=== 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: 9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # 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: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/01 job id: 49325537 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/02 job id: 49325538 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/03 job id: 49325539 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/04 job id: 49325540 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/05 job id: 49325541 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/06 job id: 49325542 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/07 job id: 49325543 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/08 job id: 49325544 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/09 job id: 49325545 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/10 job id: 49325546 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/11 job id: 49325547 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/12 job id: 49325548 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/13 job id: 49325549 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/14 job id: 49325550 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/15 job id: 49325551 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/16 job id: 49325552 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/17 job id: 49325553 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/18 job id: 49325554 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/19 job id: 49325555 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/20 job id: 49325556 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/21 job id: 49325557 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/22 job id: 49325558 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/23 job id: 49325559 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/24 job id: 49325560 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/25 job id: 49325561 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/26 job id: 49325562 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/27 job id: 49325563 --- 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/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 0.019 0.037 134.966 134.967 farming_run 1 2.0 134.247 134.250 134.915 134.920 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489465E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.025 115.234 115.235 qs_energies 1 2.0 0.000 0.000 115.033 115.036 mp2_main 1 3.0 0.000 0.000 112.991 112.994 mp2_gpw_main 1 4.0 0.027 0.033 112.001 112.004 mp2_ri_gpw_compute_in 1 5.0 0.173 0.174 92.676 93.022 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.681 55.029 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 41.080 46.124 get_2c_integrals 1 6.0 0.008 0.009 37.117 37.821 integrate_v_rspace 273 8.0 0.436 0.453 24.682 29.533 pw_transfer 6555 10.6 0.378 0.410 26.873 27.253 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.503 25.953 grid_integrate_task_list 273 9.0 20.582 25.920 20.582 25.920 fft_wrap_pw1pw2_100 2178 12.4 1.146 1.282 23.070 23.498 rpa_ri_compute_en 1 5.0 0.296 0.343 19.209 19.447 compute_2c_integrals 1 7.0 0.002 0.003 19.309 19.310 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.613 19.046 mp2_eri_2c_integrate_gpw 1 9.0 2.372 2.416 18.610 19.044 cp_fm_cholesky_decompose 12 8.2 17.734 18.427 17.734 18.427 cholesky_decomp 1 7.0 0.000 0.000 16.656 17.349 fft3d_s 5443 13.4 16.155 16.457 16.177 16.479 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.306 10.764 15.306 calculate_wavefunction 272 8.0 5.382 5.544 12.316 12.920 rpa_num_int 1 6.0 0.000 0.002 10.643 10.645 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.601 10.602 calc_mat_Q 8 8.0 0.000 0.000 9.472 9.557 contract_S_to_Q 8 9.0 0.000 0.000 8.893 8.981 calc_potential_gpw 544 9.5 0.005 0.005 8.183 8.762 parallel_gemm_fm 14 9.1 0.000 0.000 8.481 8.576 parallel_gemm_fm_cosma 14 10.1 8.481 8.576 8.481 8.576 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.084 8.403 potential_pw2rs 545 10.0 0.107 0.108 7.542 8.178 create_integ_mat 1 6.0 0.021 0.028 7.797 7.798 collocate_single_gaussian 272 10.0 0.039 0.041 7.354 7.589 array2fm 1 7.0 0.000 0.000 6.724 7.218 pw_scatter_s 2720 13.7 4.333 4.491 4.333 4.491 pw_gather_s 2722 13.2 3.471 3.712 3.471 3.712 array2fm_buffer_send 1 8.0 2.964 3.162 2.964 3.162 pw_poisson_solve 545 10.5 1.116 1.190 2.191 2.503 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.001179, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2808.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 0.024 0.038 423.486 423.488 farming_run 1 2.0 422.038 422.042 423.444 423.448 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.245766E+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 26776291. 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 0.016 0.038 217.240 217.242 qs_energies 1 2.0 0.000 0.000 216.981 216.994 scf_env_do_scf 1 3.0 0.000 0.000 114.746 114.746 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.757 113.764 rebuild_ks_matrix 4 6.0 0.000 0.000 113.756 113.762 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 113.756 113.762 hfx_ks_matrix 4 8.0 0.001 0.001 113.406 113.411 integrate_four_center 4 9.0 0.153 0.468 113.405 113.410 integrate_four_center_main 4 10.0 0.114 0.706 102.230 104.742 integrate_four_center_bin 263 11.0 102.116 104.599 102.116 104.599 mp2_main 1 3.0 0.000 0.000 101.944 101.956 mp2_gpw_main 1 4.0 0.053 0.083 101.061 101.077 init_scf_loop 1 4.0 0.000 0.000 96.523 96.524 mp2_ri_gpw_compute_in 1 5.0 0.069 0.086 74.051 75.093 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.837 54.879 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 41.495 46.612 integrate_v_rspace 95 8.0 0.394 0.560 27.955 32.890 pw_transfer 2240 10.6 0.143 0.157 29.535 29.896 mp2_ri_gpw_compute_en 1 5.0 0.060 0.073 26.828 28.935 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.527 28.901 grid_integrate_task_list 95 9.0 23.368 28.514 23.368 28.514 ao_to_mo_and_store_B_mult_1 91 7.0 10.662 27.107 10.662 27.107 fft_wrap_pw1pw2_100 730 12.4 1.305 1.489 26.270 26.616 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.899 25.111 25.121 get_2c_integrals 1 6.0 0.000 0.000 20.120 20.151 compute_2c_integrals 1 7.0 0.004 0.011 19.104 19.114 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.679 18.958 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.902 18.676 18.958 fft3d_s 1823 13.4 18.503 18.829 18.517 18.843 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.221 18.221 calculate_wavefunction 91 8.0 2.019 2.040 9.630 9.817 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.579 8.805 9.528 potential_pw2rs 186 10.0 0.033 0.034 8.437 8.997 local_gemm 172 8.0 8.248 8.974 8.248 8.974 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.516 8.076 8.658 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.194 8.553 calc_potential_gpw 182 9.5 0.002 0.002 7.940 8.147 collocate_single_gaussian 91 10.0 0.017 0.024 7.778 8.021 integrate_four_center_load 4 10.0 0.000 0.000 6.782 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.782 6.786 mp_sendrecv_dm3 2068 8.0 6.113 6.673 6.113 6.673 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.427 6.338 6.427 mp_sync 37 10.5 3.128 6.142 3.128 6.142 pw_gather_s 912 13.2 4.465 4.983 4.465 4.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.060615, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1505.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 451.661824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.050 56.612 56.614 qs_mol_dyn_low 1 2.0 0.004 0.010 56.243 56.250 qs_forces 11 3.9 0.003 0.007 56.129 56.131 qs_energies 11 4.9 0.002 0.003 54.617 54.640 scf_env_do_scf 11 5.9 0.000 0.001 47.007 47.007 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.921 44.922 dbcsr_multiply_generic 2286 12.5 0.092 0.096 35.285 35.779 qs_scf_new_mos 108 7.5 0.000 0.001 34.758 35.053 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.757 35.053 ot_scf_mini 108 9.5 0.002 0.002 33.033 33.240 multiply_cannon 2286 13.5 0.198 0.212 27.593 29.106 multiply_cannon_loop 2286 14.5 1.805 1.936 26.871 28.410 velocity_verlet 10 3.0 0.001 0.002 26.917 26.918 ot_mini 108 10.5 0.001 0.008 20.323 20.581 qs_ot_get_derivative 108 11.5 0.001 0.002 17.299 17.499 mp_waitall_1 245248 16.5 9.133 16.070 9.133 16.070 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.525 14.389 multiply_cannon_multrec 54864 15.5 3.684 5.697 8.275 12.188 qs_ot_get_p 119 10.4 0.001 0.002 7.950 8.243 rebuild_ks_matrix 119 8.3 0.000 0.001 7.907 8.051 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 7.906 8.051 mp_sum_l 7287 12.8 5.605 7.365 5.605 7.365 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.967 7.099 dbcsr_mm_accdrv_process 76910 16.1 1.829 2.892 4.505 6.545 multiply_cannon_sync_h2d 54864 15.5 5.113 6.490 5.113 6.490 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.852 6.354 init_scf_run 11 5.9 0.000 0.001 6.273 6.273 scf_env_initial_rho_setup 11 6.9 0.006 0.013 6.272 6.273 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.459 5.562 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.193 5.229 sum_up_and_integrate 119 10.3 0.001 0.003 4.468 4.474 integrate_v_rspace 119 11.3 0.002 0.004 4.457 4.465 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.244 4.412 calculate_rho_elec 119 8.7 0.011 0.017 4.243 4.411 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.347 4.348 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.149 4.149 cp_fm_redistribute_end 50 14.0 2.119 4.125 2.124 4.128 cp_fm_diag_elpa_base 50 14.0 1.999 4.031 2.002 4.040 calculate_first_density_matrix 1 7.0 0.000 0.003 3.997 4.007 jit_kernel_multiply 13 15.8 2.614 3.659 2.614 3.659 cp_dbcsr_sm_fm_multiply 37 9.5 0.048 0.386 3.206 3.207 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.952 3.136 apply_single 119 13.6 0.000 0.000 2.952 3.136 calculate_dm_sparse 119 9.5 0.000 0.000 2.966 3.122 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.911 2.991 acc_transpose_blocks 54864 15.5 0.228 0.255 2.326 2.917 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.546 2.898 ot_diis_step 108 11.5 0.009 0.027 2.756 2.757 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.527 2.578 density_rs2pw 119 9.7 0.004 0.004 2.187 2.286 wfi_extrapolate 11 7.9 0.001 0.002 2.186 2.186 grid_integrate_task_list 119 12.3 2.010 2.124 2.010 2.124 init_scf_loop 11 6.9 0.000 0.001 2.068 2.068 mp_sum_d 4135 12.0 1.333 1.985 1.333 1.985 potential_pw2rs 119 12.3 0.004 0.004 1.835 1.849 make_m2s 4572 13.5 0.053 0.055 1.786 1.838 pw_transfer 1439 11.6 0.052 0.058 1.712 1.781 make_images 4572 14.5 0.134 0.140 1.705 1.754 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.636 1.707 mp_alltoall_d11v 2130 13.8 1.415 1.592 1.415 1.592 transfer_rs2pw 487 10.6 0.005 0.006 1.400 1.534 acc_transpose_blocks_sync 164592 16.5 1.207 1.463 1.207 1.463 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.409 1.437 grid_collocate_task_list 119 9.7 1.351 1.431 1.351 1.431 mp_waitany 12084 13.8 1.299 1.429 1.299 1.429 fft3d_ps 1201 14.6 0.370 0.474 1.351 1.417 transfer_pw2rs 487 13.2 0.006 0.006 1.336 1.345 mp_alltoall_i22 627 13.8 0.972 1.333 0.972 1.333 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.252 1.323 dbcsr_dot_sd 1205 11.9 0.048 0.058 0.754 1.164 acc_transpose_blocks_kernels 54864 16.5 0.248 0.386 0.847 1.153 parallel_gemm_fm 81 9.0 0.000 0.000 1.140 1.151 parallel_gemm_fm_cosma 81 10.0 1.140 1.151 1.140 1.151 make_basis_sm 11 9.8 0.050 0.402 1.143 1.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.614000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 489.496576E+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 960653. 