=== 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: 0a22203fba5dd876101bb33c272b97e0757c22c1 ################# 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 2022.11.001.rc2, 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 (13.12.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001.rc2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/01 job id: 43491625 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/02 job id: 43491626 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/03 job id: 43491627 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/04 job id: 43491628 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/05 job id: 43491629 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/06 job id: 43491630 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/07 job id: 43491631 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/08 job id: 43491632 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/09 job id: 43491633 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/10 job id: 43491634 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/11 job id: 43491635 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/12 job id: 43491636 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/13 job id: 43491637 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/14 job id: 43491639 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/15 job id: 43491640 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/16 job id: 43491641 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/17 job id: 43491642 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/18 job id: 43491643 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/19 job id: 43491644 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/20 job id: 43491645 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/21 job id: 43491646 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/22 job id: 43491647 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/23 job id: 43491648 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/24 job id: 43491649 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/25 job id: 43491650 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/26 job id: 43491651 --- 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/0a22203fba5dd876101bb33c272b97e0757c22c1_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 0.021 0.032 134.961 134.962 farming_run 1 2.0 134.276 134.346 134.930 134.934 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459028E+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 0.020 0.036 116.670 116.670 qs_energies 1 2.0 0.000 0.000 116.404 116.406 mp2_main 1 3.0 0.000 0.000 114.101 114.103 mp2_gpw_main 1 4.0 0.024 0.079 113.069 113.071 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.903 94.431 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.500 56.028 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 41.824 47.326 get_2c_integrals 1 6.0 0.000 0.000 37.621 38.232 integrate_v_rspace 273 8.0 0.436 0.448 25.150 30.316 pw_transfer 6555 10.6 0.376 0.393 27.385 27.999 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.045 26.657 grid_integrate_task_list 273 9.0 20.956 26.601 20.956 26.601 fft_wrap_pw1pw2_100 2178 12.4 1.188 1.416 23.608 24.205 compute_2c_integrals 1 7.0 0.011 0.012 19.701 19.702 compute_2c_integrals_loop_lm 1 8.0 0.004 0.004 18.883 19.381 mp2_eri_2c_integrate_gpw 1 9.0 2.376 2.425 18.880 19.376 rpa_ri_compute_en 1 5.0 0.010 0.013 18.995 19.202 cp_fm_cholesky_decompose 12 8.2 17.951 18.576 17.951 18.576 cholesky_decomp 1 7.0 0.000 0.000 16.759 17.375 fft3d_s 5443 13.4 16.199 16.692 16.221 16.713 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.576 10.851 15.576 calculate_wavefunction 272 8.0 5.453 5.601 12.592 13.183 rpa_num_int 1 6.0 0.000 0.001 10.604 10.614 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.587 10.588 calc_mat_Q 8 8.0 0.000 0.000 9.355 9.458 contract_S_to_Q 8 9.0 0.000 0.000 8.779 8.879 calc_potential_gpw 544 9.5 0.005 0.006 8.229 8.699 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.225 8.583 parallel_gemm_fm 14 9.1 0.000 0.000 8.366 8.438 parallel_gemm_fm_cosma 14 10.1 8.366 8.438 8.366 8.438 potential_pw2rs 545 10.0 0.107 0.109 7.730 8.412 create_integ_mat 1 6.0 0.015 0.027 7.758 7.767 collocate_single_gaussian 272 10.0 0.041 0.043 7.464 7.708 array2fm 1 7.0 0.000 0.000 6.752 7.131 pw_scatter_s 2720 13.7 4.423 4.645 4.423 4.645 pw_gather_s 2722 13.2 3.865 4.229 3.865 4.229 array2fm_buffer_send 1 8.0 2.965 3.143 2.965 3.143 pw_poisson_solve 545 10.5 1.128 1.176 2.173 2.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.069378, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 0.029 0.039 397.107 397.108 farming_run 1 2.0 396.081 396.083 397.061 397.063 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227158E+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 20248069. 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 0.015 0.052 209.707 209.708 qs_energies 1 2.0 0.000 0.000 209.470 209.483 scf_env_do_scf 1 3.0 0.000 0.000 106.297 106.297 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.387 105.397 rebuild_ks_matrix 4 6.0 0.000 0.000 105.386 105.396 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.065 105.386 105.396 hfx_ks_matrix 4 8.0 0.001 0.001 105.001 105.004 integrate_four_center 4 9.0 0.144 0.455 105.000 105.003 mp2_main 1 3.0 0.000 0.000 102.877 102.890 mp2_gpw_main 1 4.0 0.044 0.094 101.890 101.907 integrate_four_center_main 4 10.0 0.100 0.545 96.542 99.487 integrate_four_center_bin 267 11.0 96.441 99.391 96.441 99.391 init_scf_loop 1 4.0 0.000 0.000 92.090 92.090 mp2_ri_gpw_compute_in 1 5.0 0.066 0.077 75.090 76.172 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.516 55.582 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 42.098 47.348 integrate_v_rspace 95 8.0 0.398 0.569 28.472 33.500 pw_transfer 2240 10.6 0.143 0.163 29.935 30.418 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.941 29.401 grid_integrate_task_list 95 9.0 23.742 28.979 23.742 28.979 ao_to_mo_and_store_B_mult_1 91 7.0 10.737 28.544 10.737 28.544 mp2_ri_gpw_compute_en 1 5.0 0.061 0.076 26.620 28.285 fft_wrap_pw1pw2_100 730 12.4 1.288 1.449 26.662 27.151 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.900 24.908 24.918 get_2c_integrals 1 6.0 0.009 0.047 20.451 20.525 compute_2c_integrals 1 7.0 0.010 0.035 19.417 19.438 compute_2c_integrals_loop_lm 1 8.0 0.004 0.011 18.912 19.199 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.877 18.909 19.190 fft3d_s 1823 13.4 18.404 18.847 18.417 18.860 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.205 14.205 calculate_wavefunction 91 8.0 2.021 2.057 9.751 9.977 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.580 8.736 9.308 potential_pw2rs 186 10.0 0.034 0.035 8.661 9.232 local_gemm 172 8.0 8.179 8.746 8.179 8.746 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.003 0.019 8.285 8.537 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.527 7.952 8.455 calc_potential_gpw 182 9.5 0.002 0.002 7.946 8.219 collocate_single_gaussian 91 10.0 0.021 0.062 7.901 8.117 mp_sendrecv_dm3 2068 8.0 5.989 6.495 5.989 6.495 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.431 6.344 6.431 mp_sync 38 10.4 3.135 5.538 3.135 5.538 pw_gather_s 912 13.2 4.921 5.432 4.921 5.432 pw_scatter_s 910 13.7 3.958 4.254 3.958 4.254 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.883143, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.661824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 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 0.057 0.077 56.304 56.306 qs_mol_dyn_low 1 2.0 0.003 0.003 55.883 55.891 qs_forces 11 3.9 0.002 0.002 55.822 55.824 qs_energies 11 4.9 0.006 0.025 54.253 54.272 scf_env_do_scf 11 5.9 0.001 0.002 47.610 47.611 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.351 45.352 dbcsr_multiply_generic 2286 12.5 0.093 0.101 35.725 36.180 qs_scf_new_mos 108 7.5 0.000 0.001 34.570 34.843 qs_scf_loop_do_ot 108 8.5 0.001 0.002 34.570 34.843 ot_scf_mini 108 9.5 0.002 0.003 32.863 33.065 multiply_cannon 2286 13.5 0.185 0.194 27.176 28.702 velocity_verlet 10 3.0 0.001 0.001 27.975 27.976 multiply_cannon_loop 2286 14.5 1.504 1.578 26.359 27.845 ot_mini 108 10.5 0.001 0.001 20.995 21.247 qs_ot_get_derivative 108 11.5 0.001 0.001 17.880 18.070 mp_waitall_1 267858 16.1 10.164 16.407 10.164 16.407 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.641 14.127 multiply_cannon_multrec 54864 15.5 4.146 6.364 7.758 11.180 rebuild_ks_matrix 119 8.3 0.000 0.000 8.362 8.512 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 8.362 8.511 mp_sum_l 7207 12.9 6.162 7.853 6.162 7.853 multiply_cannon_sync_h2d 54864 15.5 5.795 7.848 5.795 7.848 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.415 7.551 qs_ot_get_p 119 10.4 0.001 0.001 6.917 7.239 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.229 6.678 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.873 5.981 init_scf_run 11 5.9 0.000 0.001 5.112 5.113 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.112 5.113 dbcsr_mm_accdrv_process 76910 16.1 1.192 1.828 3.534 4.926 sum_up_and_integrate 119 10.3 0.012 0.015 4.855 4.863 integrate_v_rspace 119 11.3 0.002 0.002 4.843 4.852 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.485 4.594 calculate_rho_elec 119 8.7 0.011 0.016 4.484 4.594 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.876 3.959 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 2.188 3.637 rs_pw_transfer 974 11.9 0.012 0.013 3.228 3.336 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.931 3.133 apply_single 119 13.6 0.000 0.000 2.930 3.133 calculate_dm_sparse 119 9.5 0.000 0.000 3.004 3.111 jit_kernel_multiply 13 15.8 2.279 3.008 2.279 3.008 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.002 3.003 calculate_first_density_matrix 1 7.0 0.001 0.004 2.828 2.835 ot_diis_step 108 11.5 0.006 0.006 2.814 2.815 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.727 2.806 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.720 2.721 cp_fm_redistribute_end 50 14.0 2.468 2.689 2.474 2.692 cp_fm_diag_elpa_base 50 14.0 0.215 2.597 0.216 2.609 density_rs2pw 119 9.7 0.004 0.004 2.525 2.604 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.585 2.