=== 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: 7444654cf14297a8dae96b6cf94e55baac10f21c ################# 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 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, 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. # # Author: Matthias Krack (26.10.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/01 job id: 43389636 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/02 job id: 43389637 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/03 job id: 43389638 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/04 job id: 43389639 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/05 job id: 43389640 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/06 job id: 43389641 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/07 job id: 43389642 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/08 job id: 43389643 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/09 job id: 43389645 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/10 job id: 43389647 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/11 job id: 43389648 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/12 job id: 43389649 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/13 job id: 43389650 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/14 job id: 43389651 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/15 job id: 43389652 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/16 job id: 43389653 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/17 job id: 43389654 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/18 job id: 43389656 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/19 job id: 43389657 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/20 job id: 43389658 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/21 job id: 43389660 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/22 job id: 43389661 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/23 job id: 43389662 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/24 job id: 43389664 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/25 job id: 43389666 --- 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/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/26 job id: 43389667 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 344 9. 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 0.027 0.034 132.805 132.806 farming_run 1 2.0 132.317 132.318 132.770 132.773 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460625E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 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 0.010 0.027 114.947 114.947 qs_energies 1 2.0 0.000 0.000 114.727 114.730 mp2_main 1 3.0 0.000 0.000 112.694 112.697 mp2_gpw_main 1 4.0 0.020 0.027 111.738 111.741 mp2_ri_gpw_compute_in 1 5.0 0.176 0.217 93.255 93.349 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.324 55.417 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.167 41.633 46.566 get_2c_integrals 1 6.0 0.001 0.012 37.214 37.757 integrate_v_rspace 273 8.0 0.435 0.451 25.028 29.616 pw_transfer 6555 10.6 0.374 0.384 27.295 27.775 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.950 26.216 grid_integrate_task_list 273 9.0 20.863 25.888 20.863 25.888 fft_wrap_pw1pw2_100 2178 12.4 1.198 1.406 23.509 23.784 compute_2c_integrals 1 7.0 0.002 0.003 19.726 19.727 compute_2c_integrals_loop_lm 1 8.0 0.004 0.004 18.883 19.390 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.429 18.879 19.386 rpa_ri_compute_en 1 5.0 0.035 0.040 18.375 18.444 cp_fm_cholesky_decompose 12 8.2 17.487 18.097 17.487 18.097 cholesky_decomp 1 7.0 0.000 0.000 16.340 16.888 fft3d_s 5443 13.4 16.090 16.292 16.112 16.315 ao_to_mo_and_store_B_mult_1 272 7.0 10.852 15.574 10.852 15.574 calculate_wavefunction 272 8.0 5.436 5.518 12.531 13.169 rpa_num_int 1 6.0 0.001 0.009 10.574 10.575 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.499 10.538 calc_mat_Q 8 8.0 0.000 0.000 9.303 9.409 contract_S_to_Q 8 9.0 0.000 0.000 8.727 8.834 calc_potential_gpw 544 9.5 0.005 0.006 8.273 8.695 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.228 8.564 parallel_gemm_fm 14 9.1 0.000 0.000 8.306 8.417 parallel_gemm_fm_cosma 14 10.1 8.306 8.417 8.306 8.417 potential_pw2rs 545 10.0 0.107 0.109 7.654 8.308 collocate_single_gaussian 272 10.0 0.041 0.043 7.458 7.722 create_integ_mat 1 6.0 0.023 0.028 7.571 7.571 array2fm 1 7.0 0.000 0.000 6.660 7.031 pw_scatter_s 2720 13.7 4.448 4.552 4.448 4.552 pw_gather_s 2722 13.2 3.840 4.167 3.840 4.167 array2fm_buffer_send 1 8.0 2.932 3.127 2.932 3.127 pw_poisson_solve 545 10.5 1.126 1.177 2.176 2.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.738064, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 344 10. 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 0.030 0.056 397.131 397.132 farming_run 1 2.0 396.198 396.213 397.068 397.069 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225511E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.014 0.042 209.534 209.535 qs_energies 1 2.0 0.000 0.000 209.269 209.280 scf_env_do_scf 1 3.0 0.000 0.000 106.232 106.233 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.356 105.363 rebuild_ks_matrix 4 6.0 0.000 0.000 105.355 105.362 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.355 105.362 hfx_ks_matrix 4 8.0 0.001 0.001 104.972 104.976 integrate_four_center 4 9.0 0.144 0.457 104.971 104.975 mp2_main 1 3.0 0.000 0.000 102.737 102.747 mp2_gpw_main 1 4.0 0.035 0.068 101.824 101.839 integrate_four_center_main 4 10.0 0.084 0.449 96.509 99.627 integrate_four_center_bin 264 11.0 96.425 99.585 96.425 99.585 init_scf_loop 1 4.0 0.000 0.000 91.823 91.824 mp2_ri_gpw_compute_in 1 5.0 0.072 0.131 74.989 76.107 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.047 54.482 55.599 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 42.087 47.071 integrate_v_rspace 95 8.0 0.398 0.566 28.449 33.254 pw_transfer 2240 10.6 0.142 0.162 29.920 30.316 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.920 29.317 grid_integrate_task_list 95 9.0 23.730 28.717 23.730 28.717 mp2_ri_gpw_compute_en 1 5.0 0.071 0.137 26.679 28.439 ao_to_mo_and_store_B_mult_1 91 7.0 10.716 28.167 10.716 28.167 fft_wrap_pw1pw2_100 730 12.4 1.290 1.448 26.641 27.068 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.892 24.855 24.866 get_2c_integrals 1 6.0 0.004 0.045 20.401 20.442 compute_2c_integrals 1 7.0 0.008 0.040 19.377 19.389 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.915 19.198 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.860 18.912 19.196 fft3d_s 1823 13.4 18.435 18.927 18.448 18.941 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.406 14.406 calculate_wavefunction 91 8.0 2.033 2.058 9.758 10.007 potential_pw2rs 186 10.0 0.034 0.035 8.678 9.323 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.581 8.765 9.254 local_gemm 172 8.0 8.208 8.679 8.208 8.679 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.275 8.641 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.528 7.857 8.382 collocate_single_gaussian 91 10.0 0.019 0.069 7.894 8.151 calc_potential_gpw 182 9.5 0.002 0.002 7.914 8.144 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.433 6.355 6.433 mp_sendrecv_dm3 2068 8.0 5.892 6.410 5.892 6.410 mp_sync 38 10.4 3.162 5.751 3.162 5.751 pw_gather_s 912 13.2 4.910 5.530 4.910 5.530 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.829777, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 450.949120E+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 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.040 0.093 52.797 52.799 qs_mol_dyn_low 1 2.0 0.004 0.014 52.441 52.450 qs_forces 11 3.9 0.002 0.003 52.365 52.366 qs_energies 11 4.9 0.002 0.006 50.882 50.891 scf_env_do_scf 11 5.9 0.001 0.002 44.858 44.858 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.605 42.605 dbcsr_multiply_generic 2286 12.5 0.095 0.099 33.624 34.090 qs_scf_new_mos 108 7.5 0.000 0.000 32.357 32.657 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.356 32.657 ot_scf_mini 108 9.5 0.002 0.002 30.710 30.921 multiply_cannon 2286 13.5 0.187 0.196 25.760 27.304 multiply_cannon_loop 2286 14.5 1.484 1.566 24.999 26.553 velocity_verlet 10 3.0 0.001 0.002 26.154 26.154 ot_mini 108 10.5 0.001 0.001 19.681 19.926 qs_ot_get_derivative 108 11.5 0.001 0.001 16.643 16.902 mp_waitall_1 267858 16.1 8.862 15.082 8.862 15.082 multiply_cannon_metrocomm3 54864 15.5 0.066 0.070 5.916 12.920 multiply_cannon_multrec 54864 15.5 4.199 6.485 7.467 10.922 rebuild_ks_matrix 119 8.3 0.000 0.000 8.003 8.138 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 8.003 8.138 multiply_cannon_sync_h2d 54864 15.5 5.842 7.542 5.842 7.542 mp_sum_l 7207 12.9 5.653 7.492 5.653 7.492 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.075 7.196 qs_ot_get_p 119 10.4 0.001 0.001 6.407 6.699 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.501 5.954 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.577 5.678 init_scf_run 11 5.9 0.000 0.001 4.697 4.697 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.697 4.697 sum_up_and_integrate 119 10.3 0.012 0.014 4.608 4.614 integrate_v_rspace 119 11.3 0.002 0.002 4.596 4.603 dbcsr_mm_accdrv_process 76910 16.1 1.182 1.821 3.190 4.434 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.219 4.305 calculate_rho_elec 119 8.7 0.011 0.016 4.219 4.304 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.548 3.602 multiply_cannon_metrocomm1 54864 15.5 0.050 0.055 1.882 3.319 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.936 3.127 apply_single 119 13.6 0.000 0.000 2.936 3.127 rs_pw_transfer 974 11.9 0.011 0.012 2.959 3.044 calculate_dm_sparse 119 9.5 0.000 0.000 2.863 3.004 ot_diis_step 108 11.5 0.006 0.006 2.747 2.748 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.668 2.669 jit_kernel_multiply 13 15.8 1.945 2.524 1.945 2.524 calculate_first_density_matrix 1 7.0 0.000 0.003 2.486 2.489 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.424 2.484 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.413 2.414 cp_fm_redistribute_end 50 14.0 2.185 2.385 2.190 2.388 density_rs2pw 119 9.7 0.004 0.004 2.274 2.371 cp_fm_diag_elpa_base 50 14.0 0.195 2.337 0.195 2.346 init_scf_loop 11 6.9 0.000 0.003 2.233 2.234 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.225 2.227 wfi_extrapolate 11 7.9 0.001 0.001 2.151 2.151 acc_transpose_blocks 54864 15.5 0.228 0.244 1.716 2.127 grid_integrate_task_list 119 12.3 2.004 2.096 2.004 2.096 mp_sum_d 4125 12.0 1.391 2.036 1.391 2.036 make_m2s 4572 13.5 0.054 0.056 1.928 1.991 potential_pw2rs 119 12.3 0.004 0.004 1.959 1.976 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.922 1.976 make_images 4572 14.5 0.133 0.139 1.846 1.907 pw_transfer 1439 11.6 0.051 0.055 1.728 1.802 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.653 1.730 mp_alltoall_d11v 2130 13.8 1.493 1.599 1.493 1.599 fft3d_ps 1201 14.6 0.359 0.462 1.431 1.501 mp_waitany 12084 13.8 1.310 1.472 1.310 1.472 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.378 1.398 grid_collocate_task_list 119 9.7 1.286 1.358 1.286 1.358 fft_wrap_pw1pw2_140 487 13.2 0.080 0.095 1.268 1.342 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.785 1.180 prepare_preconditioner 11 7.9 0.000 0.000 1.107 1.138 make_preconditioner 11 8.9 0.000 0.002 1.107 1.137 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.041 1.079 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.799000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 487.559168E+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 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 598651. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.031 0.054 40.019 40.021 qs_mol_dyn_low 1 2.0 0.003 0.007 39.605 39.691 qs_forces 11 3.9 0.002 0.002 39.520 39.521 qs_energies 11 4.9 0.002 0.005 37.777 37.781 scf_env_do_scf 11 5.9 0.001 0.001 32.485 32.486 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.827 29.829 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.917 22.339 qs_scf_new_mos 108 7.5 0.001 0.001 20.456 20.708 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.456 20.707 ot_scf_mini 108 9.5 0.003 0.003 19.543 19.714 velocity_verlet 10 3.0 0.001 0.002 19.054 19.055 multiply_cannon 2286 13.5 0.208 0.219 16.540 18.060 multiply_cannon_loop 2286 14.5 0.893 0.966 15.297 16.964 ot_mini 108 10.5 0.001 0.001 12.201 12.437 mp_waitall_1 217478 16.2 6.479 11.703 6.479 11.703 qs_ot_get_derivative 108 11.5 0.001 0.001 9.700 9.875 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.247 9.851 multiply_cannon_multrec 27432 15.5 1.959 4.502 5.909 8.796 rebuild_ks_matrix 119 8.3 0.000 0.000 7.445 7.594 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.445 7.593 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.554 6.690 dbcsr_mm_accdrv_process 47894 16.0 2.982 5.023 3.882 5.704 qs_ot_get_p 119 10.4 0.001 0.002 4.362 4.