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 0.013 0.049 40.636 40.636 qs_mol_dyn_low 1 2.0 0.004 0.006 40.376 40.384 qs_forces 11 3.9 0.005 0.026 40.013 40.015 qs_energies 11 4.9 0.015 0.110 38.311 38.317 scf_env_do_scf 11 5.9 0.000 0.001 32.346 32.347 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.879 29.879 dbcsr_multiply_generic 2286 12.5 0.099 0.101 22.475 22.873 qs_scf_new_mos 108 7.5 0.001 0.001 21.142 21.371 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.141 21.370 ot_scf_mini 108 9.5 0.002 0.003 20.210 20.372 velocity_verlet 10 3.0 0.001 0.002 18.887 18.894 multiply_cannon 2286 13.5 0.213 0.222 17.070 18.449 multiply_cannon_loop 2286 14.5 1.191 1.246 15.869 17.407 ot_mini 108 10.5 0.001 0.006 12.542 12.769 mp_waitall_1 200699 16.5 5.787 10.799 5.787 10.799 qs_ot_get_derivative 108 11.5 0.022 0.333 10.146 10.308 multiply_cannon_metrocomm3 27432 15.5 0.070 0.073 4.163 9.475 multiply_cannon_multrec 27432 15.5 1.825 4.178 6.141 8.929 rebuild_ks_matrix 119 8.3 0.000 0.000 6.837 6.983 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.836 6.982 dbcsr_mm_accdrv_process 47894 16.0 3.445 5.820 4.233 6.319 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.014 6.148 qs_ot_get_p 119 10.4 0.001 0.002 4.707 4.921 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.842 4.879 init_scf_run 11 5.9 0.000 0.001 4.480 4.480 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.479 4.480 apply_preconditioner_dbcsr 119 12.6 0.001 0.019 3.188 4.344 apply_single 119 13.6 0.000 0.000 3.186 4.344 mp_sum_l 7287 12.8 2.275 4.167 2.275 4.167 sum_up_and_integrate 119 10.3 0.001 0.002 3.751 3.755 integrate_v_rspace 119 11.3 0.002 0.003 3.736 3.741 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.667 3.700 calculate_rho_elec 119 8.7 0.021 0.024 3.666 3.699 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.039 3.059 make_m2s 4572 13.5 0.052 0.053 2.693 3.044 make_images 4572 14.5 0.206 0.244 2.605 2.958 calculate_first_density_matrix 1 7.0 0.006 0.046 2.952 2.954 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.640 2.640 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.369 2.453 init_scf_loop 11 6.9 0.000 0.000 2.447 2.448 ot_diis_step 108 11.5 0.011 0.014 2.342 2.343 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.305 2.305 cp_fm_redistribute_end 50 14.0 1.168 2.280 1.171 2.282 cp_fm_diag_elpa_base 50 14.0 1.080 2.192 1.108 2.229 calculate_dm_sparse 119 9.5 0.000 0.001 2.143 2.229 acc_transpose_blocks 27432 15.5 0.112 0.117 1.616 2.188 multiply_cannon_sync_h2d 27432 15.5 1.697 2.174 1.697 2.174 density_rs2pw 119 9.7 0.004 0.004 1.963 2.041 jit_kernel_multiply 10 16.3 0.731 1.929 0.731 1.929 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.925 1.928 grid_integrate_task_list 119 12.3 1.843 1.928 1.843 1.928 pw_transfer 1439 11.6 0.065 0.068 1.864 1.901 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.772 1.810 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.690 1.730 make_images_data 4572 15.5 0.047 0.053 1.278 1.628 prepare_preconditioner 11 7.9 0.000 0.000 1.568 1.595 make_preconditioner 11 8.9 0.000 0.000 1.568 1.595 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.136 1.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.469 1.526 potential_pw2rs 119 12.3 0.006 0.006 1.490 1.496 wfi_extrapolate 11 7.9 0.001 0.004 1.470 1.470 fft3d_ps 1201 14.6 0.519 0.575 1.408 1.442 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.390 1.426 mp_alltoall_d11v 2130 13.8 1.234 1.378 1.234 1.378 grid_collocate_task_list 119 9.7 1.290 1.352 1.290 1.352 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.319 1.337 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.241 1.287 transfer_rs2pw 487 10.6 0.005 0.005 1.146 1.243 acc_transpose_blocks_kernels 27432 16.5 0.187 0.276 0.659 1.104 mp_allgather_i34 2286 14.5 0.622 1.064 0.622 1.064 mp_sum_d 4135 12.0 0.559 1.033 0.559 1.033 parallel_gemm_fm 81 9.0 0.000 0.000 1.004 1.016 parallel_gemm_fm_cosma 81 10.0 1.003 1.015 1.003 1.015 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.968 0.970 transfer_pw2rs 487 13.2 0.005 0.006 0.948 0.951 acc_transpose_blocks_sync 82296 16.5 0.820 0.945 0.820 0.945 make_basis_sm 11 9.8 0.005 0.016 0.939 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.880 0.893 jit_kernel_transpose 5 15.5 0.472 0.829 0.472 0.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.636000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.454545, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 519.172096E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.030 34.083 34.088 qs_mol_dyn_low 1 2.0 0.003 0.004 33.874 33.881 qs_forces 11 3.9 0.002 0.002 33.731 33.732 qs_energies 11 4.9 0.001 0.001 32.202 32.203 scf_env_do_scf 11 5.9 0.000 0.001 25.863 25.863 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.405 23.406 dbcsr_multiply_generic 2286 12.5 0.093 0.095 18.211 18.304 qs_scf_new_mos 108 7.5 0.001 0.001 15.773 15.788 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.773 15.788 multiply_cannon 2286 13.5 0.194 0.199 15.010 15.593 velocity_verlet 10 3.0 0.001 0.002 15.243 15.247 ot_scf_mini 108 9.5 0.002 0.002 15.015 15.030 multiply_cannon_loop 2286 14.5 0.855 0.887 14.246 14.949 ot_mini 108 10.5 0.001 0.001 9.383 9.399 multiply_cannon_multrec 18288 15.5 1.880 2.856 7.616 8.091 qs_ot_get_derivative 108 11.5 0.001 0.001 7.883 7.898 dbcsr_mm_accdrv_process 38222 16.0 5.552 6.601 5.641 6.667 rebuild_ks_matrix 119 8.3 0.000 0.000 5.981 6.003 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.980 6.003 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.280 5.300 init_scf_run 11 5.9 0.000 0.001 4.885 4.885 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.885 4.885 mp_waitall_1 158411 16.6 2.937 4.120 2.937 4.120 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.201 3.737 calculate_first_density_matrix 1 7.0 0.000 0.000 3.667 3.668 sum_up_and_integrate 119 10.3 0.001 0.001 3.542 3.547 integrate_v_rspace 119 11.3 0.002 0.003 3.528 3.536 qs_ot_get_p 119 10.4 0.001 0.002 3.402 3.435 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.210 3.216 calculate_rho_elec 119 8.7 0.031 0.031 3.209 3.215 multiply_cannon_metrocomm3 18288 15.5 0.047 0.050 1.763 2.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.530 2.532 init_scf_loop 11 6.9 0.000 0.000 2.442 2.443 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.371 2.381 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.947 2.354 apply_single 119 13.6 0.000 0.000 1.947 2.353 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.277 2.284 acc_transpose_blocks 18288 15.5 0.078 0.081 2.040 2.112 calculate_dm_sparse 119 9.5 0.000 0.001 1.975 1.983 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.977 1.978 make_m2s 4572 13.5 0.044 0.044 1.813 1.948 grid_integrate_task_list 119 12.3 1.807 1.898 1.807 1.898 make_images 4572 14.5 0.193 0.205 1.728 1.861 density_rs2pw 119 9.7 0.004 0.004 1.783 1.853 pw_transfer 1439 11.6 0.065 0.069 1.805 1.812 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.740 1.745 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.729 1.739 cp_fm_diag_elpa_base 50 14.0 1.706 1.718 1.727 1.737 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.712 1.721 prepare_preconditioner 11 7.9 0.000 0.000 1.667 1.669 make_preconditioner 11 8.9 0.000 0.000 1.667 1.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.526 1.609 ot_diis_step 108 11.5 0.011 0.012 1.481 1.481 mp_sum_l 7287 12.8 1.001 1.429 1.001 1.429 fft_wrap_pw1pw2_140 487 13.2 0.211 0.216 1.381 1.391 potential_pw2rs 119 12.3 0.007 0.009 1.345 1.348 grid_collocate_task_list 119 9.7 1.233 1.344 1.233 1.344 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.312 1.316 fft3d_ps 1201 14.6 0.525 0.542 1.282 1.290 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 1.200 1.210 multiply_cannon_sync_h2d 18288 15.5 1.039 1.161 1.039 1.161 wfi_extrapolate 11 7.9 0.001 0.001 1.160 1.160 parallel_gemm_fm 81 9.0 0.000 0.000 1.123 1.126 parallel_gemm_fm_cosma 81 10.0 1.123 1.126 1.123 1.126 transfer_rs2pw 487 10.6 0.005 0.005 0.958 1.030 jit_kernel_transpose 5 15.6 0.983 0.987 0.983 0.987 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.949 0.964 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.946 0.947 make_images_data 4572 15.5 0.047 0.051 0.782 0.935 hybrid_alltoall_any 4725 16.4 0.057 0.116 0.672 0.851 transfer_pw2rs 487 13.2 0.004 0.004 0.816 0.819 acc_transpose_blocks_sync 54864 16.5 0.742 0.816 0.742 0.816 make_basis_sm 11 9.8 0.001 0.001 0.785 0.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.773 mp_alltoall_d11v 2130 13.8 0.655 0.749 0.655 0.749 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.675 0.732 cp_fm_cholesky_invert 11 10.9 0.692 0.695 0.692 0.695 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.088000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.636364, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 559.652864E+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 0.023 0.382 36.588 36.927 qs_mol_dyn_low 1 2.0 0.003 0.004 36.400 36.409 qs_forces 11 3.9 0.002 0.004 36.330 36.331 qs_energies 11 4.9 0.001 0.001 34.628 34.634 scf_env_do_scf 11 5.9 0.000 0.001 29.423 29.424 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.045 26.045 dbcsr_multiply_generic 2286 12.5 0.097 0.100 20.258 20.382 velocity_verlet 10 3.0 0.011 0.021 18.396 18.398 qs_scf_new_mos 108 7.5 0.001 0.001 18.047 18.100 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.047 18.099 ot_scf_mini 108 9.5 0.002 0.003 17.017 17.063 multiply_cannon 2286 13.5 0.218 0.225 16.522 17.042 multiply_cannon_loop 2286 14.5 1.516 1.617 15.566 15.901 ot_mini 108 10.5 0.001 0.001 10.541 10.603 multiply_cannon_multrec 27432 15.5 2.569 3.244 9.204 9.462 qs_ot_get_derivative 108 11.5 0.001 0.001 8.673 8.720 dbcsr_mm_accdrv_process 47916 15.9 6.066 7.959 6.533 8.019 rebuild_ks_matrix 119 8.3 0.000 0.000 6.275 6.324 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.275 6.323 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.560 5.602 init_scf_run 11 5.9 0.000 0.001 3.850 3.851 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.850 3.850 qs_ot_get_p 119 10.4 0.001 0.001 3.622 3.689 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.117 3.524 sum_up_and_integrate 119 10.3 0.001 0.001 3.471 3.479 integrate_v_rspace 119 11.3 0.002 0.003 3.460 3.468 init_scf_loop 11 6.9 0.000 0.000 3.355 3.356 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.314 3.350 calculate_rho_elec 119 8.7 0.040 0.046 3.314 3.349 acc_transpose_blocks 27432 15.5 0.118 0.124 2.595 2.729 prepare_preconditioner 11 7.9 0.000 0.000 2.530 2.538 make_preconditioner 11 8.9 0.000 0.000 2.529 2.538 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.142 2.461 calculate_first_density_matrix 1 7.0 0.000 0.000 2.454 2.459 make_m2s 4572 13.5 0.054 0.056 2.285 2.403 mp_waitall_1 137007 16.6 1.811 2.360 1.811 2.360 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.329 2.354 make_images 4572 14.5 0.275 0.338 2.177 2.293 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.115 2.247 apply_single 119 13.6 0.000 0.000 2.115 2.247 calculate_dm_sparse 119 9.5 0.000 0.000 2.186 2.240 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.206 2.218 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.999 1.999 pw_transfer 1439 11.6 0.065 0.070 1.912 1.946 grid_integrate_task_list 119 12.3 1.824 1.905 1.824 1.905 density_rs2pw 119 9.7 0.004 0.004 1.788 1.883 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.819 1.858 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.823 1.834 ot_diis_step 108 11.5 0.012 0.012 1.825 1.825 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.804 1.804 acc_transpose_blocks_sync 82296 16.5 1.436 1.572 1.436 1.572 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.563 1.572 cp_fm_diag_elpa_base 50 14.0 1.529 1.545 1.561 1.570 fft_wrap_pw1pw2_140 487 13.2 0.245 0.258 1.500 1.541 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.817 1.414 mp_sum_l 7287 12.8 1.027 1.395 1.027 1.395 fft3d_ps 1201 14.6 0.554 0.603 1.340 1.365 wfi_extrapolate 11 7.9 0.001 0.001 1.340 1.340 grid_collocate_task_list 119 9.7 1.248 1.338 1.248 1.338 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.296 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.285 1.293 potential_pw2rs 119 12.3 0.009 0.010 1.281 1.283 cp_fm_upper_to_full 72 14.2 0.812 1.151 0.812 1.151 jit_kernel_multiply 6 15.7 0.402 1.145 0.402 1.145 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.093 1.094 dbcsr_complete_redistribute 329 12.2 0.120 0.148 0.771 1.037 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 1.012 1.028 transfer_rs2pw 487 10.6 0.004 0.005 0.900 0.996 make_images_data 4572 15.5 0.048 0.051 0.852 0.986 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.743 0.929 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.800 0.876 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.559 0.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.808 mp_alltoall_d11v 2130 13.8 0.721 0.791 0.721 0.791 jit_kernel_transpose 5 15.6 0.742 0.755 0.742 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.927000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.000000, yerr=3.541956 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 618.188800E+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 0.007 0.026 28.982 28.983 qs_mol_dyn_low 1 2.0 0.006 0.006 28.706 28.821 qs_forces 11 3.9 0.004 0.005 28.482 28.482 qs_energies 11 4.9 0.001 0.001 26.779 26.782 scf_env_do_scf 11 5.9 0.000 0.001 21.786 21.786 scf_env_do_scf_inner_loop 108 6.5 0.021 0.024 19.338 19.338 velocity_verlet 10 3.0 0.002 0.002 14.744 14.757 dbcsr_multiply_generic 2286 12.5 0.089 0.094 13.368 13.466 qs_scf_new_mos 108 7.5 0.001 0.001 11.755 11.779 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.754 11.778 multiply_cannon 2286 13.5 0.223 0.228 10.775 11.266 ot_scf_mini 108 9.5 0.002 0.002 11.041 11.065 multiply_cannon_loop 2286 14.5 0.640 0.664 9.862 10.056 multiply_cannon_multrec 9144 15.5 1.663 1.965 6.140 6.411 ot_mini 108 10.5 0.001 0.001 6.354 6.382 rebuild_ks_matrix 119 8.3 0.000 0.000 5.726 5.752 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.726 5.752 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.089 5.112 qs_ot_get_derivative 108 11.5 0.001 0.001 5.066 5.090 dbcsr_mm_accdrv_process 12550 15.8 3.680 4.389 4.363 4.431 init_scf_run 11 5.9 0.000 0.001 3.546 3.546 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.546 3.546 sum_up_and_integrate 119 10.3 0.001 0.001 3.418 3.422 integrate_v_rspace 119 11.3 0.003 0.003 3.408 3.411 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.295 3.304 calculate_rho_elec 119 8.7 0.060 0.061 3.295 3.304 qs_ot_get_p 119 10.4 0.002 0.002 2.806 2.838 init_scf_loop 11 6.9 0.000 0.000 2.427 2.428 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.428 pw_transfer 1439 11.6 0.065 0.068 1.949 1.959 calculate_dm_sparse 119 9.5 0.000 0.000 1.941 1.957 make_m2s 4572 13.5 0.034 0.035 1.801 1.955 grid_integrate_task_list 119 12.3 1.868 1.931 1.868 1.931 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.856 1.869 make_images 4572 14.5 0.271 0.301 1.712 1.864 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.826 1.829 mp_waitall_1 115863 16.7 1.368 1.824 1.368 1.824 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.723 1.757 density_rs2pw 119 9.7 0.003 0.004 1.675 1.735 prepare_preconditioner 11 7.9 0.000 0.000 1.705 1.710 make_preconditioner 11 8.9 0.000 0.000 1.705 1.710 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.685 1.686 acc_transpose_blocks 9144 15.5 0.041 0.042 1.603 1.658 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.598 1.625 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.572 1.573 jit_kernel_multiply 8 15.7 0.644 1.546 0.644 1.546 fft_wrap_pw1pw2_140 487 13.2 0.323 0.331 1.529 1.