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.255 2.322 init_scf_loop 11 6.9 0.002 0.014 2.237 2.237 mp_sum_d 4125 12.0 1.561 2.209 1.561 2.209 wfi_extrapolate 11 7.9 0.001 0.001 2.208 2.208 potential_pw2rs 119 12.3 0.004 0.004 2.177 2.197 acc_transpose_blocks 54864 15.5 0.232 0.246 1.764 2.188 make_m2s 4572 13.5 0.054 0.062 2.092 2.147 grid_integrate_task_list 119 12.3 2.016 2.120 2.016 2.120 pw_transfer 1439 11.6 0.052 0.056 1.979 2.069 make_images 4572 14.5 0.137 0.174 2.010 2.065 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.904 1.996 fft3d_ps 1201 14.6 0.358 0.461 1.681 1.764 mp_alltoall_d11v 2130 13.8 1.544 1.702 1.544 1.702 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.462 1.556 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.497 1.528 mp_waitany 12084 13.8 1.323 1.524 1.323 1.524 grid_collocate_task_list 119 9.7 1.289 1.365 1.289 1.365 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.864 1.308 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.306000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 487.628800E+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 1032117. 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 0.021 0.031 41.009 41.010 qs_mol_dyn_low 1 2.0 0.003 0.003 40.740 40.748 qs_forces 11 3.9 0.002 0.003 40.677 40.678 qs_energies 11 4.9 0.001 0.002 38.959 38.963 scf_env_do_scf 11 5.9 0.000 0.001 33.238 33.238 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.628 30.628 dbcsr_multiply_generic 2286 12.5 0.098 0.101 22.575 22.970 qs_scf_new_mos 108 7.5 0.001 0.001 21.054 21.300 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.053 21.300 ot_scf_mini 108 9.5 0.002 0.003 20.099 20.268 velocity_verlet 10 3.0 0.001 0.001 19.000 19.001 multiply_cannon 2286 13.5 0.210 0.222 17.142 18.807 multiply_cannon_loop 2286 14.5 0.896 0.973 15.911 17.557 ot_mini 108 10.5 0.001 0.001 12.335 12.570 mp_waitall_1 217478 16.2 6.881 12.397 6.881 12.397 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.459 10.010 qs_ot_get_derivative 108 11.5 0.001 0.001 9.805 9.977 multiply_cannon_multrec 27432 15.5 1.969 4.438 6.154 9.291 rebuild_ks_matrix 119 8.3 0.000 0.000 7.494 7.629 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.494 7.628 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.641 6.765 dbcsr_mm_accdrv_process 47894 16.0 3.345 5.348 4.116 6.178 qs_ot_get_p 119 10.4 0.001 0.001 4.768 5.002 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.789 4.640 sum_up_and_integrate 119 10.3 0.029 0.036 4.482 4.488 init_scf_run 11 5.9 0.000 0.001 4.466 4.467 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.466 4.466 integrate_v_rspace 119 11.3 0.002 0.002 4.453 4.465 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.089 4.148 apply_single 119 13.6 0.000 0.000 3.089 4.148 mp_sum_l 7207 12.9 2.217 4.125 2.217 4.125 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.922 3.956 calculate_rho_elec 119 8.7 0.021 0.024 3.921 3.955 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.166 3.187 rs_pw_transfer 974 11.9 0.010 0.011 2.952 3.037 make_m2s 4572 13.5 0.052 0.054 2.773 3.030 make_images 4572 14.5 0.200 0.235 2.686 2.941 multiply_cannon_sync_h2d 27432 15.5 2.179 2.876 2.179 2.876 calculate_first_density_matrix 1 7.0 0.000 0.004 2.874 2.876 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.704 2.706 jit_kernel_multiply 12 16.3 0.718 2.694 0.718 2.694 init_scf_loop 11 6.9 0.000 0.000 2.582 2.583 ot_diis_step 108 11.5 0.010 0.011 2.469 2.471 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.326 2.413 calculate_dm_sparse 119 9.5 0.000 0.001 2.271 2.345 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.329 2.331 density_rs2pw 119 9.7 0.004 0.004 2.259 2.320 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.290 2.291 cp_fm_redistribute_end 50 14.0 1.887 2.254 1.892 2.256 potential_pw2rs 119 12.3 0.006 0.006 2.189 2.211 cp_fm_diag_elpa_base 50 14.0 0.348 2.128 0.362 2.185 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.088 2.127 pw_transfer 1439 11.6 0.063 0.066 1.979 2.031 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.889 1.943 grid_integrate_task_list 119 12.3 1.841 1.938 1.841 1.938 make_images_data 4572 15.5 0.045 0.052 1.247 1.743 prepare_preconditioner 11 7.9 0.000 0.000 1.643 1.670 make_preconditioner 11 8.9 0.000 0.000 1.643 1.670 fft3d_ps 1201 14.6 0.499 0.550 1.598 1.646 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.085 1.613 acc_transpose_blocks 27432 15.5 0.108 0.113 1.275 1.603 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.542 1.600 wfi_extrapolate 11 7.9 0.001 0.001 1.542 1.542 mp_allgather_i34 2286 14.5 0.670 1.470 0.670 1.470 mp_alltoall_d11v 2130 13.8 1.342 1.452 1.342 1.452 fft_wrap_pw1pw2_140 487 13.2 0.074 0.082 1.406 1.451 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.431 1.438 grid_collocate_task_list 119 9.7 1.242 1.308 1.242 1.308 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.241 1.287 mp_sum_d 4125 12.0 0.633 1.084 0.633 1.084 make_images_sizes 4572 15.5 0.005 0.005 0.712 1.003 mp_alltoall_i44 4572 16.5 0.708 0.998 0.708 0.998 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.749 0.992 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.958 0.958 mp_alltoall_z22v 1201 16.6 0.833 0.946 0.833 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.925 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.797 0.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.010000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.818182, yerr=1.799908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 525.242368E+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 0.053 0.195 33.437 33.439 qs_mol_dyn_low 1 2.0 0.003 0.003 32.930 32.937 qs_forces 11 3.9 0.002 0.002 32.870 32.872 qs_energies 11 4.9 0.003 0.012 31.253 31.255 scf_env_do_scf 11 5.9 0.001 0.001 26.298 26.298 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.487 23.488 dbcsr_multiply_generic 2286 12.5 0.092 0.096 16.764 16.850 velocity_verlet 10 3.0 0.001 0.001 15.781 15.782 qs_scf_new_mos 108 7.5 0.001 0.001 15.255 15.293 qs_scf_loop_do_ot 108 8.5 0.010 0.052 15.255 15.292 ot_scf_mini 108 9.5 0.003 0.005 14.509 14.525 multiply_cannon 2286 13.5 0.196 0.208 13.215 13.983 multiply_cannon_loop 2286 14.5 0.632 0.665 12.386 13.177 ot_mini 108 10.5 0.001 0.001 8.947 8.965 qs_ot_get_derivative 108 11.5 0.001 0.001 7.437 7.456 multiply_cannon_multrec 18288 15.5 1.927 2.833 6.850 7.133 rebuild_ks_matrix 119 8.3 0.000 0.000 6.533 6.548 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.024 6.532 6.548 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.776 5.792 dbcsr_mm_accdrv_process 38222 16.0 3.935 5.385 4.840 5.663 mp_waitall_1 169478 16.3 3.258 4.539 3.258 4.539 sum_up_and_integrate 119 10.3 0.031 0.035 4.136 4.146 integrate_v_rspace 119 11.3 0.002 0.003 4.104 4.115 init_scf_run 11 5.9 0.000 0.001 3.735 3.736 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.735 3.736 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.832 3.449 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.409 3.423 calculate_rho_elec 119 8.7 0.030 0.031 3.408 3.422 qs_ot_get_p 119 10.4 0.001 0.001 3.375 3.398 init_scf_loop 11 6.9 0.008 0.047 2.789 2.792 rs_pw_transfer 974 11.9 0.009 0.010 2.543 2.616 calculate_first_density_matrix 1 7.0 0.001 0.002 2.462 2.464 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.356 2.402 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.050 2.365 apply_single 119 13.6 0.000 0.000 2.050 2.365 jit_kernel_multiply 12 16.1 0.854 2.284 0.854 2.284 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.187 2.192 make_m2s 4572 13.5 0.045 0.046 1.963 2.127 density_rs2pw 119 9.7 0.004 0.004 1.998 2.096 make_images 4572 14.5 0.191 0.202 1.879 2.042 potential_pw2rs 119 12.3 0.007 0.008 1.913 1.926 prepare_preconditioner 11 7.9 0.000 0.000 1.912 1.918 make_preconditioner 11 8.9 0.002 0.016 1.912 1.918 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.916 1.916 grid_integrate_task_list 119 12.3 1.796 1.908 1.796 1.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.877 1.879 calculate_dm_sparse 119 9.5 0.000 0.000 1.856 1.871 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.754 1.845 pw_transfer 1439 11.6 0.064 0.066 1.818 1.833 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.727 1.741 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.707 1.712 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.655 1.656 cp_fm_redistribute_end 50 14.0 1.233 1.632 1.235 1.632 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.605 1.618 cp_fm_diag_elpa_base 50 14.0 0.381 1.546 0.396 1.593 multiply_cannon_sync_h2d 18288 15.5 1.402 1.593 1.402 1.593 mp_sum_l 7207 12.9 1.259 1.590 1.259 1.590 ot_diis_step 108 11.5 0.011 0.011 1.490 1.491 fft3d_ps 1201 14.6 0.505 0.521 1.414 1.430 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.307 1.322 grid_collocate_task_list 119 9.7 1.205 1.282 1.205 1.282 acc_transpose_blocks 18288 15.5 0.074 0.076 1.242 1.270 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.214 1.221 wfi_extrapolate 11 7.9 0.001 0.001 1.212 1.212 make_images_data 4572 15.5 0.045 0.050 0.859 1.047 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.968 0.970 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.908 0.932 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.732 0.915 mp_alltoall_d11v 2130 13.8 0.757 0.872 0.757 0.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.842 0.843 acc_transpose_blocks_kernels 18288 16.5 0.208 0.220 0.808 0.822 mp_alltoall_z22v 1201 16.6 0.740 0.798 0.740 0.798 cp_fm_cholesky_invert 11 10.9 0.792 0.797 0.792 0.797 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.507 0.764 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.647 0.712 mp_waitany 9880 13.7 0.571 0.680 0.571 0.