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.750 4.571 sum_up_and_integrate 119 10.3 0.024 0.027 4.444 4.458 integrate_v_rspace 119 11.3 0.002 0.003 4.420 4.435 mp_sum_l 7207 12.9 2.240 4.257 2.240 4.257 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.094 4.168 apply_single 119 13.6 0.000 0.000 3.094 4.168 init_scf_run 11 5.9 0.000 0.001 4.042 4.042 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.041 4.042 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.819 3.863 calculate_rho_elec 119 8.7 0.021 0.024 3.819 3.863 make_m2s 4572 13.5 0.052 0.054 2.696 2.974 rs_pw_transfer 974 11.9 0.010 0.011 2.848 2.941 make_images 4572 14.5 0.200 0.237 2.608 2.885 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.793 2.815 multiply_cannon_sync_h2d 27432 15.5 2.167 2.757 2.167 2.757 init_scf_loop 11 6.9 0.000 0.002 2.634 2.638 calculate_first_density_matrix 1 7.0 0.001 0.003 2.471 2.472 ot_diis_step 108 11.5 0.011 0.011 2.442 2.443 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.335 2.336 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.227 2.314 density_rs2pw 119 9.7 0.004 0.004 2.166 2.246 potential_pw2rs 119 12.3 0.006 0.006 2.175 2.195 calculate_dm_sparse 119 9.5 0.000 0.000 2.092 2.172 pw_transfer 1439 11.6 0.063 0.066 1.957 2.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.953 1.956 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.940 1.941 grid_integrate_task_list 119 12.3 1.841 1.939 1.841 1.939 cp_fm_redistribute_end 50 14.0 1.595 1.912 1.598 1.913 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.867 1.913 cp_fm_diag_elpa_base 50 14.0 0.306 1.850 0.313 1.882 jit_kernel_multiply 9 16.1 0.846 1.772 0.846 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.716 1.757 prepare_preconditioner 11 7.9 0.000 0.000 1.672 1.699 make_preconditioner 11 8.9 0.000 0.001 1.672 1.699 make_images_data 4572 15.5 0.045 0.052 1.214 1.699 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.570 1.630 fft3d_ps 1201 14.6 0.500 0.552 1.574 1.616 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.061 1.569 acc_transpose_blocks 27432 15.5 0.108 0.112 1.187 1.500 wfi_extrapolate 11 7.9 0.001 0.001 1.495 1.495 mp_allgather_i34 2286 14.5 0.671 1.445 0.671 1.445 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.380 1.424 mp_alltoall_d11v 2130 13.8 1.292 1.388 1.292 1.388 grid_collocate_task_list 119 9.7 1.244 1.320 1.244 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.306 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.238 1.283 mp_sum_d 4125 12.0 0.632 1.057 0.632 1.057 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.949 0.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.959 0.959 make_images_sizes 4572 15.5 0.005 0.005 0.673 0.949 mp_alltoall_i44 4572 16.5 0.668 0.944 0.668 0.944 mp_alltoall_z22v 1201 16.6 0.812 0.910 0.812 0.910 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.605 0.833 0.883 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.658 0.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.021000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.000000, yerr=1.044466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 524.423168E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 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 0.023 0.048 32.289 32.291 qs_mol_dyn_low 1 2.0 0.003 0.003 31.991 32.000 qs_forces 11 3.9 0.002 0.003 31.881 31.883 qs_energies 11 4.9 0.002 0.006 30.330 30.332 scf_env_do_scf 11 5.9 0.001 0.001 25.599 25.599 scf_env_do_scf_inner_loop 108 6.5 0.004 0.021 23.013 23.013 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.484 16.561 velocity_verlet 10 3.0 0.001 0.002 15.176 15.179 qs_scf_new_mos 108 7.5 0.001 0.001 14.927 14.938 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.927 14.938 ot_scf_mini 108 9.5 0.002 0.003 14.205 14.215 multiply_cannon 2286 13.5 0.197 0.205 13.133 13.818 multiply_cannon_loop 2286 14.5 0.631 0.658 12.325 13.063 ot_mini 108 10.5 0.001 0.001 8.861 8.877 qs_ot_get_derivative 108 11.5 0.001 0.001 7.365 7.375 multiply_cannon_multrec 18288 15.5 1.946 2.959 6.794 7.164 rebuild_ks_matrix 119 8.3 0.000 0.000 6.412 6.425 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.412 6.424 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.676 5.688 dbcsr_mm_accdrv_process 38222 16.0 4.077 5.362 4.764 5.531 mp_waitall_1 169478 16.3 3.085 4.217 3.085 4.217 sum_up_and_integrate 119 10.3 0.031 0.032 4.020 4.026 integrate_v_rspace 119 11.3 0.002 0.002 3.989 3.999 init_scf_run 11 5.9 0.000 0.001 3.523 3.524 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.523 3.523 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.781 3.407 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.284 3.294 calculate_rho_elec 119 8.7 0.030 0.031 3.283 3.294 qs_ot_get_p 119 10.4 0.001 0.001 3.154 3.177 init_scf_loop 11 6.9 0.000 0.001 2.570 2.570 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.438 2.512 rs_pw_transfer 974 11.9 0.009 0.010 2.350 2.439 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.038 2.349 apply_single 119 13.6 0.000 0.000 2.037 2.349 calculate_first_density_matrix 1 7.0 0.001 0.003 2.312 2.314 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.056 2.061 make_m2s 4572 13.5 0.044 0.045 1.865 2.012 make_images 4572 14.5 0.191 0.203 1.779 1.924 density_rs2pw 119 9.7 0.004 0.004 1.853 1.923 grid_integrate_task_list 119 12.3 1.797 1.870 1.797 1.870 calculate_dm_sparse 119 9.5 0.000 0.000 1.814 1.824 potential_pw2rs 119 12.3 0.007 0.008 1.791 1.798 prepare_preconditioner 11 7.9 0.000 0.000 1.776 1.779 make_preconditioner 11 8.9 0.000 0.003 1.776 1.779 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.775 1.775 pw_transfer 1439 11.6 0.064 0.067 1.750 1.759 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.627 1.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.707 1.708 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.659 1.670 multiply_cannon_sync_h2d 18288 15.5 1.397 1.586 1.397 1.586 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.581 1.585 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.531 1.532 jit_kernel_multiply 8 15.9 0.634 1.529 0.634 1.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.509 1.515 cp_fm_redistribute_end 50 14.0 1.134 1.507 1.135 1.507 cp_fm_diag_elpa_base 50 14.0 0.355 1.440 0.369 1.479 ot_diis_step 108 11.5 0.011 0.011 1.474 1.474 mp_sum_l 7207 12.9 1.105 1.449 1.105 1.449 fft3d_ps 1201 14.6 0.508 0.524 1.345 1.355 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.307 1.316 grid_collocate_task_list 119 9.7 1.203 1.292 1.203 1.292 acc_transpose_blocks 18288 15.5 0.075 0.076 1.230 1.261 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.185 1.188 wfi_extrapolate 11 7.9 0.001 0.001 1.162 1.162 make_images_data 4572 15.5 0.045 0.049 0.840 1.015 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.983 0.985 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.720 0.917 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.878 0.904 mp_alltoall_d11v 2130 13.8 0.740 0.825 0.740 0.825 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 0.792 0.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.808 cp_fm_cholesky_invert 11 10.9 0.747 0.751 0.747 0.751 mp_alltoall_z22v 1201 16.6 0.683 0.748 0.683 0.748 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.439 0.708 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.705 mp_waitany 9880 13.7 0.554 0.653 0.554 0.653 dbcsr_complete_redistribute 329 12.2 0.108 0.156 0.535 0.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.291000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=3.082877 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 555.696128E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 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 0.063 0.126 36.939 36.971 qs_mol_dyn_low 1 2.0 0.003 0.007 36.552 36.583 qs_forces 11 3.9 0.007 0.041 35.425 35.453 qs_energies 11 4.9 0.002 0.006 33.642 33.671 scf_env_do_scf 11 5.9 0.001 0.002 28.577 28.602 scf_env_do_scf_inner_loop 108 6.5 0.007 0.022 24.950 24.972 velocity_verlet 10 3.0 0.002 0.002 19.027 19.091 dbcsr_multiply_generic 2286 12.5 0.101 0.103 18.437 18.545 qs_scf_new_mos 108 7.5 0.001 0.001 16.747 16.800 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.746 16.799 ot_scf_mini 108 9.5 0.003 0.005 15.794 15.844 multiply_cannon 2286 13.5 0.231 0.268 14.455 14.925 multiply_cannon_loop 2286 14.5 0.933 0.961 13.428 13.810 ot_mini 108 10.5 0.001 0.002 9.470 9.532 multiply_cannon_multrec 27432 15.5 2.333 3.031 8.617 8.959 qs_ot_get_derivative 108 11.5 0.001 0.002 7.614 7.666 dbcsr_mm_accdrv_process 47916 15.9 5.240 6.969 6.189 7.377 rebuild_ks_matrix 119 8.3 0.000 0.000 6.558 6.612 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.557 6.611 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.785 5.834 sum_up_and_integrate 119 10.3 0.036 0.039 3.831 3.844 integrate_v_rspace 119 11.3 0.002 0.003 3.795 3.809 qs_ot_get_p 119 10.4 0.001 0.001 3.648 3.731 init_scf_run 11 5.9 0.000 0.001 3.659 3.662 scf_env_initial_rho_setup 11 6.9 0.001 0.004 3.659 3.662 init_scf_loop 11 6.9 0.001 0.004 3.604 3.608 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.342 3.379 calculate_rho_elec 119 8.7 0.040 0.046 3.341 3.378 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.766 3.236 mp_waitall_1 145218 16.4 2.248 2.813 2.248 2.813 prepare_preconditioner 11 7.9 0.000 0.000 2.721 2.730 make_preconditioner 11 8.9 0.000 0.002 2.721 2.730 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.325 2.651 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.127 2.606 apply_single 119 13.6 0.000 0.000 2.127 2.606 make_m2s 4572 13.5 0.054 0.056 2.424 2.576 make_images 4572 14.5 0.329 0.397 2.316 2.466 calculate_first_density_matrix 1 7.0 0.001 0.004 2.254 2.256 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.223 2.233 rs_pw_transfer 974 11.9 0.009 0.009 2.132 2.217 calculate_dm_sparse 119 9.5 0.000 0.000 2.050 2.103 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.016 2.043 density_rs2pw 119 9.7 0.003 0.004 1.819 1.904 grid_integrate_task_list 119 12.3 1.822 1.884 1.822 1.884 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.820 1.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.816 1.818 ot_diis_step 108 11.5 0.015 0.033 1.814 1.816 jit_kernel_multiply 10 15.7 0.887 1.815 0.887 1.815 pw_transfer 1439 11.6 0.063 0.066 1.761 1.798 mp_sum_l 7207 12.9 1.184 1.719 1.184 1.719 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.671 1.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.624 1.639 potential_pw2rs 119 12.3 0.008 0.009 1.621 1.632 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.554 1.554 cp_fm_redistribute_end 50 14.0 1.022 1.527 1.024 1.527 cp_fm_diag_elpa_base 50 14.0 0.481 1.464 0.501 1.507 acc_transpose_blocks 27432 15.5 0.110 0.113 1.444 1.463 fft3d_ps 1201 14.6 0.529 0.583 1.351 1.385 fft_wrap_pw1pw2_140 487 13.2 0.085 0.093 1.331 1.374 wfi_extrapolate 11 7.9 0.001 0.003 1.354 1.355 grid_collocate_task_list 119 9.7 1.221 1.317 1.221 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.230 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.748 1.225 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.157 1.175 mp_sum_dm 438 4.9 1.082 1.139 1.082 1.139 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.122 1.124 cp_fm_upper_to_full 72 13.5 0.808 1.121 0.808 1.121 md_write_output 11 3.9 0.035 0.779 0.049 1.095 update_particle_set 20 4.0 0.000 0.000 1.044 1.092 dbcsr_complete_redistribute 329 12.2 0.123 0.149 0.810 1.083 make_images_data 4572 15.5 0.045 0.048 0.924 1.079 multiply_cannon_sync_h2d 27432 15.5 0.983 1.053 0.983 1.053 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.787 1.000 mp_alltoall_d11v 2130 13.8 0.783 0.925 0.783 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.864 0.869 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 cp_fm_cholesky_invert 11 10.9 0.863 0.866 0.863 0.866 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.594 0.859 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.822 0.836 mp_alltoall_i22 627 13.8 0.446 0.741 0.446 0.741 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.971000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.545455, yerr=3.340213 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 604.016640E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 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 0.036 0.061 31.249 31.249 qs_mol_dyn_low 1 2.0 0.003 0.004 30.594 30.978 qs_forces 11 3.9 0.002 0.002 30.487 30.489 qs_energies 11 4.9 0.001 0.002 28.705 28.709 scf_env_do_scf 11 5.9 0.000 0.001 23.725 23.726 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.771 20.773 velocity_verlet 10 3.0 0.002 0.002 16.016 16.020 dbcsr_multiply_generic 2286 12.5 0.094 0.096 13.179 13.278 qs_scf_new_mos 108 7.5 0.001 0.001 12.244 12.267 