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.527 1.535 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.348 1.362 grid_collocate_task_list 119 9.7 1.297 1.341 1.297 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.330 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.308 1.316 cp_fm_diag_elpa_base 50 14.0 1.281 1.298 1.306 1.315 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.231 1.286 apply_single 119 13.6 0.000 0.000 1.230 1.286 fft3d_ps 1201 14.6 0.556 0.570 1.270 1.282 ot_diis_step 108 11.5 0.013 0.013 1.274 1.275 potential_pw2rs 119 12.3 0.010 0.011 1.218 1.220 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.217 1.217 wfi_extrapolate 11 7.9 0.001 0.001 1.067 1.067 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.761 0.982 make_images_data 4572 15.5 0.042 0.046 0.781 0.949 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.925 mp_alltoall_d11v 2130 13.8 0.778 0.859 0.778 0.859 cp_fm_cholesky_invert 11 10.9 0.855 0.857 0.855 0.857 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.827 0.841 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.825 0.833 transfer_rs2pw 487 10.6 0.004 0.005 0.769 0.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.760 acc_transpose_blocks_sync 27432 16.5 0.720 0.759 0.720 0.759 jit_kernel_transpose 5 15.6 0.710 0.724 0.710 0.724 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.664 0.715 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.361 0.713 transfer_pw2rs 487 13.2 0.003 0.004 0.650 0.651 mp_allgather_i34 2286 14.5 0.230 0.648 0.230 0.648 mp_alltoall_z22v 1201 16.6 0.588 0.620 0.588 0.620 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.983000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.363636, yerr=5.121660 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 765.001728E+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 0.072 0.127 43.441 43.443 qs_mol_dyn_low 1 2.0 0.003 0.004 42.901 42.908 qs_forces 11 3.9 0.002 0.002 42.840 42.841 qs_energies 11 4.9 0.001 0.002 40.858 40.861 scf_env_do_scf 11 5.9 0.001 0.001 34.354 34.355 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.576 26.577 velocity_verlet 10 3.0 0.002 0.002 23.717 23.723 dbcsr_multiply_generic 2286 12.5 0.097 0.099 19.403 19.636 qs_scf_new_mos 108 7.5 0.001 0.001 17.347 17.449 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.346 17.448 multiply_cannon 2286 13.5 0.297 0.307 15.571 16.556 ot_scf_mini 108 9.5 0.002 0.002 16.192 16.295 multiply_cannon_loop 2286 14.5 0.872 0.911 14.331 15.343 ot_mini 108 10.5 0.001 0.001 9.970 10.085 multiply_cannon_multrec 9144 15.5 3.670 5.029 9.311 9.493 qs_ot_get_derivative 108 11.5 0.001 0.001 7.934 8.038 init_scf_loop 11 6.9 0.000 0.000 7.752 7.754 rebuild_ks_matrix 119 8.3 0.000 0.000 6.973 7.112 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.973 7.112 dbcsr_mm_accdrv_process 12550 15.8 4.947 6.889 5.501 6.921 prepare_preconditioner 11 7.9 0.000 0.000 6.813 6.826 make_preconditioner 11 8.9 0.000 0.000 6.813 6.826 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.336 6.707 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.281 6.408 cp_fm_upper_to_full 72 14.2 3.176 4.603 3.176 4.603 init_scf_run 11 5.9 0.000 0.001 4.467 4.467 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.467 4.467 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.961 3.966 calculate_rho_elec 119 8.7 0.118 0.121 3.960 3.966 sum_up_and_integrate 119 10.3 0.001 0.001 3.702 3.709 integrate_v_rspace 119 11.3 0.003 0.004 3.692 3.698 qs_ot_get_p 119 10.4 0.002 0.002 3.314 3.447 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.781 3.209 mp_waitall_1 94719 16.7 2.220 3.134 2.220 3.134 calculate_first_density_matrix 1 7.0 0.000 0.000 3.031 3.032 dbcsr_complete_redistribute 329 12.2 0.285 0.291 2.014 2.836 pw_transfer 1439 11.6 0.068 0.069 2.552 2.558 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.705 2.527 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.222 2.515 apply_single 119 13.6 0.000 0.000 2.221 2.514 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.454 2.461 make_m2s 4572 13.5 0.037 0.038 2.283 2.427 mp_alltoall_i22 627 13.8 1.452 2.313 1.452 2.313 make_images 4572 14.5 0.356 0.392 2.164 2.308 calculate_dm_sparse 119 9.5 0.000 0.000 2.283 2.299 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.471 2.286 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.262 2.173 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.111 2.112 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.057 2.109 fft_wrap_pw1pw2_140 487 13.2 0.576 0.581 2.080 2.088 grid_integrate_task_list 119 12.3 2.064 2.087 2.064 2.087 density_rs2pw 119 9.7 0.003 0.004 2.000 2.021 ot_diis_step 108 11.5 0.014 0.014 2.012 2.012 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.953 1.957 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.867 1.911 acc_transpose_blocks 9144 15.5 0.044 0.046 1.837 1.893 mp_sum_l 7287 12.8 1.010 1.790 1.010 1.790 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.758 1.758 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.595 1.595 fft3d_ps 1201 14.6 0.593 0.605 1.550 1.557 grid_collocate_task_list 119 9.7 1.510 1.520 1.510 1.520 cp_fm_cholesky_invert 11 10.9 1.385 1.389 1.385 1.389 wfi_extrapolate 11 7.9 0.001 0.001 1.374 1.374 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.306 1.325 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.312 1.312 cp_fm_diag_elpa_base 50 14.0 1.167 1.220 1.310 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.298 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.037 1.246 potential_pw2rs 119 12.3 0.014 0.014 1.241 1.243 acc_transpose_blocks_sync 27432 16.5 1.122 1.180 1.122 1.180 mp_alltoall_d11v 2130 13.8 1.144 1.178 1.144 1.178 make_images_data 4572 15.5 0.046 0.049 0.950 1.125 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.086 1.100 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.997 1.044 jit_kernel_multiply 5 15.5 0.525 1.030 0.525 1.030 qs_create_task_list 11 7.9 0.001 0.001 0.924 0.937 generate_qs_task_list 11 8.9 0.367 0.385 0.924 0.936 make_basis_sm 11 9.8 0.001 0.001 0.934 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.916 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.443000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.272727, yerr=8.842983 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.551104E+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 2122888. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 81.581 81.582 qs_mol_dyn_low 1 2.0 0.003 0.004 81.329 81.341 qs_forces 11 3.9 0.003 0.004 81.260 81.261 qs_energies 11 4.9 0.001 0.002 78.424 78.441 scf_env_do_scf 11 5.9 0.000 0.001 69.512 69.514 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.712 63.712 dbcsr_multiply_generic 2055 12.4 0.105 0.107 50.613 50.932 qs_scf_new_mos 99 7.5 0.000 0.001 46.738 46.842 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.737 46.841 ot_scf_mini 99 9.5 0.002 0.002 44.336 44.454 multiply_cannon 2055 13.4 0.187 0.193 42.414 43.270 velocity_verlet 10 3.0 0.001 0.002 42.802 42.802 multiply_cannon_loop 2055 14.4 1.769 1.805 41.406 42.328 ot_mini 99 10.5 0.001 0.001 26.038 26.158 qs_ot_get_derivative 99 11.5 0.001 0.001 19.244 19.368 multiply_cannon_multrec 49320 15.4 11.436 12.151 17.430 18.126 rebuild_ks_matrix 110 8.3 0.000 0.001 14.446 14.600 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.445 14.600 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.675 12.816 mp_waitall_1 220248 16.4 10.532 11.588 10.532 11.588 multiply_cannon_sync_h2d 49320 15.4 9.659 10.210 9.659 10.210 qs_ot_get_p 110 10.4 0.001 0.001 9.762 9.883 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.427 7.979 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.217 7.741 apply_single 110 13.6 0.000 0.000 7.217 7.741 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.356 7.644 sum_up_and_integrate 110 10.3 0.002 0.003 6.958 6.972 integrate_v_rspace 110 11.3 0.003 0.003 6.933 6.953 init_scf_run 11 5.9 0.000 0.001 6.889 6.889 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.889 6.889 ot_diis_step 99 11.5 0.006 0.006 6.597 6.597 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.564 6.586 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.278 6.415 calculate_rho_elec 110 8.6 0.020 0.024 6.277 6.414 dbcsr_mm_accdrv_process 87628 16.1 3.023 3.096 5.863 6.123 init_scf_loop 11 6.9 0.000 0.000 5.773 5.773 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.688 5.688 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.241 5.302 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.111 5.140 cp_fm_diag_elpa_base 48 14.0 5.098 5.126 5.109 5.138 mp_sum_l 6594 12.7 3.866 4.684 3.866 4.684 wfi_extrapolate 11 7.9 0.001 0.001 4.025 4.025 make_m2s 4110 13.4 0.060 0.065 3.874 3.998 make_images 4110 14.4 0.178 0.191 3.778 3.906 calculate_dm_sparse 110 9.5 0.001 0.001 3.765 3.852 prepare_preconditioner 11 7.9 0.000 0.000 3.620 3.640 make_preconditioner 11 8.9 0.000 0.000 3.620 3.640 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.625 3.629 density_rs2pw 110 9.6 0.004 0.005 3.304 3.519 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.402 3.448 grid_integrate_task_list 110 12.3 3.248 3.418 3.248 3.418 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.341 3.371 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.278 3.322 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.235 3.288 pw_transfer 1331 11.6 0.055 0.065 3.045 3.120 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.956 3.035 mp_alltoall_d11v 2046 13.8 2.286 2.799 2.286 2.799 calculate_first_density_matrix 1 7.0 0.000 0.000 2.766 2.768 jit_kernel_multiply 13 15.9 2.564 2.635 2.564 2.635 fft_wrap_pw1pw2_140 451 13.1 0.390 0.431 2.524 2.611 acc_transpose_blocks 49320 15.4 0.210 0.217 2.541 2.588 potential_pw2rs 110 12.3 0.005 0.006 2.557 2.580 fft3d_ps 1111 14.6 0.794 0.885 2.251 2.306 grid_collocate_task_list 110 9.6 2.160 2.283 2.160 2.283 hybrid_alltoall_any 4261 16.3 0.085 0.486 1.760 2.088 mp_waitany 14300 13.8 1.813 2.065 1.813 2.065 transfer_rs2pw 451 10.6 0.005 0.006 1.858 2.063 make_images_data 4110 15.4 0.044 0.047 1.775 1.920 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.901 1.917 cp_fm_cholesky_invert 11 10.9 1.835 1.838 1.835 1.838 mp_sum_d 3889 11.9 1.325 1.827 1.325 1.827 transfer_pw2rs 451 13.1 0.006 0.007 1.733 1.750 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.649 1.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.582000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.796101 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 594.444288E+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 2333512. 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 0.009 0.029 69.351 69.352 qs_mol_dyn_low 1 2.0 0.003 0.004 69.137 69.147 qs_forces 11 3.9 0.003 0.003 69.020 69.020 qs_energies 11 4.9 0.001 0.002 65.720 65.725 scf_env_do_scf 11 5.9 0.000 0.001 56.472 56.476 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.963 48.963 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.288 38.428 velocity_verlet 10 3.0 0.015 0.060 35.959 35.960 qs_scf_new_mos 99 7.5 0.001 0.001 33.246 33.369 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.246 33.368 multiply_cannon 2055 13.4 0.224 0.243 31.708 32.625 ot_scf_mini 99 9.5 0.003 0.003 31.574 31.698 multiply_cannon_loop 2055 14.4 1.165 1.192 30.583 31.486 ot_mini 99 10.5 0.001 0.001 18.517 18.639 multiply_cannon_multrec 24660 15.4 7.025 8.168 14.408 15.708 rebuild_ks_matrix 110 8.3 0.000 0.001 13.424 13.535 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.423 13.534 qs_ot_get_derivative 99 11.5 0.001 0.001 12.757 12.876 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.810 11.911 mp_waitall_1 176588 16.5 7.648 9.932 7.648 9.932 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.277 8.036 multiply_cannon_sync_h2d 24660 15.4 6.376 7.611 6.376 7.611 dbcsr_mm_accdrv_process 52282 16.1 5.732 6.778 7.212 7.553 init_scf_loop 11 6.9 0.000 0.000 7.475 7.476 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.435 7.071 apply_single 110 13.6 0.000 0.001 6.435 7.071 init_scf_run 11 5.9 0.000 0.001 6.747 6.748 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.747 6.748 qs_ot_get_p 110 10.4 0.001 0.001 6.416 6.561 sum_up_and_integrate 110 10.3 0.001 0.003 6.323 6.334 integrate_v_rspace 110 11.3 0.002 0.003 6.296 6.308 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.828 5.838 calculate_rho_elec 110 8.6 0.039 0.047 5.827 5.837 ot_diis_step 99 11.5 0.010 0.010 5.711 5.711 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.911 5.640 prepare_preconditioner 11 7.9 0.000 0.000 5.486 5.505 make_preconditioner 11 8.9 0.000 0.000 5.486 5.505 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.083 5.239 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.543 4.563 make_m2s 4110 13.4 0.057 0.059 4.231 4.552 make_images 4110 14.4 0.408 0.467 4.122 4.438 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.083 4.083 pw_transfer 1331 11.6 0.066 0.071 3.519 3.