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.439000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.454545, yerr=2.775334 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 546.369536E+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 0.040 0.054 37.594 37.595 qs_mol_dyn_low 1 2.0 0.003 0.003 37.350 37.358 qs_forces 11 3.9 0.002 0.003 37.270 37.271 qs_energies 11 4.9 0.005 0.022 35.482 35.492 scf_env_do_scf 11 5.9 0.001 0.002 29.935 29.937 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.035 26.036 dbcsr_multiply_generic 2286 12.5 0.098 0.100 19.041 19.164 velocity_verlet 10 3.0 0.001 0.001 18.682 18.684 qs_scf_new_mos 108 7.5 0.001 0.001 17.437 17.490 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.437 17.489 ot_scf_mini 108 9.5 0.003 0.003 16.473 16.523 multiply_cannon 2286 13.5 0.234 0.267 14.896 15.390 multiply_cannon_loop 2286 14.5 0.936 0.977 13.828 14.169 ot_mini 108 10.5 0.001 0.001 9.792 9.860 multiply_cannon_multrec 27432 15.5 2.368 3.036 8.823 9.185 qs_ot_get_derivative 108 11.5 0.001 0.001 7.919 7.973 dbcsr_mm_accdrv_process 47916 15.9 5.286 7.245 6.363 7.533 rebuild_ks_matrix 119 8.3 0.000 0.000 6.797 6.846 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.797 6.846 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.032 6.078 init_scf_run 11 5.9 0.000 0.001 4.043 4.043 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.043 4.043 qs_ot_get_p 119 10.4 0.001 0.001 3.956 4.034 sum_up_and_integrate 119 10.3 0.035 0.038 4.003 4.011 integrate_v_rspace 119 11.3 0.002 0.003 3.967 3.976 init_scf_loop 11 6.9 0.001 0.005 3.874 3.875 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.544 3.576 calculate_rho_elec 119 8.7 0.040 0.046 3.543 3.576 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.826 3.309 mp_waitall_1 145218 16.4 2.638 3.240 2.638 3.240 prepare_preconditioner 11 7.9 0.000 0.000 2.964 2.971 make_preconditioner 11 8.9 0.000 0.002 2.964 2.971 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.556 2.892 make_m2s 4572 13.5 0.054 0.056 2.506 2.666 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.191 2.643 apply_single 119 13.6 0.000 0.000 2.191 2.642 make_images 4572 14.5 0.272 0.333 2.399 2.557 calculate_first_density_matrix 1 7.0 0.001 0.003 2.544 2.545 rs_pw_transfer 974 11.9 0.009 0.009 2.371 2.500 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.482 2.492 calculate_dm_sparse 119 9.5 0.000 0.000 2.228 2.286 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.129 2.129 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.099 2.126 density_rs2pw 119 9.7 0.004 0.004 1.999 2.114 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.918 1.920 pw_transfer 1439 11.6 0.063 0.067 1.881 1.913 grid_integrate_task_list 119 12.3 1.826 1.893 1.826 1.893 ot_diis_step 108 11.5 0.012 0.012 1.832 1.832 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.791 1.826 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.795 1.798 potential_pw2rs 119 12.3 0.008 0.009 1.767 1.773 cp_fm_redistribute_end 50 14.0 1.186 1.759 1.189 1.761 mp_sum_l 7207 12.9 1.227 1.756 1.227 1.756 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.703 1.721 cp_fm_diag_elpa_base 50 14.0 0.546 1.671 0.567 1.711 jit_kernel_multiply 8 15.8 1.016 1.685 1.016 1.685 acc_transpose_blocks 27432 15.5 0.109 0.111 1.526 1.552 fft3d_ps 1201 14.6 0.533 0.589 1.472 1.495 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.413 1.450 wfi_extrapolate 11 7.9 0.001 0.001 1.428 1.428 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.406 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.820 1.363 grid_collocate_task_list 119 9.7 1.221 1.307 1.221 1.307 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.176 1.191 make_images_data 4572 15.5 0.045 0.048 1.012 1.186 dbcsr_complete_redistribute 329 12.2 0.118 0.145 0.904 1.183 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.151 1.153 cp_fm_upper_to_full 72 13.5 0.822 1.145 0.822 1.145 hybrid_alltoall_any 4725 16.4 0.061 0.149 0.867 1.087 multiply_cannon_sync_h2d 27432 15.5 0.994 1.054 0.994 1.054 mp_alltoall_d11v 2130 13.8 0.903 1.015 0.903 1.015 cp_fm_cholesky_invert 11 10.9 0.961 0.966 0.961 0.966 acc_transpose_blocks_kernels 27432 16.5 0.266 0.276 0.911 0.925 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.640 0.918 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.797 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.865 mp_alltoall_z22v 1201 16.6 0.821 0.851 0.821 0.851 mp_alltoall_i22 627 13.8 0.460 0.765 0.460 0.765 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.595000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.727273, yerr=3.492613 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 596.066304E+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 0.035 0.049 30.202 30.203 qs_mol_dyn_low 1 2.0 0.003 0.003 29.942 29.951 qs_forces 11 3.9 0.002 0.002 29.881 29.881 qs_energies 11 4.9 0.002 0.002 28.139 28.142 scf_env_do_scf 11 5.9 0.001 0.011 23.174 23.174 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.234 20.235 velocity_verlet 10 3.0 0.001 0.001 15.420 15.423 dbcsr_multiply_generic 2286 12.5 0.090 0.092 12.740 12.805 qs_scf_new_mos 108 7.5 0.001 0.001 12.194 12.229 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.194 12.228 ot_scf_mini 108 9.5 0.002 0.002 11.485 11.516 multiply_cannon 2286 13.5 0.231 0.236 9.834 10.414 multiply_cannon_loop 2286 14.5 0.328 0.339 8.837 9.011 ot_mini 108 10.5 0.001 0.001 6.229 6.263 multiply_cannon_multrec 9144 15.5 1.716 1.951 5.867 6.144 rebuild_ks_matrix 119 8.3 0.000 0.000 6.114 6.141 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.114 6.140 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.455 5.479 qs_ot_get_derivative 108 11.5 0.001 0.001 4.842 4.873 dbcsr_mm_accdrv_process 12550 15.8 2.989 3.660 4.051 4.115 sum_up_and_integrate 119 10.3 0.038 0.042 3.802 3.807 integrate_v_rspace 119 11.3 0.002 0.003 3.764 3.770 init_scf_run 11 5.9 0.000 0.001 3.501 3.502 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.501 3.501 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.457 3.464 calculate_rho_elec 119 8.7 0.059 0.061 3.456 3.464 qs_ot_get_p 119 10.4 0.001 0.001 3.379 3.418 init_scf_loop 11 6.9 0.000 0.000 2.906 2.907 mp_waitall_1 121218 16.5 2.173 2.631 2.173 2.631 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.287 2.290 calculate_first_density_matrix 1 7.0 0.000 0.000 2.265 2.266 make_m2s 4572 13.5 0.034 0.036 1.988 2.185 prepare_preconditioner 11 7.9 0.000 0.000 2.132 2.136 make_preconditioner 11 8.9 0.000 0.000 2.132 2.136 make_images 4572 14.5 0.267 0.304 1.900 2.096 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.070 2.071 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.015 2.043 rs_pw_transfer 974 11.9 0.008 0.008 1.934 2.009 grid_integrate_task_list 119 12.3 1.856 1.928 1.856 1.928 density_rs2pw 119 9.7 0.003 0.003 1.839 1.910 pw_transfer 1439 11.6 0.064 0.065 1.894 1.905 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.803 1.814 calculate_dm_sparse 119 9.5 0.000 0.001 1.741 1.761 jit_kernel_multiply 10 15.8 1.024 1.731 1.024 1.731 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.713 1.715 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.668 1.686 cp_fm_redistribute_end 50 14.0 0.857 1.679 0.859 1.680 cp_fm_diag_elpa_base 50 14.0 0.773 1.582 0.818 1.644 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.566 1.567 potential_pw2rs 119 12.3 0.010 0.010 1.537 1.544 fft3d_ps 1201 14.6 0.542 0.554 1.470 1.480 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.416 1.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.364 1.373 ot_diis_step 108 11.5 0.012 0.013 1.371 1.371 grid_collocate_task_list 119 9.7 1.276 1.350 1.276 1.350 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.262 1.279 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.237 1.270 apply_single 119 13.6 0.000 0.000 1.236 1.270 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.216 1.217 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.915 1.196 make_images_data 4572 15.5 0.039 0.042 0.941 1.191 wfi_extrapolate 11 7.9 0.001 0.001 1.189 1.189 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.140 1.146 cp_fm_cholesky_invert 11 10.9 1.141 1.144 1.141 1.144 mp_alltoall_d11v 2130 13.8 0.957 1.051 0.957 1.051 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.922 acc_transpose_blocks 9144 15.5 0.038 0.039 0.890 0.897 mp_alltoall_z22v 1201 16.6 0.801 0.835 0.801 0.835 mp_allgather_i34 2286 14.5 0.330 0.818 0.330 0.818 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.365 0.812 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.793 0.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.786 0.788 multiply_cannon_sync_h2d 9144 15.5 0.723 0.782 0.723 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.673 0.723 dbcsr_complete_redistribute 329 12.2 0.163 0.171 0.642 0.673 acc_transpose_blocks_kernels 9144 16.5 0.115 0.119 0.653 0.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.203000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=564.181818, yerr=5.424158 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 781.971456E+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 0.053 0.059 42.895 42.896 qs_mol_dyn_low 1 2.0 0.003 0.003 42.574 42.582 qs_forces 11 3.9 0.002 0.002 42.513 42.515 qs_energies 11 4.9 0.001 0.001 40.533 40.538 scf_env_do_scf 11 5.9 0.001 0.001 34.752 34.753 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.593 26.595 velocity_verlet 10 3.0 0.001 0.001 23.881 23.887 dbcsr_multiply_generic 2286 12.5 0.098 0.099 18.335 18.415 qs_scf_new_mos 108 7.5 0.001 0.001 17.294 17.395 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.294 17.394 ot_scf_mini 108 9.5 0.002 0.002 16.172 16.276 multiply_cannon 2286 13.5 0.306 0.318 13.948 14.786 multiply_cannon_loop 2286 14.5 0.342 0.348 12.595 13.471 ot_mini 108 10.5 0.001 0.001 9.641 9.765 multiply_cannon_multrec 9144 15.5 3.415 4.877 8.595 8.640 init_scf_loop 11 6.9 0.000 0.000 8.127 8.131 qs_ot_get_derivative 108 11.5 0.001 0.001 7.554 7.661 rebuild_ks_matrix 119 8.3 0.000 0.000 7.121 7.271 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.121 7.271 prepare_preconditioner 11 7.9 0.000 0.000 7.159 7.174 make_preconditioner 11 8.9 0.000 0.000 7.159 7.174 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.727 7.037 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.431 6.567 dbcsr_mm_accdrv_process 12550 15.8 4.262 5.672 5.056 6.396 cp_fm_upper_to_full 72 14.2 3.168 4.529 3.168 4.529 mp_waitall_1 97218 16.6 3.074 3.978 3.074 3.978 sum_up_and_integrate 119 10.3 0.064 0.065 3.894 3.899 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.872 3.881 calculate_rho_elec 119 8.7 0.118 0.121 3.872 3.880 qs_ot_get_p 119 10.4 0.001 0.001 3.693 3.842 integrate_v_rspace 119 11.3 0.003 0.003 3.830 3.835 init_scf_run 11 5.9 0.000 0.001 3.743 3.743 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.743 3.743 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.605 3.082 dbcsr_complete_redistribute 329 12.2 0.289 0.292 2.071 2.877 make_m2s 4572 13.5 0.038 0.038 2.552 2.776 make_images 4572 14.5 0.351 0.379 2.433 2.657 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.289 2.536 apply_single 119 13.6 0.000 0.000 2.289 2.536 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.703 2.490 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.561 2.406 mp_alltoall_i22 627 13.8 1.477 2.294 1.477 2.294 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.279 2.281 calculate_dm_sparse 119 9.5 0.000 0.000 2.221 2.240 calculate_first_density_matrix 1 7.0 0.000 0.000 2.231 2.231 pw_transfer 1439 11.6 0.066 0.068 2.208 2.213 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.428 2.210 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.112 2.116 grid_integrate_task_list 119 12.3 2.061 2.084 2.061 2.084 ot_diis_step 108 11.5 0.014 0.014 2.056 2.056 mp_sum_l 7207 12.9 1.303 2.011 1.303 2.011 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.982 1.983 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.968 1.969 density_rs2pw 119 9.7 0.003 0.003 1.933 1.944 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.867 1.921 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.754 1.755 fft3d_ps 1201 14.6 0.568 0.579 1.745 1.750 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.723 1.730 cp_fm_cholesky_invert 11 10.9 1.695 1.698 1.695 1.698 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.644 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.632 1.632 cp_fm_diag_elpa_base 50 14.0 1.483 1.536 1.629 1.629 rs_pw_transfer 974 11.9 0.009 0.009 1.543 1.563 make_images_data 4572 15.5 0.043 0.046 1.220 1.531 hybrid_alltoall_any 4725 16.4 0.086 0.148 1.247 1.519 grid_collocate_task_list 119 9.7 1.488 1.500 1.488 1.500 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 potential_pw2rs 119 12.3 0.014 0.014 1.356 1.362 mp_alltoall_d11v 2130 13.8 1.278 1.305 1.278 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.251 1.260 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.209 1.234 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.092 1.114 mp_alltoall_z22v 1201 16.6 1.042 1.064 1.042 1.064 multiply_cannon_sync_h2d 9144 15.5 1.048 1.052 1.048 1.052 jit_kernel_multiply 6 15.4 0.767 1.020 0.767 1.020 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.014 qs_create_task_list 11 7.9 0.000 0.001 0.945 0.958 generate_qs_task_list 11 8.9 0.372 0.392 0.945 0.