 1.417 1.585 1.417 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.573 1.584 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.539 1.549 cp_fm_diag_elpa_base 50 14.0 1.506 1.524 1.537 1.547 potential_pw2rs 119 12.3 0.009 0.010 1.521 1.525 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.523 1.524 grid_collocate_task_list 119 9.7 1.250 1.379 1.250 1.379 fft3d_ps 1201 14.6 0.555 0.604 1.318 1.342 wfi_extrapolate 11 7.9 0.001 0.001 1.335 1.335 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.859 1.330 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.274 1.288 mp_sum_l 7287 12.8 0.906 1.277 0.906 1.277 cp_fm_upper_to_full 72 14.2 0.805 1.146 0.805 1.146 rs_pw_transfer_RS2PW_140 130 11.5 0.108 0.111 0.853 1.098 mp_waitany 7280 13.7 0.843 1.082 0.843 1.082 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.013 1.045 dbcsr_complete_redistribute 329 12.2 0.127 0.165 0.765 1.034 make_images_data 4572 15.5 0.047 0.051 0.814 0.929 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.706 0.885 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.873 mp_alltoall_d11v 2130 13.8 0.717 0.860 0.717 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.848 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.822 0.831 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.565 0.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.223000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.000000, yerr=3.162278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 642.539520E+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 0.012 0.029 30.447 30.448 qs_mol_dyn_low 1 2.0 0.003 0.003 30.227 30.235 qs_forces 11 3.9 0.022 0.029 29.817 29.818 qs_energies 11 4.9 0.001 0.001 28.068 28.089 scf_env_do_scf 11 5.9 0.000 0.001 22.738 22.738 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.922 19.923 velocity_verlet 10 3.0 0.359 0.360 15.257 15.259 dbcsr_multiply_generic 2286 12.5 0.092 0.097 13.523 13.650 qs_scf_new_mos 108 7.5 0.001 0.001 11.866 11.892 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.866 11.891 multiply_cannon 2286 13.5 0.222 0.229 10.964 11.304 ot_scf_mini 108 9.5 0.002 0.002 11.160 11.182 multiply_cannon_loop 2286 14.5 0.643 0.678 10.044 10.300 ot_mini 108 10.5 0.001 0.001 6.416 6.444 rebuild_ks_matrix 119 8.3 0.000 0.000 6.362 6.384 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.361 6.384 multiply_cannon_multrec 9144 15.5 1.667 1.829 6.097 6.250 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.699 5.720 qs_ot_get_derivative 108 11.5 0.001 0.001 5.135 5.157 dbcsr_mm_accdrv_process 12550 15.8 3.510 4.340 4.319 4.388 init_scf_run 11 5.9 0.000 0.001 3.866 3.866 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.866 3.866 sum_up_and_integrate 119 10.3 0.038 0.042 3.691 3.696 integrate_v_rspace 119 11.3 0.003 0.003 3.654 3.659 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.563 3.572 calculate_rho_elec 119 8.7 0.060 0.061 3.562 3.572 qs_ot_get_p 119 10.4 0.001 0.001 2.865 2.901 init_scf_loop 11 6.9 0.000 0.000 2.795 2.796 calculate_first_density_matrix 1 7.0 0.000 0.000 2.715 2.717 calculate_dm_sparse 119 9.5 0.000 0.000 2.105 2.121 pw_transfer 1439 11.6 0.066 0.068 2.050 2.061 mp_waitall_1 115863 16.7 1.468 2.049 1.468 2.049 density_rs2pw 119 9.7 0.004 0.004 1.866 2.011 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.956 1.968 make_m2s 4572 13.5 0.034 0.035 1.781 1.948 grid_integrate_task_list 119 12.3 1.849 1.927 1.849 1.927 rs_pw_transfer 974 11.9 0.008 0.008 1.734 1.897 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.879 1.882 make_images 4572 14.5 0.270 0.300 1.694 1.859 acc_transpose_blocks 9144 15.5 0.042 0.044 1.746 1.803 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.720 1.740 prepare_preconditioner 11 7.9 0.000 0.000 1.720 1.724 make_preconditioner 11 8.9 0.000 0.000 1.719 1.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.704 1.705 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.637 1.637 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.608 1.634 fft_wrap_pw1pw2_140 487 13.2 0.366 0.376 1.605 1.617 jit_kernel_multiply 8 15.2 0.770 1.589 0.770 1.589 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.506 1.514 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.495 1.503 potential_pw2rs 119 12.3 0.010 0.011 1.424 1.427 grid_collocate_task_list 119 9.7 1.295 1.411 1.295 1.411 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.351 1.360 cp_fm_diag_elpa_base 50 14.0 1.323 1.341 1.350 1.358 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.337 1.350 fft3d_ps 1201 14.6 0.560 0.572 1.287 1.299 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.249 1.269 ot_diis_step 108 11.5 0.012 0.013 1.267 1.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.212 1.249 apply_single 119 13.6 0.000 0.000 1.212 1.248 wfi_extrapolate 11 7.9 0.001 0.001 1.096 1.096 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.777 1.017 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.969 0.976 make_images_data 4572 15.5 0.042 0.046 0.774 0.960 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.922 mp_alltoall_d11v 2130 13.8 0.782 0.879 0.782 0.879 cp_fm_cholesky_invert 11 10.9 0.863 0.865 0.863 0.865 jit_kernel_transpose 5 15.6 0.852 0.859 0.852 0.859 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.816 0.823 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.420 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.787 acc_transpose_blocks_sync 27432 16.5 0.718 0.775 0.718 0.775 qs_env_update_s_mstruct 11 6.9 0.017 0.025 0.695 0.747 mp_allgather_i34 2286 14.5 0.235 0.660 0.235 0.660 mp_alltoall_z22v 1201 16.6 0.602 0.628 0.602 0.628 mp_waitany 5200 13.7 0.459 0.620 0.459 0.620 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.274 0.610 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.448000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=605.363636, yerr=10.074926 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 762.318848E+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 0.017 0.034 44.302 44.303 qs_mol_dyn_low 1 2.0 0.003 0.003 44.087 44.096 qs_forces 11 3.9 0.004 0.007 43.847 43.848 qs_energies 11 4.9 0.005 0.009 41.787 41.792 scf_env_do_scf 11 5.9 0.001 0.001 35.910 35.911 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.040 28.041 velocity_verlet 10 3.0 0.002 0.002 24.533 24.561 dbcsr_multiply_generic 2286 12.5 0.100 0.104 19.826 19.946 qs_scf_new_mos 108 7.5 0.001 0.001 18.187 18.286 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.186 18.285 ot_scf_mini 108 9.5 0.002 0.002 17.033 17.136 multiply_cannon 2286 13.5 0.297 0.305 15.877 16.745 multiply_cannon_loop 2286 14.5 0.873 0.890 14.594 15.504 ot_mini 108 10.5 0.001 0.001 10.659 10.775 multiply_cannon_multrec 9144 15.5 3.671 5.247 9.206 9.591 qs_ot_get_derivative 108 11.5 0.001 0.001 8.580 8.683 init_scf_loop 11 6.9 0.000 0.000 7.843 7.844 rebuild_ks_matrix 119 8.3 0.000 0.000 7.400 7.543 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.399 7.542 dbcsr_mm_accdrv_process 12550 15.8 4.772 6.870 5.396 6.902 prepare_preconditioner 11 7.9 0.000 0.000 6.850 6.863 make_preconditioner 11 8.9 0.000 0.000 6.850 6.863 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.650 6.780 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.410 6.743 cp_fm_upper_to_full 72 14.2 3.197 4.583 3.197 4.583 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.290 4.294 calculate_rho_elec 119 8.7 0.119 0.121 4.289 4.294 sum_up_and_integrate 119 10.3 0.064 0.066 4.058 4.065 integrate_v_rspace 119 11.3 0.003 0.003 3.993 3.999 init_scf_run 11 5.9 0.000 0.001 3.795 3.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.795 3.795 mp_waitall_1 94719 16.7 2.674 3.675 2.674 3.675 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.180 3.603 qs_ot_get_p 119 10.4 0.001 0.001 3.432 3.572 dbcsr_complete_redistribute 329 12.2 0.301 0.310 2.002 2.855 pw_transfer 1439 11.6 0.068 0.069 2.658 2.661 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.686 2.575 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.560 2.562 make_m2s 4572 13.5 0.037 0.037 2.385 2.557 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.245 2.531 apply_single 119 13.6 0.000 0.000 2.245 2.531 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.692 2.521 make_images 4572 14.5 0.354 0.387 2.266 2.438 mp_alltoall_i22 627 13.8 1.465 2.313 1.465 2.313 calculate_first_density_matrix 1 7.0 0.000 0.000 2.293 2.295 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.435 2.254 calculate_dm_sparse 119 9.5 0.000 0.000 2.220 2.233 density_rs2pw 119 9.7 0.004 0.004 2.169 2.191 fft_wrap_pw1pw2_140 487 13.2 0.617 0.620 2.165 2.169 grid_integrate_task_list 119 12.3 2.093 2.151 2.093 2.151 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.069 2.122 ot_diis_step 108 11.5 0.014 0.015 2.054 2.055 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.034 2.035 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.990 1.992 acc_transpose_blocks 9144 15.5 0.044 0.045 1.862 1.894 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.783 1.786 mp_sum_l 7287 12.8 1.022 1.772 1.022 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.664 1.683 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.675 1.676 fft3d_ps 1201 14.6 0.595 0.603 1.577 1.580 grid_collocate_task_list 119 9.7 1.532 1.565 1.532 1.565 rs_pw_transfer 974 11.9 0.009 0.009 1.480 1.509 cp_fm_cholesky_invert 11 10.9 1.445 1.448 1.445 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.433 1.434 potential_pw2rs 119 12.3 0.014 0.014 1.385 1.389 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.384 1.384 cp_fm_diag_elpa_base 50 14.0 1.238 1.290 1.382 1.382 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.086 1.350 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.330 make_images_data 4572 15.5 0.045 0.048 1.031 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.236 mp_alltoall_d11v 2130 13.8 1.123 1.174 1.123 1.174 jit_kernel_multiply 5 15.6 0.595 1.155 0.595 1.155 acc_transpose_blocks_sync 27432 16.5 1.110 1.139 1.110 1.139 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.106 1.120 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.000 1.061 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.960 0.972 qs_create_task_list 11 7.9 0.005 0.006 0.935 0.946 generate_qs_task_list 11 8.9 0.369 0.388 0.930 0.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.303000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.000000, yerr=13.810405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 500.588544E+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 2196943. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.046 81.928 81.930 qs_mol_dyn_low 1 2.0 0.003 0.004 81.547 81.556 qs_forces 11 3.9 0.004 0.015 81.435 81.437 qs_energies 11 4.9 0.005 0.009 78.539 78.556 scf_env_do_scf 11 5.9 0.000 0.001 69.632 69.635 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.078 64.079 dbcsr_multiply_generic 2055 12.4 0.105 0.107 50.482 50.717 qs_scf_new_mos 99 7.5 0.001 0.001 46.674 46.809 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.674 46.809 ot_scf_mini 99 9.5 0.002 0.002 44.297 44.435 multiply_cannon 2055 13.4 0.182 0.187 42.363 43.077 velocity_verlet 10 3.0 0.002 0.005 42.951 42.952 multiply_cannon_loop 2055 14.4 1.793 1.835 41.412 42.078 ot_mini 99 10.5 0.001 0.001 26.205 26.334 qs_ot_get_derivative 99 11.5 0.001 0.001 19.385 19.509 multiply_cannon_multrec 49320 15.4 11.478 12.058 17.488 18.192 rebuild_ks_matrix 110 8.3 0.000 0.001 14.703 14.882 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.022 14.703 14.881 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.880 13.032 mp_waitall_1 220248 16.4 10.426 11.301 10.426 11.301 multiply_cannon_sync_h2d 49320 15.4 9.696 10.306 9.696 10.306 qs_ot_get_p 110 10.4 0.001 0.001 9.638 9.784 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.547 8.100 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.258 7.817 apply_single 110 13.6 0.000 0.000 7.257 7.817 multiply_cannon_metrocomm3 49320 15.4 0.084 0.089 6.330 7.415 sum_up_and_integrate 110 10.3 0.036 0.045 7.124 7.139 integrate_v_rspace 110 11.3 0.003 0.007 7.087 7.111 qs_rho_update_rho_low 110 7.6 0.001 0.006 6.667 6.813 calculate_rho_elec 110 8.6 0.022 0.030 6.666 6.813 init_scf_run 11 5.9 0.000 0.001 6.802 6.802 scf_env_initial_rho_setup 11 6.9 0.005 0.043 6.802 6.802 ot_diis_step 99 11.5 0.005 0.005 6.630 6.630 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.532 6.550 dbcsr_mm_accdrv_process 87628 16.1 3.024 3.114 5.879 6.202 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.713 5.713 init_scf_loop 11 6.9 0.000 0.000 5.528 5.528 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.278 5.345 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.138 5.159 cp_fm_diag_elpa_base 48 14.0 5.126 5.147 5.136 5.157 mp_sum_l 6594 12.7 3.809 4.490 3.809 4.490 rs_pw_transfer 902 11.9 0.011 0.013 3.721 4.295 wfi_extrapolate 11 7.9 0.001 0.001 4.049 4.050 make_m2s 4110 13.4 0.060 0.066 3.861 3.958 density_rs2pw 110 9.6 0.004 0.005 3.387 3.879 make_images 4110 14.4 0.177 0.189 3.766 3.867 calculate_dm_sparse 110 9.5 0.001 0.001 3.742 3.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.522 3.526 grid_integrate_task_list 110 12.3 3.241 3.364 3.241 3.364 prepare_preconditioner 11 7.9 0.000 0.000 3.300 3.317 make_preconditioner 11 8.9 0.000 0.000 3.300 3.317 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.257 3.262 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.195 3.245 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.135 3.219 pw_transfer 1331 11.6 0.055 0.070 3.103 3.186 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.079 3.124 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.014 3.099 fft_wrap_pw1pw2_140 451 13.1 0.458 0.504 2.576 2.676 potential_pw2rs 110 12.3 0.006 0.007 2.650 2.670 jit_kernel_multiply 13 15.9 2.581 2.636 2.581 2.636 calculate_first_density_matrix 1 7.0 0.000 0.000 2.609 2.613 acc_transpose_blocks 49320 15.4 0.210 0.217 2.531 2.582 mp_alltoall_d11v 2046 13.8 2.062 2.549 2.062 2.549 grid_collocate_task_list 110 9.6 2.155 2.374 2.155 2.374 mp_waitany 14300 13.8 1.827 2.360 1.827 2.360 fft3d_ps 1111 14.6 0.784 0.867 2.201 2.259 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.954 1.981 mp_sum_d 3889 11.9 1.400 1.955 1.400 1.955 make_images_data 4110 15.4 0.044 0.047 1.758 1.884 hybrid_alltoall_any 4261 16.3 0.085 0.491 1.524 1.809 cp_fm_cholesky_invert 11 10.9 1.793 1.797 1.793 1.797 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.641 1.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.930000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 592.719872E+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 2762017. 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 0.029 0.111 71.211 71.211 qs_mol_dyn_low 1 2.0 0.003 0.006 70.852 70.881 qs_forces 11 3.9 0.003 0.005 70.782 70.783 qs_energies 11 4.9 0.009 0.024 67.350 67.353 scf_env_do_scf 11 5.9 0.000 0.001 57.120 57.124 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 49.638 49.639 dbcsr_multiply_generic 2055 12.4 0.116 0.123 38.200 38.400 velocity_verlet 10 3.0 0.001 0.002 36.520 36.522 qs_scf_new_mos 99 7.5 0.001 0.001 33.344 33.483 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.343 33.482 multiply_cannon 2055 13.4 0.223 0.241 31.661 32.795 ot_scf_mini 99 9.5 0.003 0.004 31.682 31.801 multiply_cannon_loop 2055 14.4 1.156 1.179 30.456 31.465 ot_mini 99 10.5 0.001 0.001 18.679 18.810 multiply_cannon_multrec 24660 15.4 7.052 8.734 14.180 16.047 rebuild_ks_matrix 110 8.3 0.000 0.000 13.854 13.969 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.854 13.969 qs_ot_get_derivative 99 11.5 0.001 0.002 12.910 13.030 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.114 12.217 mp_waitall_1 176588 16.5 7.798 10.077 7.798 10.077 multiply_cannon_metrocomm3 24660 15.4 0.071 0.075 5.423 8.239 multiply_cannon_sync_h2d 24660 15.4 6.355 7.667 6.355 7.667 dbcsr_mm_accdrv_process 52282 16.1 5.540 6.569 6.957 7.542 init_scf_loop 11 6.9 0.000 0.002 7.448 7.450 init_scf_run 11 5.9 0.000 0.001 7.439 7.439 scf_env_initial_rho_setup 11 6.9 0.001 0.005 7.438 7.439 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.465 7.069 apply_single 110 13.6 0.000 0.001 6.465 7.069 sum_up_and_integrate 110 10.3 0.052 0.063 6.696 6.716 integrate_v_rspace 110 11.3 0.002 0.003 6.643 6.667 qs_ot_get_p 110 10.4 0.001 0.002 6.401 6.579 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.179 6.185 calculate_rho_elec 110 8.6 0.040 0.048 6.178 6.185 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.083 5.772 ot_diis_step 99 11.5 0.010 0.013 5.721 5.721 prepare_preconditioner 11 7.9 0.000 0.000 5.405 5.426 make_preconditioner 11 8.9 0.000 0.001 5.405 5.426 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.002 5.154 make_m2s 4110 13.4 0.057 0.060 4.146 4.573 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.506 4.526 make_images 4110 14.4 0.411 0.471 4.039 4.460 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.048 4.048 wfi_extrapolate 11 7.9 0.060 0.474 3.968 3.968 pw_transfer 1331 11.6 0.067 0.074 3.674 3.824 density_rs2pw 110 9.6 0.005 0.005 3.268 3.817 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.567 3.720 rs_pw_transfer 902 11.9 0.012 0.015 2.993 3.548 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.492 3.505 cp_fm_diag_elpa_base 48 14.0 3.445 3.461 3.490 3.502 grid_integrate_task_list 110 12.3 3.160 3.398 3.160 3.398 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.376 3.378 calculate_first_density_matrix 1 7.0 0.035 0.279 3.368 3.370 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.253 3.320 fft_wrap_pw1pw2_140 451 13.1 0.523 0.541 3.076 3.227 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.965 3.020 calculate_dm_sparse 110 9.5 0.001 0.001 2.968 2.995 hybrid_alltoall_any 4261 16.3 0.108 0.464 1.978 2.707 make_images_data 4110 15.4 0.049 0.054 2.279 2.705 fft3d_ps 1111 14.6 1.106 1.332 2.537 2.704 mp_sum_l 6594 12.7 1.818 2.554 1.818 2.554 grid_collocate_task_list 110 9.6 2.094 2.536 2.094 2.536 potential_pw2rs 110 12.3 0.008 0.009 2.476 2.500 cp_fm_cholesky_invert 11 10.9 2.493 2.499 2.493 2.499 qs_energies_init_hamiltonians 11 5.9 0.036 0.055 2.236 2.240 jit_kernel_multiply 11 16.3 1.061 2.174 1.061 2.174 acc_transpose_blocks 24660 15.4 0.115 0.118 1.969 1.993 mp_alltoall_d11v 2046 13.8 1.678 1.985 1.678 1.985 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.952 1.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.897 1.915 mp_waitany 10164 13.8 1.279 1.891 1.279 1.891 mp_sum_d 3889 11.9 1.441 1.822 1.441 1.822 multiply_cannon_metrocomm4 22605 15.4 0.081 0.084 0.777 1.631 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.219 1.012 1.556 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.549 mp_allgather_i34 2055 14.4 0.513 1.516 0.513 1.516 mp_irecv_dv 57340 16.2 0.647 1.510 0.647 1.510 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.492 dbcsr_complete_redistribute 325 12.2 0.307 0.401 1.168 1.460 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.211000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.454545, yerr=7.062882 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 667.103232E+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 0.089 0.474 61.333 61.349 qs_mol_dyn_low 1 2.0 0.003 0.004 60.502 60.539 qs_forces 11 3.9 0.003 0.009 60.411 60.413 qs_energies 11 4.9 0.002 0.009 57.180 57.184 scf_env_do_scf 11 5.9 0.001 0.002 48.874 48.875 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.368 40.369 velocity_verlet 10 3.0 0.001 0.002 32.749 32.752 dbcsr_multiply_generic 2055 12.4 0.106 0.111 29.115 29.388 qs_scf_new_mos 99 7.5 0.001 0.001 25.266 25.384 qs_scf_loop_do_ot 99 8.5 0.001 0.003 25.266 25.384 ot_scf_mini 99 9.5 0.004 0.014 24.016 24.134 multiply_cannon 2055 13.4 0.211 0.217 22.862 23.944 multiply_cannon_loop 2055 14.4 0.813 0.844 21.709 22.835 ot_mini 99 10.5 0.001 0.001 13.861 13.983 rebuild_ks_matrix 110 8.3 0.000 0.000 12.436 12.615 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 12.436 12.615 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.930 11.092 multiply_cannon_multrec 16440 15.4 3.610 4.832 9.836 10.916 mp_waitall_1 139946 16.5 7.301 10.343 7.301 10.343 qs_ot_get_derivative 99 11.5 0.001 0.001 9.359 9.480 init_scf_loop 11 6.9 0.003 0.019 8.472 8.473 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.613 7.553 prepare_preconditioner 11 7.9 0.000 0.001 6.680 6.698 make_preconditioner 11 8.9 0.000 0.002 6.680 6.698 sum_up_and_integrate 110 10.3 0.060 0.061 6.559 6.573 integrate_v_rspace 110 11.3 0.003 0.003 6.499 6.513 dbcsr_mm_accdrv_process 34862 16.1 5.369 5.724 6.069 6.