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.606 3.608 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.412 3.547 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.508 3.522 cp_fm_diag_elpa_base 48 14.0 3.462 3.481 3.505 3.519 wfi_extrapolate 11 7.9 0.001 0.001 3.461 3.461 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.287 3.356 calculate_dm_sparse 110 9.5 0.001 0.001 3.318 3.345 grid_integrate_task_list 110 12.3 3.167 3.312 3.167 3.312 density_rs2pw 110 9.6 0.004 0.004 3.091 3.312 calculate_first_density_matrix 1 7.0 0.000 0.000 3.193 3.194 fft_wrap_pw1pw2_140 451 13.1 0.461 0.475 2.927 3.068 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.882 2.911 fft3d_ps 1111 14.6 1.110 1.333 2.507 2.658 make_images_data 4110 15.4 0.049 0.053 2.306 2.648 hybrid_alltoall_any 4261 16.3 0.106 0.460 1.968 2.583 cp_fm_cholesky_invert 11 10.9 2.568 2.575 2.568 2.575 mp_sum_l 6594 12.7 1.717 2.367 1.717 2.367 grid_collocate_task_list 110 9.6 2.168 2.310 2.168 2.310 jit_kernel_multiply 10 16.2 1.118 2.300 1.118 2.300 potential_pw2rs 110 12.3 0.008 0.009 2.209 2.217 acc_transpose_blocks 24660 15.4 0.114 0.117 2.008 2.200 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.980 2.005 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.994 1.995 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.881 1.893 mp_alltoall_d11v 2046 13.8 1.678 1.880 1.678 1.880 dbcsr_complete_redistribute 325 12.2 0.632 0.713 1.527 1.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.770 1.781 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.773 1.735 transfer_rs2pw 451 10.6 0.007 0.007 1.396 1.552 mp_irecv_dv 57340 16.2 0.646 1.519 0.646 1.519 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.488 mp_allgather_i34 2055 14.4 0.437 1.460 0.437 1.460 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.352000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=562.818182, yerr=7.309000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.169344E+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 0.030 0.074 60.992 60.992 qs_mol_dyn_low 1 2.0 0.004 0.004 60.596 60.606 qs_forces 11 3.9 0.003 0.004 60.441 60.443 qs_energies 11 4.9 0.074 0.580 57.048 57.274 scf_env_do_scf 11 5.9 0.005 0.034 48.953 48.993 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.187 40.187 velocity_verlet 10 3.0 0.001 0.002 32.324 32.328 dbcsr_multiply_generic 2055 12.4 0.106 0.110 29.148 29.407 qs_scf_new_mos 99 7.5 0.001 0.001 25.587 25.682 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.586 25.681 ot_scf_mini 99 9.5 0.002 0.003 24.339 24.446 multiply_cannon 2055 13.4 0.214 0.223 22.967 24.230 multiply_cannon_loop 2055 14.4 0.815 0.847 21.822 22.834 ot_mini 99 10.5 0.001 0.001 14.199 14.307 rebuild_ks_matrix 110 8.3 0.000 0.000 12.109 12.230 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.108 12.230 multiply_cannon_multrec 16440 15.4 3.757 5.337 10.366 11.987 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.658 10.773 mp_waitall_1 139946 16.5 6.843 9.969 6.843 9.969 qs_ot_get_derivative 99 11.5 0.001 0.001 9.811 9.919 init_scf_loop 11 6.9 0.001 0.003 8.729 8.774 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.309 7.284 prepare_preconditioner 11 7.9 0.000 0.001 6.658 6.676 make_preconditioner 11 8.9 0.000 0.001 6.658 6.676 dbcsr_mm_accdrv_process 34862 16.1 5.751 6.162 6.452 6.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.988 6.354 sum_up_and_integrate 110 10.3 0.001 0.002 6.269 6.284 integrate_v_rspace 110 11.3 0.003 0.003 6.243 6.258 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.738 5.747 calculate_rho_elec 110 8.6 0.059 0.059 5.737 5.747 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.943 5.470 apply_single 110 13.6 0.000 0.000 4.943 5.470 qs_ot_get_p 110 10.4 0.001 0.002 5.323 5.469 init_scf_run 11 5.9 0.000 0.001 5.431 5.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.431 5.431 make_m2s 4110 13.4 0.049 0.051 4.116 4.477 make_images 4110 14.4 0.399 0.526 4.001 4.360 ot_diis_step 99 11.5 0.011 0.011 4.359 4.359 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.343 4.016 multiply_cannon_sync_h2d 16440 15.4 3.265 3.907 3.265 3.907 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.728 3.730 pw_transfer 1331 11.6 0.065 0.073 3.556 3.565 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.449 3.459 grid_integrate_task_list 110 12.3 3.202 3.392 3.202 3.392 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.365 3.365 density_rs2pw 110 9.6 0.004 0.005 2.931 3.150 fft_wrap_pw1pw2_140 451 13.1 0.579 0.586 2.971 2.983 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.821 2.830 cp_fm_diag_elpa_base 48 14.0 2.755 2.789 2.820 2.829 make_images_data 4110 15.4 0.045 0.050 2.316 2.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.772 2.773 hybrid_alltoall_any 4261 16.3 0.109 0.376 2.101 2.709 cp_fm_cholesky_invert 11 10.9 2.575 2.582 2.575 2.582 calculate_dm_sparse 110 9.5 0.001 0.001 2.518 2.552 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.494 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.391 2.446 fft3d_ps 1111 14.6 1.088 1.096 2.391 2.401 grid_collocate_task_list 110 9.6 2.216 2.394 2.216 2.394 calculate_first_density_matrix 1 7.0 0.001 0.006 2.386 2.387 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.873 2.334 mp_irecv_dv 48980 15.7 0.799 2.204 0.799 2.204 mp_sum_l 6594 12.7 1.443 2.132 1.443 2.132 potential_pw2rs 110 12.3 0.011 0.012 2.108 2.111 mp_alltoall_d11v 2046 13.8 1.805 1.984 1.805 1.984 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 1.971 1.975 dbcsr_complete_redistribute 325 12.2 0.332 0.367 1.425 1.905 cp_fm_upper_to_full 70 14.2 1.410 1.797 1.410 1.797 acc_transpose_blocks 16440 15.4 0.076 0.077 1.584 1.664 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.645 1.659 cp_fm_cholesky_decompose 22 10.9 1.563 1.581 1.563 1.581 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.377 1.475 transfer_rs2pw 451 10.6 0.005 0.006 1.255 1.472 mp_allgather_i34 2055 14.4 0.459 1.470 0.459 1.470 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.469 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.980 1.435 mp_waitany 17072 13.8 1.135 1.400 1.135 1.400 mp_alltoall_i22 605 13.7 0.850 1.381 0.850 1.381 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.171 1.272 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.260 1.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.992000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.818182, yerr=7.395554 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 741.085184E+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 0.013 0.036 66.972 66.974 qs_mol_dyn_low 1 2.0 0.003 0.004 66.683 66.693 qs_forces 11 3.9 0.004 0.011 66.615 66.615 qs_energies 11 4.9 0.013 0.093 63.231 63.235 scf_env_do_scf 11 5.9 0.001 0.003 54.543 54.556 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.562 42.562 velocity_verlet 10 3.0 0.002 0.002 36.980 36.982 dbcsr_multiply_generic 2055 12.4 0.112 0.117 30.929 31.183 qs_scf_new_mos 99 7.5 0.001 0.001 27.730 27.840 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.729 27.839 ot_scf_mini 99 9.5 0.003 0.003 26.046 26.154 multiply_cannon 2055 13.4 0.238 0.256 23.927 24.825 multiply_cannon_loop 2055 14.4 1.395 1.454 22.617 23.416 ot_mini 99 10.5 0.001 0.001 15.094 15.234 multiply_cannon_multrec 24660 15.4 4.091 6.938 13.036 14.065 rebuild_ks_matrix 110 8.3 0.000 0.000 12.000 12.160 qs_ks_build_kohn_sham_matrix 110 9.3 0.027 0.132 11.999 12.160 init_scf_loop 11 6.9 0.005 0.038 11.941 11.952 qs_ot_get_derivative 99 11.5 0.001 0.001 10.903 11.014 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.628 10.771 prepare_preconditioner 11 7.9 0.000 0.001 9.953 9.988 make_preconditioner 11 8.9 0.002 0.015 9.953 9.988 dbcsr_mm_accdrv_process 52304 16.0 7.648 9.156 8.788 9.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.186 9.667 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.180 6.239 calculate_rho_elec 110 8.6 0.078 0.081 6.180 6.239 sum_up_and_integrate 110 10.3 0.001 0.002 6.183 6.197 integrate_v_rspace 110 11.3 0.003 0.007 6.157 6.170 mp_waitall_1 121746 16.5 4.225 6.101 4.225 6.101 qs_ot_get_p 110 10.4 0.001 0.002 5.880 6.034 init_scf_run 11 5.9 0.000 0.001 5.738 5.738 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.738 5.738 make_m2s 4110 13.4 0.059 0.061 5.384 5.643 make_images 4110 14.4 0.579 0.695 5.243 5.495 cp_fm_upper_to_full 70 14.2 3.330 4.836 3.330 4.836 ot_diis_step 99 11.5 0.011 0.011 4.152 4.152 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.058 4.134 apply_single 110 13.6 0.000 0.000 4.058 4.133 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.026 4.043 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.859 3.910 dbcsr_complete_redistribute 325 12.2 0.417 0.462 2.647 3.777 pw_transfer 1331 11.6 0.066 0.075 3.628 3.660 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.521 3.558 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.460 3.460 grid_integrate_task_list 110 12.3 3.288 3.442 3.288 3.442 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.346 3.409 multiply_cannon_metrocomm3 24660 15.4 0.038 0.038 1.534 3.409 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.111 3.225 density_rs2pw 110 9.6 0.004 0.004 2.962 3.172 calculate_dm_sparse 110 9.5 0.001 0.001 3.051 3.089 fft_wrap_pw1pw2_140 451 13.1 0.610 0.631 3.023 3.062 acc_transpose_blocks 24660 15.4 0.112 0.115 2.799 3.029 wfi_extrapolate 11 7.9 0.001 0.001 2.989 2.989 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.924 2.934 cp_fm_diag_elpa_base 48 14.0 2.779 2.837 2.922 2.932 mp_alltoall_i22 605 13.7 1.737 2.913 1.737 2.913 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.909 2.911 make_images_data 4110 15.4 0.049 0.053 2.632 2.901 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.756 2.851 hybrid_alltoall_any 4261 16.3 0.122 0.459 2.317 2.832 cp_fm_cholesky_invert 11 10.9 2.645 2.653 2.645 2.653 calculate_first_density_matrix 1 7.0 0.001 0.004 2.605 2.607 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.546 2.586 multiply_cannon_sync_h2d 24660 15.4 2.373 2.566 2.373 2.566 grid_collocate_task_list 110 9.6 2.264 2.443 2.264 2.443 fft3d_ps 1111 14.6 1.087 1.118 2.403 2.418 mp_alltoall_d11v 2046 13.8 2.088 2.281 2.088 2.281 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 2.267 2.268 jit_kernel_multiply 11 15.8 0.797 2.079 0.797 2.079 potential_pw2rs 110 12.3 0.013 0.014 2.009 2.013 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.794 1.829 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.714 cp_fm_cholesky_decompose 22 10.9 1.642 1.676 1.642 1.676 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.622 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.612 1.629 mp_sum_l 6594 12.7 0.940 1.562 0.940 1.562 transfer_rs2pw 451 10.6 0.005 0.007 1.245 1.519 acc_transpose_blocks_sync 73980 16.4 1.381 1.471 1.381 1.471 multiply_cannon_metrocomm4 20550 15.4 0.060 0.064 0.863 1.454 mp_waitany 13376 13.8 1.092 1.414 1.092 1.414 acc_transpose_blocks_kernels 24660 16.4 0.306 0.311 1.265 1.405 mp_irecv_dv 62702 16.1 0.760 1.368 0.760 1.