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.918 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.851 0.897 acc_transpose_blocks 9144 15.5 0.037 0.038 0.892 0.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.896000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=733.454545, yerr=14.278806 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 501.444608E+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 1496160. 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 0.025 0.046 84.511 84.513 qs_mol_dyn_low 1 2.0 0.004 0.006 84.111 84.121 qs_forces 11 3.9 0.002 0.003 84.037 84.039 qs_energies 11 4.9 0.002 0.008 81.192 81.207 scf_env_do_scf 11 5.9 0.001 0.001 72.202 72.204 scf_env_do_scf_inner_loop 99 6.5 0.005 0.015 66.412 66.413 dbcsr_multiply_generic 2055 12.4 0.103 0.106 52.634 52.916 qs_scf_new_mos 99 7.5 0.000 0.001 49.101 49.248 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.101 49.247 ot_scf_mini 99 9.5 0.002 0.003 46.661 46.760 multiply_cannon 2055 13.4 0.178 0.188 42.734 43.428 velocity_verlet 10 3.0 0.001 0.001 42.756 42.756 multiply_cannon_loop 2055 14.4 1.553 1.596 41.705 42.370 ot_mini 99 10.5 0.001 0.001 28.192 28.298 qs_ot_get_derivative 99 11.5 0.001 0.001 21.338 21.434 multiply_cannon_multrec 49320 15.4 12.093 12.760 17.153 17.899 rebuild_ks_matrix 110 8.3 0.000 0.001 14.654 14.746 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.653 14.746 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.909 12.994 mp_waitall_1 241148 16.1 12.225 12.936 12.225 12.936 multiply_cannon_sync_h2d 49320 15.4 9.918 10.533 9.918 10.533 qs_ot_get_p 110 10.4 0.001 0.001 9.907 10.004 multiply_cannon_metrocomm3 49320 15.4 0.080 0.085 7.002 7.872 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.221 7.766 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.682 apply_single 110 13.6 0.000 0.001 7.171 7.681 sum_up_and_integrate 110 10.3 0.037 0.044 7.198 7.213 integrate_v_rspace 110 11.3 0.003 0.003 7.161 7.186 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.016 7.089 init_scf_run 11 5.9 0.000 0.001 6.855 6.856 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.855 6.856 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.425 6.602 calculate_rho_elec 110 8.6 0.020 0.024 6.424 6.601 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.562 6.594 ot_diis_step 99 11.5 0.005 0.005 6.559 6.560 mp_sum_l 6514 12.8 5.350 6.103 5.350 6.103 init_scf_loop 11 6.9 0.001 0.005 5.759 5.760 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.523 5.524 dbcsr_mm_accdrv_process 87628 16.1 2.063 2.162 4.939 5.214 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.987 4.989 cp_fm_redistribute_end 48 14.0 4.347 4.954 4.351 4.956 cp_fm_diag_elpa_base 48 14.0 0.598 4.816 0.602 4.848 rs_pw_transfer 902 11.9 0.012 0.013 4.045 4.314 make_m2s 4110 13.4 0.060 0.064 4.104 4.196 make_images 4110 14.4 0.177 0.191 4.009 4.106 wfi_extrapolate 11 7.9 0.001 0.001 4.090 4.090 calculate_dm_sparse 110 9.5 0.000 0.001 3.830 3.939 density_rs2pw 110 9.6 0.004 0.004 3.463 3.708 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 2.820 3.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.579 3.584 prepare_preconditioner 11 7.9 0.000 0.000 3.551 3.569 make_preconditioner 11 8.9 0.002 0.009 3.551 3.569 grid_integrate_task_list 110 12.3 3.248 3.411 3.248 3.411 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.322 3.383 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.333 3.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.219 3.247 pw_transfer 1331 11.6 0.055 0.061 3.019 3.083 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.931 2.994 potential_pw2rs 110 12.3 0.006 0.007 2.785 2.823 mp_alltoall_d11v 2046 13.8 2.188 2.757 2.188 2.757 calculate_first_density_matrix 1 7.0 0.001 0.003 2.671 2.680 jit_kernel_multiply 13 15.9 2.593 2.630 2.593 2.630 fft3d_ps 1111 14.6 0.747 0.831 2.467 2.518 fft_wrap_pw1pw2_140 451 13.1 0.169 0.189 2.417 2.478 acc_transpose_blocks 49320 15.4 0.221 0.231 2.178 2.244 grid_collocate_task_list 110 9.6 2.087 2.192 2.087 2.192 mp_waitany 14300 13.8 1.875 2.087 1.875 2.087 mp_sum_d 3879 11.9 1.556 2.079 1.556 2.079 make_images_data 4110 15.4 0.044 0.048 1.849 1.964 cp_fm_cholesky_invert 11 10.9 1.953 1.957 1.953 1.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.887 1.903 hybrid_alltoall_any 4261 16.3 0.083 0.485 1.619 1.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.513000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.545455, yerr=2.230517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 587.157504E+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 2469272. 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 0.026 0.033 69.268 69.269 qs_mol_dyn_low 1 2.0 0.003 0.003 68.992 69.001 qs_forces 11 3.9 0.002 0.002 68.923 68.924 qs_energies 11 4.9 0.001 0.002 65.611 65.615 scf_env_do_scf 11 5.9 0.000 0.001 57.005 57.008 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.144 49.144 dbcsr_multiply_generic 2055 12.4 0.110 0.116 37.878 38.017 velocity_verlet 10 3.0 0.001 0.001 36.453 36.454 qs_scf_new_mos 99 7.5 0.001 0.001 33.040 33.180 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.039 33.180 multiply_cannon 2055 13.4 0.223 0.246 31.094 32.186 ot_scf_mini 99 9.5 0.003 0.003 31.375 31.498 multiply_cannon_loop 2055 14.4 0.919 0.940 29.726 30.570 ot_mini 99 10.5 0.001 0.001 18.349 18.476 multiply_cannon_multrec 24660 15.4 7.675 9.335 13.935 15.418 rebuild_ks_matrix 110 8.3 0.000 0.001 13.736 13.809 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.736 13.809 qs_ot_get_derivative 99 11.5 0.001 0.001 12.542 12.670 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.125 12.191 mp_waitall_1 186928 16.3 7.945 10.097 7.945 10.097 multiply_cannon_sync_h2d 24660 15.4 7.036 8.130 7.036 8.130 init_scf_loop 11 6.9 0.000 0.000 7.820 7.821 multiply_cannon_metrocomm3 24660 15.4 0.069 0.070 5.076 7.781 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.521 7.176 apply_single 110 13.6 0.000 0.001 6.520 7.176 sum_up_and_integrate 110 10.3 0.055 0.058 6.668 6.678 integrate_v_rspace 110 11.3 0.002 0.002 6.613 6.625 qs_ot_get_p 110 10.4 0.001 0.001 6.353 6.510 dbcsr_mm_accdrv_process 52282 16.1 4.625 5.534 6.099 6.387 init_scf_run 11 5.9 0.000 0.001 6.222 6.222 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.222 6.222 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.926 5.935 calculate_rho_elec 110 8.6 0.039 0.047 5.925 5.935 prepare_preconditioner 11 7.9 0.000 0.000 5.802 5.819 make_preconditioner 11 8.9 0.000 0.000 5.802 5.819 ot_diis_step 99 11.5 0.010 0.010 5.757 5.758 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.371 5.530 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.688 5.430 make_m2s 4110 13.4 0.057 0.060 4.328 4.804 make_images 4110 14.4 0.399 0.445 4.224 4.698 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.375 4.394 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.928 3.929 pw_transfer 1331 11.6 0.066 0.070 3.414 3.556 wfi_extrapolate 11 7.9 0.001 0.001 3.541 3.541 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.308 3.449 density_rs2pw 110 9.6 0.004 0.004 3.235 3.430 rs_pw_transfer 902 11.9 0.012 0.014 3.154 3.364 grid_integrate_task_list 110 12.3 3.157 3.352 3.157 3.352 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.338 3.340 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.330 3.332 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.257 3.328 cp_fm_redistribute_end 48 14.0 2.488 3.296 2.491 3.299 cp_fm_diag_elpa_base 48 14.0 0.771 3.142 0.803 3.231 calculate_dm_sparse 110 9.5 0.001 0.001 2.999 3.029 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.982 3.025 make_images_data 4110 15.4 0.047 0.051 2.381 2.863 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.692 2.835 hybrid_alltoall_any 4261 16.3 0.103 0.447 2.082 2.821 fft3d_ps 1111 14.6 1.075 1.273 2.666 2.799 cp_fm_cholesky_invert 11 10.9 2.726 2.733 2.726 2.733 calculate_first_density_matrix 1 7.0 0.000 0.000 2.589 2.591 potential_pw2rs 110 12.3 0.008 0.009 2.529 2.542 mp_sum_l 6514 12.8 1.854 2.469 1.854 2.469 jit_kernel_multiply 12 16.4 1.125 2.388 1.125 2.388 grid_collocate_task_list 110 9.6 2.117 2.284 2.117 2.284 mp_alltoall_d11v 2046 13.8 1.802 2.011 1.802 2.011 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.974 1.994 qs_energies_init_hamiltonians 11 5.9 0.000 0.006 1.852 1.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.776 1.789 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.776 1.770 mp_allgather_i34 2055 14.4 0.693 1.669 0.693 1.669 mp_irecv_dv 57340 16.2 0.653 1.655 0.653 1.655 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.573 1.586 acc_transpose_blocks 24660 15.4 0.110 0.113 1.547 1.567 dbcsr_complete_redistribute 325 12.2 0.242 0.306 1.223 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.478 mp_waitany 10164 13.8 1.216 1.455 1.216 1.455 cp_fm_cholesky_decompose 22 10.9 1.403 1.410 1.403 1.410 mp_alltoall_z22v 1111 16.6 1.317 1.392 1.317 1.