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.243 12.267 ot_scf_mini 108 9.5 0.002 0.002 11.512 11.532 multiply_cannon 2286 13.5 0.228 0.235 9.859 10.400 multiply_cannon_loop 2286 14.5 0.329 0.341 8.872 9.118 rebuild_ks_matrix 119 8.3 0.000 0.000 6.416 6.449 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.416 6.449 ot_mini 108 10.5 0.001 0.001 6.347 6.373 multiply_cannon_multrec 9144 15.5 1.583 1.828 5.769 6.000 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.718 5.746 qs_ot_get_derivative 108 11.5 0.001 0.001 4.960 4.981 dbcsr_mm_accdrv_process 12550 15.8 3.194 4.098 4.085 4.165 sum_up_and_integrate 119 10.3 0.037 0.041 4.001 4.008 integrate_v_rspace 119 11.3 0.002 0.003 3.964 3.970 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.692 3.707 calculate_rho_elec 119 8.7 0.059 0.061 3.691 3.707 init_scf_run 11 5.9 0.000 0.001 3.490 3.490 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.490 3.490 qs_ot_get_p 119 10.4 0.001 0.001 3.280 3.319 init_scf_loop 11 6.9 0.000 0.000 2.921 2.923 mp_waitall_1 121218 16.5 2.454 2.905 2.454 2.905 make_m2s 4572 13.5 0.035 0.036 2.255 2.424 make_images 4572 14.5 0.269 0.301 2.166 2.333 pw_transfer 1439 11.6 0.064 0.066 2.193 2.208 calculate_first_density_matrix 1 7.0 0.000 0.000 2.204 2.205 rs_pw_transfer 974 11.9 0.008 0.008 2.029 2.120 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.102 2.117 density_rs2pw 119 9.7 0.003 0.004 2.028 2.108 prepare_preconditioner 11 7.9 0.000 0.000 2.069 2.073 make_preconditioner 11 8.9 0.000 0.000 2.069 2.073 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.921 1.960 grid_integrate_task_list 119 12.3 1.849 1.958 1.849 1.958 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.945 1.948 calculate_dm_sparse 119 9.5 0.000 0.000 1.788 1.811 jit_kernel_multiply 10 15.7 0.852 1.791 0.852 1.791 fft3d_ps 1201 14.6 0.547 0.560 1.769 1.780 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.747 1.766 potential_pw2rs 119 12.3 0.010 0.010 1.703 1.710 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.702 1.703 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.656 1.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.663 1.665 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.401 1.412 ot_diis_step 108 11.5 0.012 0.013 1.368 1.370 grid_collocate_task_list 119 9.7 1.277 1.350 1.277 1.350 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.333 1.348 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.313 1.315 make_images_data 4572 15.5 0.039 0.043 1.040 1.282 cp_fm_redistribute_end 50 14.0 0.647 1.272 0.649 1.274 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.990 1.265 cp_fm_diag_elpa_base 50 14.0 0.579 1.187 0.622 1.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.237 1.239 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.210 1.238 apply_single 119 13.6 0.000 0.000 1.210 1.237 wfi_extrapolate 11 7.9 0.001 0.001 1.228 1.228 mp_alltoall_d11v 2130 13.8 1.138 1.223 1.138 1.223 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.162 1.168 mp_alltoall_z22v 1201 16.6 1.094 1.129 1.094 1.129 cp_fm_cholesky_invert 11 10.9 1.034 1.037 1.034 1.037 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.917 acc_transpose_blocks 9144 15.5 0.038 0.039 0.894 0.902 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.839 0.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.829 dbcsr_complete_redistribute 329 12.2 0.174 0.208 0.785 0.826 mp_allgather_i34 2286 14.5 0.320 0.813 0.320 0.813 make_images_sizes 4572 15.5 0.005 0.005 0.552 0.807 mp_alltoall_i44 4572 16.5 0.547 0.802 0.547 0.802 mp_sum_l 7207 12.9 0.633 0.777 0.633 0.777 multiply_cannon_sync_h2d 9144 15.5 0.707 0.774 0.707 0.774 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.694 0.743 arnoldi_extremal 119 11.4 0.001 0.001 0.675 0.711 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.620 0.711 arnoldi_normal_ev 119 12.4 0.026 0.068 0.674 0.710 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.654 0.661 yz_to_x 606 15.1 0.053 0.056 0.638 0.657 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.630 0.641 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.249000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.818182, yerr=6.220467 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 767.963136E+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 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.026 0.044 46.340 46.341 qs_mol_dyn_low 1 2.0 0.003 0.003 46.096 46.103 qs_forces 11 3.9 0.001 0.001 45.119 45.120 qs_energies 11 4.9 0.001 0.001 43.049 43.053 scf_env_do_scf 11 5.9 0.001 0.001 37.009 37.010 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 28.589 28.590 velocity_verlet 10 3.0 0.002 0.002 26.119 26.239 dbcsr_multiply_generic 2286 12.5 0.102 0.103 19.421 19.539 qs_scf_new_mos 108 7.5 0.001 0.001 18.493 18.586 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.492 18.585 ot_scf_mini 108 9.5 0.002 0.002 17.304 17.406 multiply_cannon 2286 13.5 0.303 0.312 14.570 15.586 multiply_cannon_loop 2286 14.5 0.343 0.348 13.086 13.976 ot_mini 108 10.5 0.001 0.001 10.112 10.226 multiply_cannon_multrec 9144 15.5 3.685 5.223 8.789 8.857 init_scf_loop 11 6.9 0.000 0.000 8.385 8.386 qs_ot_get_derivative 108 11.5 0.001 0.001 7.827 7.929 rebuild_ks_matrix 119 8.3 0.000 0.000 7.629 7.759 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.629 7.759 prepare_preconditioner 11 7.9 0.000 0.000 7.368 7.383 make_preconditioner 11 8.9 0.000 0.000 7.368 7.383 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.901 7.248 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.895 7.014 dbcsr_mm_accdrv_process 12550 15.8 4.120 6.356 4.977 6.386 mp_waitall_1 97218 16.6 3.669 4.735 3.669 4.735 cp_fm_upper_to_full 72 14.2 3.178 4.579 3.178 4.579 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.351 4.409 calculate_rho_elec 119 8.7 0.118 0.121 4.350 4.408 qs_ot_get_p 119 10.4 0.001 0.001 4.104 4.239 sum_up_and_integrate 119 10.3 0.064 0.066 4.218 4.224 integrate_v_rspace 119 11.3 0.003 0.003 4.154 4.162 init_scf_run 11 5.9 0.000 0.001 3.926 3.927 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.926 3.926 make_m2s 4572 13.5 0.038 0.038 2.903 3.183 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.728 3.159 make_images 4572 14.5 0.353 0.388 2.782 3.061 dbcsr_complete_redistribute 329 12.2 0.285 0.288 2.128 2.961 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.842 2.737 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.390 2.636 apply_single 119 13.6 0.000 0.000 2.390 2.636 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.757 2.589 pw_transfer 1439 11.6 0.066 0.067 2.539 2.544 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.539 2.542 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.443 2.448 mp_alltoall_i22 627 13.8 1.531 2.391 1.531 2.391 calculate_dm_sparse 119 9.5 0.000 0.000 2.338 2.387 calculate_first_density_matrix 1 7.0 0.000 0.000 2.290 2.351 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.462 2.289 density_rs2pw 119 9.7 0.003 0.004 2.236 2.275 ot_diis_step 108 11.5 0.014 0.014 2.238 2.239 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.224 2.224 grid_integrate_task_list 119 12.3 2.095 2.152 2.095 2.152 mp_sum_l 7207 12.9 1.422 2.096 1.422 2.096 fft3d_ps 1201 14.6 0.571 0.582 2.077 2.080 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.016 2.053 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.990 1.991 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.965 1.969 cp_fm_cholesky_invert 11 10.9 1.838 1.842 1.838 1.842 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.841 1.841 cp_fm_diag_elpa_base 50 14.0 1.686 1.743 1.839 1.839 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.820 1.821 make_images_data 4572 15.5 0.043 0.046 1.442 1.817 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.445 1.813 rs_pw_transfer 974 11.9 0.009 0.009 1.750 1.781 jit_kernel_multiply 10 15.6 0.829 1.729 0.829 1.729 mp_alltoall_d11v 2130 13.8 1.604 1.704 1.604 1.704 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.661 grid_collocate_task_list 119 9.7 1.515 1.547 1.515 1.547 potential_pw2rs 119 12.3 0.014 0.015 1.520 1.524 wfi_extrapolate 11 7.9 0.001 0.001 1.516 1.516 mp_alltoall_z22v 1201 16.6 1.369 1.386 1.369 1.386 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.307 1.363 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.330 1.352 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.152 1.170 mp_allgather_i34 2286 14.5 0.461 1.067 0.461 1.067 multiply_cannon_sync_h2d 9144 15.5 1.047 1.053 1.047 1.053 mp_sum_dm 438 4.9 0.898 1.029 0.898 1.029 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.027 qs_create_task_list 11 7.9 0.001 0.001 0.981 0.991 generate_qs_task_list 11 8.9 0.373 0.393 0.980 0.990 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.957 0.967 md_write_output 11 3.9 0.111 0.888 0.122 0.962 update_particle_set 20 4.0 0.000 0.000 0.839 0.959 md_output 10 3.0 0.000 0.000 0.122 0.958 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.341000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.090909, yerr=13.180251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 501.678080E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1715453. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 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 0.013 0.030 84.538 84.539 qs_mol_dyn_low 1 2.0 0.003 0.004 84.279 84.289 qs_forces 11 3.9 0.003 0.004 83.788 83.788 qs_energies 11 4.9 0.002 0.003 80.933 80.945 scf_env_do_scf 11 5.9 0.000 0.001 71.788 71.790 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.944 65.945 dbcsr_multiply_generic 2055 12.4 0.110 0.133 52.747 53.045 qs_scf_new_mos 99 7.5 0.000 0.001 48.778 48.909 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.777 48.909 ot_scf_mini 99 9.5 0.002 0.002 46.329 46.437 multiply_cannon 2055 13.4 0.182 0.189 42.801 43.666 velocity_verlet 10 3.0 0.002 0.002 43.564 43.570 multiply_cannon_loop 2055 14.4 1.542 1.581 41.745 42.561 ot_mini 99 10.5 0.001 0.001 28.141 28.281 qs_ot_get_derivative 99 11.5 0.001 0.001 21.278 21.429 multiply_cannon_multrec 49320 15.4 12.092 12.801 17.141 17.630 rebuild_ks_matrix 110 8.3 0.000 0.000 14.585 14.728 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.584 14.727 mp_waitall_1 241148 16.1 12.254 12.947 12.254 12.947 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.807 12.928 multiply_cannon_sync_h2d 49320 15.4 9.880 10.626 9.880 10.626 qs_ot_get_p 110 10.4 0.001 0.001 9.579 9.754 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.078 8.192 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.414 7.972 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.216 7.713 apply_single 110 13.6 0.000 0.001 7.215 7.713 sum_up_and_integrate 110 10.3 0.036 0.042 7.107 7.123 integrate_v_rspace 110 11.3 0.002 0.003 7.071 7.095 init_scf_run 11 5.9 0.000 0.001 7.065 7.066 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.065 7.066 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.702 6.785 ot_diis_step 99 11.5 0.005 0.006 6.587 6.588 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.354 6.491 calculate_rho_elec 110 8.6 0.020 0.024 6.353 6.490 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.162 6.203 mp_sum_l 6514 12.8 5.384 6.158 5.384 6.158 init_scf_loop 11 6.9 0.000 0.000 5.813 5.813 dbcsr_mm_accdrv_process 87628 16.1 2.105 2.209 4.928 5.234 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.151 5.151 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.624 4.626 cp_fm_redistribute_end 48 14.0 4.022 4.597 4.027 4.598 cp_fm_diag_elpa_base 48 14.0 0.564 4.517 0.567 4.542 make_m2s 4110 13.4 0.060 0.065 4.099 4.234 make_images 4110 14.4 0.176 0.191 4.004 4.140 wfi_extrapolate 11 7.9 0.001 0.001 4.102 4.102 rs_pw_transfer 902 11.9 0.011 0.012 3.887 4.033 multiply_cannon_metrocomm1 49320 15.4 0.056 0.059 2.870 3.988 calculate_dm_sparse 110 9.5 0.000 0.001 3.825 3.913 prepare_preconditioner 11 7.9 0.000 0.000 3.598 3.613 make_preconditioner 11 8.9 0.000 0.000 3.598 3.613 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.606 3.610 density_rs2pw 110 9.6 0.004 0.004 3.393 3.539 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.383 3.432 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.370 3.430 grid_integrate_task_list 110 12.3 3.210 3.375 3.210 3.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.258 3.298 pw_transfer 1331 11.6 0.054 0.062 2.976 3.043 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.889 2.954 calculate_first_density_matrix 1 7.0 0.000 0.000 2.867 2.881 potential_pw2rs 110 12.3 0.006 0.007 2.694 2.732 mp_alltoall_d11v 2046 13.8 2.237 2.610 2.237 2.610 jit_kernel_multiply 13 15.9 2.535 2.549 2.535 2.549 fft3d_ps 1111 14.6 0.752 0.838 2.426 2.485 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.411 2.477 acc_transpose_blocks 49320 15.4 0.222 0.231 2.175 2.249 grid_collocate_task_list 110 9.6 2.087 2.215 2.087 2.215 mp_waitany 14300 13.8 1.836 2.075 1.836 2.075 cp_fm_cholesky_invert 11 10.9 2.017 2.021 2.017 2.021 make_images_data 4110 15.4 0.043 0.046 1.841 1.977 mp_sum_d 3879 11.9 1.502 1.976 1.502 1.