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.001 6.356 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.021 6.032 calculate_rho_elec 110 8.6 0.059 0.059 6.021 6.031 init_scf_run 11 5.9 0.000 0.001 5.809 5.809 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.809 5.809 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.021 5.572 apply_single 110 13.6 0.000 0.000 5.021 5.572 qs_ot_get_p 110 10.4 0.003 0.016 5.326 5.470 make_m2s 4110 13.4 0.050 0.053 4.155 4.535 ot_diis_step 99 11.5 0.011 0.011 4.470 4.471 make_images 4110 14.4 0.400 0.521 4.042 4.421 density_rs2pw 110 9.6 0.004 0.005 3.093 4.336 rs_pw_transfer 902 11.9 0.010 0.011 2.724 3.965 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.275 3.943 multiply_cannon_sync_h2d 16440 15.4 3.238 3.896 3.238 3.896 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.716 3.719 pw_transfer 1331 11.6 0.066 0.074 3.640 3.647 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.532 3.542 grid_integrate_task_list 110 12.3 3.181 3.416 3.181 3.416 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.351 3.351 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.094 3.096 fft_wrap_pw1pw2_140 451 13.1 0.641 0.649 3.068 3.077 wfi_extrapolate 11 7.9 0.001 0.001 2.984 2.984 make_images_data 4110 15.4 0.045 0.050 2.363 2.844 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.821 2.831 cp_fm_diag_elpa_base 48 14.0 2.754 2.785 2.820 2.830 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.722 2.778 calculate_first_density_matrix 1 7.0 0.001 0.006 2.730 2.731 hybrid_alltoall_any 4261 16.3 0.110 0.383 2.131 2.699 cp_fm_cholesky_invert 11 10.9 2.606 2.612 2.606 2.612 mp_sum_l 6594 12.7 1.852 2.572 1.852 2.572 calculate_dm_sparse 110 9.5 0.001 0.001 2.523 2.560 mp_waitany 17072 13.8 1.249 2.554 1.249 2.554 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.898 2.539 grid_collocate_task_list 110 9.6 2.120 2.523 2.120 2.523 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 2.417 2.494 mp_irecv_dv 48980 15.7 0.823 2.403 0.823 2.403 fft3d_ps 1111 14.6 1.090 1.104 2.351 2.367 potential_pw2rs 110 12.3 0.011 0.012 2.270 2.276 mp_alltoall_d11v 2046 13.8 1.775 2.262 1.775 2.262 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.183 0.969 2.207 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.035 2.037 dbcsr_complete_redistribute 325 12.2 0.335 0.373 1.465 1.905 cp_fm_upper_to_full 70 14.2 1.371 1.748 1.371 1.748 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.718 acc_transpose_blocks 16440 15.4 0.078 0.081 1.581 1.635 mp_allgather_i34 2055 14.4 0.465 1.574 0.465 1.574 cp_fm_cholesky_decompose 22 10.9 1.535 1.554 1.535 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.475 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.468 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.985 1.430 rs_gather_matrices 110 12.3 0.232 0.260 0.970 1.383 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.275 1.285 qs_env_update_s_mstruct 11 6.9 0.001 0.006 1.118 1.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.349000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.090909, yerr=7.798071 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 738.516992E+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 0.026 0.070 67.857 67.858 qs_mol_dyn_low 1 2.0 0.003 0.004 67.497 67.507 qs_forces 11 3.9 0.003 0.004 67.431 67.433 qs_energies 11 4.9 0.022 0.086 63.952 63.956 scf_env_do_scf 11 5.9 0.001 0.001 54.482 54.485 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.899 42.900 velocity_verlet 10 3.0 0.002 0.006 37.437 37.440 dbcsr_multiply_generic 2055 12.4 0.113 0.119 31.233 31.500 qs_scf_new_mos 99 7.5 0.001 0.001 27.750 27.856 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.749 27.856 ot_scf_mini 99 9.5 0.003 0.003 26.067 26.179 multiply_cannon 2055 13.4 0.238 0.254 24.118 25.487 multiply_cannon_loop 2055 14.4 1.396 1.451 22.758 23.363 ot_mini 99 10.5 0.001 0.001 15.102 15.235 multiply_cannon_multrec 24660 15.4 4.134 6.682 13.134 14.371 rebuild_ks_matrix 110 8.3 0.000 0.000 12.323 12.445 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.322 12.445 init_scf_loop 11 6.9 0.001 0.003 11.544 11.545 qs_ot_get_derivative 99 11.5 0.001 0.001 10.928 11.044 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.858 10.966 dbcsr_mm_accdrv_process 52304 16.0 7.697 9.509 8.841 9.805 prepare_preconditioner 11 7.9 0.000 0.000 9.760 9.777 make_preconditioner 11 8.9 0.000 0.001 9.760 9.777 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.013 9.463 sum_up_and_integrate 110 10.3 0.068 0.071 6.558 6.569 integrate_v_rspace 110 11.3 0.003 0.003 6.490 6.500 mp_waitall_1 121746 16.5 4.417 6.191 4.417 6.191 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.168 6.176 calculate_rho_elec 110 8.6 0.078 0.082 6.167 6.176 init_scf_run 11 5.9 0.000 0.001 6.154 6.154 scf_env_initial_rho_setup 11 6.9 0.001 0.004 6.153 6.154 qs_ot_get_p 110 10.4 0.001 0.001 5.860 6.005 make_m2s 4110 13.4 0.059 0.061 5.456 5.741 make_images 4110 14.4 0.581 0.700 5.317 5.597 cp_fm_upper_to_full 70 14.2 3.324 4.818 3.324 4.818 ot_diis_step 99 11.5 0.011 0.013 4.137 4.137 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.052 4.136 apply_single 110 13.6 0.000 0.000 4.052 4.136 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.008 4.025 pw_transfer 1331 11.6 0.066 0.075 3.827 3.859 dbcsr_complete_redistribute 325 12.2 0.465 0.552 2.697 3.824 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.721 3.756 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.555 3.611 density_rs2pw 110 9.6 0.004 0.004 3.057 3.577 grid_integrate_task_list 110 12.3 3.273 3.449 3.273 3.449 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.438 3.438 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.343 3.403 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.310 3.313 copy_fm_to_dbcsr 174 11.2 0.002 0.007 2.199 3.310 rs_pw_transfer 902 11.9 0.010 0.013 2.686 3.294 fft_wrap_pw1pw2_140 451 13.1 0.678 0.758 3.210 3.249 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.518 3.092 calculate_first_density_matrix 1 7.0 0.001 0.008 3.045 3.048 make_images_data 4110 15.4 0.048 0.052 2.726 3.035 calculate_dm_sparse 110 9.5 0.001 0.001 2.984 3.027 hybrid_alltoall_any 4261 16.3 0.124 0.464 2.302 3.010 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.007 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.910 2.920 cp_fm_diag_elpa_base 48 14.0 2.762 2.820 2.908 2.919 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.854 2.894 acc_transpose_blocks 24660 15.4 0.110 0.111 2.746 2.866 mp_alltoall_i22 605 13.7 1.683 2.851 1.683 2.851 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.738 2.838 qs_energies_init_hamiltonians 11 5.9 0.002 0.009 2.711 2.712 cp_fm_cholesky_invert 11 10.9 2.627 2.635 2.627 2.635 multiply_cannon_sync_h2d 24660 15.4 2.394 2.539 2.394 2.539 fft3d_ps 1111 14.6 1.085 1.114 2.471 2.492 grid_collocate_task_list 110 9.6 2.224 2.492 2.224 2.492 potential_pw2rs 110 12.3 0.013 0.013 2.188 2.196 mp_alltoall_d11v 2046 13.8 1.728 2.049 1.728 2.049 mp_waitany 13376 13.8 1.295 1.841 1.295 1.841 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.779 1.820 qs_env_update_s_mstruct 11 6.9 0.021 0.034 1.586 1.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.707 1.720 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.710 cp_fm_cholesky_decompose 22 10.9 1.649 1.709 1.649 1.709 rs_pw_transfer_RS2PW_140 121 11.5 0.196 0.217 1.072 1.685 mp_sum_l 6594 12.7 1.011 1.634 1.011 1.634 jit_kernel_multiply 8 15.9 0.804 1.566 0.804 1.566 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.850 1.566 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.559 mp_allgather_i34 2055 14.4 0.443 1.548 0.443 1.548 mp_irecv_dv 62702 16.1 0.747 1.484 0.747 1.484 acc_transpose_blocks_sync 73980 16.4 1.363 1.469 1.363 1.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.858000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.545455, yerr=11.436494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 856.625152E+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 0.018 0.031 56.008 56.009 qs_mol_dyn_low 1 2.0 0.003 0.003 55.748 55.759 qs_forces 11 3.9 0.003 0.003 55.682 55.683 qs_energies 11 4.9 0.001 0.001 51.967 51.971 scf_env_do_scf 11 5.9 0.000 0.001 43.409 43.409 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.536 35.537 velocity_verlet 10 3.0 0.002 0.002 31.218 31.221 dbcsr_multiply_generic 2055 12.4 0.105 0.110 23.584 23.720 qs_scf_new_mos 99 7.5 0.001 0.001 20.528 20.581 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.527 20.581 multiply_cannon 2055 13.4 0.235 0.254 18.117 19.312 ot_scf_mini 99 9.5 0.002 0.002 19.285 19.307 multiply_cannon_loop 2055 14.4 0.605 0.629 16.849 17.102 rebuild_ks_matrix 110 8.3 0.000 0.000 12.140 12.163 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.017 12.140 12.163 ot_mini 99 10.5 0.001 0.001 10.834 10.850 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.795 10.817 multiply_cannon_multrec 8220 15.4 3.387 4.639 7.884 8.746 init_scf_loop 11 6.9 0.000 0.000 7.825 7.826 mp_waitall_1 103326 16.6 5.685 7.454 5.685 7.454 qs_ot_get_derivative 99 11.5 0.001 0.001 7.155 7.178 sum_up_and_integrate 110 10.3 0.079 0.081 6.501 6.517 integrate_v_rspace 110 11.3 0.003 0.003 6.421 6.438 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.277 6.297 calculate_rho_elec 110 8.6 0.115 0.116 6.277 6.296 prepare_preconditioner 11 7.9 0.000 0.000 5.922 5.926 make_preconditioner 11 8.9 0.000 0.000 5.922 5.926 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.493 5.568 dbcsr_mm_accdrv_process 17442 15.9 3.133 4.406 4.360 5.486 init_scf_run 11 5.9 0.000 0.001 5.416 5.416 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.415 5.416 qs_ot_get_p 110 10.4 0.001 0.001 4.520 4.529 make_m2s 4110 13.4 0.038 0.039 4.174 4.463 make_images 4110 14.4 0.654 0.717 4.046 4.336 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.945 4.249 pw_transfer 1331 11.6 0.066 0.071 3.879 3.888 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.772 3.785 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.719 3.766 apply_single 110 13.6 0.000 0.000 3.719 3.765 ot_diis_step 99 11.5 0.012 0.012 3.656 3.656 grid_integrate_task_list 110 12.3 3.363 3.538 3.363 3.538 density_rs2pw 110 9.6 0.004 0.004 2.987 3.359 fft_wrap_pw1pw2_140 451 13.1 0.835 0.847 3.273 3.289 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.200 3.204 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.865 2.866 cp_fm_cholesky_invert 11 10.9 2.808 2.811 2.808 2.811 calculate_dm_sparse 110 9.5 0.001 0.001 2.761 2.802 hybrid_alltoall_any 4261 16.3 0.202 0.867 2.203 2.722 wfi_extrapolate 11 7.9 0.001 0.001 2.718 2.718 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.690 2.691 make_images_data 4110 15.4 0.041 0.046 2.239 2.624 calculate_first_density_matrix 1 7.0 0.000 