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.269000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.363636, yerr=7.087412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 656.424960E+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 0.031 0.043 62.025 62.027 qs_mol_dyn_low 1 2.0 0.003 0.007 61.644 61.654 qs_forces 11 3.9 0.004 0.015 61.572 61.575 qs_energies 11 4.9 0.004 0.016 58.276 58.283 scf_env_do_scf 11 5.9 0.001 0.001 50.212 50.212 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.031 41.033 velocity_verlet 10 3.0 0.001 0.001 33.476 33.478 dbcsr_multiply_generic 2055 12.4 0.106 0.108 29.032 29.336 qs_scf_new_mos 99 7.5 0.001 0.001 26.193 26.293 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.192 26.292 ot_scf_mini 99 9.5 0.003 0.004 24.943 25.062 multiply_cannon 2055 13.4 0.212 0.225 22.294 23.624 multiply_cannon_loop 2055 14.4 0.614 0.632 21.013 22.131 ot_mini 99 10.5 0.001 0.001 14.039 14.157 rebuild_ks_matrix 110 8.3 0.000 0.000 12.461 12.604 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 12.461 12.603 mp_waitall_1 146766 16.3 7.756 11.217 7.756 11.217 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.924 11.055 multiply_cannon_multrec 16440 15.4 3.824 4.685 9.556 10.296 qs_ot_get_derivative 99 11.5 0.001 0.001 9.478 9.595 init_scf_loop 11 6.9 0.002 0.013 9.143 9.150 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.576 7.809 prepare_preconditioner 11 7.9 0.000 0.000 7.328 7.346 make_preconditioner 11 8.9 0.000 0.001 7.328 7.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.650 7.022 sum_up_and_integrate 110 10.3 0.061 0.063 6.665 6.693 integrate_v_rspace 110 11.3 0.002 0.003 6.604 6.632 qs_ot_get_p 110 10.4 0.001 0.002 5.995 6.147 qs_rho_update_rho_low 110 7.6 0.001 0.004 5.732 5.741 calculate_rho_elec 110 8.6 0.058 0.059 5.732 5.741 dbcsr_mm_accdrv_process 34862 16.1 4.385 5.283 5.589 5.701 init_scf_run 11 5.9 0.000 0.001 5.549 5.550 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.549 5.550 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.053 5.495 apply_single 110 13.6 0.000 0.000 5.052 5.495 make_m2s 4110 13.4 0.050 0.052 4.419 4.861 make_images 4110 14.4 0.394 0.510 4.307 4.747 ot_diis_step 99 11.5 0.011 0.011 4.523 4.523 multiply_cannon_sync_h2d 16440 15.4 3.634 4.208 3.634 4.208 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.197 4.202 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.224 3.855 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.829 3.830 pw_transfer 1331 11.6 0.065 0.072 3.409 3.434 grid_integrate_task_list 110 12.3 3.201 3.405 3.201 3.405 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.302 3.328 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.254 3.257 cp_fm_redistribute_end 48 14.0 2.027 3.220 2.030 3.220 density_rs2pw 110 9.6 0.004 0.004 2.991 3.176 rs_pw_transfer 902 11.9 0.011 0.012 2.931 3.173 cp_fm_diag_elpa_base 48 14.0 1.123 3.052 1.184 3.165 make_images_data 4110 15.4 0.044 0.048 2.560 3.089 wfi_extrapolate 11 7.9 0.001 0.001 3.025 3.025 cp_fm_cholesky_invert 11 10.9 3.014 3.023 3.014 3.023 hybrid_alltoall_any 4261 16.3 0.106 0.380 2.306 2.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.809 2.811 fft_wrap_pw1pw2_140 451 13.1 0.209 0.213 2.778 2.805 fft3d_ps 1111 14.6 1.065 1.074 2.617 2.643 calculate_dm_sparse 110 9.5 0.001 0.001 2.531 2.554 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.481 2.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.456 2.519 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.898 2.504 potential_pw2rs 110 12.3 0.011 0.011 2.474 2.504 calculate_first_density_matrix 1 7.0 0.001 0.003 2.437 2.440 mp_irecv_dv 48980 15.7 0.829 2.382 0.829 2.382 mp_sum_l 6514 12.8 1.716 2.356 1.716 2.356 grid_collocate_task_list 110 9.6 2.182 2.355 2.182 2.355 mp_alltoall_d11v 2046 13.8 1.877 2.219 1.877 2.219 jit_kernel_multiply 11 16.4 0.817 2.089 0.817 2.089 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.996 2.001 dbcsr_complete_redistribute 325 12.2 0.325 0.359 1.500 1.954 cp_fm_upper_to_full 70 13.6 1.409 1.881 1.409 1.881 cp_fm_cholesky_decompose 22 10.9 1.754 1.777 1.754 1.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.739 1.754 mp_allgather_i34 2055 14.4 0.604 1.736 0.604 1.736 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.497 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.478 1.491 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.004 1.462 mp_waitany 17072 13.8 1.213 1.437 1.213 1.437 mp_alltoall_z22v 1111 16.6 1.337 1.377 1.337 1.377 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.348 1.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.027000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.090909, yerr=8.162604 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 729.452544E+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 0.048 0.151 67.147 67.150 qs_mol_dyn_low 1 2.0 0.003 0.004 66.706 66.716 qs_forces 11 3.9 0.002 0.003 66.633 66.633 qs_energies 11 4.9 0.027 0.068 63.213 63.215 scf_env_do_scf 11 5.9 0.001 0.001 54.566 54.569 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 42.301 42.303 velocity_verlet 10 3.0 0.001 0.001 37.803 37.805 dbcsr_multiply_generic 2055 12.4 0.112 0.115 29.772 29.962 qs_scf_new_mos 99 7.5 0.001 0.001 27.502 27.612 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.501 27.611 ot_scf_mini 99 9.5 0.003 0.004 25.851 25.960 multiply_cannon 2055 13.4 0.243 0.264 22.371 23.448 multiply_cannon_loop 2055 14.4 0.883 0.902 20.934 21.509 ot_mini 99 10.5 0.001 0.001 14.416 14.547 multiply_cannon_multrec 24660 15.4 4.227 6.840 12.630 13.791 rebuild_ks_matrix 110 8.3 0.000 0.001 12.203 12.307 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.203 12.306 init_scf_loop 11 6.9 0.001 0.003 12.215 12.222 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.794 10.888 prepare_preconditioner 11 7.9 0.000 0.000 10.419 10.434 make_preconditioner 11 8.9 0.000 0.001 10.419 10.434 qs_ot_get_derivative 99 11.5 0.001 0.001 10.221 10.332 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.664 10.096 dbcsr_mm_accdrv_process 52304 16.0 6.633 8.171 8.256 9.166 mp_waitall_1 126806 16.4 4.779 6.575 4.779 6.575 qs_ot_get_p 110 10.4 0.001 0.002 6.402 6.566 sum_up_and_integrate 110 10.3 0.068 0.071 6.549 6.561 integrate_v_rspace 110 11.3 0.003 0.003 6.481 6.497 make_m2s 4110 13.4 0.059 0.060 5.695 5.939 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.838 5.848 calculate_rho_elec 110 8.6 0.077 0.081 5.838 5.847 make_images 4110 14.4 0.577 0.695 5.555 5.796 init_scf_run 11 5.9 0.000 0.001 5.762 5.769 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.761 5.769 cp_fm_upper_to_full 70 13.8 3.260 4.635 3.260 4.635 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.420 4.434 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.155 apply_single 110 13.6 0.000 0.000 4.045 4.154 ot_diis_step 99 11.5 0.011 0.011 4.152 4.153 dbcsr_complete_redistribute 325 12.2 0.432 0.565 2.864 4.003 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.933 3.933 pw_transfer 1331 11.6 0.065 0.073 3.456 3.485 grid_integrate_task_list 110 12.3 3.284 3.451 3.284 3.451 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.387 3.436 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.271 3.393 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.350 3.383 multiply_cannon_sync_h2d 24660 15.4 3.181 3.325 3.181 3.325 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.295 3.297 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.373 3.287 cp_fm_redistribute_end 48 14.0 1.645 3.252 1.647 3.253 density_rs2pw 110 9.6 0.004 0.004 2.963 3.223 cp_fm_diag_elpa_base 48 14.0 1.522 3.104 1.602 3.215 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.141 3.202 make_images_data 4110 15.4 0.047 0.050 2.874 3.152 wfi_extrapolate 11 7.9 0.001 0.001 3.103 3.103 hybrid_alltoall_any 4261 16.3 0.121 0.460 2.447 3.086 cp_fm_cholesky_invert 11 10.9 3.075 3.084 3.075 3.084 calculate_dm_sparse 110 9.5 0.001 0.001 2.945 2.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.951 2.952 rs_pw_transfer 902 11.9 0.010 0.011 2.657 2.920 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.746 2.851 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.770 2.804 mp_alltoall_i22 605 13.7 1.643 2.801 1.643 2.801 fft3d_ps 1111 14.6 1.069 1.100 2.654 2.672 calculate_first_density_matrix 1 7.0 0.001 0.003 2.559 2.567 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.469 grid_collocate_task_list 110 9.6 2.220 2.369 2.220 2.369 potential_pw2rs 110 12.3 0.012 0.013 2.301 2.316 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 2.294 2.308 mp_alltoall_d11v 2046 13.8 1.983 2.184 1.983 2.184 jit_kernel_multiply 12 15.8 1.290 2.182 1.290 2.182 cp_fm_cholesky_decompose 22 10.9 1.854 1.904 1.854 1.904 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.736 1.768 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.596 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.657 1.668 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.538 1.552 acc_transpose_blocks 24660 15.4 0.103 0.105 1.491 1.512 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.847 1.502 mp_sum_l 6514 12.8 1.031 1.482 1.031 1.482 mp_alltoall_z22v 1111 16.6 1.369 1.428 1.369 1.428 mp_waitany 13376 13.8 1.093 1.425 1.093 1.425 mp_irecv_dv 62702 16.1 0.749 1.424 0.749 1.424 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.150000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.181818, yerr=8.407985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 829.874176E+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 0.020 0.034 56.619 56.620 qs_mol_dyn_low 1 2.0 0.003 0.003 56.316 56.325 qs_forces 11 3.9 0.002 0.002 56.250 56.250 qs_energies 11 4.9 0.001 0.002 52.622 52.626 scf_env_do_scf 11 5.9 0.000 0.001 44.193 44.193 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.941 35.941 velocity_verlet 10 3.0 0.001 0.001 31.431 31.434 dbcsr_multiply_generic 2055 12.4 0.103 0.104 23.307 23.414 qs_scf_new_mos 99 7.5 0.001 0.001 21.581 21.617 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.580 21.616 ot_scf_mini 99 9.5 0.002 0.002 20.323 20.342 multiply_cannon 2055 13.4 0.246 0.272 17.561 19.055 multiply_cannon_loop 2055 14.4 0.320 0.330 16.154 16.497 rebuild_ks_matrix 110 8.3 0.000 0.000 11.582 11.614 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.582 11.614 ot_mini 99 10.5 0.001 0.001 10.762 10.779 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.320 10.349 multiply_cannon_multrec 8220 15.4 3.217 4.561 7.675 8.671 init_scf_loop 11 6.9 0.000 0.000 8.202 8.203 mp_waitall_1 106626 16.5 6.270 7.852 6.270 7.852 qs_ot_get_derivative 99 11.5 0.001 0.001 6.917 6.936 prepare_preconditioner 11 7.9 0.000 0.000 6.558 6.564 make_preconditioner 11 8.9 0.000 0.000 6.558 6.564 sum_up_and_integrate 110 10.3 0.080 0.082 6.287 6.303 integrate_v_rspace 110 11.3 0.003 0.003 6.206 6.