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.918 1.943 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.593 1.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.539000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=2.987578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 587.173888E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1711873. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 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 0.034 0.046 91.991 91.993 qs_mol_dyn_low 1 2.0 0.003 0.006 91.613 91.623 qs_forces 11 3.9 0.002 0.002 91.216 91.217 qs_energies 11 4.9 0.001 0.001 77.122 77.125 scf_env_do_scf 11 5.9 0.000 0.001 68.153 68.156 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 60.162 60.163 velocity_verlet 10 3.0 0.001 0.002 58.710 58.718 dbcsr_multiply_generic 2055 12.4 0.114 0.119 48.657 49.259 multiply_cannon 2055 13.4 0.222 0.244 41.705 43.283 multiply_cannon_loop 2055 14.4 0.919 0.943 40.369 41.625 rebuild_ks_matrix 110 8.3 0.000 0.000 34.888 35.406 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 34.887 35.406 qs_scf_new_mos 99 7.5 0.001 0.001 33.178 33.319 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.177 33.318 ot_scf_mini 99 9.5 0.003 0.003 31.509 31.636 qs_ks_update_qs_env 110 7.6 0.001 0.001 22.949 23.016 mp_waitall_1 186928 16.3 18.449 21.264 18.449 21.264 ot_mini 99 10.5 0.001 0.001 18.523 18.659 multiply_cannon_metrocomm3 24660 15.4 0.069 0.070 10.504 18.068 sum_up_and_integrate 110 10.3 0.052 0.059 17.496 17.508 integrate_v_rspace 110 11.3 0.002 0.002 17.444 17.458 multiply_cannon_multrec 24660 15.4 7.609 9.300 14.185 15.788 mp_alltoall_d11v 2046 13.8 12.674 13.061 12.674 13.061 qs_ot_get_derivative 99 11.5 0.001 0.001 12.716 12.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 12.106 12.569 rs_gather_matrices 110 12.3 0.130 0.141 11.680 12.059 mp_sum_dm 438 4.9 0.881 11.194 0.881 11.194 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 5.118 11.128 multiply_cannon_sync_h2d 24660 15.4 6.997 8.215 6.997 8.215 init_scf_loop 11 6.9 0.000 0.000 7.953 7.954 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.489 7.176 apply_single 110 13.6 0.000 0.001 6.489 7.176 dbcsr_mm_accdrv_process 52282 16.1 4.830 5.947 6.415 6.680 init_scf_run 11 5.9 0.000 0.001 6.520 6.520 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.519 6.520 qs_ot_get_p 110 10.4 0.001 0.001 6.309 6.502 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.016 6.025 calculate_rho_elec 110 8.6 0.039 0.047 6.016 6.025 prepare_preconditioner 11 7.9 0.000 0.000 5.904 5.921 make_preconditioner 11 8.9 0.000 0.000 5.904 5.921 ot_diis_step 99 11.5 0.010 0.010 5.758 5.758 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.485 5.639 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.789 5.504 make_m2s 4110 13.4 0.057 0.060 4.482 4.942 make_images 4110 14.4 0.407 0.483 4.373 4.829 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.281 4.302 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.788 3.788 pw_transfer 1331 11.6 0.066 0.070 3.486 3.639 wfi_extrapolate 11 7.9 0.001 0.001 3.589 3.589 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.379 3.532 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.468 3.469 density_rs2pw 110 9.6 0.004 0.004 3.285 3.453 rs_pw_transfer 902 11.9 0.012 0.014 3.231 3.428 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.338 grid_integrate_task_list 110 12.3 3.135 3.268 3.135 3.268 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.177 3.180 calculate_dm_sparse 110 9.5 0.001 0.001 3.135 3.164 cp_fm_redistribute_end 48 14.0 2.362 3.144 2.364 3.146 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.115 3.146 cp_fm_diag_elpa_base 48 14.0 0.747 3.027 0.778 3.113 make_images_data 4110 15.4 0.047 0.051 2.472 2.926 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.753 2.902 fft3d_ps 1111 14.6 1.073 1.263 2.738 2.876 hybrid_alltoall_any 4261 16.3 0.103 0.446 2.152 2.849 calculate_first_density_matrix 1 7.0 0.000 0.002 2.844 2.845 cp_fm_cholesky_invert 11 10.9 2.770 2.777 2.770 2.777 jit_kernel_multiply 12 16.2 1.232 2.674 1.232 2.674 potential_pw2rs 110 12.3 0.008 0.009 2.574 2.595 mp_sum_l 6514 12.8 1.849 2.468 1.849 2.468 grid_collocate_task_list 110 9.6 2.118 2.298 2.118 2.298 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.986 2.011 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.910 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=91.993000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=6.873112 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 658.939904E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 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 0.043 0.088 63.525 63.526 qs_mol_dyn_low 1 2.0 0.003 0.003 63.062 63.072 qs_forces 11 3.9 0.003 0.014 62.282 62.283 qs_energies 11 4.9 0.003 0.007 58.998 59.003 scf_env_do_scf 11 5.9 0.001 0.002 50.672 50.672 scf_env_do_scf_inner_loop 99 6.5 0.007 0.026 41.282 41.283 velocity_verlet 10 3.0 0.001 0.002 34.696 34.722 dbcsr_multiply_generic 2055 12.4 0.110 0.115 29.691 30.063 qs_scf_new_mos 99 7.5 0.001 0.001 26.394 26.498 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.394 26.498 ot_scf_mini 99 9.5 0.003 0.003 25.127 25.249 multiply_cannon 2055 13.4 0.211 0.224 22.868 24.238 multiply_cannon_loop 2055 14.4 0.612 0.632 21.615 23.044 ot_mini 99 10.5 0.001 0.001 14.290 14.422 rebuild_ks_matrix 110 8.3 0.000 0.000 12.516 12.692 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.515 12.691 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.999 11.161 mp_waitall_1 146766 16.3 8.166 11.104 8.166 11.104 multiply_cannon_multrec 16440 15.4 3.690 4.541 9.637 10.443 qs_ot_get_derivative 99 11.5 0.001 0.001 9.626 9.738 init_scf_loop 11 6.9 0.001 0.004 9.349 9.351 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 5.053 8.118 prepare_preconditioner 11 7.9 0.000 0.000 7.461 7.480 make_preconditioner 11 8.9 0.000 0.001 7.461 7.480 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.770 7.128 sum_up_and_integrate 110 10.3 0.060 0.061 6.648 6.676 integrate_v_rspace 110 11.3 0.002 0.003 6.588 6.616 qs_ot_get_p 110 10.4 0.001 0.001 5.900 6.056 dbcsr_mm_accdrv_process 34862 16.1 4.771 5.339 5.805 5.952 init_scf_run 11 5.9 0.000 0.001 5.862 5.862 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.862 5.862 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.710 5.726 calculate_rho_elec 110 8.6 0.058 0.058 5.709 5.726 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.144 5.639 apply_single 110 13.6 0.000 0.000 5.144 5.638 make_m2s 4110 13.4 0.050 0.052 4.447 4.771 make_images 4110 14.4 0.391 0.512 4.332 4.656 ot_diis_step 99 11.5 0.011 0.011 4.613 4.614 multiply_cannon_sync_h2d 16440 15.4 3.629 4.150 3.629 4.150 qs_ot_p2m_diag 48 11.0 0.041 0.044 4.077 4.084 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.306 3.944 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.711 3.711 pw_transfer 1331 11.6 0.065 0.070 3.379 3.407 grid_integrate_task_list 110 12.3 3.174 3.359 3.174 3.359 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.273 3.302 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.138 3.140 cp_fm_cholesky_invert 11 10.9 3.113 3.121 3.113 3.121 density_rs2pw 110 9.6 0.004 0.004 2.919 3.107 cp_fm_redistribute_end 48 14.0 1.946 3.103 1.948 3.104 cp_fm_diag_elpa_base 48 14.0 1.091 2.963 1.151 3.072 wfi_extrapolate 11 7.9 0.001 0.001 3.071 3.072 make_images_data 4110 15.4 0.044 0.047 2.600 3.062 rs_pw_transfer 902 11.9 0.010 0.011 2.827 3.030 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.295 2.957 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.946 2.948 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.761 2.795 calculate_first_density_matrix 1 7.0 0.000 0.002 2.704 2.705 calculate_dm_sparse 110 9.5 0.001 0.001 2.649 2.679 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.594 2.647 fft3d_ps 1111 14.6 1.069 1.097 2.588 2.622 mp_sum_l 6514 12.8 1.776 2.612 1.776 2.612 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.909 2.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.473 2.535 potential_pw2rs 110 12.3 0.010 0.010 2.437 2.460 mp_irecv_dv 48980 15.7 0.841 2.437 0.841 2.437 grid_collocate_task_list 110 9.6 2.172 2.325 2.172 2.325 mp_alltoall_d11v 2046 13.8 1.943 2.290 1.943 2.290 dbcsr_complete_redistribute 325 12.2 0.329 0.383 1.551 2.014 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.979 1.980 cp_fm_upper_to_full 70 13.6 1.403 1.889 1.403 1.889 cp_fm_cholesky_decompose 22 10.9 1.814 1.842 1.814 1.842 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.723 1.746 mp_allgather_i34 2055 14.4 0.580 1.740 0.580 1.740 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.572 1.586 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.037 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.358 1.477 mp_waitany 17072 13.8 1.208 1.425 1.208 1.425 mp_alltoall_z22v 1111 16.6 1.305 1.336 1.305 1.336 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.324 1.332 rs_gather_matrices 110 12.3 0.138 0.152 0.924 1.294 acc_transpose_blocks 16440 15.4 0.071 0.073 1.273 1.285 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.526000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.636364, yerr=8.875633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 736.153600E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 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 0.050 0.100 67.519 67.522 qs_mol_dyn_low 1 2.0 0.005 0.018 67.032 67.051 qs_forces 11 3.9 0.004 0.015 66.930 66.932 qs_energies 11 4.9 0.010 0.031 63.491 63.502 scf_env_do_scf 11 5.9 0.004 0.010 54.432 54.435 scf_env_do_scf_inner_loop 99 6.5 0.012 0.030 42.104 42.108 velocity_verlet 10 3.0 0.003 0.008 37.710 37.714 dbcsr_multiply_generic 2055 12.4 0.127 0.131 30.124 30.347 qs_scf_new_mos 99 7.5 0.001 0.001 27.463 27.554 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.462 27.553 ot_scf_mini 99 9.5 0.003 0.004 25.808 25.899 multiply_cannon 2055 13.4 0.245 0.272 22.658 23.712 multiply_cannon_loop 2055 14.4 0.884 0.907 21.212 21.804 ot_mini 99 10.5 0.001 0.001 14.569 14.670 multiply_cannon_multrec 24660 15.4 4.184 6.654 12.791 14.021 init_scf_loop 11 6.9 0.001 0.003 12.281 12.282 rebuild_ks_matrix 110 8.3 0.000 0.000 12.166 12.280 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 12.166 12.280 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.770 10.873 prepare_preconditioner 11 7.9 0.000 0.000 10.411 10.429 make_preconditioner 11 8.9 0.000 0.001 10.411 10.429 qs_ot_get_derivative 99 11.5 0.001 0.002 10.298 10.388 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.649 10.099 dbcsr_mm_accdrv_process 52304 16.0 6.826 8.308 8.460 9.405 mp_waitall_1 126806 16.4 4.744 6.656 4.744 6.656 sum_up_and_integrate 110 10.3 0.068 0.074 6.485 6.543 integrate_v_rspace 110 11.3 0.002 0.003 6.416 6.477 qs_ot_get_p 110 10.4 0.001 0.001 6.175 6.310 init_scf_run 11 5.9 0.000 0.001 5.949 5.949 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.948 5.949 make_m2s 4110 13.4 0.060 0.061 5.656 5.881 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.756 5.769 calculate_rho_elec 110 8.6 0.077 0.081 5.755 5.769 make_images 4110 14.4 0.575 0.698 5.515 5.736 cp_fm_upper_to_full 70 13.8 3.255 4.635 3.255 4.635 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.250 4.264 ot_diis_step 99 11.5 0.011 0.014 4.218 4.218 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.083 4.165 apply_single 110 13.6 0.000 0.000 4.083 4.165 dbcsr_complete_redistribute 325 12.2 0.443 0.552 2.800 3.920 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.752 3.753 pw_transfer 1331 11.6 0.065 0.073 3.433 3.520 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.414 3.468 grid_integrate_task_list 110 12.3 3.266 3.440 3.266 3.440 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.328 3.417 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.490 3.368 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.231 3.345 multiply_cannon_sync_h2d 24660 15.4 3.160 3.326 3.160 3.326 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.181 3.225 calculate_dm_sparse 110 9.5 0.001 0.001 3.101 3.128 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.101 3.103 make_images_data 4110 15.4 0.046 0.050 2.848 3.100 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.479 3.098 cp_fm_redistribute_end 48 14.0 1.539 3.061 1.541 3.062 wfi_extrapolate 11 7.9 0.001 0.003 3.057 3.057 cp_fm_cholesky_invert 11 10.9 3.031 3.040 3.031 3.040 cp_fm_diag_elpa_base 48 14.0 1.438 2.925 1.518 3.039 density_rs2pw 110 9.6 0.004 0.004 2.853 3.004 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.947 2.949 fft_wrap_pw1pw2_140 451 13.1 0.223 0.274 2.779 2.871 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.753 2.847 rs_pw_transfer 902 11.9 0.010 0.011 2.637 2.836 mp_alltoall_i22 605 13.7 1.658 2.830 1.658 2.830 calculate_first_density_matrix 1 7.0 0.001 0.004 2.802 2.804 fft3d_ps 1111 14.6 1.062 1.096 2.610 2.677 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.484 2.530 qs_energies_init_hamiltonians 11 5.9 0.012 0.054 2.490 2.491 grid_collocate_task_list 110 9.6 2.225 2.410 2.225 2.410 jit_kernel_multiply 12 15.8 1.301 2.363 1.301 2.363 mp_alltoall_d11v 2046 13.8 2.025 2.329 2.025 2.329 potential_pw2rs 110 12.3 0.012 0.013 2.266 2.317 cp_fm_cholesky_decompose 22 10.9 1.869 1.922 1.869 1.922 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.745 1.778 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.691 1.707 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.644 1.655 mp_sum_l 6514 12.8 1.167 1.599 1.167 1.599 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.853 1.595 acc_transpose_blocks 24660 15.4 0.103 0.105 1.540 1.568 mp_irecv_dv 62702 16.1 0.756 1.520 0.756 1.520 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.369 1.457 mp_alltoall_z22v 1111 16.6 1.332 1.449 1.332 1.449 mp_waitany 13376 13.8 1.178 1.389 1.178 1.389 mp_allgather_i34 2055 14.4 0.549 1.361 0.549 1.