0.000 2.596 2.598 grid_collocate_task_list 110 9.6 2.315 2.575 2.315 2.575 rs_pw_transfer 902 11.9 0.010 0.010 2.161 2.560 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.550 2.551 multiply_cannon_sync_h2d 8220 15.4 2.376 2.471 2.376 2.471 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.372 2.380 cp_fm_diag_elpa_base 48 14.0 2.316 2.334 2.371 2.378 fft3d_ps 1111 14.6 1.128 1.138 2.316 2.328 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.163 2.172 potential_pw2rs 110 12.3 0.015 0.016 2.045 2.048 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.013 2.041 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 2.012 mp_alltoall_d11v 2046 13.8 1.612 1.919 1.612 1.919 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.884 1.896 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.720 1.734 acc_transpose_blocks 8220 15.4 0.039 0.041 1.659 1.692 cp_fm_cholesky_decompose 22 10.9 1.669 1.682 1.669 1.682 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.675 1.677 jit_kernel_multiply 8 15.5 0.920 1.654 0.920 1.654 qs_env_update_s_mstruct 11 6.9 0.003 0.006 1.511 1.633 dbcsr_complete_redistribute 325 12.2 0.602 0.620 1.455 1.562 mp_allgather_i34 2055 14.4 0.430 1.503 0.430 1.503 mp_waitany 9240 13.8 1.073 1.500 1.073 1.500 qs_create_task_list 11 7.9 0.000 0.001 1.215 1.320 generate_qs_task_list 11 8.9 0.376 0.445 1.214 1.320 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.172 0.808 1.207 rs_gather_matrices 110 12.3 0.323 0.368 0.925 1.160 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.116 1.132 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.009000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.454545, yerr=9.930336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 1.388401E+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 0.032 0.072 87.807 87.816 qs_mol_dyn_low 1 2.0 0.003 0.003 87.451 87.476 qs_forces 11 3.9 0.003 0.003 87.368 87.369 qs_energies 11 4.9 0.003 0.016 83.173 83.175 scf_env_do_scf 11 5.9 0.000 0.001 73.063 73.063 velocity_verlet 10 3.0 0.002 0.002 55.846 55.852 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.021 45.024 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.574 29.669 init_scf_loop 11 6.9 0.000 0.000 27.971 27.975 qs_scf_new_mos 99 7.5 0.001 0.001 26.811 26.858 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.811 26.858 prepare_preconditioner 11 7.9 0.000 0.000 25.929 25.938 make_preconditioner 11 8.9 0.000 0.000 25.929 25.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.136 25.410 ot_scf_mini 99 9.5 0.002 0.002 25.019 25.058 multiply_cannon 2055 13.4 0.326 0.352 22.663 23.312 multiply_cannon_loop 2055 14.4 0.813 0.834 20.907 21.183 cp_fm_upper_to_full 70 14.2 12.522 18.042 12.522 18.042 ot_mini 99 10.5 0.001 0.001 14.258 14.298 rebuild_ks_matrix 110 8.3 0.001 0.001 14.125 14.178 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.125 14.177 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.773 12.821 dbcsr_complete_redistribute 325 12.2 1.045 1.066 7.366 10.562 multiply_cannon_multrec 8220 15.4 4.291 4.411 9.969 10.104 qs_ot_get_derivative 99 11.5 0.001 0.001 9.643 9.686 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.383 9.580 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.777 8.933 mp_alltoall_i22 605 13.7 5.412 8.625 5.412 8.625 mp_waitall_1 84994 16.7 7.452 8.290 7.452 8.290 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.686 7.720 calculate_rho_elec 110 8.6 0.227 0.228 7.685 7.719 sum_up_and_integrate 110 10.3 0.150 0.152 7.292 7.305 integrate_v_rspace 110 11.3 0.004 0.004 7.141 7.155 init_scf_run 11 5.9 0.000 0.001 5.795 5.796 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.795 5.795 dbcsr_mm_accdrv_process 11614 15.7 3.840 4.126 5.531 5.745 make_m2s 4110 13.4 0.043 0.043 5.178 5.648 qs_ot_get_p 110 10.4 0.001 0.001 5.480 5.538 make_images 4110 14.4 0.891 0.946 4.992 5.460 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.755 5.210 apply_single 110 13.6 0.000 0.000 4.755 5.210 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.921 5.171 cp_fm_cholesky_invert 11 10.9 5.132 5.136 5.132 5.136 pw_transfer 1331 11.6 0.075 0.076 4.981 4.986 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.864 4.868 ot_diis_step 99 11.5 0.015 0.015 4.594 4.594 fft_wrap_pw1pw2_140 451 13.1 1.335 1.342 4.254 4.260 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.950 3.957 grid_integrate_task_list 110 12.3 3.687 3.756 3.687 3.756 density_rs2pw 110 9.6 0.004 0.005 3.690 3.742 qs_energies_init_hamiltonians 11 5.9 0.007 0.012 3.687 3.688 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.986 3.454 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.439 3.439 hybrid_alltoall_any 4261 16.3 0.263 0.567 2.809 3.429 wfi_extrapolate 11 7.9 0.001 0.001 3.329 3.329 make_images_data 4110 15.4 0.044 0.048 2.672 3.268 calculate_dm_sparse 110 9.5 0.001 0.001 3.157 3.190 multiply_cannon_sync_h2d 8220 15.4 3.142 3.171 3.142 3.171 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.940 2.945 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.867 2.868 cp_fm_diag_elpa_base 48 14.0 2.333 2.529 2.866 2.866 fft3d_ps 1111 14.6 1.301 1.308 2.796 2.806 grid_collocate_task_list 110 9.6 2.679 2.717 2.679 2.717 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.620 2.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.388 2.397 potential_pw2rs 110 12.3 0.021 0.022 2.378 2.380 calculate_first_density_matrix 1 7.0 0.000 0.000 2.344 2.344 rs_pw_transfer 902 11.9 0.010 0.011 2.266 2.300 qs_env_update_s_mstruct 11 6.9 0.002 0.008 2.186 2.234 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.180 mp_alltoall_d11v 2046 13.8 1.940 2.041 1.940 2.041 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.949 1.955 cp_fm_cholesky_decompose 22 10.9 1.941 1.952 1.941 1.952 qs_create_task_list 11 7.9 0.001 0.001 1.883 1.931 generate_qs_task_list 11 8.9 0.735 0.788 1.882 1.930 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.762 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.816000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1246.000000, yerr=44.441995 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 634.187776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2084567. 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 0.014 0.038 207.136 207.137 qs_mol_dyn_low 1 2.0 0.003 0.003 206.689 206.702 qs_forces 11 3.9 0.005 0.005 206.591 206.592 qs_energies 11 4.9 0.025 0.027 200.867 200.879 scf_env_do_scf 11 5.9 0.001 0.001 183.459 183.463 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.800 162.803 dbcsr_multiply_generic 2507 12.6 0.177 0.180 123.614 124.256 velocity_verlet 10 3.0 0.001 0.001 124.050 124.051 qs_scf_new_mos 117 7.6 0.001 0.001 122.171 122.402 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.171 122.402 ot_scf_mini 117 9.6 0.003 0.003 115.499 115.774 multiply_cannon 2507 13.6 0.239 0.253 101.259 102.533 multiply_cannon_loop 2507 14.6 2.330 2.452 99.050 100.533 ot_mini 117 10.6 0.001 0.001 64.372 64.632 multiply_cannon_multrec 60168 15.6 32.423 35.111 42.467 44.179 qs_ot_get_derivative 117 11.6 0.001 0.001 39.719 39.941 rebuild_ks_matrix 128 8.3 0.001 0.001 34.105 34.316 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.105 34.316 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.558 30.750 mp_waitall_1 267128 16.5 27.358 30.139 27.358 30.139 qs_ot_get_p 128 10.4 0.001 0.001 29.195 29.407 multiply_cannon_sync_h2d 60168 15.6 27.037 28.404 27.037 28.404 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.307 24.930 apply_single 128 13.6 0.001 0.001 24.307 24.929 ot_diis_step 117 11.6 0.008 0.008 24.456 24.457 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.642 22.707 init_scf_loop 11 6.9 0.000 0.001 20.583 20.584 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.153 20.154 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 17.921 18.100 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.127 17.151 cp_fm_diag_elpa_base 83 14.4 17.050 17.092 17.123 17.148 multiply_cannon_metrocomm3 60168 15.6 0.118 0.126 14.947 16.682 prepare_preconditioner 11 7.9 0.000 0.000 15.948 15.992 make_preconditioner 11 8.9 0.000 0.000 15.948 15.992 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.178 15.366 sum_up_and_integrate 128 10.3 0.089 0.108 14.527 14.539 integrate_v_rspace 128 11.3 0.004 0.004 14.438 14.454 make_m2s 5014 13.6 0.104 0.112 13.918 14.328 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.145 14.242 calculate_rho_elec 128 8.7 0.046 0.065 14.145 14.241 make_images 5014 14.6 0.400 0.422 13.742 14.158 init_scf_run 11 5.9 0.000 0.001 13.003 13.003 scf_env_initial_rho_setup 11 6.9 0.002 0.003 13.003 13.003 density_rs2pw 128 9.7 0.006 0.007 7.407 10.797 dbcsr_mm_accdrv_process 124484 16.2 4.413 4.783 9.586 10.307 rs_pw_transfer 1046 11.9 0.016 0.018 6.176 9.547 wfi_extrapolate 11 7.9 0.001 0.001 9.257 9.257 cp_fm_cholesky_invert 11 10.9 9.140 9.148 9.140 9.148 calculate_dm_sparse 128 9.5 0.001 0.001 8.493 8.594 mp_sum_l 7950 12.9 7.158 8.237 7.158 8.237 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.098 8.219 pw_transfer 1547 11.6 0.074 0.087 7.994 8.184 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.791 7.983 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.808 7.887 make_images_data 5014 15.6 0.069 0.074 6.797 7.731 grid_integrate_task_list 128 12.3 7.109 7.594 7.109 7.594 multiply_cannon_metrocomm1 60168 15.6 0.096 0.101 5.736 7.384 hybrid_alltoall_any 5200 16.5 0.298 2.286 5.968 7.173 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.978 6.988 fft_wrap_pw1pw2_140 523 13.2 1.278 1.325 6.748 6.960 mp_waitany 16020 13.9 2.753 6.255 2.753 6.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.090 6.188 grid_collocate_task_list 128 9.7 4.675 5.956 4.675 5.956 fft3d_ps 1291 14.7 2.134 2.779 5.550 5.882 rs_pw_transfer_RS2PW_140 139 11.5 0.314 0.359 2.210 5.610 mp_alltoall_d11v 2415 14.1 4.277 5.176 4.277 5.176 potential_pw2rs 128 12.3 0.009 0.011 4.833 4.882 cp_fm_cholesky_decompose 22 10.9 4.632 4.645 4.632 4.645 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.137000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.000000, yerr=6.714976 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/16/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 2.183246E+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 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 838.606848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5736082. 