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.101 6.182 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.770 5.789 calculate_rho_elec 110 8.6 0.115 0.115 5.769 5.788 qs_ot_get_p 110 10.4 0.001 0.001 5.608 5.640 dbcsr_mm_accdrv_process 17442 15.9 2.813 3.914 4.328 5.333 init_scf_run 11 5.9 0.000 0.001 5.300 5.301 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.300 5.300 make_m2s 4110 13.4 0.039 0.040 4.352 4.652 make_images 4110 14.4 0.637 0.695 4.225 4.525 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.921 4.220 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.169 4.174 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.947 apply_single 110 13.6 0.000 0.000 3.776 3.947 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.856 3.857 ot_diis_step 99 11.5 0.012 0.012 3.820 3.820 grid_integrate_task_list 110 12.3 3.365 3.460 3.365 3.460 pw_transfer 1331 11.6 0.065 0.070 3.387 3.397 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.293 3.294 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.280 3.292 cp_fm_redistribute_end 48 14.0 0.837 3.248 0.845 3.250 cp_fm_cholesky_invert 11 10.9 3.206 3.210 3.206 3.210 cp_fm_diag_elpa_base 48 14.0 2.215 3.040 2.396 3.209 multiply_cannon_sync_h2d 8220 15.4 2.932 3.002 2.932 3.002 make_images_data 4110 15.4 0.039 0.044 2.481 2.937 density_rs2pw 110 9.6 0.004 0.004 2.805 2.886 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.400 2.833 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.788 2.804 wfi_extrapolate 11 7.9 0.001 0.001 2.747 2.747 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.663 2.664 calculate_dm_sparse 110 9.5 0.001 0.001 2.541 2.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.572 2.574 fft3d_ps 1111 14.6 1.119 1.140 2.537 2.552 calculate_first_density_matrix 1 7.0 0.000 0.000 2.460 2.461 grid_collocate_task_list 110 9.6 2.328 2.456 2.328 2.456 rs_pw_transfer 902 11.9 0.010 0.010 2.301 2.374 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.199 2.260 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.209 2.218 potential_pw2rs 110 12.3 0.015 0.015 2.010 2.019 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.991 mp_alltoall_d11v 2046 13.8 1.766 1.938 1.766 1.938 jit_kernel_multiply 10 15.8 1.209 1.905 1.209 1.905 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.868 1.881 cp_fm_cholesky_decompose 22 10.9 1.852 1.875 1.852 1.875 mp_allgather_i34 2055 14.4 0.590 1.743 0.590 1.743 dbcsr_complete_redistribute 325 12.2 0.558 0.577 1.545 1.648 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.510 1.627 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.596 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.489 1.498 qs_create_task_list 11 7.9 0.001 0.001 1.233 1.332 generate_qs_task_list 11 8.9 0.380 0.450 1.232 1.331 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.195 1.216 mp_alltoall_z22v 1111 16.6 1.184 1.198 1.184 1.198 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.793 1.156 mp_waitany 9240 13.8 1.063 1.156 1.063 1.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.620000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.272727, yerr=11.378536 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 1.339830E+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 0.030 0.093 89.931 89.932 qs_mol_dyn_low 1 2.0 0.003 0.003 89.545 89.553 qs_forces 11 3.9 0.002 0.002 89.463 89.463 qs_energies 11 4.9 0.002 0.002 85.328 85.330 scf_env_do_scf 11 5.9 0.000 0.001 75.194 75.194 velocity_verlet 10 3.0 0.001 0.001 57.224 57.232 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 45.593 45.595 dbcsr_multiply_generic 2055 12.4 0.118 0.119 29.762 29.874 init_scf_loop 11 6.9 0.000 0.000 29.524 29.529 qs_scf_new_mos 99 7.5 0.001 0.001 28.628 28.662 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.628 28.661 prepare_preconditioner 11 7.9 0.000 0.000 27.520 27.524 make_preconditioner 11 8.9 0.000 0.000 27.520 27.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.855 26.971 ot_scf_mini 99 9.5 0.002 0.002 26.800 26.842 multiply_cannon 2055 13.4 0.352 0.376 21.986 23.227 multiply_cannon_loop 2055 14.4 0.340 0.347 20.079 20.554 cp_fm_upper_to_full 70 14.2 12.675 18.029 12.675 18.029 ot_mini 99 10.5 0.001 0.001 14.493 14.534 rebuild_ks_matrix 110 8.3 0.001 0.001 13.494 13.559 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.494 13.559 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.253 12.309 dbcsr_complete_redistribute 325 12.2 1.029 1.054 7.412 10.577 qs_ot_get_derivative 99 11.5 0.001 0.001 9.760 9.797 multiply_cannon_multrec 8220 15.4 4.381 4.546 9.604 9.749 mp_waitall_1 87304 16.6 8.623 9.480 8.623 9.480 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.291 9.443 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.649 8.776 mp_alltoall_i22 605 13.7 5.295 8.456 5.295 8.456 qs_ot_get_p 110 10.4 0.001 0.001 6.942 6.981 sum_up_and_integrate 110 10.3 0.150 0.152 6.855 6.870 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.792 6.819 calculate_rho_elec 110 8.6 0.227 0.227 6.791 6.819 integrate_v_rspace 110 11.3 0.003 0.003 6.705 6.720 cp_fm_cholesky_invert 11 10.9 6.374 6.379 6.374 6.379 make_m2s 4110 13.4 0.043 0.044 5.702 6.250 make_images 4110 14.4 0.876 0.922 5.516 6.064 init_scf_run 11 5.9 0.000 0.001 5.865 5.865 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.865 5.865 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.095 5.639 dbcsr_mm_accdrv_process 11614 15.7 3.327 3.655 5.081 5.340 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.773 5.232 apply_single 110 13.6 0.000 0.000 4.773 5.232 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.112 5.120 ot_diis_step 99 11.5 0.015 0.015 4.696 4.697 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.629 4.630 pw_transfer 1331 11.6 0.073 0.074 4.004 4.011 multiply_cannon_sync_h2d 8220 15.4 3.953 3.960 3.953 3.960 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.932 3.932 cp_fm_diag_elpa_base 48 14.0 3.390 3.586 3.929 3.930 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.888 3.895 hybrid_alltoall_any 4261 16.3 0.256 0.555 3.076 3.859 make_images_data 4110 15.4 0.042 0.045 3.152 3.837 grid_integrate_task_list 110 12.3 3.678 3.727 3.678 3.727 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.643 3.644 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.131 3.577 wfi_extrapolate 11 7.9 0.001 0.001 3.421 3.421 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.315 3.321 calculate_dm_sparse 110 9.5 0.001 0.001 3.270 3.299 density_rs2pw 110 9.6 0.004 0.004 3.227 3.258 fft3d_ps 1111 14.6 1.264 1.271 3.097 3.104 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.944 2.948 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.645 2.676 grid_collocate_task_list 110 9.6 2.639 2.663 2.639 2.663 rs_pw_transfer 902 11.9 0.010 0.011 2.354 2.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.371 2.384 calculate_first_density_matrix 1 7.0 0.000 0.000 2.325 2.331 mp_alltoall_d11v 2046 13.8 2.274 2.324 2.274 2.324 cp_fm_cholesky_decompose 22 10.9 2.277 2.294 2.277 2.294 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.180 2.245 potential_pw2rs 110 12.3 0.021 0.021 2.225 2.229 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.165 qs_create_task_list 11 7.9 0.001 0.001 1.911 1.957 generate_qs_task_list 11 8.9 0.741 0.795 1.911 1.956 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.900 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.853 1.862 mp_sum_l 6514 12.8 1.238 1.821 1.238 1.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.932000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1207.090909, yerr=57.887751 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 628.105216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56756. MP_Allreduce 11253 786. MP_Sync 168 MP_Alltoall 2210 2530908. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.120 205.936 205.938 qs_mol_dyn_low 1 2.0 0.004 0.006 205.263 205.276 qs_forces 11 3.9 0.003 0.004 205.169 205.170 qs_energies 11 4.9 0.002 0.002 199.488 199.507 scf_env_do_scf 11 5.9 0.001 0.002 182.518 182.522 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 160.372 160.374 dbcsr_multiply_generic 2485 12.5 0.170 0.175 123.316 124.173 velocity_verlet 10 3.0 0.001 0.001 123.871 123.873 qs_scf_new_mos 116 7.6 0.001 0.001 120.912 121.136 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.912 121.135 ot_scf_mini 116 9.6 0.003 0.003 114.420 114.617 multiply_cannon 2485 13.5 0.234 0.242 99.697 101.825 multiply_cannon_loop 2485 14.5 2.103 2.171 97.427 99.327 ot_mini 116 10.6 0.001 0.001 65.491 65.678 multiply_cannon_multrec 59640 15.5 32.614 34.563 41.005 42.958 qs_ot_get_derivative 116 11.6 0.001 0.001 40.827 41.044 rebuild_ks_matrix 127 8.3 0.001 0.001 33.634 33.906 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 33.633 33.905 mp_waitall_1 288940 16.2 29.396 32.689 29.396 32.689 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.234 30.487 multiply_cannon_sync_h2d 59640 15.5 26.998 28.882 26.998 28.882 qs_ot_get_p 127 10.4 0.001 0.001 27.196 27.470 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.022 24.733 apply_single 127 13.6 0.001 0.001 24.022 24.733 ot_diis_step 116 11.6 0.007 0.008 24.304 24.306 init_scf_loop 11 6.9 0.000 0.000 22.054 22.055 qs_ot_p2m_diag 82 11.4 0.076 0.090 20.347 20.401 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.716 18.882 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.790 17.792 prepare_preconditioner 11 7.9 0.000 0.000 17.401 17.452 make_preconditioner 11 8.9 0.000 0.000 17.401 17.452 multiply_cannon_metrocomm3 59640 15.5 0.115 0.120 15.087 17.249 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.609 16.819 make_m2s 4970 13.5 0.102 0.108 14.382 14.723 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.654 14.661 cp_fm_redistribute_end 82 14.4 11.575 14.561 11.588 14.565 make_images 4970 14.5 0.405 0.422 14.211 14.561 sum_up_and_integrate 127 10.3 0.089 0.108 14.494 14.516 integrate_v_rspace 127 11.3 0.003 0.004 14.405 14.429 cp_fm_diag_elpa_base 82 14.4 2.931 14.143 2.963 14.275 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.245 13.356 calculate_rho_elec 127 8.7 0.044 0.063 13.245 13.355 init_scf_run 11 5.9 0.000 0.001 12.625 12.626 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.625 12.625 cp_fm_cholesky_invert 11 10.9 10.015 10.024 10.015 10.024 wfi_extrapolate 11 7.9 0.001 0.001 9.348 9.348 mp_sum_l 7804 13.0 8.006 9.054 8.006 9.054 dbcsr_mm_accdrv_process 123452 16.2 3.250 3.419 7.960 8.505 calculate_dm_sparse 127 9.5 0.001 0.001 8.312 8.389 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.043 8.164 multiply_cannon_metrocomm1 59640 15.5 0.085 0.090 6.295 8.011 make_images_data 4970 15.5 0.070 0.076 7.109 7.996 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.815 7.887 grid_integrate_task_list 127 12.3 7.084 7.520 7.084 7.520 pw_transfer 1535 11.6 0.075 0.105 7.224 7.520 density_rs2pw 127 9.7 0.006 0.007 7.068 7.513 hybrid_alltoall_any 5155 16.4 0.292 2.236 6.218 7.317 fft_wrap_pw1pw2 1281 12.7 0.010 0.014 7.021 7.295 rs_pw_transfer 1038 11.9 0.017 0.019 6.569 7.111 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 6.750 6.758 fft3d_ps 1281 14.7 2.079 2.551 5.822 6.037 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.847 5.966 fft_wrap_pw1pw2_140 519 13.2 0.440 0.509 5.736 5.905 mp_alltoall_d11v 2401 14.1 4.337 5.572 4.337 5.572 cp_fm_cholesky_decompose 22 10.9 5.179 5.197 5.179 5.197 potential_pw2rs 127 12.3 0.009 0.011 5.032 5.068 grid_collocate_task_list 127 9.7 4.688 5.025 4.688 5.025 mp_sum_d 4442 12.2 3.787 4.460 3.787 4.460 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.938000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.454545, yerr=7.511974 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 830.861312E+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 4808418. 