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.522000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.000000, yerr=11.962061 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 818.728960E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 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 0.037 0.054 56.044 56.044 qs_mol_dyn_low 1 2.0 0.003 0.004 55.769 55.778 qs_forces 11 3.9 0.002 0.002 55.083 55.084 qs_energies 11 4.9 0.001 0.002 51.472 51.478 scf_env_do_scf 11 5.9 0.000 0.001 43.080 43.080 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.113 35.114 velocity_verlet 10 3.0 0.016 0.017 31.522 31.563 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.102 23.274 qs_scf_new_mos 99 7.5 0.001 0.001 20.901 20.942 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.900 20.941 ot_scf_mini 99 9.5 0.002 0.002 19.653 19.675 multiply_cannon 2055 13.4 0.245 0.254 17.459 18.969 multiply_cannon_loop 2055 14.4 0.321 0.348 16.074 16.421 rebuild_ks_matrix 110 8.3 0.000 0.000 11.616 11.636 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.615 11.635 ot_mini 99 10.5 0.001 0.001 10.582 10.600 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.368 10.388 multiply_cannon_multrec 8220 15.4 3.270 4.646 7.582 8.531 mp_waitall_1 106626 16.5 6.286 7.938 6.286 7.938 init_scf_loop 11 6.9 0.000 0.000 7.920 7.920 qs_ot_get_derivative 99 11.5 0.001 0.001 6.775 6.796 sum_up_and_integrate 110 10.3 0.079 0.081 6.221 6.233 prepare_preconditioner 11 7.9 0.000 0.000 6.199 6.204 make_preconditioner 11 8.9 0.000 0.000 6.199 6.204 integrate_v_rspace 110 11.3 0.003 0.003 6.141 6.153 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.777 5.857 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.678 5.700 calculate_rho_elec 110 8.6 0.115 0.115 5.677 5.699 init_scf_run 11 5.9 0.000 0.001 5.265 5.265 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.264 5.265 dbcsr_mm_accdrv_process 17442 15.9 2.817 4.090 4.182 5.188 qs_ot_get_p 110 10.4 0.001 0.001 5.098 5.122 make_m2s 4110 13.4 0.039 0.040 4.339 4.637 make_images 4110 14.4 0.635 0.693 4.210 4.508 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.952 4.414 ot_diis_step 99 11.5 0.012 0.012 3.786 3.786 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.724 3.728 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.698 3.720 apply_single 110 13.6 0.000 0.000 3.697 3.720 grid_integrate_task_list 110 12.3 3.376 3.498 3.376 3.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.421 3.421 pw_transfer 1331 11.6 0.065 0.070 3.315 3.321 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.208 3.218 multiply_cannon_sync_h2d 8220 15.4 2.914 3.072 2.914 3.072 cp_fm_cholesky_invert 11 10.9 2.998 3.002 2.998 3.002 make_images_data 4110 15.4 0.039 0.043 2.511 2.971 density_rs2pw 110 9.6 0.004 0.004 2.683 2.904 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.892 2.893 cp_fm_redistribute_end 48 14.0 0.729 2.861 0.734 2.861 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.405 2.839 cp_fm_diag_elpa_base 48 14.0 1.945 2.672 2.120 2.830 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.749 2.758 wfi_extrapolate 11 7.9 0.001 0.001 2.717 2.717 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.659 2.659 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.597 2.598 calculate_dm_sparse 110 9.5 0.001 0.001 2.498 2.558 fft3d_ps 1111 14.6 1.112 1.122 2.467 2.477 grid_collocate_task_list 110 9.6 2.328 2.469 2.328 2.469 calculate_first_density_matrix 1 7.0 0.000 0.000 2.435 2.462 rs_pw_transfer 902 11.9 0.010 0.011 2.192 2.424 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.223 2.233 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.995 potential_pw2rs 110 12.3 0.015 0.016 1.961 1.965 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.931 1.949 mp_alltoall_d11v 2046 13.8 1.688 1.912 1.688 1.912 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.861 1.873 cp_fm_cholesky_decompose 22 10.9 1.757 1.780 1.757 1.780 mp_allgather_i34 2055 14.4 0.572 1.743 0.572 1.743 jit_kernel_multiply 9 15.5 1.057 1.633 1.057 1.633 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.504 1.619 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.581 1.584 dbcsr_complete_redistribute 325 12.2 0.563 0.586 1.478 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.457 1.495 qs_create_task_list 11 7.9 0.001 0.001 1.227 1.323 generate_qs_task_list 11 8.9 0.380 0.448 1.226 1.322 mp_waitany 9240 13.8 1.059 1.309 1.059 1.309 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.846 1.206 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.150 1.176 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.480 1.160 mp_alltoall_z22v 1111 16.6 1.123 1.138 1.123 1.138 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.044000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=773.363636, yerr=11.749450 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 1.415365E+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 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 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 0.019 0.038 87.274 87.275 qs_mol_dyn_low 1 2.0 0.003 0.003 86.893 86.902 qs_forces 11 3.9 0.002 0.003 86.721 86.722 qs_energies 11 4.9 0.001 0.001 82.675 82.678 scf_env_do_scf 11 5.9 0.001 0.001 71.977 71.977 velocity_verlet 10 3.0 0.006 0.006 55.245 55.264 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 43.363 43.364 dbcsr_multiply_generic 2055 12.4 0.131 0.146 29.208 29.343 init_scf_loop 11 6.9 0.000 0.000 28.539 28.539 qs_scf_new_mos 99 7.5 0.001 0.001 26.866 26.909 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.866 26.908 prepare_preconditioner 11 7.9 0.000 0.000 26.581 26.589 make_preconditioner 11 8.9 0.000 0.000 26.581 26.589 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.904 26.056 ot_scf_mini 99 9.5 0.002 0.002 25.089 25.123 multiply_cannon 2055 13.4 0.343 0.364 21.954 22.886 multiply_cannon_loop 2055 14.4 0.341 0.344 20.134 20.439 cp_fm_upper_to_full 70 14.2 12.544 17.937 12.544 17.937 ot_mini 99 10.5 0.001 0.001 13.735 13.772 rebuild_ks_matrix 110 8.3 0.001 0.001 13.178 13.224 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.178 13.223 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.992 12.032 dbcsr_complete_redistribute 325 12.2 1.040 1.076 7.336 10.471 multiply_cannon_multrec 8220 15.4 4.432 4.607 9.930 10.256 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.273 9.400 qs_ot_get_derivative 99 11.5 0.001 0.001 9.206 9.237 mp_waitall_1 87304 16.6 8.034 8.953 8.034 8.953 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.662 8.752 mp_alltoall_i22 605 13.7 5.301 8.439 5.301 8.439 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.582 6.751 calculate_rho_elec 110 8.6 0.227 0.227 6.581 6.750 sum_up_and_integrate 110 10.3 0.151 0.153 6.699 6.707 integrate_v_rspace 110 11.3 0.003 0.004 6.549 6.557 init_scf_run 11 5.9 0.000 0.001 6.443 6.443 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.443 6.443 qs_ot_get_p 110 10.4 0.001 0.001 6.147 6.186 make_m2s 4110 13.4 0.043 0.044 5.397 5.914 dbcsr_mm_accdrv_process 11614 15.7 3.380 3.686 5.354 5.819 make_images 4110 14.4 0.880 0.927 5.210 5.727 cp_fm_cholesky_invert 11 10.9 5.700 5.705 5.700 5.705 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.612 5.091 apply_single 110 13.6 0.000 0.000 4.612 5.091 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.795 5.067 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.514 4.521 ot_diis_step 99 11.5 0.015 0.015 4.507 4.507 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.062 4.062 multiply_cannon_sync_h2d 8220 15.4 3.951 3.958 3.951 3.958 pw_transfer 1331 11.6 0.074 0.075 3.781 3.792 grid_integrate_task_list 110 12.3 3.674 3.726 3.674 3.726 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.664 3.675 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.632 3.633 calculate_dm_sparse 110 9.5 0.001 0.001 3.578 3.630 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.917 3.626 make_images_data 4110 15.4 0.042 0.044 2.931 3.581 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.428 3.428 cp_fm_diag_elpa_base 48 14.0 2.895 3.090 3.426 3.426 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.903 3.358 wfi_extrapolate 11 7.9 0.001 0.001 3.296 3.296 fft_wrap_pw1pw2_140 451 13.1 0.216 0.223 3.141 3.150 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.059 3.061 density_rs2pw 110 9.6 0.004 0.004 3.039 3.059 calculate_first_density_matrix 1 7.0 0.000 0.000 3.020 3.044 fft3d_ps 1111 14.6 1.265 1.275 2.874 2.885 grid_collocate_task_list 110 9.6 2.639 2.667 2.639 2.667 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.506 2.523 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.471 2.492 rs_pw_transfer 902 11.9 0.010 0.011 2.232 2.268 mp_alltoall_d11v 2046 13.8 2.117 2.229 2.117 2.229 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.225 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.094 2.170 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.098 2.138 cp_fm_cholesky_decompose 22 10.9 2.112 2.129 2.112 2.129 potential_pw2rs 110 12.3 0.021 0.022 2.120 2.129 jit_kernel_multiply 10 15.4 1.774 2.100 1.774 2.100 qs_create_task_list 11 7.9 0.001 0.001 1.915 1.964 generate_qs_task_list 11 8.9 0.742 0.795 1.914 1.963 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.841 1.880 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.797 1.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.275000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1227.545455, yerr=83.704419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.170752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3920228. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.029 245.199 245.200 qs_mol_dyn_low 1 2.0 0.003 0.003 244.797 244.811 qs_forces 11 3.9 0.003 0.003 244.704 244.705 qs_energies 11 4.9 0.001 0.002 239.157 239.171 scf_env_do_scf 11 5.9 0.001 0.001 222.196 222.200 scf_env_do_scf_inner_loop 117 6.6 0.002 0.007 200.413 200.415 velocity_verlet 10 3.0 0.001 0.004 163.285 163.286 dbcsr_multiply_generic 2507 12.6 0.190 0.194 157.165 157.864 qs_scf_new_mos 117 7.6 0.001 0.001 154.469 154.718 qs_scf_loop_do_ot 117 8.6 0.001 0.001 154.468 154.717 ot_scf_mini 117 9.6 0.003 0.003 147.835 148.076 multiply_cannon 2507 13.6 0.235 0.243 133.114 135.584 multiply_cannon_loop 2507 14.6 2.123 2.211 130.712 132.669 ot_mini 117 10.6 0.001 0.001 66.252 66.521 mp_waitall_1 291448 16.2 61.486 64.311 61.486 64.311 qs_ot_get_p 128 10.4 0.001 0.001 59.586 59.841 qs_ot_p2m_diag 83 11.4 0.077 0.091 52.676 52.744 multiply_cannon_metrocomm3 60168 15.6 0.115 0.121 20.641 50.043 multiply_cannon_multrec 60168 15.6 32.918 35.432 41.364 43.160 qs_ot_get_derivative 117 11.6 0.001 0.001 41.499 41.740 multiply_cannon_metrocomm1 60168 15.6 0.082 0.086 33.328 39.876 rebuild_ks_matrix 128 8.3 0.001 0.001 33.664 33.983 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.664 33.982 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.281 30.598 multiply_cannon_sync_h2d 60168 15.6 27.327 28.940 27.327 28.940 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.282 24.840 apply_single 128 13.6 0.001 0.001 24.281 24.840 ot_diis_step 117 11.6 0.007 0.008 24.413 24.415 init_scf_loop 11 6.9 0.000 0.000 21.703 21.704 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.503 19.738 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.833 17.835 prepare_preconditioner 11 7.9 0.000 0.000 17.090 17.146 make_preconditioner 11 8.9 0.000 0.000 17.090 17.146 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.263 16.431 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.864 14.872 cp_fm_redistribute_end 83 14.4 11.713 14.780 11.727 14.785 make_m2s 5014 13.6 0.104 0.111 14.330 14.701 cp_fm_diag_elpa_base 83 14.4 3.014 14.505 3.045 14.627 make_images 5014 14.6 0.398 0.417 14.149 14.534 sum_up_and_integrate 128 10.3 0.089 0.107 14.284 14.302 integrate_v_rspace 128 11.3 0.003 0.004 14.195 14.216 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.190 13.311 calculate_rho_elec 128 8.7 0.044 0.063 13.189 13.310 init_scf_run 11 5.9 0.000 0.001 12.758 12.