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 0.025 0.072 192.187 192.188 qs_mol_dyn_low 1 2.0 0.003 0.003 191.712 191.727 qs_forces 11 3.9 0.005 0.015 191.625 191.627 qs_energies 11 4.9 0.004 0.026 184.781 184.795 scf_env_do_scf 11 5.9 0.001 0.001 166.568 166.578 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.567 133.570 velocity_verlet 10 3.0 0.001 0.002 119.513 119.515 dbcsr_multiply_generic 2507 12.6 0.190 0.196 97.366 98.658 qs_scf_new_mos 117 7.6 0.001 0.001 93.916 94.528 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.915 94.528 ot_scf_mini 117 9.6 0.004 0.004 89.090 89.762 multiply_cannon 2507 13.6 0.506 0.569 77.174 82.215 multiply_cannon_loop 2507 14.6 1.557 1.623 73.835 76.537 ot_mini 117 10.6 0.001 0.001 50.306 50.916 mp_waitall_1 214728 16.6 24.188 39.897 24.188 39.897 multiply_cannon_multrec 30084 15.6 20.885 25.961 31.394 36.934 rebuild_ks_matrix 128 8.3 0.001 0.001 32.986 33.726 qs_ks_build_kohn_sham_matrix 128 9.3 0.031 0.090 32.985 33.725 init_scf_loop 11 6.9 0.001 0.009 32.912 32.914 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.643 30.312 multiply_cannon_metrocomm3 30084 15.6 0.095 0.100 15.910 29.852 qs_ot_get_derivative 117 11.6 0.001 0.002 28.098 28.776 prepare_preconditioner 11 7.9 0.000 0.000 28.305 28.393 make_preconditioner 11 8.9 0.001 0.008 28.305 28.393 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.980 27.527 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.316 23.434 apply_single 128 13.6 0.001 0.001 22.315 23.434 ot_diis_step 117 11.6 0.014 0.015 22.023 22.026 qs_ot_get_p 128 10.4 0.001 0.002 20.872 21.637 multiply_cannon_sync_h2d 30084 15.6 17.828 20.326 17.828 20.326 cp_fm_cholesky_invert 11 10.9 16.536 16.548 16.536 16.548 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.037 16.076 make_m2s 5014 13.6 0.094 0.101 14.104 15.735 make_images 5014 14.6 1.185 1.387 13.892 15.521 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.836 14.837 sum_up_and_integrate 128 10.3 0.122 0.173 14.684 14.712 integrate_v_rspace 128 11.3 0.004 0.007 14.562 14.600 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.298 14.327 calculate_rho_elec 128 8.7 0.089 0.106 14.297 14.326 init_scf_run 11 5.9 0.000 0.001 12.831 12.833 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.831 12.833 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.701 11.729 cp_fm_diag_elpa_base 83 14.4 11.448 11.541 11.695 11.723 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.159 11.633 multiply_cannon_metrocomm4 27577 15.6 0.110 0.121 3.923 11.023 mp_irecv_dv 69486 16.3 3.714 10.620 3.714 10.620 dbcsr_mm_accdrv_process 62242 16.2 5.432 6.115 9.958 10.507 make_images_data 5014 15.6 0.068 0.078 8.325 10.319 density_rs2pw 128 9.7 0.007 0.007 7.580 10.023 hybrid_alltoall_any 5200 16.5 0.352 1.564 7.166 9.598 pw_transfer 1547 11.6 0.086 0.099 8.763 8.818 rs_pw_transfer 1046 11.9 0.015 0.016 6.276 8.795 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.539 8.596 wfi_extrapolate 11 7.9 0.001 0.001 8.372 8.372 fft_wrap_pw1pw2_140 523 13.2 1.337 1.356 7.575 7.656 grid_integrate_task_list 128 12.3 7.186 7.573 7.186 7.573 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.230 7.006 cp_fm_cholesky_decompose 22 10.9 6.870 6.968 6.870 6.968 calculate_dm_sparse 128 9.5 0.001 0.001 6.485 6.633 mp_sum_l 7950 12.9 4.295 6.528 4.295 6.528 grid_collocate_task_list 128 9.7 4.808 6.187 4.808 6.187 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.108 6.118 fft3d_ps 1291 14.7 2.810 2.983 5.843 5.889 mp_waitany 11748 13.9 3.105 5.625 3.105 5.625 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.401 5.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.213 5.407 rs_pw_transfer_RS2PW_140 139 11.5 0.353 0.376 2.719 5.205 mp_allgather_i34 2507 14.6 1.846 4.966 1.846 4.966 potential_pw2rs 128 12.3 0.016 0.019 4.933 4.953 mp_alltoall_d11v 2415 14.1 4.082 4.614 4.082 4.614 mp_sum_d 4470 12.1 2.705 4.149 2.705 4.149 calculate_first_density_matrix 1 7.0 0.001 0.005 4.053 4.056 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 3.967 3.968 dbcsr_complete_redistribute 395 12.7 0.787 0.876 3.123 3.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.188000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 949.121024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58278. MP_Allreduce 11070 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.121 175.654 175.655 qs_mol_dyn_low 1 2.0 0.003 0.003 174.706 174.719 qs_forces 11 3.9 0.004 0.004 174.611 174.615 qs_energies 11 4.9 0.005 0.015 167.936 167.952 scf_env_do_scf 11 5.9 0.001 0.001 152.219 152.219 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.394 117.395 velocity_verlet 10 3.0 0.006 0.010 111.795 111.797 dbcsr_multiply_generic 2485 12.5 0.178 0.184 81.260 82.503 qs_scf_new_mos 116 7.6 0.001 0.001 79.245 79.557 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.244 79.556 ot_scf_mini 116 9.6 0.003 0.004 75.074 75.431 multiply_cannon 2485 13.5 0.503 0.530 61.807 66.402 multiply_cannon_loop 2485 14.5 1.110 1.181 58.760 61.531 ot_mini 116 10.6 0.001 0.001 42.097 42.456 init_scf_loop 11 6.9 0.001 0.011 34.722 34.722 mp_waitall_1 169034 16.6 24.296 33.445 24.296 33.445 rebuild_ks_matrix 127 8.3 0.001 0.001 30.746 31.238 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 30.746 31.237 prepare_preconditioner 11 7.9 0.000 0.000 30.636 30.692 make_preconditioner 11 8.9 0.000 0.001 30.636 30.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.267 29.703 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.625 28.068 multiply_cannon_multrec 19880 15.5 13.063 15.810 22.299 25.084 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 15.094 24.809 qs_ot_get_derivative 116 11.6 0.001 0.002 22.741 23.115 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.461 20.518 apply_single 127 13.6 0.001 0.001 19.461 20.518 ot_diis_step 116 11.6 0.017 0.018 19.244 19.244 qs_ot_get_p 127 10.4 0.001 0.001 18.404 18.819 make_m2s 4970 13.5 0.081 0.086 14.477 15.798 make_images 4970 14.5 1.146 1.241 14.247 15.565 multiply_cannon_sync_h2d 19880 15.5 13.643 15.471 13.643 15.471 sum_up_and_integrate 127 10.3 0.133 0.144 14.491 14.514 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.387 14.414 calculate_rho_elec 127 8.7 0.131 0.146 14.386 14.413 integrate_v_rspace 127 11.3 0.004 0.009 14.358 14.386 cp_fm_cholesky_invert 11 10.9 14.251 14.260 14.251 14.260 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.197 14.206 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.153 13.154 init_scf_run 11 5.9 0.000 0.001 10.602 10.602 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.601 10.602 make_images_data 4970 15.5 0.062 0.073 8.697 10.390 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.006 10.024 cp_fm_diag_elpa_base 82 14.4 9.595 9.740 10.003 10.021 hybrid_alltoall_any 5155 16.4 0.447 2.030 7.528 9.558 density_rs2pw 127 9.7 0.006 0.007 7.102 9.325 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.929 9.197 multiply_cannon_metrocomm4 17395 15.5 0.065 0.076 3.426 9.136 pw_transfer 1535 11.6 0.085 0.103 8.807 8.915 mp_irecv_dv 49801 16.2 3.300 8.883 3.300 8.883 dbcsr_mm_accdrv_process 41158 16.2 5.646 5.920 8.699 8.818 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.585 8.697 grid_integrate_task_list 127 12.3 7.285 7.745 7.285 7.745 fft_wrap_pw1pw2_140 519 13.2 1.404 1.432 7.607 7.736 wfi_extrapolate 11 7.9 0.001 0.001 7.507 7.507 cp_fm_cholesky_decompose 22 10.9 7.308 7.344 7.308 7.344 rs_pw_transfer 1038 11.9 0.013 0.014 5.055 7.290 cp_fm_upper_to_full 104 14.8 5.662 7.232 5.662 7.232 dbcsr_complete_redistribute 393 12.7 1.183 1.212 4.593 6.343 calculate_dm_sparse 127 9.5 0.001 0.001 5.764 5.886 fft3d_ps 1281 14.7 2.721 2.951 5.725 5.806 grid_collocate_task_list 127 9.7 4.949 5.682 4.949 5.682 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.473 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.589 5.238 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.443 5.187 mp_alltoall_d11v 2401 14.1 4.309 4.846 4.309 4.846 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.560 4.718 mp_allgather_i34 2485 14.5 1.546 4.689 1.546 4.689 mp_sum_l 7884 12.9 3.159 4.687 3.159 4.687 potential_pw2rs 127 12.3 0.021 0.024 4.638 4.652 mp_waitany 11660 13.9 2.330 4.535 2.330 4.535 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.346 1.923 4.147 transfer_fm_to_dbcsr 11 9.9 0.017 0.021 2.350 4.072 mp_alltoall_i22 712 14.1 1.978 3.932 1.978 3.932 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.878 3.908 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 3.782 3.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.615 3.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.655000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=901.000000, yerr=9.844242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.150939E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509632. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.062 192.237 192.240 qs_mol_dyn_low 1 2.0 0.003 0.003 191.695 191.718 qs_forces 11 3.9 0.013 0.015 191.588 191.593 qs_energies 11 4.9 0.018 0.079 184.332 184.345 scf_env_do_scf 11 5.9 0.002 0.018 166.686 166.700 velocity_verlet 10 3.0 0.002 0.002 126.456 126.459 scf_env_do_scf_inner_loop 118 6.6 0.011 0.074 119.638 119.639 dbcsr_multiply_generic 2529 12.6 0.257 0.265 82.122 82.947 qs_scf_new_mos 118 7.6 0.001 0.001 81.592 81.900 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.591 81.900 ot_scf_mini 118 9.6 0.003 0.004 76.963 77.272 multiply_cannon 2529 13.6 0.558 0.588 57.072 60.133 multiply_cannon_loop 2529 14.6 1.823 1.929 53.299 55.266 init_scf_loop 11 6.9 0.023 0.185 46.926 46.935 ot_mini 118 10.6 0.001 0.001 43.726 44.036 prepare_preconditioner 11 7.9 0.001 0.007 42.342 42.556 make_preconditioner 11 8.9 0.018 0.143 42.341 42.556 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.988 41.212 multiply_cannon_multrec 30348 15.6 13.772 18.843 27.267 32.061 rebuild_ks_matrix 129 8.3 0.001 0.001 30.589 30.925 qs_ks_build_kohn_sham_matrix 129 9.3 0.209 0.245 30.589 30.925 mp_waitall_1 149172 16.7 17.413 28.454 17.413 28.454 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.652 27.938 qs_ot_get_derivative 118 11.6 0.002 0.002 23.861 24.169 make_m2s 5058 13.6 0.098 0.104 20.513 21.717 make_images 5058 14.6 1.963 2.282 20.203 21.416 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.203 19.774 apply_single 129 13.6 0.001 0.001 19.203 19.774 ot_diis_step 118 11.6 0.018 0.019 19.727 19.729 qs_ot_get_p 129 10.4 0.001 0.001 19.246 19.604 cp_fm_upper_to_full 106 14.8 11.446 16.853 