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.974000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.727273, yerr=11.701522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 866.623488E+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 0.009 0.030 59.660 59.662 qs_mol_dyn_low 1 2.0 0.003 0.004 59.406 59.418 qs_forces 11 3.9 0.003 0.003 57.279 57.279 qs_energies 11 4.9 0.001 0.001 53.589 53.594 scf_env_do_scf 11 5.9 0.000 0.001 44.214 44.214 scf_env_do_scf_inner_loop 99 6.5 0.049 0.058 35.970 35.971 velocity_verlet 10 3.0 0.002 0.002 33.968 34.100 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.867 24.013 qs_scf_new_mos 99 7.5 0.001 0.001 21.215 21.284 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.215 21.284 ot_scf_mini 99 9.5 0.002 0.002 19.937 20.000 multiply_cannon 2055 13.4 0.238 0.250 18.039 19.389 multiply_cannon_loop 2055 14.4 0.600 0.622 16.745 17.040 rebuild_ks_matrix 110 8.3 0.000 0.000 11.732 11.780 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.732 11.779 ot_mini 99 10.5 0.001 0.001 10.873 10.936 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.432 10.476 multiply_cannon_multrec 8220 15.4 3.131 4.466 7.613 8.875 init_scf_loop 11 6.9 0.000 0.000 8.193 8.195 mp_waitall_1 103326 16.6 6.214 8.010 6.214 8.010 qs_ot_get_derivative 99 11.5 0.001 0.001 7.093 7.156 prepare_preconditioner 11 7.9 0.000 0.000 6.391 6.403 make_preconditioner 11 8.9 0.000 0.000 6.391 6.403 sum_up_and_integrate 110 10.3 0.001 0.002 6.297 6.309 integrate_v_rspace 110 11.3 0.003 0.003 6.270 6.282 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.053 6.065 calculate_rho_elec 110 8.6 0.114 0.114 6.053 6.065 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.964 6.043 init_scf_run 11 5.9 0.000 0.001 5.799 5.800 scf_env_initial_rho_setup 11 6.9 0.034 0.037 5.799 5.799 dbcsr_mm_accdrv_process 17442 15.9 3.129 4.165 4.341 5.303 qs_ot_get_p 110 10.4 0.134 0.134 5.173 5.246 make_m2s 4110 13.4 0.038 0.040 4.393 4.639 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.110 4.595 make_images 4110 14.4 0.650 0.709 4.263 4.505 pw_transfer 1331 11.6 0.066 0.071 3.937 3.956 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.798 3.853 apply_single 110 13.6 0.000 0.000 3.798 3.853 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.829 3.851 ot_diis_step 99 11.5 0.012 0.012 3.756 3.756 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.593 3.598 grid_integrate_task_list 110 12.3 3.383 3.496 3.383 3.496 fft_wrap_pw1pw2_140 451 13.1 0.778 0.788 3.281 3.307 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.242 3.243 density_rs2pw 110 9.6 0.004 0.004 3.015 3.217 cp_fm_cholesky_invert 11 10.9 3.144 3.148 3.144 3.148 qs_energies_init_hamiltonians 11 5.9 0.087 0.088 3.102 3.102 calculate_dm_sparse 110 9.5 0.001 0.001 2.889 2.932 calculate_first_density_matrix 1 7.0 0.000 0.000 2.916 2.922 wfi_extrapolate 11 7.9 0.001 0.001 2.753 2.753 hybrid_alltoall_any 4261 16.3 0.199 0.848 2.341 2.737 make_images_data 4110 15.4 0.041 0.046 2.357 2.683 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.648 2.656 cp_fm_diag_elpa_base 48 14.0 2.588 2.614 2.644 2.652 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.542 2.544 fft3d_ps 1111 14.6 1.136 1.162 2.494 2.510 grid_collocate_task_list 110 9.6 2.370 2.475 2.370 2.475 multiply_cannon_sync_h2d 8220 15.4 2.339 2.450 2.339 2.450 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.130 2.180 mp_sum_dm 438 4.9 2.031 2.169 2.031 2.169 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.136 2.151 md_output 10 3.0 0.000 0.000 0.135 2.117 md_write_output 11 3.9 0.129 2.056 0.134 2.115 update_particle_set 20 4.0 0.000 0.000 1.979 2.111 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.031 2.065 potential_pw2rs 110 12.3 0.015 0.016 2.055 2.061 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.790 2.000 qs_env_update_s_mstruct 11 6.9 0.011 0.021 1.859 1.965 mp_alltoall_d11v 2046 13.8 1.817 1.951 1.817 1.951 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.818 1.830 cp_fm_cholesky_decompose 22 10.9 1.772 1.789 1.772 1.789 acc_transpose_blocks 8220 15.4 0.039 0.040 1.505 1.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.630 1.634 mp_allgather_i34 2055 14.4 0.467 1.606 0.467 1.606 dbcsr_complete_redistribute 325 12.2 0.564 0.598 1.521 1.604 jit_kernel_multiply 8 15.5 0.899 1.562 0.899 1.562 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.473 1.406 transfer_rs2pw 451 10.6 0.005 0.005 1.177 1.400 mp_irecv_dv 24056 15.7 0.447 1.358 0.447 1.358 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.892 1.341 mp_waitany 9240 13.8 1.072 1.313 1.072 1.313 qs_create_task_list 11 7.9 0.001 0.001 1.210 1.308 generate_qs_task_list 11 8.9 0.372 0.439 1.209 1.307 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.254 1.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.662000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=819.727273, yerr=11.917069 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 1.366565E+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 0.015 0.050 87.102 87.120 qs_mol_dyn_low 1 2.0 0.003 0.004 86.803 86.813 qs_forces 11 3.9 0.003 0.003 86.719 86.719 qs_energies 11 4.9 0.001 0.001 82.632 82.633 scf_env_do_scf 11 5.9 0.001 0.001 72.205 72.205 velocity_verlet 10 3.0 0.004 0.016 55.304 55.311 scf_env_do_scf_inner_loop 99 6.5 0.053 0.057 43.964 43.967 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.563 29.647 init_scf_loop 11 6.9 0.000 0.000 28.171 28.177 qs_scf_new_mos 99 7.5 0.001 0.001 26.769 26.815 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.769 26.814 prepare_preconditioner 11 7.9 0.000 0.000 26.052 26.059 make_preconditioner 11 8.9 0.000 0.000 26.052 26.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.403 25.543 ot_scf_mini 99 9.5 0.002 0.002 24.967 25.007 multiply_cannon 2055 13.4 0.341 0.355 22.568 23.177 multiply_cannon_loop 2055 14.4 0.835 0.857 20.807 21.172 cp_fm_upper_to_full 70 14.2 12.612 17.996 12.612 17.996 ot_mini 99 10.5 0.001 0.001 14.138 14.180 rebuild_ks_matrix 110 8.3 0.000 0.001 13.568 13.612 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.568 13.611 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.320 12.360 dbcsr_complete_redistribute 325 12.2 1.025 1.046 7.200 10.313 multiply_cannon_multrec 8220 15.4 4.031 4.192 9.770 9.874 qs_ot_get_derivative 99 11.5 0.001 0.001 9.514 9.555 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.200 9.309 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.633 8.716 mp_waitall_1 84994 16.7 7.538 8.354 7.538 8.354 mp_alltoall_i22 605 13.7 5.262 8.352 5.262 8.352 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.026 7.062 calculate_rho_elec 110 8.6 0.224 0.224 7.026 7.061 sum_up_and_integrate 110 10.3 0.002 0.002 6.739 6.765 integrate_v_rspace 110 11.3 0.004 0.004 6.711 6.737 init_scf_run 11 5.9 0.000 0.001 6.180 6.180 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.180 6.180 dbcsr_mm_accdrv_process 11614 15.7 3.849 4.108 5.591 5.787 make_m2s 4110 13.4 0.043 0.044 5.257 5.725 qs_ot_get_p 110 10.4 0.065 0.065 5.602 5.662 make_images 4110 14.4 0.888 0.946 5.067 5.535 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.972 5.296 cp_fm_cholesky_invert 11 10.9 5.278 5.283 5.278 5.283 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.792 5.260 apply_single 110 13.6 0.000 0.000 4.792 5.259 pw_transfer 1331 11.6 0.074 0.075 4.933 4.960 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.815 4.843 ot_diis_step 99 11.5 0.015 0.015 4.604 4.605 fft_wrap_pw1pw2_140 451 13.1 1.278 1.282 4.208 4.234 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.960 3.967 grid_integrate_task_list 110 12.3 3.679 3.754 3.679 3.754 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.645 3.645 density_rs2pw 110 9.6 0.004 0.004 3.491 3.513 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.017 3.503 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.440 3.440 hybrid_alltoall_any 4261 16.3 0.263 0.561 2.807 3.425 make_images_data 4110 15.4 0.045 0.049 2.722 3.320 wfi_extrapolate 11 7.9 0.001 0.001 3.284 3.284 calculate_dm_sparse 110 9.5 0.001 0.001 3.171 3.196 multiply_cannon_sync_h2d 8220 15.4 3.103 3.149 3.103 3.149 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.967 2.970 fft3d_ps 1111 14.6 1.303 1.319 2.870 2.899 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.852 2.853 cp_fm_diag_elpa_base 48 14.0 2.310 2.505 2.850 2.850 calculate_first_density_matrix 1 7.0 0.000 0.000 2.783 2.783 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.658 2.686 grid_collocate_task_list 110 9.6 2.668 2.682 2.668 2.682 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.455 2.470 potential_pw2rs 110 12.3 0.022 0.022 2.273 2.295 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.165 2.224 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.176 mp_alltoall_d11v 2046 13.8 2.005 2.061 2.005 2.061 cp_fm_cholesky_decompose 22 10.9 2.007 2.025 2.007 2.025 qs_create_task_list 11 7.9 0.001 0.001 1.875 1.922 generate_qs_task_list 11 8.9 0.730 0.783 1.875 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.849 1.853 acc_transpose_blocks 8220 15.4 0.041 0.042 1.764 1.805 jit_kernel_multiply 10 15.2 1.547 1.795 1.547 1.795 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.729 1.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.120000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1242.818182, yerr=49.119184 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 628.789248E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 1587544. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.098 210.311 210.312 qs_mol_dyn_low 1 2.0 0.012 0.072 209.357 209.706 qs_forces 11 3.9 0.007 0.024 209.095 209.096 qs_energies 11 4.9 0.002 0.008 203.484 203.507 scf_env_do_scf 11 5.9 0.001 0.006 186.919 186.922 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 166.243 166.245 dbcsr_multiply_generic 2527 12.6 0.182 0.192 127.824 128.338 qs_scf_new_mos 118 7.6 0.001 0.001 126.534 126.884 qs_scf_loop_do_ot 118 8.6 0.001 0.001 126.533 126.883 velocity_verlet 10 3.0 0.001 0.003 125.083 125.086 ot_scf_mini 118 9.6 0.003 0.004 119.760 120.074 multiply_cannon 2527 13.6 0.242 0.255 103.742 105.666 multiply_cannon_loop 2527 14.6 2.408 2.468 101.558 103.559 ot_mini 118 10.6 0.001 0.001 67.916 68.310 multiply_cannon_multrec 60648 15.6 32.056 34.056 43.488 45.551 qs_ot_get_derivative 118 11.6 0.001 0.002 42.903 43.229 rebuild_ks_matrix 129 8.3 0.001 0.001 33.770 34.441 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.018 33.770 34.441 mp_waitall_1 269254 16.5 28.802 32.958 28.802 32.958 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.362 30.996 qs_ot_get_p 129 10.4 0.010 0.028 29.455 29.753 multiply_cannon_sync_h2d 60648 15.6 26.626 28.556 26.626 28.556 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.478 25.326 apply_single 129 13.6 0.001 0.001 24.478 25.326 ot_diis_step 118 11.6 0.008 0.008 24.739 24.740 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.795 22.843 init_scf_loop 11 6.9 0.014 0.108 20.600 20.601 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.882 19.883 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.500 19.769 multiply_cannon_metrocomm3 60648 15.6 0.118 0.123 15.970 18.064 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.893 16.928 cp_fm_diag_elpa_base 83 14.4 16.826 16.859 16.889 16.922 prepare_preconditioner 11 7.9 0.000 0.000 15.872 15.941 make_preconditioner 11 8.9 0.000 0.001 15.872 15.941 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.110 15.316 make_m2s 5054 13.6 0.106 0.114 13.920 14.373 make_images 5054 14.6 0.401 0.421 13.734 14.198 sum_up_and_integrate 129 10.3 0.002 0.005 14.107 14.125 integrate_v_rspace 129 11.3 0.004 0.005 14.047 14.067 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.226 13.388 calculate_rho_elec 129 8.7 0.045 0.064 13.225 13.388 init_scf_run 11 5.9 0.000 0.001 12.368 12.369 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.368 12.369 dbcsr_mm_accdrv_process 125468 16.2 4.757 4.885 10.997 11.486 mp_sum_l 8010 12.9 8.934 10.541 8.934 10.541 multiply_cannon_metrocomm1 60648 15.6 0.098 0.101 6.310 9.328 cp_fm_cholesky_invert 11 10.9 9.081 9.089 9.081 9.089 wfi_extrapolate 11 7.9 0.001 0.001 9.067 9.067 calculate_dm_sparse 129 9.5 0.001 0.001 8.608 8.689 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.393 8.551 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.157 8.206 pw_transfer 1559 11.6 0.075 0.088 7.671 7.848 make_images_data 5054 15.6 0.067 0.072 6.791 7.758 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.466 7.642 grid_integrate_task_list 129 12.3 7.115 7.478 7.115 7.478 density_rs2pw 129 9.7 0.006 0.007 6.804 7.391 hybrid_alltoall_any 5240 16.5 0.297 2.287 5.956 7.193 fft_wrap_pw1pw2_140 527 13.2 1.134 1.188 6.538 6.710 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.687 6.700 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.807 5.909 fft3d_ps 1301 14.7 2.208 2.869 5.456 5.792 mp_alltoall_d11v 2423 14.1 4.337 5.686 4.337 5.686 grid_collocate_task_list 129 9.7 4.875 5.288 4.875 5.288 cp_fm_cholesky_decompose 22 10.9 4.611 4.625 4.611 4.625 potential_pw2rs 129 12.3 0.009 0.010 4.552 4.576 mp_sum_d 4499 12.1 3.514 4.373 3.514 4.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.312000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.000000, yerr=5.736645 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 844.992512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6187730. 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 0.008 0.027 188.864 188.865 qs_mol_dyn_low 1 2.0 0.003 0.004 188.524 188.539 qs_forces 11 3.9 0.004 0.005 188.440 188.441 qs_energies 11 4.9 0.001 0.002 181.698 181.708 scf_env_do_scf 11 5.9 0.001 0.001 164.878 164.888 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.015 132.017 velocity_verlet 10 3.0 0.001 0.001 118.764 118.766 dbcsr_multiply_generic 2507 12.6 0.189 0.195 97.894 99.149 qs_scf_new_mos 117 7.6 0.001 0.001 93.642 94.204 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.641 94.203 ot_scf_mini 117 9.6 0.004 0.004 88.829 89.393 multiply_cannon 2507 13.6 0.503 0.562 77.802 83.005 multiply_cannon_loop 2507 14.6 1.573 1.641 74.340 76.976 ot_mini 117 10.6 0.001 0.001 50.189 50.770 mp_waitall_1 214728 16.6 24.107 40.242 24.107 40.242 multiply_cannon_multrec 30084 15.6 21.038 25.483 31.788 36.826 rebuild_ks_matrix 128 8.3 0.001 0.001 32.176 32.860 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.175 32.860 init_scf_loop 11 6.9 0.000 0.000 32.776 32.777 multiply_cannon_metrocomm3 30084 15.6 0.095 0.099 15.783 29.879 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.944 29.565 qs_ot_get_derivative 117 11.6 0.001 0.001 28.201 28.734 prepare_preconditioner 11 7.9 0.000 0.000 28.451 28.518 make_preconditioner 11 8.9 0.000 0.000 28.451 28.