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.758 12.759 cp_fm_cholesky_invert 11 10.9 9.832 9.840 9.832 9.840 mp_sum_l 7870 13.0 8.445 9.608 8.445 9.608 wfi_extrapolate 11 7.9 0.001 0.001 9.342 9.342 calculate_dm_sparse 128 9.5 0.001 0.001 8.466 8.545 dbcsr_mm_accdrv_process 124484 16.2 3.357 3.530 8.014 8.531 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.036 8.185 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.947 8.041 make_images_data 5014 15.6 0.067 0.073 7.051 8.004 pw_transfer 1547 11.6 0.075 0.107 7.307 7.609 density_rs2pw 128 9.7 0.006 0.007 6.872 7.499 grid_integrate_task_list 128 12.3 7.093 7.460 7.093 7.460 hybrid_alltoall_any 5200 16.5 0.292 2.260 6.155 7.429 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 7.103 7.383 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.794 6.802 mp_bcast_b 1983 13.7 6.589 6.767 6.589 6.767 external_control 127 7.2 0.067 6.410 6.617 6.725 rs_pw_transfer 1046 11.9 0.016 0.018 5.936 6.516 fft3d_ps 1291 14.7 2.095 2.577 5.901 6.122 fft_wrap_pw1pw2_140 523 13.2 0.443 0.508 5.929 6.121 mp_alltoall_d11v 2415 14.1 4.605 5.994 4.605 5.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.836 5.929 grid_collocate_task_list 128 9.7 4.730 5.073 4.730 5.073 cp_fm_cholesky_decompose 22 10.9 4.968 4.981 4.968 4.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=245.200000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.181818, yerr=6.873112 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 827.305984E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 6066927. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 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 0.053 0.072 236.174 236.176 qs_mol_dyn_low 1 2.0 0.003 0.004 235.739 235.753 qs_forces 11 3.9 0.003 0.004 235.653 235.655 qs_energies 11 4.9 0.009 0.016 228.837 228.849 scf_env_do_scf 11 5.9 0.001 0.001 212.113 212.123 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 177.286 177.289 velocity_verlet 10 3.0 0.002 0.002 162.747 162.749 dbcsr_multiply_generic 2507 12.6 0.187 0.194 131.588 132.936 qs_scf_new_mos 117 7.6 0.001 0.001 131.097 131.629 qs_scf_loop_do_ot 117 8.6 0.001 0.001 131.097 131.628 ot_scf_mini 117 9.6 0.004 0.005 126.211 126.773 ot_mini 117 10.6 0.001 0.001 83.672 84.288 multiply_cannon 2507 13.6 0.478 0.534 78.068 82.817 multiply_cannon_loop 2507 14.6 1.241 1.290 74.589 77.760 mp_waitall_1 226760 16.4 35.213 74.137 35.213 74.137 qs_ot_get_derivative 117 11.6 0.001 0.002 61.310 61.865 make_m2s 5014 13.6 0.089 0.094 47.257 48.778 make_images 5014 14.6 1.147 1.337 47.047 48.568 make_images_data 5014 15.6 0.064 0.072 17.185 43.051 hybrid_alltoall_any 5200 16.5 0.343 1.519 11.133 42.125 multiply_cannon_multrec 30084 15.6 21.463 25.554 31.099 35.646 init_scf_loop 11 6.9 0.006 0.015 34.725 34.727 make_images_sizes 5014 15.6 0.007 0.007 25.582 34.714 mp_alltoall_i44 5014 16.6 25.575 34.707 25.575 34.707 rebuild_ks_matrix 128 8.3 0.001 0.001 33.457 34.142 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.456 34.142 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.149 30.773 multiply_cannon_metrocomm3 30084 15.6 0.093 0.099 17.094 30.430 prepare_preconditioner 11 7.9 0.000 0.000 30.230 30.315 make_preconditioner 11 8.9 0.000 0.001 30.230 30.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.836 29.377 qs_ot_get_p 128 10.4 0.001 0.001 24.205 24.877 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.356 23.448 apply_single 128 13.6 0.001 0.001 22.355 23.448 ot_diis_step 117 11.6 0.014 0.015 22.157 22.160 multiply_cannon_sync_h2d 30084 15.6 18.784 20.845 18.784 20.845 qs_ot_p2m_diag 83 11.4 0.187 0.215 18.935 18.976 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.715 17.717 cp_fm_cholesky_invert 11 10.9 17.680 17.693 17.680 17.693 sum_up_and_integrate 128 10.3 0.116 0.132 14.807 14.834 integrate_v_rspace 128 11.3 0.003 0.003 14.690 14.722 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.414 14.425 cp_fm_redistribute_end 83 14.4 8.385 14.336 8.400 14.341 cp_fm_diag_elpa_base 83 14.4 5.694 13.838 5.920 14.218 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.554 13.583 calculate_rho_elec 128 8.7 0.087 0.105 13.553 13.583 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.541 11.987 init_scf_run 11 5.9 0.000 0.001 11.843 11.846 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.843 11.845 multiply_cannon_metrocomm4 27577 15.6 0.098 0.112 3.875 10.689 mp_irecv_dv 69486 16.3 3.680 10.303 3.680 10.303 dbcsr_mm_accdrv_process 62242 16.2 4.533 5.460 9.116 9.709 wfi_extrapolate 11 7.9 0.001 0.001 8.682 8.682 pw_transfer 1547 11.6 0.085 0.112 8.293 8.376 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.068 8.155 density_rs2pw 128 9.7 0.006 0.006 7.267 7.790 cp_fm_cholesky_decompose 22 10.9 7.607 7.692 7.607 7.692 grid_integrate_task_list 128 12.3 7.175 7.530 7.175 7.530 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.408 7.202 fft_wrap_pw1pw2_140 523 13.2 0.469 0.521 6.935 7.034 mp_bcast_b 1983 13.7 6.593 7.023 6.593 7.023 mp_sum_l 7870 13.0 4.504 6.768 4.504 6.768 external_control 127 7.2 0.135 6.470 6.617 6.740 calculate_dm_sparse 128 9.5 0.001 0.001 6.558 6.682 rs_pw_transfer 1046 11.9 0.014 0.016 6.106 6.625 fft3d_ps 1291 14.7 2.777 2.929 6.441 6.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.262 6.272 mp_alltoall_d11v 2415 14.1 4.582 5.586 4.582 5.586 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.492 5.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.321 5.537 mp_allgather_i34 2507 14.6 2.055 5.335 2.055 5.335 grid_collocate_task_list 128 9.7 4.915 5.249 4.915 5.249 potential_pw2rs 128 12.3 0.014 0.016 5.128 5.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=236.176000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.000000, yerr=2.215647 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 938.627072E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57222. MP_Allreduce 11246 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 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 0.040 0.087 214.254 214.256 qs_mol_dyn_low 1 2.0 0.003 0.003 213.689 213.703 qs_forces 11 3.9 0.002 0.003 213.585 213.589 qs_energies 11 4.9 0.002 0.007 206.988 206.996 scf_env_do_scf 11 5.9 0.001 0.002 191.318 191.319 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 155.474 155.477 velocity_verlet 10 3.0 0.001 0.002 150.333 150.335 dbcsr_multiply_generic 2485 12.5 0.181 0.186 81.621 82.674 qs_scf_new_mos 116 7.6 0.001 0.001 81.170 81.538 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.169 81.537 ot_scf_mini 116 9.6 0.004 0.005 77.027 77.430 mp_waitall_1 176908 16.5 57.800 67.503 57.800 67.503 multiply_cannon 2485 13.5 0.497 0.528 61.852 65.957 multiply_cannon_loop 2485 14.5 0.845 0.870 58.527 61.233 qs_rho_update_rho_low 127 7.7 0.001 0.001 45.048 45.089 calculate_rho_elec 127 8.7 0.131 0.145 45.048 45.089 ot_mini 116 10.6 0.001 0.001 42.538 42.938 density_rs2pw 127 9.7 0.005 0.006 38.550 39.123 rs_pw_transfer 1038 11.9 0.013 0.014 37.203 37.670 init_scf_loop 11 6.9 0.001 0.003 35.734 35.741 rs_pw_transfer_RS2PW_20 127 11.7 0.141 0.145 32.413 32.430 prepare_preconditioner 11 7.9 0.000 0.000 31.687 31.739 make_preconditioner 11 8.9 0.000 0.002 31.687 31.739 rebuild_ks_matrix 127 8.3 0.001 0.001 30.204 30.736 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.203 30.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.304 30.688 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.189 27.678 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 15.616 25.055 multiply_cannon_multrec 19880 15.5 13.249 15.896 21.908 24.565 qs_ot_get_derivative 116 11.6 0.001 0.002 22.713 23.114 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.973 21.072 apply_single 127 13.6 0.001 0.001 19.973 21.071 qs_ot_get_p 127 10.4 0.001 0.002 19.841 20.352 ot_diis_step 116 11.6 0.018 0.032 19.720 19.720 make_m2s 4970 13.5 0.080 0.085 14.742 15.929 multiply_cannon_sync_h2d 19880 15.5 14.119 15.845 14.119 15.845 make_images 4970 14.5 1.158 1.236 14.511 15.698 qs_ot_p2m_diag 82 11.4 0.262 0.269 15.624 15.639 cp_fm_cholesky_invert 11 10.9 14.906 14.915 14.906 14.915 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.604 14.606 sum_up_and_integrate 127 10.3 0.132 0.144 14.269 14.318 integrate_v_rspace 127 11.3 0.003 0.004 14.136 14.190 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.582 11.583 cp_fm_redistribute_end 82 14.4 4.349 11.521 4.366 11.523 cp_fm_diag_elpa_base 82 14.4 6.723 10.962 7.135 11.431 make_images_data 4970 15.5 0.060 0.066 9.124 10.747 init_scf_run 11 5.9 0.000 0.001 10.588 10.588 scf_env_initial_rho_setup 11 6.9 0.002 0.005 10.588 10.588 hybrid_alltoall_any 5155 16.4 0.428 1.953 8.017 9.808 multiply_cannon_metrocomm4 17395 15.5 0.059 0.069 3.470 9.630 mp_irecv_dv 49801 16.2 3.350 9.384 3.350 9.384 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.840 9.135 dbcsr_mm_accdrv_process 41158 16.2 4.380 5.062 8.121 8.295 pw_transfer 1535 11.6 0.084 0.101 7.888 8.017 fft_wrap_pw1pw2 1281 12.7 0.009 0.010 7.666 7.802 grid_integrate_task_list 127 12.3 7.235 7.734 7.235 7.734 cp_fm_cholesky_decompose 22 10.9 7.556 7.585 7.556 7.585 cp_fm_upper_to_full 104 14.5 5.783 7.482 5.783 7.482 wfi_extrapolate 11 7.9 0.001 0.001 7.468 7.468 fft_wrap_pw1pw2_140 519 13.2 0.474 0.530 6.719 6.865 dbcsr_complete_redistribute 393 12.7 1.171 1.202 4.567 6.359 mp_bcast_b 1987 13.8 5.776 6.313 5.776 6.313 fft3d_ps 1281 14.7 2.673 2.881 5.951 6.033 external_control 126 7.2 0.180 5.759 5.864 5.963 calculate_dm_sparse 127 9.5 0.001 0.001 5.741 5.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.549 5.556 grid_collocate_task_list 127 9.7 5.033 5.426 5.033 5.426 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.600 5.270 mp_alltoall_d11v 2401 14.1 4.327 5.250 4.327 5.250 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.357 5.149 mp_allgather_i34 2485 14.5 1.865 4.942 1.865 4.942 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.622 4.756 potential_pw2rs 127 12.3 0.020 0.022 4.660 4.700 mp_sum_l 7804 13.0 3.253 4.621 3.253 4.621 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=214.256000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.181818, yerr=9.388757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 1.144418E+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 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 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 0.041 0.107 226.177 226.178 qs_mol_dyn_low 1 2.0 0.003 0.003 225.700 225.715 qs_forces 11 3.9 0.003 0.012 225.585 225.591 qs_energies 11 4.9 0.005 0.035 218.546 218.559 scf_env_do_scf 11 5.9 0.001 0.005 201.686 201.701 velocity_verlet 10 3.0 0.004 0.006 162.287 162.290 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 122.381 122.382 qs_scf_new_mos 117 7.6 0.001 0.001 81.531 81.835 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.530 81.835 dbcsr_multiply_generic 2507 12.6 0.193 0.214 79.347 80.193 init_scf_loop 11 6.9 0.004 0.031 79.177 79.180 ot_scf_mini 117 9.6 0.003 0.004 77.045 77.335 prepare_preconditioner 11 7.9 0.000 0.000 74.952 74.971 make_preconditioner 11 8.9 0.000 0.001 74.952 74.971 make_full_inverse_cholesky 11 9.9 0.000 0.000 48.480 73.536 multiply_cannon 2507 13.6 0.554 0.583 54.582 57.730 multiply_cannon_loop 2507 14.6 1.182 1.219 50.848 52.888 cp_fm_upper_to_full 105 14.7 23.012 48.413 23.012 48.413 dbcsr_complete_redistribute 395 12.7 1.513 1.634 29.212 45.036 copy_fm_to_dbcsr 209 11.7 0.002 0.002 27.787 43.596 ot_mini 117 10.6 0.001 0.001 42.343 42.621 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 26.450 42.108 mp_alltoall_i22 716 14.1 25.702 41.563 25.702 41.563 multiply_cannon_multrec 30084 15.6 14.162 19.724 26.095 31.088 rebuild_ks_matrix 128 8.3 0.001 0.001 29.055 29.311 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.063 29.054 29.310 mp_waitall_1 153770 16.5 17.354 26.749 17.354 26.749 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.299 26.530 qs_ot_get_derivative 117 11.6 0.001 0.002 22.689 22.971 make_m2s 5014 13.6 0.095 0.099 20.321 21.454 qs_ot_get_p 128 10.4 0.001 0.001 20.915 21.231 make_images 5014 14.6 1.953 2.232 20.018 21.153 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.988 19.553 apply_single 128 13.6 0.001 0.001 18.988 19.553 ot_diis_step 117 11.6 0.017 0.018 19.527 19.528 qs_ot_p2m_diag 83 11.4 0.342 0.390 16.775 16.826 cp_fm_cholesky_invert 11 10.9 16.580 16.589 16.580 16.589 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.479 15.480 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.190 14.724 sum_up_and_integrate 128 10.3 0.140 0.151 14.089 14.119 integrate_v_rspace 128 11.3 0.003 0.003 13.949 13.983 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.158 13.192 calculate_rho_elec 128 8.7 0.175 0.190 13.158 13.192 make_images_data 5014 15.6 0.064 0.068 10.877 12.701 multiply_cannon_sync_h2d 30084 15.6 11.781 12.590 11.781 12.590 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.310 12.313 cp_fm_redistribute_end 83 14.4 2.095 12.227 2.109 12.233 cp_fm_diag_elpa_base 83 14.4 9.499 11.640 10.086 12.115 dbcsr_mm_accdrv_process 62264 16.2 7.392 8.414 11.506 12.008 hybrid_alltoall_any 5200 16.5 0.525 2.220 9.706 11.734 init_scf_run 11 5.9 0.000 0.001 10.870 10.871 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.869 10.871 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.416 9.613 pw_transfer 1547 11.6 0.085 0.099 7.889 7.975 grid_integrate_task_list 128 12.3 7.500 7.895 7.500 7.895 cp_fm_cholesky_decompose 22 10.9 7.761 7.847 7.761 7.847 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.664 7.752 wfi_extrapolate 11 7.9 0.001 0.001 7.625 7.625 multiply_cannon_metrocomm4 25070 15.6 0.075 0.084 2.772 7.248 mp_irecv_dv 76098 16.2 2.629 6.987 2.629 6.987 fft_wrap_pw1pw2_140 523 13.2 0.482 0.488 6.729 6.830 density_rs2pw 128 9.7 0.005 0.006 6.289 6.703 calculate_dm_sparse 128 9.5 0.001 0.001 6.188 6.273 fft3d_ps 1291 14.7 2.803 2.885 5.894 5.973 mp_bcast_b 1983 13.8 5.320 5.895 5.320 5.895 mp_alltoall_d11v 2415 14.1 4.929 5.662 4.929 5.