11.446 16.853 cp_fm_cholesky_invert 11 10.9 16.079 16.088 16.079 16.088 multiply_cannon_metrocomm3 30348 15.6 0.050 0.053 6.446 15.421 qs_ot_p2m_diag 84 11.4 0.347 0.395 15.054 15.109 sum_up_and_integrate 129 10.3 0.142 0.153 14.848 14.875 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.730 14.784 calculate_rho_elec 129 8.7 0.177 0.194 14.729 14.783 integrate_v_rspace 129 11.3 0.004 0.004 14.706 14.738 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.643 13.644 dbcsr_mm_accdrv_process 62780 16.2 8.452 9.338 13.065 13.593 make_images_data 5058 15.6 0.066 0.072 11.096 12.991 dbcsr_complete_redistribute 397 12.7 1.567 1.680 9.108 12.827 hybrid_alltoall_any 5245 16.5 0.536 2.246 9.747 11.986 multiply_cannon_sync_h2d 30348 15.6 10.639 11.758 10.639 11.758 init_scf_run 11 5.9 0.000 0.001 11.648 11.649 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.647 11.649 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.748 11.480 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.380 10.400 cp_fm_diag_elpa_base 84 14.4 9.409 9.744 10.373 10.392 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.898 10.126 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.312 9.944 mp_alltoall_i22 720 14.1 5.636 9.312 5.636 9.312 pw_transfer 1559 11.6 0.087 0.101 9.218 9.304 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.989 9.083 density_rs2pw 129 9.7 0.006 0.007 7.049 8.625 fft_wrap_pw1pw2_140 527 13.2 1.573 1.613 7.995 8.091 grid_integrate_task_list 129 12.3 7.566 8.087 7.566 8.087 wfi_extrapolate 11 7.9 0.001 0.001 7.761 7.761 cp_fm_cholesky_decompose 22 10.9 7.510 7.608 7.510 7.608 multiply_cannon_metrocomm4 25290 15.6 0.085 0.096 2.829 7.342 mp_irecv_dv 76751 16.2 2.672 7.059 2.672 7.059 calculate_dm_sparse 129 9.5 0.001 0.001 6.357 6.441 rs_pw_transfer 1054 12.0 0.013 0.015 4.675 6.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.069 6.125 grid_collocate_task_list 129 9.7 5.183 6.043 5.183 6.043 fft3d_ps 1301 14.7 2.860 2.955 5.891 5.978 mp_alltoall_d11v 2429 14.1 4.962 5.510 4.962 5.510 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.940 5.041 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.794 4.898 potential_pw2rs 129 12.3 0.023 0.024 4.556 4.581 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 4.468 4.469 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.273 4.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.240000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.909091, yerr=17.911167 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.536406E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.073 170.840 170.849 qs_mol_dyn_low 1 2.0 0.003 0.003 169.986 170.010 qs_forces 11 3.9 0.004 0.005 169.880 169.883 qs_energies 11 4.9 0.001 0.002 162.306 162.315 scf_env_do_scf 11 5.9 0.001 0.001 144.664 144.667 velocity_verlet 10 3.0 0.002 0.002 111.760 111.763 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.411 109.413 dbcsr_multiply_generic 2507 12.6 0.181 0.194 72.305 72.836 qs_scf_new_mos 117 7.6 0.001 0.001 70.989 71.084 qs_scf_loop_do_ot 117 8.6 0.001 0.001 70.988 71.084 ot_scf_mini 117 9.6 0.003 0.004 66.585 66.677 multiply_cannon 2507 13.6 0.561 0.582 54.008 56.863 multiply_cannon_loop 2507 14.6 0.810 0.848 50.882 51.952 ot_mini 117 10.6 0.001 0.001 37.587 37.672 init_scf_loop 11 6.9 0.000 0.000 35.107 35.109 prepare_preconditioner 11 7.9 0.000 0.000 31.273 31.301 make_preconditioner 11 8.9 0.000 0.000 31.273 31.301 mp_waitall_1 125778 16.7 24.619 31.124 24.619 31.124 rebuild_ks_matrix 128 8.3 0.001 0.001 29.733 29.851 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.018 29.733 29.850 make_full_inverse_cholesky 11 9.9 0.016 0.026 29.218 29.476 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.971 27.075 multiply_cannon_multrec 10028 15.6 10.344 16.058 18.004 22.298 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 13.011 20.523 qs_ot_get_derivative 117 11.6 0.002 0.002 20.304 20.393 cp_fm_cholesky_invert 11 10.9 17.843 17.848 17.843 17.848 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.322 17.566 apply_single 128 13.6 0.001 0.001 17.322 17.565 ot_diis_step 117 11.6 0.020 0.020 17.214 17.215 qs_ot_get_p 128 10.4 0.001 0.001 16.304 16.432 make_m2s 5014 13.6 0.066 0.073 14.524 15.489 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.343 15.374 calculate_rho_elec 128 8.7 0.259 0.269 15.342 15.373 make_images 5014 14.6 2.193 2.627 14.220 15.181 sum_up_and_integrate 128 10.3 0.180 0.189 14.900 14.942 integrate_v_rspace 128 11.3 0.004 0.004 14.720 14.770 qs_ot_p2m_diag 83 11.4 0.496 0.502 12.638 12.656 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.492 11.493 multiply_cannon_sync_h2d 10028 15.6 10.738 11.193 10.738 11.193 init_scf_run 11 5.9 0.000 0.001 10.849 10.849 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.849 10.849 make_images_data 5014 15.6 0.056 0.065 8.290 9.732 pw_transfer 1547 11.6 0.086 0.094 9.398 9.426 hybrid_alltoall_any 5200 16.5 0.846 3.795 8.172 9.322 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.175 9.208 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.674 8.686 cp_fm_diag_elpa_base 83 14.4 8.440 8.499 8.668 8.679 grid_integrate_task_list 128 12.3 7.736 8.184 7.736 8.184 fft_wrap_pw1pw2_140 523 13.2 1.916 1.949 8.143 8.173 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.948 8.011 dbcsr_mm_accdrv_process 20762 16.1 3.168 4.456 7.312 7.974 cp_fm_cholesky_decompose 22 10.9 7.787 7.916 7.787 7.916 density_rs2pw 128 9.7 0.006 0.006 6.825 7.741 wfi_extrapolate 11 7.9 0.001 0.001 7.331 7.331 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.125 7.250 calculate_dm_sparse 128 9.5 0.001 0.001 6.408 6.488 grid_collocate_task_list 128 9.7 5.470 6.166 5.470 6.166 multiply_cannon_metrocomm4 7521 15.6 0.026 0.030 1.867 5.856 mp_irecv_dv 28860 15.9 1.827 5.763 1.827 5.763 mp_alltoall_d11v 2415 14.1 4.835 5.763 4.835 5.763 fft3d_ps 1291 14.7 2.765 2.845 5.650 5.684 dbcsr_complete_redistribute 395 12.7 2.111 2.182 5.064 5.459 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.281 5.281 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.204 5.216 rs_pw_transfer 1046 11.9 0.012 0.013 4.052 4.981 mp_allgather_i34 2507 14.6 1.297 4.518 1.297 4.518 potential_pw2rs 128 12.3 0.027 0.028 4.336 4.349 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.150 4.207 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.581 3.876 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.634 3.681 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.658 3.672 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.573 3.656 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.269 3.585 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.382 3.479 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.849000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1448.181818, yerr=23.032855 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.151843E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3943 59990. MP_Allreduce 10935 1507. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.053 295.291 295.295 qs_mol_dyn_low 1 2.0 0.003 0.003 294.740 294.764 qs_forces 11 3.9 0.005 0.005 294.638 294.642 qs_energies 11 4.9 0.007 0.014 285.136 285.144 scf_env_do_scf 11 5.9 0.001 0.001 261.943 261.952 velocity_verlet 10 3.0 0.011 0.012 212.075 212.085 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 138.151 138.153 init_scf_loop 11 6.9 0.000 0.000 123.522 123.528 prepare_preconditioner 11 7.9 0.000 0.000 118.268 118.290 make_preconditioner 11 8.9 0.000 0.000 118.268 118.290 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.270 115.392 qs_scf_new_mos 116 7.6 0.001 0.001 88.725 88.784 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.725 88.784 ot_scf_mini 116 9.6 0.003 0.004 83.908 83.937 dbcsr_multiply_generic 2485 12.5 0.211 0.223 81.952 82.411 cp_fm_upper_to_full 104 14.8 52.042 74.907 52.042 74.907 multiply_cannon 2485 13.5 0.673 0.717 58.620 59.970 multiply_cannon_loop 2485 14.5 1.011 1.030 54.626 55.896 ot_mini 116 10.6 0.001 0.001 44.311 44.355 dbcsr_complete_redistribute 393 12.7 4.014 4.074 29.945 42.945 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.419 39.399 rebuild_ks_matrix 127 8.3 0.001 0.001 37.597 37.624 qs_ks_build_kohn_sham_matrix 127 9.3 0.210 0.210 37.597 37.624 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.954 36.771 mp_alltoall_i22 712 14.1 21.713 34.808 21.713 34.808 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.477 34.519 cp_fm_cholesky_invert 11 10.9 33.141 33.147 33.141 33.147 mp_waitall_1 102768 16.8 28.039 32.150 28.039 32.150 qs_ot_get_p 127 10.4 0.001 0.001 24.894 24.970 qs_ot_get_derivative 116 11.6 0.002 0.002 24.513 24.541 qs_ot_p2m_diag 82 11.4 0.869 0.875 20.898 20.930 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.298 20.304 calculate_rho_elec 127 8.7 0.479 0.479 20.298 20.304 make_m2s 4970 13.5 0.075 0.079 18.767 20.213 multiply_cannon_metrocomm3 9940 15.5 0.024 0.027 18.574 19.834 ot_diis_step 116 11.6 0.022 0.022 19.761 19.762 make_images 4970 14.5 3.048 3.248 18.292 19.740 sum_up_and_integrate 127 10.3 0.319 0.322 19.409 19.489 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.114 19.237 apply_single 127 13.6 0.001 0.001 19.113 19.237 integrate_v_rspace 127 11.3 0.004 0.004 19.090 19.171 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.140 19.142 multiply_cannon_multrec 9940 15.5 10.282 12.085 18.179 18.264 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.034 16.035 cp_fm_diag_elpa_base 82 14.4 11.757 13.304 16.031 16.031 multiply_cannon_sync_h2d 9940 15.5 14.242 14.258 14.242 14.258 hybrid_alltoall_any 5155 16.4 1.300 3.015 10.896 13.006 make_images_data 4970 15.5 0.063 0.067 10.712 12.899 init_scf_run 11 5.9 0.000 0.001 12.676 12.678 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.676 12.678 pw_transfer 1535 11.6 0.092 0.093 11.358 11.363 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.125 11.132 fft_wrap_pw1pw2_140 519 13.2 3.097 3.152 9.925 9.933 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.558 9.578 dbcsr_mm_accdrv_process 20590 16.0 4.201 6.188 7.651 9.564 wfi_extrapolate 11 7.9 0.001 0.001 9.175 9.176 cp_fm_cholesky_decompose 22 10.9 8.799 8.814 8.799 8.814 grid_integrate_task_list 127 12.3 8.526 8.756 8.526 8.756 mp_alltoall_d11v 2401 14.1 7.922 8.582 7.922 8.582 qs_energies_init_hamiltonians 11 5.9 0.159 0.167 8.328 8.329 density_rs2pw 127 9.7 0.005 0.005 7.904 8.045 calculate_dm_sparse 127 9.5 0.001 0.001 6.849 6.