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.080 27.627 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.074 23.242 apply_single 128 13.6 0.001 0.001 22.074 23.241 ot_diis_step 117 11.6 0.014 0.015 21.808 21.809 qs_ot_get_p 128 10.4 0.001 0.001 20.630 21.249 multiply_cannon_sync_h2d 30084 15.6 18.062 20.637 18.062 20.637 cp_fm_cholesky_invert 11 10.9 16.539 16.551 16.539 16.551 make_m2s 5014 13.6 0.089 0.099 14.169 15.964 make_images 5014 14.6 1.166 1.368 13.956 15.749 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.707 15.742 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.509 14.510 sum_up_and_integrate 128 10.3 0.002 0.004 13.885 13.913 integrate_v_rspace 128 11.3 0.003 0.004 13.825 13.853 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.266 13.293 calculate_rho_elec 128 8.7 0.087 0.104 13.265 13.293 init_scf_run 11 5.9 0.000 0.001 12.035 12.036 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.035 12.036 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.195 11.598 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.363 11.395 cp_fm_diag_elpa_base 83 14.4 11.110 11.211 11.357 11.385 dbcsr_mm_accdrv_process 62242 16.2 5.483 6.596 10.196 10.961 multiply_cannon_metrocomm4 27577 15.6 0.102 0.116 3.852 10.841 mp_irecv_dv 69486 16.3 3.649 10.440 3.649 10.440 make_images_data 5014 15.6 0.066 0.076 8.338 10.387 hybrid_alltoall_any 5200 16.5 0.349 1.553 7.108 9.769 pw_transfer 1547 11.6 0.085 0.097 8.502 8.566 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.279 8.347 wfi_extrapolate 11 7.9 0.001 0.001 8.344 8.344 density_rs2pw 128 9.7 0.006 0.007 6.998 7.603 grid_integrate_task_list 128 12.3 7.180 7.485 7.180 7.485 fft_wrap_pw1pw2_140 523 13.2 1.213 1.234 7.350 7.440 calculate_dm_sparse 128 9.5 0.001 0.001 6.908 7.051 cp_fm_cholesky_decompose 22 10.9 6.970 7.050 6.970 7.050 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.250 7.043 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.204 6.212 mp_sum_l 7950 12.9 4.118 6.091 4.118 6.091 fft3d_ps 1291 14.7 2.834 3.009 5.842 5.883 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.358 5.449 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.266 5.417 grid_collocate_task_list 128 9.7 5.001 5.326 5.001 5.326 mp_allgather_i34 2507 14.6 1.992 5.091 1.992 5.091 mp_alltoall_d11v 2415 14.1 4.154 4.613 4.154 4.613 potential_pw2rs 128 12.3 0.016 0.019 4.386 4.397 dbcsr_complete_redistribute 395 12.7 0.809 0.917 3.237 4.066 mp_sum_d 4465 12.1 2.618 3.938 2.618 3.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.865000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.636364, yerr=2.739367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 962.682880E+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 4007 58135. MP_Allreduce 11095 999. 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 0.036 0.101 174.285 174.288 qs_mol_dyn_low 1 2.0 0.004 0.004 173.729 173.759 qs_forces 11 3.9 0.004 0.005 173.623 173.625 qs_energies 11 4.9 0.008 0.058 167.053 167.065 scf_env_do_scf 11 5.9 0.001 0.003 151.457 151.462 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.563 116.563 velocity_verlet 10 3.0 0.002 0.004 110.825 110.828 dbcsr_multiply_generic 2485 12.5 0.178 0.183 82.430 83.560 qs_scf_new_mos 116 7.6 0.001 0.001 80.046 80.382 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.045 80.381 ot_scf_mini 116 9.6 0.003 0.004 75.841 76.241 multiply_cannon 2485 13.5 0.505 0.524 62.763 68.223 multiply_cannon_loop 2485 14.5 1.118 1.182 59.844 62.275 ot_mini 116 10.6 0.001 0.001 43.252 43.660 init_scf_loop 11 6.9 0.001 0.005 34.792 34.793 mp_waitall_1 169034 16.6 24.526 34.377 24.526 34.377 prepare_preconditioner 11 7.9 0.000 0.001 30.798 30.839 make_preconditioner 11 8.9 0.000 0.002 30.798 30.839 rebuild_ks_matrix 127 8.3 0.001 0.001 29.849 30.285 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.849 30.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.428 29.886 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.869 27.269 multiply_cannon_multrec 19880 15.5 13.136 15.858 23.360 26.048 qs_ot_get_derivative 116 11.6 0.001 0.002 23.979 24.381 multiply_cannon_metrocomm3 19880 15.5 0.061 0.065 15.151 24.344 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.403 20.347 apply_single 127 13.6 0.001 0.001 19.403 20.346 ot_diis_step 116 11.6 0.017 0.018 19.163 19.164 qs_ot_get_p 127 10.4 0.001 0.002 18.099 18.575 make_m2s 4970 13.5 0.078 0.083 14.557 15.712 make_images 4970 14.5 1.135 1.230 14.324 15.468 multiply_cannon_sync_h2d 19880 15.5 13.524 15.103 13.524 15.103 cp_fm_cholesky_invert 11 10.9 14.210 14.219 14.210 14.219 qs_ot_p2m_diag 82 11.4 0.262 0.273 13.816 13.826 sum_up_and_integrate 127 10.3 0.002 0.003 13.733 13.756 integrate_v_rspace 127 11.3 0.003 0.003 13.674 13.697 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.333 13.364 calculate_rho_elec 127 8.7 0.129 0.144 13.333 13.364 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.785 12.786 init_scf_run 11 5.9 0.000 0.001 10.553 10.555 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.553 10.555 make_images_data 4970 15.5 0.061 0.071 8.780 10.335 hybrid_alltoall_any 5155 16.4 0.445 2.016 7.686 10.100 dbcsr_mm_accdrv_process 41158 16.2 6.036 7.028 9.687 9.954 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.676 9.699 cp_fm_diag_elpa_base 82 14.4 9.272 9.409 9.673 9.696 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.008 9.288 multiply_cannon_metrocomm4 17395 15.5 0.064 0.073 3.404 9.222 mp_irecv_dv 49801 16.2 3.278 8.970 3.278 8.970 pw_transfer 1535 11.6 0.084 0.102 8.530 8.633 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.308 8.417 grid_integrate_task_list 127 12.3 7.256 7.676 7.256 7.676 fft_wrap_pw1pw2_140 519 13.2 1.290 1.317 7.401 7.519 wfi_extrapolate 11 7.9 0.001 0.001 7.410 7.410 cp_fm_upper_to_full 104 14.8 5.850 7.371 5.850 7.371 cp_fm_cholesky_decompose 22 10.9 7.298 7.336 7.298 7.336 density_rs2pw 127 9.7 0.006 0.006 6.737 7.060 dbcsr_complete_redistribute 393 12.7 1.174 1.208 4.565 6.344 calculate_dm_sparse 127 9.5 0.001 0.001 5.927 6.038 fft3d_ps 1281 14.7 2.739 2.958 5.708 5.795 grid_collocate_task_list 127 9.7 5.123 5.548 5.123 5.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.444 5.450 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.702 5.417 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.423 5.195 mp_alltoall_d11v 2401 14.1 4.247 5.031 4.247 5.031 mp_sum_l 7884 12.9 3.327 4.968 3.327 4.968 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.543 4.660 mp_allgather_i34 2485 14.5 1.437 4.649 1.437 4.649 potential_pw2rs 127 12.3 0.021 0.023 4.228 4.239 transfer_fm_to_dbcsr 11 9.9 0.020 0.027 2.350 4.108 mp_alltoall_i22 712 14.1 1.968 3.919 1.968 3.919 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.855 3.883 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 3.716 3.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.476 3.513 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.288000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=911.636364, yerr=13.158286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.155285E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3994 58319. MP_Allreduce 11062 1083. MP_Sync 87 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 187.922 187.923 qs_mol_dyn_low 1 2.0 0.003 0.004 187.532 187.545 qs_forces 11 3.9 0.004 0.005 187.411 187.425 qs_energies 11 4.9 0.002 0.006 180.315 180.329 scf_env_do_scf 11 5.9 0.001 0.001 163.203 163.218 velocity_verlet 10 3.0 0.002 0.002 124.599 124.603 scf_env_do_scf_inner_loop 117 6.6 0.023 0.166 116.886 116.887 dbcsr_multiply_generic 2507 12.6 0.187 0.192 81.079 81.894 qs_scf_new_mos 117 7.6 0.001 0.001 80.526 80.859 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.525 80.858 ot_scf_mini 117 9.6 0.004 0.006 75.950 76.300 multiply_cannon 2507 13.6 0.554 0.589 56.090 59.329 multiply_cannon_loop 2507 14.6 1.814 1.908 52.420 54.560 init_scf_loop 11 6.9 0.001 0.004 46.186 46.187 ot_mini 117 10.6 0.001 0.001 43.238 43.593 prepare_preconditioner 11 7.9 0.000 0.000 42.132 42.158 make_preconditioner 11 8.9 0.000 0.001 42.132 42.158 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.775 40.746 multiply_cannon_multrec 30084 15.6 13.580 18.554 26.769 31.506 rebuild_ks_matrix 128 8.3 0.001 0.001 29.208 29.532 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.207 29.531 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.329 26.613 mp_waitall_1 147882 16.7 17.308 26.590 17.308 26.590 qs_ot_get_derivative 117 11.6 0.001 0.002 23.636 23.995 make_m2s 5014 13.6 0.096 0.100 20.454 21.461 make_images 5014 14.6 1.946 2.317 20.142 21.150 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.957 19.530 apply_single 128 13.6 0.001 0.001 18.957 19.530 ot_diis_step 117 11.6 0.018 0.018 19.472 19.473 qs_ot_get_p 128 10.4 0.002 0.002 18.865 19.225 cp_fm_upper_to_full 105 14.8 11.448 16.815 11.448 16.815 cp_fm_cholesky_invert 11 10.9 16.070 16.082 16.070 16.082 multiply_cannon_metrocomm3 30084 15.6 0.049 0.053 6.394 15.323 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.611 14.664 sum_up_and_integrate 128 10.3 0.002 0.003 13.947 13.973 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.892 13.951 calculate_rho_elec 128 8.7 0.173 0.188 13.892 13.950 integrate_v_rspace 128 11.3 0.003 0.004 13.888 13.915 dbcsr_mm_accdrv_process 62264 16.2 8.704 9.541 12.761 13.277 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.209 13.211 dbcsr_complete_redistribute 395 12.7 1.519 1.653 9.034 12.758 make_images_data 5014 15.6 0.066 0.072 10.980 12.603 multiply_cannon_sync_h2d 30084 15.6 10.468 11.812 10.468 11.812 hybrid_alltoall_any 5200 16.5 0.528 2.205 9.741 11.628 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.642 11.364 init_scf_run 11 5.9 0.000 0.001 11.111 11.113 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.111 11.112 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.944 10.207 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.121 10.132 cp_fm_diag_elpa_base 83 14.4 9.163 9.484 10.115 10.127 transfer_fm_to_dbcsr 11 9.9 0.000 0.002 6.335 9.979 mp_alltoall_i22 716 14.1 5.560 9.430 5.560 9.430 pw_transfer 1547 11.6 0.085 0.102 8.929 8.996 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.704 8.777 fft_wrap_pw1pw2_140 523 13.2 1.445 1.481 7.761 7.850 grid_integrate_task_list 128 12.3 7.480 7.810 7.480 7.810 wfi_extrapolate 11 7.9 0.001 0.001 7.721 7.721 cp_fm_cholesky_decompose 22 10.9 7.537 7.626 7.537 7.626 density_rs2pw 128 9.7 0.006 0.006 6.781 7.187 multiply_cannon_metrocomm4 25070 15.6 0.082 0.092 2.842 6.701 calculate_dm_sparse 128 9.5 0.001 0.001 6.527 6.618 mp_irecv_dv 76098 16.2 2.687 6.424 2.687 6.424 fft3d_ps 1291 14.7 2.843 2.901 5.885 5.931 grid_collocate_task_list 128 9.7 5.282 5.680 5.282 5.680 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.545 5.613 mp_alltoall_d11v 2415 14.1 5.018 5.501 5.018 5.501 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.503 4.594 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.406 4.516 qs_energies_init_hamiltonians 11 5.9 0.011 0.027 4.505 4.506 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.231 4.315 potential_pw2rs 128 12.3 0.023 0.023 4.208 4.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.923000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.454545, yerr=18.729479 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/19/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 5.820059E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.555919E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3975 58590. MP_Allreduce 11015 1169. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.058 171.346 171.363 qs_mol_dyn_low 1 2.0 0.018 0.023 170.421 170.453 qs_forces 11 3.9 0.020 0.023 170.077 170.083 qs_energies 11 4.9 0.002 0.002 162.607 162.613 scf_env_do_scf 11 5.9 0.001 0.001 143.978 143.992 velocity_verlet 10 3.0 0.244 0.245 110.672 110.677 scf_env_do_scf_inner_loop 116 6.6 0.004 0.010 108.169 108.171 dbcsr_multiply_generic 2485 12.5 0.179 0.189 73.445 73.817 qs_scf_new_mos 116 7.6 0.001 0.001 71.812 71.901 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.811 71.900 ot_scf_mini 116 9.6 0.003 0.003 67.422 67.483 multiply_cannon 2485 13.5 0.558 0.594 54.691 58.308 multiply_cannon_loop 2485 14.5 0.795 0.823 51.785 52.568 ot_mini 116 10.6 0.001 0.001 37.778 37.840 init_scf_loop 11 6.9 0.000 0.000 35.659 35.660 prepare_preconditioner 11 7.9 0.000 0.000 31.839 31.865 make_preconditioner 11 8.9 0.000 0.000 31.839 31.865 mp_waitall_1 124680 16.7 24.871 31.196 24.871 31.196 make_full_inverse_cholesky 11 9.9 0.013 0.028 29.704 29.982 rebuild_ks_matrix 127 8.3 0.001 0.001 28.770 28.855 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.770 28.855 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.187 26.262 multiply_cannon_multrec 9940 15.5 10.293 14.514 18.664 21.694 qs_ot_get_derivative 116 11.6 0.001 0.001 20.934 20.999 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.437 19.282 cp_fm_cholesky_invert 11 10.9 18.195 18.201 18.195 18.201 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.018 17.447 apply_single 127 13.6 0.001 0.001 17.018 17.447 qs_ot_get_p 127 10.4 0.002 0.002 17.042 17.107 ot_diis_step 116 11.6 0.020 0.020 16.779 16.780 make_m2s 4970 13.5 0.064 0.068 14.776 15.576 make_images 4970 14.5 2.154 2.584 14.471 15.271 sum_up_and_integrate 127 10.3 0.002 0.002 13.985 14.031 integrate_v_rspace 127 11.3 0.004 0.004 13.926 13.973 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.911 13.952 calculate_rho_elec 127 8.7 0.254 0.264 13.910 13.952 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.294 13.312 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.131 12.133 init_scf_run 11 5.9 0.000 0.001 11.864 11.864 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.864 11.864 multiply_cannon_sync_h2d 9940 15.5 10.809 11.240 10.809 11.240 make_images_data 4970 15.5 0.055 0.064 8.518 9.824 hybrid_alltoall_any 5155 16.4 0.834 3.739 8.343 9.755 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.265 9.276 cp_fm_diag_elpa_base 82 14.4 9.028 9.119 9.261 9.273 pw_transfer 1535 11.6 0.085 0.094 9.236 9.266 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.014 9.049 dbcsr_mm_accdrv_process 20590 16.1 3.945 5.294 8.022 8.676 cp_fm_cholesky_decompose 22 10.9 7.976 8.085 7.976 8.085 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.989 8.022 grid_integrate_task_list 127 12.3 7.697 8.006 7.697 8.006 fft_wrap_pw1pw2_140 519 13.2 1.789 1.825 7.971 8.004 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.664 7.615 calculate_dm_sparse 127 9.5 0.001 0.001 7.542 7.604 wfi_extrapolate 11 7.9 0.001 0.001 7.354 7.354 density_rs2pw 127 9.7 0.005 0.006 6.658 6.934 fft3d_ps 1281 14.7 2.747 2.830 5.768 5.786 grid_collocate_task_list 127 9.7 5.524 5.763 5.524 5.763 dbcsr_complete_redistribute 393 12.7 2.106 2.153 5.258 5.632 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.211 5.212 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.131 5.143 mp_alltoall_d11v 2401 14.1 4.656 5.088 4.656 5.088 mp_allgather_i34 2485 14.5 1.119 4.597 1.119 4.597 calculate_first_density_matrix 1 7.0 0.000 0.000 4.258 4.259 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.990 4.221 potential_pw2rs 127 12.3 0.026 0.028 4.190 4.196 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.030 4.061 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.596 3.929 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.825 3.840 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.422 3.754 multiply_cannon_metrocomm4 7455 15.5 0.026 0.027 1.731 3.670 mp_irecv_dv 28618 15.9 1.692 3.603 1.692 3.603 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.555 3.594 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.414 3.512 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.481 3.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.363000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1460.272727, yerr=36.349543 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.193311E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 58741. MP_Allreduce 11152 1501. MP_Sync 87 MP_Alltoall 1724 36993717. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.056 292.011 292.025 qs_mol_dyn_low 1 2.0 0.003 0.004 291.471 291.487 qs_forces 11 3.9 0.004 0.005 291.375 291.380 qs_energies 11 4.9 0.002 0.002 282.583 282.594 scf_env_do_scf 11 5.9 0.001 0.001 260.029 260.038 velocity_verlet 10 3.0 0.002 0.002 211.088 211.096 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 135.139 135.141 init_scf_loop 11 6.9 0.000 0.000 124.616 124.617 prepare_preconditioner 11 7.9 0.000 0.000 119.860 119.879 make_preconditioner 11 8.9 0.000 0.000 119.860 119.879 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.551 117.015 qs_scf_new_mos 118 7.6 0.001 0.001 91.089 91.137 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.088 91.136 ot_scf_mini 118 9.6 0.004 0.004 86.178 86.183 dbcsr_multiply_generic 2527 12.6 0.213 0.225 83.090 83.477 cp_fm_upper_to_full 105 14.8 53.136 76.416 53.136 76.416 multiply_cannon 2527 13.6 0.684 0.736 59.855 61.447 multiply_cannon_loop 2527 14.6 1.045 1.052 55.824 57.002 ot_mini 118 10.6 0.001 0.001 44.914 44.923 dbcsr_complete_redistribute 395 12.7 3.987 4.030 30.246 43.458 copy_fm_to_dbcsr 209 11.7 0.001 0.001 26.744 39.928 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.266 37.249 mp_alltoall_i22 718 14.1 22.029 35.331 22.029 35.331 rebuild_ks_matrix 129 8.3 0.001 0.001 34.228 34.237 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.228 34.237 cp_fm_cholesky_invert 11 10.9 33.248 33.254 33.248 33.254 mp_waitall_1 104500 16.8 28.223 32.587 28.223 32.587 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.820 31.848 qs_ot_get_p 129 10.4 0.002 0.002 25.999 26.042 qs_ot_get_derivative 118 11.6 0.002 0.002 25.431 25.438 qs_ot_p2m_diag 83 11.4 0.879 0.886 21.878 21.906 make_m2s 5054 13.6 0.076 0.078 18.735 20.223 multiply_cannon_metrocomm3 10108 15.6 0.025 0.026 18.962 20.169 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.080 20.081 make_images 5054 14.6 3.094 3.281 18.251 19.741 ot_diis_step 118 11.6 0.022 0.022 19.458 19.459 multiply_cannon_multrec 10108 15.6 10.526 12.575 18.825 19.034 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.745 18.821 apply_single 129 13.6 0.001 0.001 18.744 18.821 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.898 16.898 cp_fm_diag_elpa_base 83 14.4 12.521 14.121 16.893 16.893 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.756 16.765 calculate_rho_elec 129 8.7 0.485 0.485 16.755 16.765 sum_up_and_integrate 129 10.3 0.002 0.002 16.311 16.404 integrate_v_rspace 129 11.3 0.004 0.005 16.248 16.341 multiply_cannon_sync_h2d 10108 15.6 14.430 14.470 14.430 14.470 hybrid_alltoall_any 5240 16.5 1.314 3.056 10.765 13.113 make_images_data 5054 15.6 0.065 0.071 10.535 12.759 init_scf_run 11 5.9 0.000 0.001 12.238 12.239 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.237 12.239 pw_transfer 1559 11.6 0.096 0.096 11.545 11.552 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.305 11.312 dbcsr_mm_accdrv_process 20926 16.1 4.583 6.700 8.048 10.103 fft_wrap_pw1pw2_140 527 13.2 3.064 3.113 10.088 10.099 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.751 9.758 cp_fm_cholesky_decompose 22 10.9 8.979 9.004 8.979 9.004 wfi_extrapolate 11 7.9 0.001 0.001 8.943 8.943 grid_integrate_task_list 129 12.3 8.615 8.805 8.615 8.805 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.148 8.149 density_rs2pw 129 9.7 0.005 0.005 7.645 7.702 mp_alltoall_d11v 2423 14.1 7.085 7.251 7.085 7.251 calculate_dm_sparse 129 9.5 0.001 0.001 6.938 7.002 fft3d_ps 1301 14.7 2.877 2.888 6.520 6.560 grid_collocate_task_list 129 9.7 6.463 6.534 6.463 6.534 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.361 6.405 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.147 6.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.025000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2873.818182, yerr=164.430045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 1.259852E+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 16861011. 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 0.012 0.030 85.628 85.629 qs_energies 1 2.0 0.000 0.000 84.785 84.798 ls_scf 1 3.0 0.000 0.000 83.807 83.820 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.344 72.508 multiply_cannon 111 7.7 0.018 0.020 55.784 57.172 multiply_cannon_loop 111 8.7 0.226 0.240 52.392 53.952 ls_scf_main 1 4.0 0.000 0.000 52.174 52.174 density_matrix_trs4 2 5.0 0.002 0.003 46.652 46.773 ls_scf_init_scf 1 4.0 0.000 0.000 28.300 28.300 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.180 27.230 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.024 25.045 mp_waitall_1 11031 10.9 22.096 24.329 22.096 24.329 multiply_cannon_multrec 2664 9.7 8.177 8.968 15.560 17.374 multiply_cannon_sync_h2d 2664 9.7 13.704 15.255 13.704 15.255 make_m2s 222 7.7 0.009 0.011 12.983 13.474 make_images 222 8.7 0.099 0.110 12.961 13.454 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.520 12.199 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.348 8.097 make_images_data 222 9.7 0.004 0.005 7.556 8.094 dbcsr_mm_accdrv_process 4760 10.4 0.589 0.728 7.000 7.977 hybrid_alltoall_any 227 10.6 0.216 1.845 6.527 7.666 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.101 6.212 7.101 calculate_norms 4752 9.8 5.536 6.163 5.536 6.163 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.011 5.187 mp_sum_l 887 5.1 3.038 4.388 3.038 4.388 make_images_sizes 222 9.7 0.000 0.000 0.698 3.354 mp_alltoall_i44 222 10.7 0.698 3.354 0.698 3.354 ls_scf_post 1 4.0 0.000 0.000 3.333 3.346 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.254 3.280 arnoldi_extremal 4 6.8 0.000 0.000 3.227 3.256 arnoldi_normal_ev 4 7.8 0.001 0.003 3.227 3.256 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.048 3.195 mp_irecv_dv 6231 10.9 2.030 3.168 2.030 3.168 build_subspace 16 8.4 0.009 0.012 3.131 3.133 ls_scf_store_result 1 5.0 0.000 0.000 2.829 2.867 dbcsr_special_finalize 555 9.7 0.005 0.006 2.374 2.827 dbcsr_merge_single_wm 555 10.7 0.459 0.604 2.365 2.818 make_images_pack 222 9.7 2.208 2.590 2.210 2.591 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.315 2.555 dbcsr_sort_data 658 11.4 2.162 2.531 2.162 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.458 2.070 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.295 2.374 buffer_matrices_ensure_size 222 8.7 1.746 2.096 1.746 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.762 1.763 rebuild_ks_matrix 3 7.3 0.000 0.000 1.753 1.754 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.753 1.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.629000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 2.167394E+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 0.065 0.243 90.617 90.620 qs_energies 1 2.0 0.002 0.016 89.396 89.418 ls_scf 1 3.0 0.002 0.016 87.955 87.963 dbcsr_multiply_generic 111 6.7 0.016 0.022 73.927 74.173 multiply_cannon 111 7.7 0.028 0.039 52.618 56.663 ls_scf_main 1 4.0 0.000 0.004 54.244 54.249 multiply_cannon_loop 111 8.7 0.135 0.147 49.803 52.604 density_matrix_trs4 2 5.0 0.002 0.005 48.642 48.833 ls_scf_init_scf 1 4.0 0.000 0.002 29.851 29.852 mp_waitall_1 9105 10.9 20.573 28.921 20.573 28.921 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.677 28.761 multiply_cannon_multrec 1332 9.7 13.335 17.126 22.622 27.534 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.205 26.224 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.395 19.743 make_m2s 222 7.7 0.006 0.008 15.069 15.651 make_images 222 8.7 1.367 1.701 15.038 15.622 dbcsr_mm_accdrv_process 4041 10.4 0.352 0.524 8.880 10.420 dbcsr_mm_accdrv_process_sort 4041 11.4 8.397 9.898 8.397 9.898 make_images_data 222 9.7 0.004 0.005 8.682 9.570 hybrid_alltoall_any 227 10.6 0.543 2.547 8.049 9.412 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.171 7.681 mp_irecv_dv 3311 11.0 3.150 7.627 3.150 7.627 mp_sum_l 887 5.1 4.857 7.398 4.857 7.398 calculate_norms 2376 9.8 6.069 6.761 6.069 6.761 multiply_cannon_sync_h2d 1332 9.7 4.913 6.156 4.913 6.156 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.607 5.819 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.939 5.165 arnoldi_extremal 4 6.8 0.000 0.000 4.681 4.706 arnoldi_normal_ev 4 7.8 0.001 0.005 4.681 4.706 build_subspace 16 8.4 0.014 0.021 4.409 4.412 ls_scf_post 1 4.0 0.001 0.009 3.858 3.863 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.485 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.171 3.414 dbcsr_matrix_vector_mult_local 304 10.0 2.785 3.276 2.788 3.278 mp_allgather_i34 111 8.7 0.805 3.055 0.805 3.055 ls_scf_dm_to_ks 2 5.0 0.000 0.009 2.613 2.690 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.145 2.441 dbcsr_data_new 4174 10.1 2.117 2.413 2.117 2.413 make_images_pack 222 9.7 1.855 2.176 1.857 2.179 dbcsr_sort_data 436 11.2 1.829 2.119 1.829 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.846 1.850 rebuild_ks_matrix 3 7.3 0.000 0.000 1.833 1.837 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.833 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.620000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 2.851164E+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 0.029 0.047 92.488 92.490 qs_energies 1 2.0 0.000 0.001 91.753 91.757 ls_scf 1 3.0 0.000 0.001 89.910 89.939 dbcsr_multiply_generic 111 6.7 0.018 0.034 74.206 74.466 ls_scf_main 1 4.0 0.000 0.001 56.192 56.197 multiply_cannon 111 7.7 0.030 0.067 51.228 55.832 multiply_cannon_loop 111 8.7 0.116 0.129 48.510 52.277 density_matrix_trs4 2 5.0 0.009 0.057 50.035 50.194 mp_waitall_1 7281 11.0 22.736 31.983 22.736 31.983 ls_scf_init_scf 1 4.0 0.000 0.002 30.072 30.102 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.865 28.987 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.511 26.524 multiply_cannon_multrec 888 9.7 12.736 15.655 21.309 24.710 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.340 21.558 make_m2s 222 7.7 0.006 0.007 16.411 17.088 make_images 222 8.7 1.585 1.857 16.373 17.051 make_images_data 222 9.7 0.004 0.004 9.547 10.540 hybrid_alltoall_any 227 10.6 0.642 2.959 9.017 10.237 dbcsr_mm_accdrv_process 3754 10.4 0.298 0.494 8.095 9.337 dbcsr_mm_accdrv_process_sort 3754 11.4 7.660 8.843 7.660 8.843 mp_sum_l 887 5.1 4.879 8.088 4.879 8.088 multiply_cannon_sync_h2d 888 9.7 6.118 7.258 6.118 7.258 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.451 7.148 mp_irecv_dv 2335 11.1 2.436 7.105 2.436 7.105 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.493 6.355 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.695 6.297 arnoldi_extremal 4 6.8 0.000 0.000 5.177 5.193 arnoldi_normal_ev 4 7.8 0.001 0.005 5.177 5.193 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.720 4.912 build_subspace 16 8.4 0.014 0.020 4.873 4.879 calculate_norms 1584 9.8 4.337 4.772 4.337 4.772 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.499 3.833 mp_allgather_i34 111 8.7 0.840 3.807 0.840 3.807 ls_scf_post 1 4.0 0.000 0.001 3.646 3.650 dbcsr_matrix_vector_mult_local 304 10.0 3.075 3.641 3.078 3.643 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.468 ls_scf_dm_to_ks 2 5.0 0.044 0.386 3.130 3.239 dbcsr_data_new 4116 9.9 2.111 2.448 2.111 2.448 dbcsr_sort_data 325 11.1 1.896 2.164 1.896 2.164 make_images_sizes 222 9.7 0.000 0.000 1.010 2.058 mp_alltoall_i44 222 10.7 1.009 2.058 1.009 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.885 1.887 make_images_pack 222 9.7 1.633 1.883 1.635 1.886 dbcsr_finalize 304 7.8 0.026 0.032 1.610 1.876 rebuild_ks_matrix 3 7.3 0.000 0.000 1.867 1.869 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.867 1.