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 0.029 0.043 193.624 193.625 qs_mol_dyn_low 1 2.0 0.003 0.003 193.038 193.052 qs_forces 11 3.9 0.003 0.004 192.954 192.957 qs_energies 11 4.9 0.002 0.009 186.229 186.244 scf_env_do_scf 11 5.9 0.001 0.002 169.689 169.699 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.735 135.737 velocity_verlet 10 3.0 0.001 0.001 121.497 121.499 dbcsr_multiply_generic 2507 12.6 0.182 0.189 98.131 99.326 qs_scf_new_mos 117 7.6 0.001 0.001 96.625 97.102 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.624 97.101 ot_scf_mini 117 9.6 0.004 0.005 91.788 92.320 multiply_cannon 2507 13.6 0.481 0.533 77.497 82.802 multiply_cannon_loop 2507 14.6 1.240 1.282 74.198 76.826 ot_mini 117 10.6 0.001 0.001 50.393 50.861 mp_waitall_1 226760 16.4 26.052 39.344 26.052 39.344 multiply_cannon_multrec 30084 15.6 21.773 26.431 31.438 36.282 init_scf_loop 11 6.9 0.001 0.005 33.858 33.859 rebuild_ks_matrix 128 8.3 0.001 0.001 32.995 33.419 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.995 33.418 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.769 30.160 prepare_preconditioner 11 7.9 0.000 0.000 29.454 29.510 make_preconditioner 11 8.9 0.000 0.002 29.454 29.510 qs_ot_get_derivative 117 11.6 0.001 0.002 28.233 28.763 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.130 28.657 multiply_cannon_metrocomm3 30084 15.6 0.092 0.098 16.257 28.458 qs_ot_get_p 128 10.4 0.001 0.001 23.066 23.669 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 22.209 23.376 apply_single 128 13.6 0.001 0.001 22.208 23.375 ot_diis_step 117 11.6 0.014 0.015 21.984 21.986 multiply_cannon_sync_h2d 30084 15.6 18.980 21.841 18.980 21.841 qs_ot_p2m_diag 83 11.4 0.187 0.216 18.162 18.209 cp_fm_cholesky_invert 11 10.9 17.335 17.350 17.335 17.350 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.964 16.965 make_m2s 5014 13.6 0.091 0.098 14.750 16.191 make_images 5014 14.6 1.163 1.354 14.549 15.992 sum_up_and_integrate 128 10.3 0.117 0.135 14.525 14.556 integrate_v_rspace 128 11.3 0.003 0.004 14.408 14.445 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.735 13.746 cp_fm_redistribute_end 83 14.4 8.013 13.652 8.025 13.653 cp_fm_diag_elpa_base 83 14.4 5.387 13.120 5.606 13.491 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.224 13.251 calculate_rho_elec 128 8.7 0.088 0.105 13.223 13.250 init_scf_run 11 5.9 0.000 0.001 11.691 11.692 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.690 11.692 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.191 11.530 make_images_data 5014 15.6 0.066 0.072 9.053 10.719 multiply_cannon_metrocomm4 27577 15.6 0.098 0.114 3.822 10.488 mp_irecv_dv 69486 16.3 3.627 10.101 3.627 10.101 hybrid_alltoall_any 5200 16.5 0.347 1.555 7.734 9.787 dbcsr_mm_accdrv_process 62242 16.2 4.436 5.469 9.141 9.735 wfi_extrapolate 11 7.9 0.001 0.001 8.450 8.451 pw_transfer 1547 11.6 0.085 0.100 7.933 8.000 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.708 7.774 density_rs2pw 128 9.7 0.006 0.006 6.997 7.616 grid_integrate_task_list 128 12.3 7.139 7.605 7.139 7.605 cp_fm_cholesky_decompose 22 10.9 7.188 7.268 7.188 7.268 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.242 6.991 fft_wrap_pw1pw2_140 523 13.2 0.470 0.520 6.686 6.763 calculate_dm_sparse 128 9.5 0.001 0.001 6.510 6.673 rs_pw_transfer 1046 11.9 0.014 0.016 5.890 6.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.231 6.241 fft3d_ps 1291 14.7 2.778 2.933 6.075 6.103 mp_sum_l 7870 13.0 4.112 6.071 4.112 6.071 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.370 5.493 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.395 5.461 grid_collocate_task_list 128 9.7 4.909 5.338 4.909 5.338 mp_alltoall_d11v 2415 14.1 4.302 5.328 4.302 5.328 mp_allgather_i34 2507 14.6 1.865 5.198 1.865 5.198 potential_pw2rs 128 12.3 0.015 0.017 4.944 4.973 dbcsr_complete_redistribute 395 12.7 0.771 0.886 3.282 4.151 mp_sum_d 4455 12.2 2.698 4.015 2.698 4.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.625000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.181818, yerr=6.057575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 939.298816E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.118 0.158 181.164 181.165 qs_mol_dyn_low 1 2.0 0.003 0.003 180.527 180.540 qs_forces 11 3.9 0.003 0.004 180.418 180.421 qs_energies 11 4.9 0.010 0.054 173.877 173.887 scf_env_do_scf 11 5.9 0.001 0.003 158.112 158.115 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 121.518 121.519 velocity_verlet 10 3.0 0.001 0.001 115.679 115.681 qs_scf_new_mos 117 7.6 0.001 0.001 84.473 84.834 qs_scf_loop_do_ot 117 8.6 0.001 0.002 84.472 84.833 dbcsr_multiply_generic 2507 12.6 0.177 0.181 83.165 84.341 ot_scf_mini 117 9.6 0.004 0.004 80.252 80.646 multiply_cannon 2507 13.6 0.497 0.520 62.593 67.625 multiply_cannon_loop 2507 14.6 0.851 0.882 59.232 62.073 ot_mini 117 10.6 0.001 0.001 43.557 43.987 init_scf_loop 11 6.9 0.001 0.004 36.488 36.489 mp_waitall_1 178456 16.5 26.713 36.214 26.713 36.214 prepare_preconditioner 11 7.9 0.000 0.000 32.397 32.467 make_preconditioner 11 8.9 0.000 0.003 32.397 32.467 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.964 31.423 rebuild_ks_matrix 128 8.3 0.001 0.001 30.442 31.045 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 30.442 31.044 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.511 28.051 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 15.927 25.373 multiply_cannon_multrec 20056 15.6 13.420 16.331 22.115 25.053 qs_ot_get_derivative 117 11.6 0.001 0.002 23.460 23.872 qs_ot_get_p 128 10.4 0.001 0.001 21.791 22.285 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.191 21.261 apply_single 128 13.6 0.001 0.001 20.190 21.261 ot_diis_step 117 11.6 0.018 0.018 19.987 19.987 qs_ot_p2m_diag 83 11.4 0.266 0.273 17.336 17.349 make_m2s 5014 13.6 0.082 0.087 15.287 16.383 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.293 16.294 make_images 5014 14.6 1.182 1.269 15.060 16.153 multiply_cannon_sync_h2d 20056 15.6 14.183 15.686 14.183 15.686 cp_fm_cholesky_invert 11 10.9 15.305 15.314 15.305 15.314 sum_up_and_integrate 128 10.3 0.133 0.145 14.363 14.386 integrate_v_rspace 128 11.3 0.003 0.004 14.230 14.256 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.247 13.278 calculate_rho_elec 128 8.7 0.132 0.147 13.246 13.277 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.001 13.003 cp_fm_redistribute_end 83 14.4 4.921 12.911 4.944 12.915 cp_fm_diag_elpa_base 83 14.4 7.517 12.242 7.944 12.730 make_images_data 5014 15.6 0.062 0.068 9.527 11.076 init_scf_run 11 5.9 0.000 0.001 10.626 10.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.626 10.627 hybrid_alltoall_any 5200 16.5 0.434 1.987 8.288 10.371 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.243 9.518 multiply_cannon_metrocomm4 17549 15.6 0.060 0.071 3.475 9.267 mp_irecv_dv 50230 16.2 3.355 9.021 3.355 9.021 dbcsr_mm_accdrv_process 41502 16.2 4.559 5.241 8.157 8.292 pw_transfer 1547 11.6 0.085 0.105 7.835 7.942 cp_fm_cholesky_decompose 22 10.9 7.780 7.846 7.780 7.846 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.611 7.724 grid_integrate_task_list 128 12.3 7.302 7.686 7.302 7.686 cp_fm_upper_to_full 105 14.5 5.819 7.685 5.819 7.685 wfi_extrapolate 11 7.9 0.001 0.001 7.511 7.511 density_rs2pw 128 9.7 0.006 0.006 6.719 7.021 fft_wrap_pw1pw2_140 523 13.2 0.477 0.528 6.627 6.749 dbcsr_complete_redistribute 395 12.7 1.176 1.208 4.825 6.667 calculate_dm_sparse 128 9.5 0.001 0.001 5.906 5.981 fft3d_ps 1291 14.7 2.694 2.897 5.876 5.938 rs_pw_transfer 1046 11.9 0.014 0.015 5.424 5.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.508 5.514 grid_collocate_task_list 128 9.7 5.074 5.437 5.074 5.437 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.567 5.415 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.688 5.331 mp_sum_l 7870 13.0 3.488 5.148 3.488 5.148 mp_alltoall_d11v 2415 14.1 4.439 5.134 4.439 5.134 mp_allgather_i34 2507 14.6 1.875 5.037 1.875 5.037 potential_pw2rs 128 12.3 0.020 0.022 4.714 4.730 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.552 4.704 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.414 4.234 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.140 4.176 mp_alltoall_i22 716 14.1 2.011 4.076 2.011 4.076 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.768 3.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=181.165000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.181818, yerr=7.196188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/18/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 4.353788E+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 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 1.132671E+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 665379345520 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 12503107. 