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.468 6.500 grid_collocate_task_list 127 9.7 6.385 6.473 6.385 6.473 fft3d_ps 1281 14.7 2.831 2.838 6.197 6.265 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.172 6.229 rs_scatter_matrices 138 9.7 3.944 4.832 5.992 6.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.295000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2820.363636, yerr=164.420897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 1.259614E+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 37157619. 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 0.030 0.043 85.456 85.458 qs_energies 1 2.0 0.000 0.000 84.864 84.874 ls_scf 1 3.0 0.000 0.001 83.891 83.902 dbcsr_multiply_generic 111 6.7 0.015 0.018 72.811 72.948 multiply_cannon 111 7.7 0.017 0.020 55.993 57.158 multiply_cannon_loop 111 8.7 0.227 0.243 52.547 53.717 ls_scf_main 1 4.0 0.000 0.000 52.442 52.460 density_matrix_trs4 2 5.0 0.002 0.003 46.889 46.988 ls_scf_init_scf 1 4.0 0.000 0.001 28.406 28.416 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.256 27.297 mp_waitall_1 11031 10.9 22.556 25.382 22.556 25.382 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.083 25.104 multiply_cannon_multrec 2664 9.7 8.191 8.993 15.594 17.364 multiply_cannon_sync_h2d 2664 9.7 13.586 14.983 13.586 14.983 make_m2s 222 7.7 0.010 0.012 13.214 13.727 make_images 222 8.7 0.100 0.111 13.192 13.706 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.758 13.090 make_images_data 222 9.7 0.005 0.006 7.763 8.372 hybrid_alltoall_any 227 10.6 0.220 1.850 6.697 8.253 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.709 7.019 7.966 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.371 7.905 dbcsr_mm_accdrv_process_sort 4760 11.4 6.224 7.112 6.224 7.112 calculate_norms 4752 9.8 5.519 6.162 5.519 6.162 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.185 mp_sum_l 887 5.1 3.091 4.328 3.091 4.328 make_images_sizes 222 9.7 0.000 0.000 0.767 3.618 mp_alltoall_i44 222 10.7 0.766 3.617 0.766 3.617 arnoldi_extremal 4 6.8 0.000 0.000 3.193 3.217 arnoldi_normal_ev 4 7.8 0.001 0.003 3.193 3.217 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.037 3.212 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.285 3.209 mp_irecv_dv 6231 10.9 2.019 3.179 2.019 3.179 build_subspace 16 8.4 0.009 0.012 3.104 3.105 ls_scf_post 1 4.0 0.000 0.000 3.042 3.051 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.320 2.752 dbcsr_merge_single_wm 555 10.7 0.458 0.593 2.311 2.742 make_images_pack 222 9.7 2.214 2.630 2.216 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.311 2.552 dbcsr_sort_data 658 11.4 2.111 2.487 2.111 2.487 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.458 2.069 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.320 2.398 buffer_matrices_ensure_size 222 8.7 1.757 2.104 1.757 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.796 1.812 rebuild_ks_matrix 3 7.3 0.000 0.000 1.787 1.803 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.023 1.787 1.803 mp_allgather_i34 111 8.7 0.754 1.713 0.754 1.713 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.458000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 2.172207E+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 0.058 0.098 91.056 91.058 qs_energies 1 2.0 0.000 0.000 90.078 90.092 ls_scf 1 3.0 0.002 0.011 88.736 88.749 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.789 75.052 multiply_cannon 111 7.7 0.027 0.039 52.879 57.634 ls_scf_main 1 4.0 0.004 0.045 55.068 55.071 multiply_cannon_loop 111 8.7 0.135 0.145 50.056 53.379 density_matrix_trs4 2 5.0 0.019 0.078 49.291 49.549 ls_scf_init_scf 1 4.0 0.000 0.001 29.994 29.996 mp_waitall_1 9105 10.9 20.895 29.640 20.895 29.640 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.668 28.764 multiply_cannon_multrec 1332 9.7 13.405 16.983 22.731 27.361 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.287 26.301 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.483 20.661 make_m2s 222 7.7 0.007 0.009 15.204 15.798 make_images 222 8.7 1.375 1.700 15.173 15.768 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.540 8.919 10.505 dbcsr_mm_accdrv_process_sort 4041 11.4 8.437 9.965 8.437 9.965 make_images_data 222 9.7 0.005 0.005 8.838 9.745 hybrid_alltoall_any 227 10.6 0.543 2.539 8.234 9.589 mp_sum_l 887 5.1 5.256 8.696 5.256 8.696 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.207 7.823 mp_irecv_dv 3311 11.0 3.186 7.768 3.186 7.768 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.907 6.911 calculate_norms 2376 9.8 6.070 6.707 6.070 6.707 multiply_cannon_sync_h2d 1332 9.7 4.847 6.378 4.847 6.378 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.246 arnoldi_extremal 4 6.8 0.000 0.000 4.610 4.630 arnoldi_normal_ev 4 7.8 0.001 0.007 4.610 4.630 build_subspace 16 8.4 0.014 0.021 4.352 4.355 ls_scf_post 1 4.0 0.002 0.009 3.671 3.682 ls_scf_store_result 1 5.0 0.000 0.000 3.432 3.484 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.127 3.363 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.225 2.741 3.227 mp_allgather_i34 111 8.7 0.803 3.044 0.803 3.044 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.225 2.746 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.629 2.717 dbcsr_data_new 4174 10.1 2.116 2.404 2.116 2.404 make_images_pack 222 9.7 1.820 2.138 1.822 2.140 dbcsr_sort_data 436 11.2 1.823 2.074 1.823 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.987 1.989 rebuild_ks_matrix 3 7.3 0.000 0.000 1.974 1.976 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.020 1.974 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.058000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1773.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 2.929631E+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 0.034 0.051 92.872 92.875 qs_energies 1 2.0 0.000 0.000 91.962 91.966 ls_scf 1 3.0 0.000 0.000 90.486 90.490 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.944 75.250 multiply_cannon 111 7.7 0.030 0.038 51.934 56.398 ls_scf_main 1 4.0 0.000 0.000 56.170 56.174 multiply_cannon_loop 111 8.7 0.116 0.128 48.730 52.478 density_matrix_trs4 2 5.0 0.002 0.003 50.176 50.357 mp_waitall_1 7281 11.0 22.937 32.454 22.937 32.454 ls_scf_init_scf 1 4.0 0.000 0.000 30.632 30.634 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.417 29.479 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.071 27.088 multiply_cannon_multrec 888 9.7 12.823 15.614 21.431 25.022 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.407 21.994 make_m2s 222 7.7 0.007 0.008 16.460 17.149 make_images 222 8.7 1.591 1.875 16.422 17.112 make_images_data 222 9.7 0.004 0.005 9.594 10.537 hybrid_alltoall_any 227 10.6 0.645 2.962 9.099 10.219 dbcsr_mm_accdrv_process 3754 10.4 0.311 0.487 8.123 9.349 dbcsr_mm_accdrv_process_sort 3754 11.4 7.689 8.861 7.689 8.861 mp_sum_l 887 5.1 4.842 7.785 4.842 7.785 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.450 7.217 mp_irecv_dv 2335 11.1 2.435 7.168 2.435 7.168 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.584 7.124 multiply_cannon_sync_h2d 888 9.7 5.987 7.048 5.987 7.048 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.644 6.020 arnoldi_extremal 4 6.8 0.000 0.000 5.191 5.209 arnoldi_normal_ev 4 7.8 0.001 0.005 5.191 5.209 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.818 5.044 build_subspace 16 8.4 0.014 0.020 4.885 4.892 calculate_norms 1584 9.8 4.408 4.887 4.408 4.887 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.491 3.849 mp_allgather_i34 111 8.7 0.838 3.807 0.838 3.807 ls_scf_post 1 4.0 0.000 0.000 3.685 3.689 dbcsr_matrix_vector_mult_local 304 10.0 3.020 3.594 3.022 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.432 3.507 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.897 3.008 dbcsr_data_new 4116 9.9 2.111 2.438 2.111 2.438 dbcsr_sort_data 325 11.1 1.894 2.209 1.894 2.209 make_images_sizes 222 9.7 0.000 0.000 1.008 2.146 mp_alltoall_i44 222 10.7 1.008 2.146 1.008 2.146 dbcsr_finalize 304 7.8 0.027 0.032 1.621 1.953 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.922 1.923 rebuild_ks_matrix 3 7.3 0.000 0.000 1.904 1.906 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.904 1.905 make_images_pack 222 9.7 1.624 1.883 1.627 1.886 dbcsr_merge_all 275 8.9 0.242 0.305 1.538 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.875000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2244.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 3.323519E+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 0.033 0.051 97.929 97.931 qs_energies 1 2.0 0.000 0.000 96.942 96.961 ls_scf 1 3.0 0.000 0.000 95.273 95.291 dbcsr_multiply_generic 111 6.7 0.017 0.021 78.642 78.938 ls_scf_main 1 4.0 0.000 0.000 58.967 58.968 multiply_cannon 111 7.7 0.040 0.089 51.587 56.200 density_matrix_trs4 2 5.0 0.002 0.003 52.753 52.899 multiply_cannon_loop 111 8.7 0.153 0.166 46.519 50.039 ls_scf_init_scf 1 4.0 0.000 0.000 32.814 32.817 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.582 31.652 mp_waitall_1 6369 11.0 22.665 29.513 22.665 29.513 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.104 29.115 multiply_cannon_multrec 1332 9.7 14.115 17.627 22.089 25.375 make_m2s 222 7.7 0.008 0.009 21.268 22.686 make_images 222 8.7 3.144 3.629 21.217 22.637 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.042 16.998 make_images_data 222 9.7 0.004 0.005 11.923 13.608 hybrid_alltoall_any 227 10.6 0.800 3.770 11.240 13.029 dbcsr_mm_accdrv_process 3641 10.4 0.311 0.504 7.604 9.114 dbcsr_mm_accdrv_process_sort 3641 11.4 7.154 8.617 7.154 8.617 mp_sum_l 887 5.1 4.199 7.668 4.199 7.668 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.104 6.169 mp_irecv_dv 3229 10.9 2.078 6.075 2.078 6.075 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.267 5.964 multiply_cannon_sync_h2d 1332 9.7 5.445 5.959 5.445 5.959 arnoldi_extremal 4 6.8 0.000 0.000 5.252 5.266 arnoldi_normal_ev 4 7.8 0.001 0.005 5.252 5.266 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.536 5.121 build_subspace 16 8.4 0.014 0.021 4.912 4.918 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.589 4.757 mp_allgather_i34 111 8.7 2.173 4.556 2.173 4.556 calculate_norms 2376 9.8 4.193 4.521 4.193 4.521 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.611 3.888 dbcsr_matrix_vector_mult_local 304 10.0 3.186 3.681 3.188 3.682 ls_scf_post 1 4.0 0.000 0.000 3.491 3.510 dbcsr_sort_data 658 11.4 3.081 3.346 3.081 3.346 dbcsr_special_finalize 555 9.7 0.006 0.007 2.831 3.170 dbcsr_merge_single_wm 555 10.7 0.538 0.656 2.823 3.161 