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.490000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2267.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 3.347112E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.049 98.145 98.150 qs_energies 1 2.0 0.000 0.000 97.451 97.457 ls_scf 1 3.0 0.000 0.000 95.788 95.794 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.680 78.876 ls_scf_main 1 4.0 0.000 0.000 58.787 58.788 multiply_cannon 111 7.7 0.183 0.233 51.962 56.347 density_matrix_trs4 2 5.0 0.002 0.003 52.698 52.804 multiply_cannon_loop 111 8.7 0.152 0.170 46.748 49.387 ls_scf_init_scf 1 4.0 0.000 0.000 33.746 33.748 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.487 32.566 matrix_sqrt_Newton_Schulz 2 6.5 0.017 0.024 29.260 29.272 mp_waitall_1 6369 11.0 22.832 28.961 22.832 28.961 multiply_cannon_multrec 1332 9.7 14.108 17.307 22.007 24.713 make_m2s 222 7.7 0.006 0.007 21.166 22.594 make_images 222 8.7 3.142 3.602 21.116 22.545 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.296 16.850 make_images_data 222 9.7 0.004 0.004 11.883 13.479 hybrid_alltoall_any 227 10.6 0.797 3.758 11.440 13.036 dbcsr_mm_accdrv_process 3641 10.4 0.269 0.478 7.538 9.063 dbcsr_mm_accdrv_process_sort 3641 11.4 7.100 8.569 7.100 8.569 mp_sum_l 887 5.1 3.950 7.754 3.950 7.754 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.022 6.146 multiply_cannon_sync_h2d 1332 9.7 5.460 6.139 5.460 6.139 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.091 6.012 mp_irecv_dv 3229 10.9 2.067 5.955 2.067 5.955 arnoldi_extremal 4 6.8 0.000 0.000 5.282 5.299 arnoldi_normal_ev 4 7.8 0.030 0.050 5.281 5.299 build_subspace 16 8.4 0.014 0.021 4.896 4.902 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.481 4.863 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.531 4.720 mp_allgather_i34 111 8.7 2.200 4.602 2.200 4.602 calculate_norms 2376 9.8 4.178 4.501 4.178 4.501 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.613 3.915 dbcsr_matrix_vector_mult_local 304 10.0 3.240 3.730 3.242 3.732 dbcsr_sort_data 658 11.4 3.113 3.379 3.113 3.379 ls_scf_post 1 4.0 0.000 0.000 3.255 3.261 dbcsr_special_finalize 555 9.7 0.006 0.007 2.856 3.166 dbcsr_merge_single_wm 555 10.7 0.542 0.658 2.848 3.158 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.058 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.994 3.027 dbcsr_data_release 10477 10.7 1.573 2.488 1.573 2.488 dbcsr_finalize 304 7.8 0.049 0.061 1.813 2.031 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.150000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2758.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 4.733186E+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 0.035 0.059 91.695 91.699 qs_energies 1 2.0 0.000 0.000 90.574 90.580 ls_scf 1 3.0 0.000 0.000 88.592 88.597 dbcsr_multiply_generic 111 6.7 0.018 0.019 70.264 70.465 ls_scf_main 1 4.0 0.000 0.000 55.958 55.959 multiply_cannon 111 7.7 0.095 0.210 52.293 55.339 multiply_cannon_loop 111 8.7 0.088 0.094 49.750 50.860 density_matrix_trs4 2 5.0 0.002 0.003 49.068 49.137 ls_scf_init_scf 1 4.0 0.000 0.000 29.221 29.221 mp_waitall_1 5436 11.0 24.054 28.549 24.054 28.549 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.993 28.032 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.872 25.884 multiply_cannon_multrec 444 9.7 13.666 16.606 20.753 22.158 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.598 15.307 make_m2s 222 7.7 0.005 0.005 13.351 14.267 make_images 222 8.7 2.039 2.475 13.284 14.197 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.158 13.662 make_images_data 222 9.7 0.003 0.004 8.112 9.624 hybrid_alltoall_any 227 10.6 0.800 3.817 8.033 9.442 multiply_cannon_sync_h2d 444 9.7 6.764 7.928 6.764 7.928 dbcsr_mm_accdrv_process 3003 10.4 0.363 0.400 6.765 7.926 dbcsr_mm_accdrv_process_sort 3003 11.4 6.402 7.529 6.402 7.529 arnoldi_extremal 4 6.8 0.000 0.000 5.793 5.806 arnoldi_normal_ev 4 7.8 0.002 0.004 5.793 5.806 build_subspace 16 8.4 0.015 0.020 5.401 5.410 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.430 4.659 mp_sum_l 887 5.1 2.624 4.415 2.624 4.415 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.171 4.341 dbcsr_matrix_vector_mult_local 304 10.0 3.722 4.146 3.724 4.148 calculate_norms 792 9.8 3.624 3.750 3.624 3.750 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.521 3.631 mp_irecv_dv 1241 11.2 1.508 3.611 1.508 3.611 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.833 3.608 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.487 3.537 mp_allgather_i34 111 8.7 1.122 3.472 1.122 3.472 ls_scf_post 1 4.0 0.000 0.000 3.413 3.418 make_images_sizes 222 9.7 0.000 0.000 0.857 3.256 mp_alltoall_i44 222 10.7 0.857 3.255 0.857 3.255 ls_scf_store_result 1 5.0 0.000 0.000 3.204 3.239 dbcsr_finalize 304 7.8 0.062 0.076 2.192 2.280 dbcsr_data_new 4608 9.7 1.772 2.203 1.772 2.203 dbcsr_merge_all 275 8.9 0.480 0.529 2.051 2.127 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.011 2.011 rebuild_ks_matrix 3 7.3 0.000 0.000 1.979 1.979 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.979 1.979 qs_energies_init_hamiltonians 1 3.0 0.003 0.008 1.966 1.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.699000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3776.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_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 8.852705E+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 0.142 0.158 105.326 105.328 qs_energies 1 2.0 0.000 0.000 103.795 103.799 ls_scf 1 3.0 0.000 0.000 100.687 100.690 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.343 74.483 ls_scf_main 1 4.0 0.000 0.000 63.455 63.455 density_matrix_trs4 2 5.0 0.002 0.003 54.512 54.574 multiply_cannon 111 7.7 0.181 0.263 48.032 50.092 multiply_cannon_loop 111 8.7 0.097 0.098 45.075 45.734 ls_scf_init_scf 1 4.0 0.000 0.000 33.448 33.449 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.951 31.978 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.017 29.253 29.259 mp_waitall_1 4527 11.1 21.582 25.178 21.582 25.178 make_m2s 222 7.7 0.005 0.005 22.669 23.746 make_images 222 8.7 3.579 3.872 22.561 23.636 multiply_cannon_multrec 444 9.7 17.817 18.517 22.449 23.185 hybrid_alltoall_any 227 10.6 1.651 3.617 12.826 15.804 make_images_data 222 9.7 0.003 0.004 13.076 15.187 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.915 10.321 multiply_cannon_sync_h2d 444 9.7 8.793 8.846 8.793 8.846 arnoldi_extremal 4 6.8 0.000 0.000 7.440 7.453 arnoldi_normal_ev 4 7.8 0.004 0.014 7.440 7.453 build_subspace 16 8.4 0.026 0.036 6.868 6.881 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.517 5.672 dbcsr_matrix_vector_mult_local 304 10.0 5.107 5.406 5.109 5.408 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.089 5.184 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.835 5.091 dbcsr_mm_accdrv_process 1814 10.4 0.266 0.357 4.445 4.563 dbcsr_mm_accdrv_process_sort 1814 11.4 4.110 4.242 4.110 4.242 ls_scf_post 1 4.0 0.000 0.000 3.784 3.788 make_images_sizes 222 9.7 0.000 0.000 1.463 3.609 mp_alltoall_i44 222 10.7 1.462 3.609 1.462 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.489 3.524 mp_allgather_i34 111 8.7 1.083 3.461 1.083 3.461 calculate_norms 792 9.8 3.232 3.274 3.232 3.274 dbcsr_finalize 304 7.8 0.082 0.090 3.071 3.167 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 3.076 3.077 dbcsr_merge_all 275 8.9 0.883 0.925 2.853 2.947 dbcsr_complete_redistribute 5 7.6 1.429 1.471 2.738 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.409 2.543 dbcsr_sort_data 325 11.1 2.436 2.489 2.436 2.489 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.384 2.386 dbcsr_new_transposed 4 7.5 0.241 0.252 2.303 2.318 rebuild_ks_matrix 3 7.3 0.000 0.000 2.314 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.314 2.317 dbcsr_data_new 6591 9.6 1.807 2.294 1.807 2.294 dbcsr_frobenius_norm 74 6.6 2.055 2.133 2.204 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.207 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.206 dbcsr_data_release 12724 10.6 1.978 2.172 1.978 2.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.328000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7011.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.096704E+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 4802442. 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 0.033 0.206 231.643 231.645 qs_mol_dyn_low 1 2.0 0.004 0.021 230.042 230.125 qs_forces 5 3.8 0.007 0.095 229.914 229.924 qs_energies 5 4.8 0.002 0.043 226.870 226.905 scf_env_do_scf 5 5.8 0.001 0.009 210.814 210.817 scf_env_do_scf_inner_loop 105 6.6 0.002 0.012 184.229 184.238 qs_scf_new_mos 105 7.6 0.001 0.001 144.100 144.274 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.099 144.273 dbcsr_multiply_generic 1445 12.2 0.125 0.132 134.808 135.244 ot_scf_mini 105 9.6 0.003 0.004 134.272 134.395 multiply_cannon 1445 13.2 0.276 0.287 115.442 117.408 multiply_cannon_loop 1445 14.2 2.832 2.988 113.710 114.972 velocity_verlet 4 3.0 0.003 0.015 106.207 106.208 ot_mini 105 10.6 0.010 0.313 60.176 60.654 multiply_cannon_multrec 69360 15.2 29.705 34.915 39.801 45.392 qs_ot_get_p 112 10.4 0.001 0.001 43.731 44.028 mp_waitall_1 488190 16.1 35.021 41.400 35.021 41.400 qs_ot_get_derivative 55 11.6 0.001 0.001 38.407 38.538 multiply_cannon_sync_h2d 69360 15.2 29.124 33.112 29.124 33.112 qs_ot_p2m_diag 40 11.0 0.020 0.031 32.728 32.839 multiply_cannon_metrocomm3 69360 15.2 0.202 0.213 25.707 32.705 cp_dbcsr_syevd 40 12.0 0.002 0.002 29.484 29.485 rebuild_ks_matrix 110 8.4 0.000 0.000 29.035 29.201 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.020 29.034 29.200 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.654 26.807 init_scf_loop 7 6.6 0.000 0.008 26.553 26.554 cp_fm_syevd 40 13.0 0.000 0.000 24.201 24.331 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.061 23.379 apply_single 62 13.6 0.000 0.000 23.060 23.379 prepare_preconditioner 7 7.6 0.000 0.001 21.721 21.756 make_preconditioner 7 8.6 0.000 0.002 21.721 21.756 ot_new_cg_direction 55 11.6 0.001 0.002 21.066 21.067 cp_fm_redistribute_end 40 14.0 9.726 19.401 9.731 19.403 cp_fm_syevd_base 40 14.0 9.665 19.346 9.665 19.346 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.903 17.296 calculate_rho_elec 110 8.6 0.030 0.032 16.903 17.296 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.736 14.861 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.467 14.526 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.751 13.878 mp_sum_l 4764 12.2 12.492 13.266 12.492 13.266 init_scf_run 5 5.8 0.000 0.001 13.192 13.193 scf_env_initial_rho_setup 5 6.8 0.002 0.003 13.192 13.193 pw_transfer 1645 12.4 0.079 0.108 11.660 11.903 calculate_dm_sparse 110 9.5 0.000 0.001 11.721 11.857 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 11.521 11.770 density_rs2pw 110 9.6 0.005 0.007 11.182 11.545 dbcsr_mm_accdrv_process 154766 15.8 6.218 6.405 9.964 11.061 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.536 10.607 qs_vxc_create 110 10.4 0.002 0.004 10.434 10.483 fft_wrap_pw1pw2_240 915 15.0 1.137 1.237 10.110 10.344 cp_fm_cholesky_invert 7 10.6 10.087 10.095 10.087 10.095 check_diag 80 13.5 8.684 8.940 9.653 9.803 acc_transpose_blocks 69360 15.2 0.351 0.367 8.521 9.081 sum_up_and_integrate 60 10.3 0.001 0.004 8.608 8.620 integrate_v_rspace 60 11.3 0.001 0.003 8.591 8.602 fft3d_pb 915 16.0 2.391 2.680 8.322 8.589 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.446 8.463 calculate_first_density_matrix 1 7.0 0.000 0.004 8.313 8.334 transfer_rs2pw 445 10.6 0.007 0.008 7.877 8.330 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.967 8.027 multiply_cannon_metrocomm1 69360 15.2 0.100 0.106 4.625 7.855 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.299 7.569 make_m2s 2890 13.2 0.079 0.089 6.456 7.033 xc_vxc_pw_create 60 11.3 0.039 0.050 6.974 7.023 make_full_single_inverse 7 9.6 0.001 0.006 6.984 7.019 make_images 2890 14.2 0.240 0.261 6.349 6.928 xc_pw_derive 510 13.4 0.005 0.007 6.181 6.269 acc_transpose_blocks_kernels 69360 16.2 0.847 0.894 5.755 6.206 mp_alltoall_z22v 2340 17.7 5.365 5.608 5.365 5.608 jit_kernel_transpose 5 15.0 4.908 5.332 4.908 5.332 mp_waitany 7680 13.5 4.489 5.199 4.489 5.199 multiply_cannon_metrocomm4 67915 15.2 0.181 0.197 2.011 4.866 potential_pw2rs 60 12.3 0.003 0.003 4.828 4.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=231.645000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9a6ee7dcc1ca014ef5c6145201ef38f59b3b5ca6 Summary: empty Status: OK