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.525 5.575 grid_collocate_task_list 128 9.7 5.198 5.524 5.198 5.524 external_control 127 7.2 0.223 5.342 5.422 5.487 rs_pw_transfer 1046 11.9 0.013 0.014 4.681 5.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=226.178000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.090909, yerr=21.262235 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/19/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 5.820059E+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 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 1.492713E+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 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828160. MP_ISendRecv 7620 122880. MP_Wait 19810 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 0.019 0.036 210.175 210.176 qs_mol_dyn_low 1 2.0 0.003 0.004 209.751 209.786 qs_forces 11 3.9 0.003 0.003 209.646 209.650 qs_energies 11 4.9 0.002 0.003 202.343 202.349 scf_env_do_scf 11 5.9 0.001 0.001 184.995 185.011 velocity_verlet 10 3.0 0.002 0.002 149.835 149.839 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.345 116.347 qs_scf_new_mos 116 7.6 0.001 0.001 75.834 75.911 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.834 75.911 dbcsr_multiply_generic 2485 12.5 0.183 0.189 74.327 74.749 ot_scf_mini 116 9.6 0.003 0.004 71.444 71.502 init_scf_loop 11 6.9 0.000 0.000 68.491 68.492 prepare_preconditioner 11 7.9 0.000 0.000 64.654 64.683 make_preconditioner 11 8.9 0.000 0.000 64.654 64.683 make_full_inverse_cholesky 11 9.9 0.000 0.000 62.496 62.779 multiply_cannon 2485 13.5 0.576 0.594 54.358 59.062 cp_fm_cholesky_invert 11 10.9 50.801 50.808 50.801 50.808 multiply_cannon_loop 2485 14.5 0.441 0.452 49.734 50.529 ot_mini 116 10.6 0.001 0.001 39.644 39.696 mp_waitall_1 128490 16.6 26.527 33.586 26.527 33.586 rebuild_ks_matrix 127 8.3 0.001 0.001 28.529 28.605 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.528 28.605 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.036 26.101 multiply_cannon_multrec 9940 15.5 10.374 14.132 17.701 20.925 qs_ot_get_derivative 116 11.6 0.001 0.002 19.836 19.897 ot_diis_step 116 11.6 0.019 0.020 19.732 19.733 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.367 19.633 apply_single 127 13.6 0.001 0.001 19.367 19.633 qs_ot_get_p 127 10.4 0.001 0.001 18.879 18.967 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.226 18.941 make_m2s 4970 13.5 0.066 0.073 16.331 18.386 make_images 4970 14.5 2.276 2.565 16.027 18.085 qs_ot_p2m_diag 82 11.4 0.489 0.494 14.998 15.015 sum_up_and_integrate 127 10.3 0.178 0.188 14.061 14.111 integrate_v_rspace 127 11.3 0.004 0.004 13.882 13.935 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.846 13.847 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.231 13.284 calculate_rho_elec 127 8.7 0.256 0.267 13.231 13.283 make_images_data 4970 15.5 0.052 0.061 9.999 12.517 hybrid_alltoall_any 5155 16.4 0.810 3.598 9.737 12.170 multiply_cannon_sync_h2d 9940 15.5 11.520 12.086 11.520 12.086 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.779 10.780 cp_fm_diag_elpa_base 82 14.4 10.523 10.608 10.772 10.772 init_scf_run 11 5.9 0.000 0.001 10.515 10.516 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.515 10.515 cp_fm_cholesky_decompose 22 10.9 8.120 8.222 8.120 8.222 grid_integrate_task_list 127 12.3 7.669 8.036 7.669 8.036 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.958 8.008 pw_transfer 1535 11.6 0.083 0.091 7.937 7.960 mp_allgather_i34 2485 14.5 2.855 7.854 2.855 7.854 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.718 7.747 dbcsr_mm_accdrv_process 20590 16.1 2.560 3.296 6.966 7.601 multiply_cannon_metrocomm1 9940 15.5 0.028 0.028 4.677 7.513 wfi_extrapolate 11 7.9 0.001 0.001 7.436 7.437 fft_wrap_pw1pw2_140 519 13.2 0.501 0.519 6.741 6.778 density_rs2pw 127 9.7 0.005 0.006 6.056 6.325 calculate_dm_sparse 127 9.5 0.001 0.001 6.143 6.206 fft3d_ps 1281 14.7 2.711 2.829 5.887 5.918 grid_collocate_task_list 127 9.7 5.443 5.791 5.443 5.791 dbcsr_complete_redistribute 393 12.7 2.120 2.183 5.342 5.760 mp_alltoall_d11v 2401 14.1 4.893 5.594 4.893 5.594 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.242 5.267 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.246 5.247 mp_bcast_b 1988 13.8 4.709 5.195 4.709 5.195 external_control 126 7.2 0.305 4.870 4.940 5.049 rs_pw_transfer 1038 11.9 0.012 0.013 4.147 4.442 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=210.176000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1400.272727, yerr=42.935662 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.963067E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59820. MP_Allreduce 10885 1518. MP_Sync 86 MP_Alltoall 1700 36954426. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.041 324.442 324.442 qs_mol_dyn_low 1 2.0 0.003 0.004 323.852 323.865 qs_forces 11 3.9 0.003 0.003 323.750 323.754 qs_energies 11 4.9 0.002 0.002 315.171 315.182 scf_env_do_scf 11 5.9 0.001 0.001 293.561 293.571 velocity_verlet 10 3.0 0.013 0.014 244.048 244.056 init_scf_loop 11 6.9 0.000 0.000 156.378 156.383 prepare_preconditioner 11 7.9 0.000 0.000 151.759 151.789 make_preconditioner 11 8.9 0.000 0.000 151.759 151.789 make_full_inverse_cholesky 11 9.9 0.000 0.000 103.707 148.882 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 136.926 136.929 cp_fm_upper_to_full 104 14.8 60.845 107.798 60.845 107.798 qs_scf_new_mos 116 7.6 0.001 0.001 90.579 90.688 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.578 90.687 ot_scf_mini 116 9.6 0.004 0.004 85.838 85.877 dbcsr_multiply_generic 2485 12.5 0.213 0.219 82.407 82.948 dbcsr_complete_redistribute 393 12.7 3.977 4.015 53.687 74.398 copy_fm_to_dbcsr 208 11.6 0.001 0.002 50.312 71.030 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 48.005 68.569 mp_alltoall_i22 712 14.1 45.792 66.662 45.792 66.662 multiply_cannon 2485 13.5 0.702 0.752 58.034 58.667 multiply_cannon_loop 2485 14.5 0.466 0.477 54.474 55.783 ot_mini 116 10.6 0.001 0.001 44.056 44.092 cp_fm_cholesky_invert 11 10.9 33.553 33.561 33.553 33.561 rebuild_ks_matrix 127 8.3 0.001 0.001 33.366 33.406 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.366 33.406 mp_waitall_1 104546 16.7 28.020 31.887 28.020 31.887 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.141 31.183 qs_ot_get_p 127 10.4 0.001 0.001 26.505 26.514 qs_ot_get_derivative 116 11.6 0.002 0.002 24.316 24.355 qs_ot_p2m_diag 82 11.4 0.868 0.874 22.440 22.468 make_m2s 4970 13.5 0.076 0.081 19.839 20.742 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.733 20.734 make_images 4970 14.5 3.702 3.820 19.363 20.266 ot_diis_step 116 11.6 0.022 0.024 19.715 19.715 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 18.439 19.548 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.953 19.157 apply_single 127 13.6 0.001 0.001 18.953 19.157 multiply_cannon_multrec 9940 15.5 10.543 12.425 17.972 18.023 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.494 17.495 cp_fm_diag_elpa_base 82 14.4 13.168 14.740 17.490 17.490 sum_up_and_integrate 127 10.3 0.319 0.320 15.667 15.760 multiply_cannon_sync_h2d 9940 15.5 15.534 15.549 15.534 15.549 integrate_v_rspace 127 11.3 0.004 0.004 15.348 15.442 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.045 15.058 calculate_rho_elec 127 8.7 0.479 0.480 15.045 15.057 hybrid_alltoall_any 5155 16.4 1.293 3.013 10.731 12.602 make_images_data 4970 15.5 0.060 0.064 10.587 12.280 init_scf_run 11 5.9 0.000 0.001 11.738 11.738 scf_env_initial_rho_setup 11 6.9 0.023 0.024 11.738 11.738 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.397 9.433 cp_fm_cholesky_decompose 22 10.9 9.007 9.066 9.007 9.066 dbcsr_mm_accdrv_process 20590 16.0 3.852 5.815 7.196 9.044 grid_integrate_task_list 127 12.3 8.482 8.659 8.482 8.659 wfi_extrapolate 11 7.9 0.001 0.001 8.627 8.627 pw_transfer 1535 11.6 0.090 0.091 8.585 8.593 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.353 8.360 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.920 7.921 fft_wrap_pw1pw2_140 519 13.2 0.536 0.539 7.359 7.374 mp_bcast_b 1964 13.8 4.616 7.163 4.616 7.163 mp_alltoall_d11v 2401 14.1 6.973 7.101 6.973 7.101 calculate_dm_sparse 127 9.5 0.001 0.001 6.572 6.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=324.442000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2633.272727, yerr=168.832941 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 1.262137E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.063 0.098 85.609 85.616 qs_energies 1 2.0 0.002 0.016 84.805 84.838 ls_scf 1 3.0 0.003 0.023 83.836 83.854 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.822 73.042 multiply_cannon 111 7.7 0.017 0.020 55.995 57.428 multiply_cannon_loop 111 8.7 0.210 0.223 52.561 54.043 ls_scf_main 1 4.0 0.000 0.001 52.526 52.538 density_matrix_trs4 2 5.0 0.002 0.022 46.978 47.069 ls_scf_init_scf 1 4.0 0.001 0.002 28.249 28.274 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.062 27.118 mp_waitall_1 11316 10.9 22.693 26.319 22.693 26.319 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.959 24.987 multiply_cannon_multrec 2664 9.7 8.178 8.883 15.453 17.135 multiply_cannon_sync_h2d 2664 9.7 13.667 15.483 13.667 15.483 make_m2s 222 7.7 0.008 0.010 13.125 13.609 make_images 222 8.7 0.098 0.107 13.103 13.590 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.876 13.552 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.438 9.345 make_images_data 222 9.7 0.004 0.005 7.700 8.287 hybrid_alltoall_any 227 10.6 0.215 1.829 6.520 8.163 dbcsr_mm_accdrv_process 4760 10.4 0.504 0.601 6.894 7.874 dbcsr_mm_accdrv_process_sort 4760 11.4 6.187 7.073 6.187 7.073 calculate_norms 4752 9.8 5.491 6.135 5.491 6.135 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.130 5.267 mp_sum_l 807 5.4 3.184 4.865 3.184 4.865 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.039 3.607 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.339 3.596 mp_irecv_dv 6231 10.9 2.021 3.575 2.021 3.575 make_images_sizes 222 9.7 0.000 0.000 0.795 3.346 mp_alltoall_i44 222 10.7 0.795 3.346 0.795 3.346 arnoldi_extremal 4 6.8 0.000 0.000 3.253 3.277 arnoldi_normal_ev 4 7.8 0.001 0.005 3.253 3.277 build_subspace 16 8.4 0.009 0.012 3.151 3.153 ls_scf_post 1 4.0 0.002 0.015 3.058 3.077 ls_scf_store_result 1 5.0 0.000 0.000 2.874 2.914 dbcsr_special_finalize 555 9.7 0.005 0.006 2.278 2.769 dbcsr_merge_single_wm 555 10.7 0.451 0.564 2.271 2.761 make_images_pack 222 9.7 2.207 2.647 2.209 2.648 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.316 2.553 dbcsr_sort_data 658 11.4 2.077 2.521 2.077 2.521 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.451 2.066 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.006 2.190 2.312 buffer_matrices_ensure_size 222 8.7 1.741 2.067 1.741 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.773 1.799 rebuild_ks_matrix 3 7.3 0.000 0.000 1.764 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.034 1.764 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.616000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 2.104836E+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 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.027 0.051 91.035 91.045 qs_energies 1 2.0 0.000 0.000 90.602 90.608 ls_scf 1 3.0 0.000 0.000 89.288 89.293 dbcsr_multiply_generic 111 6.7 0.017 0.034 75.457 75.757 