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.052 3.124 ls_scf_store_result 1 5.0 0.000 0.000 2.950 3.015 dbcsr_data_release 10477 10.7 1.574 2.431 1.574 2.431 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.931000, 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/aa921764592e607de23a23fbfafae42bb144baa4_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 4.733428E+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 0.040 0.057 92.773 92.775 qs_energies 1 2.0 0.000 0.000 91.955 91.958 ls_scf 1 3.0 0.000 0.000 89.996 89.999 dbcsr_multiply_generic 111 6.7 0.018 0.019 71.189 71.401 ls_scf_main 1 4.0 0.000 0.000 56.567 56.569 multiply_cannon 111 7.7 0.081 0.152 52.692 56.375 multiply_cannon_loop 111 8.7 0.090 0.099 50.136 51.558 density_matrix_trs4 2 5.0 0.002 0.003 49.543 49.614 ls_scf_init_scf 1 4.0 0.000 0.000 30.028 30.029 mp_waitall_1 5436 11.0 24.450 28.961 24.450 28.961 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.565 28.599 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.421 26.431 multiply_cannon_multrec 444 9.7 13.729 16.231 20.907 23.943 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.952 15.200 make_m2s 222 7.7 0.005 0.006 13.678 14.603 make_images 222 8.7 2.052 2.494 13.609 14.532 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.940 13.572 hybrid_alltoall_any 227 10.6 0.801 3.836 8.164 9.813 make_images_data 222 9.7 0.003 0.004 8.377 9.654 multiply_cannon_sync_h2d 444 9.7 6.712 8.285 6.712 8.285 dbcsr_mm_accdrv_process 3003 10.4 0.384 0.760 6.876 8.130 dbcsr_mm_accdrv_process_sort 3003 11.4 6.424 7.497 6.424 7.497 arnoldi_extremal 4 6.8 0.000 0.000 5.866 5.877 arnoldi_normal_ev 4 7.8 0.002 0.005 5.866 5.877 build_subspace 16 8.4 0.015 0.019 5.473 5.484 mp_sum_l 887 5.1 2.814 5.176 2.814 5.176 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.394 4.572 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.172 4.384 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.021 4.218 dbcsr_matrix_vector_mult_local 304 10.0 3.685 4.157 3.687 4.160 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.510 3.940 mp_irecv_dv 1241 11.2 1.498 3.910 1.498 3.910 calculate_norms 792 9.8 3.624 3.741 3.624 3.741 mp_allgather_i34 111 8.7 1.122 3.631 1.122 3.631 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.586 3.631 ls_scf_post 1 4.0 0.000 0.000 3.400 3.402 make_images_sizes 222 9.7 0.000 0.000 0.892 3.388 mp_alltoall_i44 222 10.7 0.892 3.387 0.892 3.387 ls_scf_store_result 1 5.0 0.000 0.000 3.189 3.218 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.320 dbcsr_data_new 4608 9.7 1.790 2.235 1.790 2.235 dbcsr_merge_all 275 8.9 0.480 0.525 2.057 2.159 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.090 2.091 rebuild_ks_matrix 3 7.3 0.000 0.000 2.057 2.058 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.003 2.057 2.058 qs_energies_init_hamiltonians 1 3.0 0.005 0.019 1.944 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.775000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3768.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 8.899891E+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 0.486 0.532 109.680 109.692 qs_energies 1 2.0 0.000 0.000 107.586 107.589 ls_scf 1 3.0 0.000 0.000 104.643 104.646 dbcsr_multiply_generic 111 6.7 0.024 0.028 78.292 78.422 ls_scf_main 1 4.0 0.000 0.000 63.811 63.812 density_matrix_trs4 2 5.0 0.002 0.003 54.772 54.853 multiply_cannon 111 7.7 0.125 0.163 51.853 53.689 multiply_cannon_loop 111 8.7 0.098 0.099 48.831 49.605 ls_scf_init_scf 1 4.0 0.000 0.000 37.038 37.039 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.469 35.481 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.513 32.524 mp_waitall_1 4527 11.1 22.302 25.899 22.302 25.899 make_m2s 222 7.7 0.006 0.006 22.616 23.712 make_images 222 8.7 3.581 3.871 22.506 23.600 multiply_cannon_multrec 444 9.7 17.909 18.501 22.598 23.191 hybrid_alltoall_any 227 10.6 1.654 3.614 12.732 15.665 make_images_data 222 9.7 0.004 0.004 12.988 15.126 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.722 14.178 multiply_cannon_sync_h2d 444 9.7 8.794 8.844 8.794 8.844 arnoldi_extremal 4 6.8 0.000 0.000 7.290 7.307 arnoldi_normal_ev 4 7.8 0.003 0.009 7.290 7.307 build_subspace 16 8.4 0.026 0.036 6.725 6.740 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.397 5.553 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.097 5.358 dbcsr_matrix_vector_mult_local 304 10.0 5.017 5.300 5.019 5.303 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.164 5.261 dbcsr_mm_accdrv_process 1814 10.4 0.198 0.354 4.498 4.615 dbcsr_mm_accdrv_process_sort 1814 11.4 4.130 4.256 4.130 4.256 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.898 3.905 ls_scf_post 1 4.0 0.000 0.000 3.793 3.796 make_images_sizes 222 9.7 0.000 0.000 1.478 3.706 mp_alltoall_i44 222 10.7 1.477 3.705 1.477 3.705 mp_allgather_i34 111 8.7 1.108 3.615 1.108 3.615 ls_scf_store_result 1 5.0 0.000 0.000 3.503 3.538 acc_transpose_blocks 444 9.7 0.003 0.003 3.000 3.401 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 2.917 3.316 jit_kernel_transpose 1 13.0 2.911 3.310 2.911 3.310 calculate_norms 792 9.8 3.228 3.281 3.228 3.281 dbcsr_finalize 304 7.8 0.082 0.089 3.083 3.167 dbcsr_merge_all 275 8.9 0.884 0.914 2.866 2.945 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.913 2.913 dbcsr_complete_redistribute 5 7.6 1.428 1.468 2.742 2.859 matrix_ls_to_qs 2 6.0 0.000 0.000 2.410 2.532 dbcsr_sort_data 325 11.1 2.439 2.497 2.439 2.497 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.471 2.472 rebuild_ks_matrix 3 7.3 0.000 0.000 2.405 2.406 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.405 2.406 dbcsr_new_transposed 4 7.5 0.253 0.278 2.321 2.338 dbcsr_data_new 6591 9.6 1.868 2.311 1.868 2.311 dbcsr_frobenius_norm 74 6.6 2.057 2.135 2.190 2.231 mp_sum_l 887 5.1 1.421 2.227 1.421 2.227 dbcsr_add_d 103 6.2 0.000 0.000 2.127 2.203 dbcsr_add_anytype 103 7.2 0.858 0.890 2.127 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.692000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7083.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aa921764592e607de23a23fbfafae42bb144baa4_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 589.627392E+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 992158847. 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 0.049 0.148 229.262 229.264 qs_mol_dyn_low 1 2.0 0.004 0.053 228.131 228.163 qs_forces 5 3.8 0.005 0.030 227.995 227.997 qs_energies 5 4.8 0.004 0.059 224.998 225.021 scf_env_do_scf 5 5.8 0.002 0.056 210.218 210.221 scf_env_do_scf_inner_loop 105 6.6 0.008 0.192 182.781 182.963 qs_scf_new_mos 105 7.6 0.001 0.001 141.241 141.428 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.240 141.428 dbcsr_multiply_generic 1445 12.2 0.129 0.137 133.239 133.696 ot_scf_mini 105 9.6 0.003 0.004 131.437 131.576 multiply_cannon 1445 13.2 0.274 0.283 114.158 116.309 multiply_cannon_loop 1445 14.2 2.825 2.975 112.556 113.763 velocity_verlet 4 3.0 0.001 0.011 107.364 107.364 ot_mini 105 10.6 0.001 0.002 59.939 60.033 multiply_cannon_multrec 69360 15.2 29.804 34.840 39.672 45.008 qs_ot_get_p 112 10.4 0.001 0.003 41.470 41.734 mp_waitall_1 488190 16.1 34.574 41.437 34.574 41.437 qs_ot_get_derivative 55 11.6 0.001 0.001 38.226 38.324 multiply_cannon_sync_h2d 69360 15.2 29.244 33.071 29.244 33.071 multiply_cannon_metrocomm3 69360 15.2 0.203 0.215 25.536 32.444 qs_ot_p2m_diag 40 11.0 0.020 0.031 30.475 30.614 rebuild_ks_matrix 110 8.4 0.000 0.001 29.685 29.886 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.018 29.685 29.886 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.294 27.473 init_scf_loop 7 6.6 0.000 0.005 27.395 27.404 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.187 27.188 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.002 23.256 apply_single 62 13.6 0.000 0.000 23.002 23.256 cp_fm_syevd 40 13.0 0.000 0.001 22.141 22.289 prepare_preconditioner 7 7.6 0.000 0.000 22.241 22.270 make_preconditioner 7 8.6 0.000 0.002 22.241 22.270 ot_new_cg_direction 55 11.6 0.001 0.007 20.991 20.992 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.648 18.040 calculate_rho_elec 110 8.6 0.030 0.034 17.648 18.039 cp_fm_redistribute_end 40 14.0 8.757 17.475 8.763 17.476 cp_fm_syevd_base 40 14.0 8.705 17.422 8.705 17.422 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.140 15.236 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.822 15.003 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.812 13.906 rs_pw_transfer 690 11.5 0.011 0.013 11.802 12.976 mp_sum_l 4764 12.2 12.207 12.896 12.207 12.896 density_rs2pw 110 9.6 0.006 0.007 11.681 12.879 pw_transfer 1645 12.4 0.082 0.100 12.642 12.815 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.501 12.677 init_scf_run 5 5.8 0.000 0.002 11.993 11.994 scf_env_initial_rho_setup 5 6.8 0.002 0.003 11.992 11.994 calculate_dm_sparse 110 9.5 0.000 0.001 11.402 11.588 fft_wrap_pw1pw2_240 915 15.0 1.193 1.269 10.715 10.923 qs_vxc_create 110 10.4 0.002 0.009 10.730 10.764 dbcsr_mm_accdrv_process 154766 15.8 6.164 6.363 9.735 10.699 cp_fm_cholesky_invert 7 10.6 10.671 10.678 10.671 10.678 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.336 10.400 check_diag 80 13.5 8.598 8.888 9.365 9.514 fft3d_pb 915 16.0 2.374 2.552 8.783 8.952 sum_up_and_integrate 60 10.3 0.028 0.032 8.903 8.915 integrate_v_rspace 60 11.3 0.001 0.002 8.875 8.887 acc_transpose_blocks 69360 15.2 0.354 0.372 7.649 8.173 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.545 7.788 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.577 7.594 xc_vxc_pw_create 60 11.3 0.039 0.049 7.200 7.234 calculate_first_density_matrix 1 7.0 0.000 0.003 7.157 7.168 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.108 7.167 make_m2s 2890 13.2 0.082 0.093 6.405 6.967 make_images 2890 14.2 0.239 0.261 6.296 6.857 multiply_cannon_metrocomm1 69360 15.2 0.096 0.103 4.545 6.826 make_full_single_inverse 7 9.6 0.001 0.005 6.791 6.821 xc_pw_derive 510 13.4 0.005 0.006 6.498 6.570 mp_alltoall_z22v 2340 17.7 5.863 6.186 5.863 6.186 mp_waitany 7680 13.5 4.636 5.977 4.636 5.977 acc_transpose_blocks_kernels 69360 16.2 0.845 0.894 4.910 5.298 potential_pw2rs 60 12.3 0.003 0.003 5.073 5.098 multiply_cannon_metrocomm4 67915 15.2 0.186 0.200 2.010 4.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.264000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.800000, yerr=2.400000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: aa921764592e607de23a23fbfafae42bb144baa4 Summary: empty Status: OK