multiply_cannon 111 7.7 0.028 0.040 53.210 57.734 ls_scf_main 1 4.0 0.000 0.000 55.044 55.057 multiply_cannon_loop 111 8.7 0.116 0.123 49.962 53.759 density_matrix_trs4 2 5.0 0.002 0.003 49.377 49.584 ls_scf_init_scf 1 4.0 0.000 0.002 30.696 30.698 mp_waitall_1 9246 10.9 21.172 29.727 21.172 29.727 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.582 29.657 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.191 27.202 multiply_cannon_multrec 1332 9.7 13.091 16.450 22.269 26.744 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.900 20.415 make_m2s 222 7.7 0.006 0.007 15.193 15.795 make_images 222 8.7 1.561 1.930 15.163 15.767 dbcsr_mm_accdrv_process 4041 10.4 0.284 0.458 8.780 10.372 dbcsr_mm_accdrv_process_sort 4041 11.4 8.372 9.914 8.372 9.914 make_images_data 222 9.7 0.004 0.004 8.660 9.635 mp_sum_l 807 5.4 5.752 9.165 5.752 9.165 hybrid_alltoall_any 227 10.6 0.519 2.443 8.130 9.058 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.211 7.781 mp_irecv_dv 3311 11.0 3.191 7.720 3.191 7.720 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.513 7.426 calculate_norms 2376 9.8 5.989 6.619 5.989 6.619 multiply_cannon_sync_h2d 1332 9.7 4.791 6.504 4.791 6.504 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.062 5.315 arnoldi_extremal 4 6.8 0.000 0.000 4.692 4.710 arnoldi_normal_ev 4 7.8 0.001 0.004 4.692 4.710 build_subspace 16 8.4 0.014 0.021 4.433 4.436 ls_scf_post 1 4.0 0.000 0.000 3.548 3.554 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.163 3.379 ls_scf_store_result 1 5.0 0.000 0.000 3.257 3.374 dbcsr_matrix_vector_mult_local 304 10.0 2.754 3.236 2.756 3.238 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.235 2.801 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.505 2.608 make_images_pack 222 9.7 2.027 2.402 2.029 2.404 mp_allgather_i34 111 8.7 0.933 2.265 0.933 2.265 dbcsr_sort_data 436 11.2 1.799 1.996 1.799 1.996 dbcsr_data_new 4174 10.1 1.613 1.846 1.613 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.045000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 2.729771E+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 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.053 0.075 93.743 93.744 qs_energies 1 2.0 0.002 0.010 93.071 93.081 ls_scf 1 3.0 0.000 0.001 91.646 91.652 dbcsr_multiply_generic 111 6.7 0.016 0.018 76.293 76.546 ls_scf_main 1 4.0 0.000 0.001 57.288 57.293 multiply_cannon 111 7.7 0.048 0.144 52.829 57.194 multiply_cannon_loop 111 8.7 0.099 0.105 49.137 52.333 density_matrix_trs4 2 5.0 0.003 0.019 51.365 51.582 mp_waitall_1 7374 11.0 23.584 32.416 23.584 32.416 ls_scf_init_scf 1 4.0 0.000 0.002 30.787 30.789 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.569 29.651 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.194 27.209 multiply_cannon_multrec 888 9.7 12.602 15.173 21.109 24.356 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.001 21.789 make_m2s 222 7.7 0.006 0.007 16.793 18.082 make_images 222 8.7 1.967 2.257 16.754 18.045 make_images_data 222 9.7 0.004 0.004 9.542 10.704 hybrid_alltoall_any 227 10.6 0.619 2.862 9.207 10.501 mp_sum_l 807 5.4 5.103 9.375 5.103 9.375 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.417 8.020 9.188 dbcsr_mm_accdrv_process_sort 3754 11.4 7.648 8.771 7.648 8.771 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.899 7.406 multiply_cannon_sync_h2d 888 9.7 6.029 7.213 6.029 7.213 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.443 6.945 mp_irecv_dv 2335 11.1 2.428 6.904 2.428 6.904 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.658 6.334 arnoldi_extremal 4 6.8 0.000 0.000 5.218 5.238 arnoldi_normal_ev 4 7.8 0.001 0.005 5.218 5.238 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.970 5.214 build_subspace 16 8.4 0.014 0.020 4.900 4.906 calculate_norms 1584 9.8 4.280 4.639 4.280 4.639 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.488 3.859 mp_allgather_i34 111 8.7 1.527 3.799 1.527 3.799 dbcsr_matrix_vector_mult_local 304 10.0 3.033 3.611 3.035 3.613 ls_scf_post 1 4.0 0.001 0.004 3.571 3.578 ls_scf_store_result 1 5.0 0.000 0.000 3.309 3.391 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.737 2.840 make_images_pack 222 9.7 1.815 2.134 1.818 2.137 make_images_sizes 222 9.7 0.000 0.000 0.952 2.129 mp_alltoall_i44 222 10.7 0.952 2.128 0.952 2.128 dbcsr_sort_data 325 11.1 1.870 2.123 1.870 2.123 dbcsr_data_release 9322 10.9 1.324 1.961 1.324 1.961 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.931 1.934 rebuild_ks_matrix 3 7.3 0.000 0.000 1.913 1.915 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.913 1.915 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.744000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 3.332899E+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 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 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 0.056 0.078 97.237 97.238 qs_energies 1 2.0 0.000 0.001 96.481 96.489 ls_scf 1 3.0 0.000 0.000 94.780 94.794 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.531 78.831 ls_scf_main 1 4.0 0.000 0.001 58.860 58.861 multiply_cannon 111 7.7 0.054 0.154 51.477 56.396 density_matrix_trs4 2 5.0 0.002 0.008 52.789 52.946 multiply_cannon_loop 111 8.7 0.115 0.126 46.456 49.840 ls_scf_init_scf 1 4.0 0.003 0.015 32.685 32.696 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.490 31.575 mp_waitall_1 6438 11.0 22.893 29.982 22.893 29.982 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.970 28.982 multiply_cannon_multrec 1332 9.7 14.119 17.706 21.943 25.051 make_m2s 222 7.7 0.006 0.007 21.342 22.746 make_images 222 8.7 3.134 3.600 21.292 22.698 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.158 17.561 make_images_data 222 9.7 0.004 0.004 11.956 13.560 hybrid_alltoall_any 227 10.6 0.799 3.843 11.358 13.168 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.400 7.456 8.950 dbcsr_mm_accdrv_process_sort 3641 11.4 7.082 8.532 7.082 8.532 mp_sum_l 807 5.4 4.123 7.331 4.123 7.331 multiply_cannon_sync_h2d 1332 9.7 5.527 6.100 5.527 6.100 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.083 6.024 mp_irecv_dv 3229 10.9 2.060 5.940 2.060 5.940 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.150 5.702 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.608 5.476 arnoldi_extremal 4 6.8 0.000 0.000 5.251 5.264 arnoldi_normal_ev 4 7.8 0.006 0.019 5.251 5.264 build_subspace 16 8.4 0.014 0.021 4.896 4.905 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.573 4.770 calculate_norms 2376 9.8 4.194 4.587 4.194 4.587 mp_allgather_i34 111 8.7 2.146 4.390 2.146 4.390 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.595 3.877 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.695 3.201 3.697 dbcsr_sort_data 658 11.4 3.126 3.432 3.126 3.432 ls_scf_post 1 4.0 0.000 0.001 3.235 3.242 dbcsr_special_finalize 555 9.7 0.006 0.007 2.865 3.174 dbcsr_merge_single_wm 555 10.7 0.535 0.647 2.856 3.166 ls_scf_store_result 1 5.0 0.000 0.000 2.984 3.036 ls_scf_dm_to_ks 2 5.0 0.001 0.011 2.926 2.989 dbcsr_data_release 10477 10.7 1.580 2.481 1.580 2.481 dbcsr_finalize 304 7.8 0.049 0.061 1.806 2.049 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.238000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2698.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 4.635271E+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 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.060 0.071 98.768 98.778 qs_energies 1 2.0 0.000 0.000 97.950 97.955 ls_scf 1 3.0 0.000 0.000 96.017 96.022 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.735 77.932 ls_scf_main 1 4.0 0.000 0.000 61.639 61.640 multiply_cannon 111 7.7 0.090 0.231 54.968 59.843 density_matrix_trs4 2 5.0 0.002 0.003 54.635 54.707 multiply_cannon_loop 111 8.7 0.069 0.079 50.416 52.122 mp_waitall_1 5481 11.0 26.436 32.667 26.436 32.667 ls_scf_init_scf 1 4.0 0.001 0.002 30.729 30.731 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.544 29.585 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.370 27.378 multiply_cannon_multrec 444 9.7 13.953 16.776 20.951 24.756 make_m2s 222 7.7 0.004 0.005 17.808 20.325 make_images 222 8.7 3.718 4.422 17.746 20.265 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.070 15.792 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.066 14.597 make_images_data 222 9.7 0.003 0.004 10.093 12.557 hybrid_alltoall_any 227 10.6 0.787 3.773 9.791 12.337 multiply_cannon_sync_h2d 444 9.7 6.479 8.708 6.479 8.708 dbcsr_mm_accdrv_process 3003 10.4 0.166 0.342 6.704 7.820 dbcsr_mm_accdrv_process_sort 3003 11.4 6.387 7.472 6.387 7.472 mp_allgather_i34 111 8.7 2.789 6.968 2.789 6.968 arnoldi_extremal 4 6.8 0.000 0.000 5.846 5.855 arnoldi_normal_ev 4 7.8 0.002 0.005 5.846 5.855 mp_sum_l 807 5.4 2.986 5.626 2.986 5.626 build_subspace 16 8.4 0.015 0.020 5.457 5.466 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.629 4.823 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.173 4.377 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.660 4.328 mp_irecv_dv 1241 11.2 1.641 4.308 1.641 4.308 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.079 4.294 dbcsr_matrix_vector_mult_local 304 10.0 3.693 4.142 3.695 4.144 calculate_norms 792 9.8 3.542 3.676 3.542 3.676 ls_scf_post 1 4.0 0.000 0.000 3.649 3.654 make_images_sizes 222 9.7 0.000 0.000 1.077 3.574 mp_alltoall_i44 222 10.7 1.077 3.573 1.077 3.573 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.380 3.491 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.465 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.331 dbcsr_merge_all 275 8.9 0.474 0.534 2.053 2.160 dbcsr_data_release 10123 10.8 1.337 1.986 1.337 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.778000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3625.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7444654cf14297a8dae96b6cf94e55baac10f21c_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 8.692957E+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 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 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 0.091 0.111 108.728 108.735 qs_energies 1 2.0 0.000 0.000 107.328 107.339 ls_scf 1 3.0 0.000 0.000 104.435 104.447 dbcsr_multiply_generic 111 6.7 0.024 0.026 78.354 78.481 ls_scf_main 1 4.0 0.000 0.000 66.322 66.333 density_matrix_trs4 2 5.0 0.002 0.003 57.374 57.425 multiply_cannon 111 7.7 0.145 0.280 50.708 52.862 multiply_cannon_loop 111 8.7 0.067 0.069 47.204 48.150 ls_scf_init_scf 1 4.0 0.001 0.001 34.406 34.408 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.929 32.938 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.109 30.121 mp_waitall_1 4569 11.1 23.025 26.459 23.025 26.459 make_m2s 222 7.7 0.005 0.005 24.090 25.138 make_images 222 8.7 4.585 4.977 23.984 25.029 multiply_cannon_multrec 444 9.7 17.845 18.522 22.431 23.037 hybrid_alltoall_any 227 10.6 1.661 3.640 12.982 15.816 make_images_data 222 9.7 0.003 0.003 13.246 15.804 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.143 11.852 multiply_cannon_sync_h2d 444 9.7 8.846 8.891 8.846 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.453 7.470 arnoldi_normal_ev 4 7.8 0.003 0.009 7.453 7.470 build_subspace 16 8.4 0.026 0.036 6.904 6.916 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.526 5.651 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.073 5.313 dbcsr_matrix_vector_mult_local 304 10.0 5.034 5.295 5.036 5.297 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.824 4.914 dbcsr_mm_accdrv_process 1814 10.4 0.212 0.335 4.416 4.558 dbcsr_mm_accdrv_process_sort 1814 11.4 4.110 4.241 4.110 4.241 ls_scf_post 1 4.0 0.000 0.000 3.707 3.719 make_images_sizes 222 9.7 0.000 0.000 1.520 3.686 mp_alltoall_i44 222 10.7 1.520 3.686 1.520 3.686 mp_allgather_i34 111 8.7 1.160 3.610 1.160 3.610 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.427 calculate_norms 792 9.8 3.242 3.273 3.242 3.273 dbcsr_finalize 304 7.8 0.082 0.090 3.086 3.143 dbcsr_complete_redistribute 5 7.6 1.440 1.471 2.824 2.951 dbcsr_merge_all 275 8.9 0.891 0.919 2.865 2.909 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.862 2.862 dbcsr_data_release 12724 10.6 2.335 2.849 2.335 2.849 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.562 dbcsr_sort_data 325 11.1 2.442 2.500 2.442 2.500 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.302 2.304 dbcsr_new_transposed 4 7.5 0.243 0.250 2.236 2.249 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.212 2.249 rebuild_ks_matrix 3 7.3 0.000 0.000 2.236 2.238 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.012 2.236 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.157 2.227 dbcsr_add_anytype 103 7.2 0.860 0.891 2.157 2.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.735000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7035.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7444654cf14297a8dae96b6cf94e55baac10f21c Summary: empty Status: OK