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 0.052 0.146 191.194 191.196 qs_mol_dyn_low 1 2.0 0.010 0.057 190.558 190.571 qs_forces 11 3.9 0.003 0.004 190.378 190.387 qs_energies 11 4.9 0.002 0.004 183.216 183.227 scf_env_do_scf 11 5.9 0.001 0.002 166.322 166.334 velocity_verlet 10 3.0 0.001 0.001 126.489 126.492 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 118.492 118.494 qs_scf_new_mos 117 7.6 0.001 0.001 82.241 82.593 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.240 82.592 dbcsr_multiply_generic 2507 12.6 0.186 0.191 80.010 80.602 ot_scf_mini 117 9.6 0.004 0.005 77.699 78.003 multiply_cannon 2507 13.6 0.556 0.589 54.817 57.301 multiply_cannon_loop 2507 14.6 1.178 1.206 51.046 52.582 init_scf_loop 11 6.9 0.001 0.002 47.692 47.696 prepare_preconditioner 11 7.9 0.000 0.000 43.564 43.590 make_preconditioner 11 8.9 0.000 0.002 43.564 43.590 ot_mini 117 10.6 0.001 0.001 42.718 43.007 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.158 42.121 multiply_cannon_multrec 30084 15.6 14.004 18.509 25.890 30.589 rebuild_ks_matrix 128 8.3 0.001 0.001 29.580 29.880 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.580 29.880 mp_waitall_1 153770 16.5 18.266 28.635 18.266 28.635 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.662 26.924 qs_ot_get_derivative 117 11.6 0.001 0.002 22.961 23.269 make_m2s 5014 13.6 0.095 0.100 20.767 21.644 qs_ot_get_p 128 10.4 0.001 0.002 21.042 21.359 make_images 5014 14.6 1.998 2.297 20.466 21.345 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.177 19.740 apply_single 128 13.6 0.001 0.001 19.177 19.740 ot_diis_step 117 11.6 0.017 0.018 19.621 19.623 cp_fm_cholesky_invert 11 10.9 17.039 17.049 17.039 17.049 qs_ot_p2m_diag 83 11.4 0.343 0.391 16.823 16.878 cp_fm_upper_to_full 105 14.7 10.927 16.215 10.927 16.215 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.474 15.477 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.685 15.402 sum_up_and_integrate 128 10.3 0.140 0.151 14.496 14.532 integrate_v_rspace 128 11.3 0.003 0.004 14.356 14.396 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.458 13.491 calculate_rho_elec 128 8.7 0.175 0.190 13.457 13.490 dbcsr_complete_redistribute 395 12.7 1.523 1.629 9.230 12.983 make_images_data 5014 15.6 0.065 0.069 11.180 12.662 multiply_cannon_sync_h2d 30084 15.6 11.689 12.590 11.689 12.590 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.222 12.225 cp_fm_redistribute_end 83 14.4 2.119 12.119 2.138 12.123 cp_fm_diag_elpa_base 83 14.4 9.363 11.470 9.945 11.947 dbcsr_mm_accdrv_process 62264 16.2 7.285 8.293 11.461 11.923 hybrid_alltoall_any 5200 16.5 0.529 2.214 10.164 11.509 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.756 11.467 init_scf_run 11 5.9 0.000 0.001 10.893 10.895 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.892 10.894 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.383 10.025 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.540 9.753 mp_alltoall_i22 716 14.1 5.545 9.277 5.545 9.277 pw_transfer 1547 11.6 0.085 0.100 8.229 8.316 cp_fm_cholesky_decompose 22 10.9 8.050 8.155 8.050 8.155 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.003 8.095 multiply_cannon_metrocomm4 25070 15.6 0.076 0.088 2.842 7.888 grid_integrate_task_list 128 12.3 7.504 7.803 7.504 7.803 wfi_extrapolate 11 7.9 0.001 0.001 7.785 7.785 mp_irecv_dv 76098 16.2 2.697 7.618 2.697 7.618 density_rs2pw 128 9.7 0.005 0.006 6.525 7.227 fft_wrap_pw1pw2_140 523 13.2 0.481 0.492 7.002 7.104 fft3d_ps 1291 14.7 2.805 2.893 6.231 6.305 calculate_dm_sparse 128 9.5 0.001 0.001 6.222 6.299 mp_alltoall_d11v 2415 14.1 5.173 5.861 5.173 5.861 grid_collocate_task_list 128 9.7 5.212 5.642 5.212 5.642 rs_pw_transfer 1046 11.9 0.013 0.015 4.938 5.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 5.548 5.604 potential_pw2rs 128 12.3 0.022 0.023 4.638 4.664 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.412 4.513 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.413 4.498 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.420 4.420 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.267 4.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.196000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1068.090909, yerr=25.314224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.510593E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516877E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592388592 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096701. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.077 177.943 177.943 qs_mol_dyn_low 1 2.0 0.003 0.003 177.509 177.522 qs_forces 11 3.9 0.003 0.003 177.412 177.416 qs_energies 11 4.9 0.002 0.002 170.084 170.090 scf_env_do_scf 11 5.9 0.001 0.002 152.897 152.908 velocity_verlet 10 3.0 0.001 0.001 116.294 116.298 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 115.798 115.800 qs_scf_new_mos 118 7.6 0.001 0.001 79.410 79.489 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.409 79.488 dbcsr_multiply_generic 2534 12.6 0.181 0.187 75.423 75.763 ot_scf_mini 118 9.6 0.003 0.004 74.888 74.931 multiply_cannon 2534 13.6 0.591 0.632 55.253 59.832 multiply_cannon_loop 2534 14.6 0.450 0.460 50.536 51.131 ot_mini 118 10.6 0.001 0.001 40.528 40.579 init_scf_loop 11 6.9 0.000 0.000 36.939 36.941 mp_waitall_1 130986 16.6 26.902 34.155 26.902 34.155 prepare_preconditioner 11 7.9 0.000 0.000 33.071 33.101 make_preconditioner 11 8.9 0.000 0.000 33.071 33.101 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.934 31.189 rebuild_ks_matrix 129 8.3 0.001 0.001 28.971 29.023 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.970 29.022 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.493 26.547 qs_ot_get_p 129 10.4 0.001 0.001 21.241 21.290 multiply_cannon_multrec 10136 15.6 10.600 14.279 17.917 20.918 qs_ot_get_derivative 118 11.6 0.002 0.002 20.459 20.500 multiply_cannon_metrocomm3 10136 15.6 0.023 0.025 12.762 20.048 ot_diis_step 118 11.6 0.020 0.020 19.989 19.990 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.580 19.789 apply_single 129 13.6 0.001 0.001 19.580 19.789 cp_fm_cholesky_invert 11 10.9 19.119 19.125 19.119 19.125 make_m2s 5068 13.6 0.068 0.072 16.422 18.679 make_images 5068 14.6 2.306 2.686 16.118 18.370 qs_ot_p2m_diag 84 11.4 0.502 0.507 17.252 17.270 cp_dbcsr_syevd 84 12.4 0.005 0.005 16.053 16.055 sum_up_and_integrate 129 10.3 0.183 0.194 14.355 14.399 integrate_v_rspace 129 11.3 0.004 0.004 14.171 14.224 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.499 13.525 calculate_rho_elec 129 8.7 0.260 0.271 13.498 13.525 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.775 12.786 cp_fm_diag_elpa_base 84 14.4 12.518 12.599 12.769 12.779 make_images_data 5068 15.6 0.055 0.061 9.931 12.525 hybrid_alltoall_any 5255 16.5 0.842 3.825 9.769 12.171 multiply_cannon_sync_h2d 10136 15.6 11.759 12.146 11.759 12.146 init_scf_run 11 5.9 0.000 0.001 10.336 10.336 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.335 10.336 cp_fm_cholesky_decompose 22 10.9 8.297 8.384 8.297 8.384 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.315 8.350 grid_integrate_task_list 129 12.3 7.811 8.172 7.811 8.172 pw_transfer 1559 11.6 0.084 0.092 8.010 8.038 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.788 7.813 dbcsr_mm_accdrv_process 20954 16.1 2.614 3.373 6.947 7.552 wfi_extrapolate 11 7.9 0.001 0.001 7.356 7.356 mp_allgather_i34 2534 14.6 2.899 7.328 2.899 7.328 multiply_cannon_metrocomm1 10136 15.6 0.028 0.029 4.509 7.315 fft_wrap_pw1pw2_140 527 13.2 0.507 0.530 6.758 6.799 density_rs2pw 129 9.7 0.005 0.006 6.266 6.527 calculate_dm_sparse 129 9.5 0.001 0.001 6.182 6.252 fft3d_ps 1301 14.7 2.742 2.811 5.934 5.964 dbcsr_complete_redistribute 397 12.7 2.111 2.191 5.410 5.802 grid_collocate_task_list 129 9.7 5.525 5.731 5.525 5.731 mp_alltoall_d11v 2429 14.1 4.907 5.473 4.907 5.473 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.261 5.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.141 5.151 rs_pw_transfer 1054 12.0 0.013 0.013 4.451 4.690 potential_pw2rs 129 12.3 0.027 0.027 4.301 4.326 multiply_cannon_metrocomm4 7602 15.6 0.023 0.026 1.847 4.066 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.028 4.058 mp_irecv_dv 29142 15.9 1.813 3.991 1.813 3.991 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.859 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.820 3.852 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.470 3.817 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.719 3.776 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.582 3.600 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.943000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.454545, yerr=50.469367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 11.787405E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.056435E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.064 291.966 291.966 qs_mol_dyn_low 1 2.0 0.003 0.003 291.376 291.388 qs_forces 11 3.9 0.003 0.003 291.274 291.277 qs_energies 11 4.9 0.002 0.002 282.667 282.687 scf_env_do_scf 11 5.9 0.001 0.002 260.582 260.592 velocity_verlet 10 3.0 0.001 0.001 210.785 210.794 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.903 135.905 init_scf_loop 11 6.9 0.000 0.000 124.420 124.424 prepare_preconditioner 11 7.9 0.000 0.000 119.643 119.681 make_preconditioner 11 8.9 0.000 0.000 119.643 119.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.895 116.750 qs_scf_new_mos 118 7.6 0.001 0.001 93.505 93.695 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.504 93.694 ot_scf_mini 118 9.6 0.004 0.004 88.646 88.767 dbcsr_multiply_generic 2534 12.6 0.212 0.218 85.172 85.802 cp_fm_upper_to_full 106 14.8 52.444 75.163 52.444 75.163 multiply_cannon 2534 13.6 0.709 0.752 60.104 60.753 multiply_cannon_loop 2534 14.6 0.475 0.485 56.323 57.987 ot_mini 118 10.6 0.001 0.001 45.537 45.677 dbcsr_complete_redistribute 397 12.7 3.997 4.034 29.457 42.287 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.068 38.832 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.702 36.371 mp_alltoall_i22 720 14.1 21.528 34.199 21.528 34.199 cp_fm_cholesky_invert 11 10.9 34.031 34.037 34.031 34.037 rebuild_ks_matrix 129 8.3 0.001 0.001 33.867 33.952 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.867 33.951 mp_waitall_1 106586 16.7 29.124 33.404 29.124 33.404 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.623 31.699 qs_ot_get_p 129 10.4 0.001 0.001 27.306 27.443 qs_ot_get_derivative 118 11.6 0.002 0.002 24.937 25.057 qs_ot_p2m_diag 84 11.4 0.890 0.895 23.050 23.085 make_m2s 5068 13.6 0.078 0.081 20.249 21.268 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.254 21.255 multiply_cannon_metrocomm3 10136 15.6 0.024 0.024 19.369 20.880 make_images 5068 14.6 3.816 3.976 19.772 20.796 ot_diis_step 118 11.6 0.022 0.023 20.561 20.562 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.691 19.959 apply_single 129 13.6 0.001 0.001 19.691 19.959 multiply_cannon_multrec 10136 15.6 10.694 12.515 18.514 18.698 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.965 17.968 cp_fm_diag_elpa_base 84 14.4 13.558 15.181 17.961 17.961 sum_up_and_integrate 129 10.3 0.324 0.326 15.866 15.962 multiply_cannon_sync_h2d 10136 15.6 15.786 15.797 15.786 15.797 integrate_v_rspace 129 11.3 0.004 0.004 15.541 15.640 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.041 15.071 calculate_rho_elec 129 8.7 0.487 0.488 15.040 15.070 hybrid_alltoall_any 5255 16.5 1.313 3.066 10.974 12.929 make_images_data 5068 15.6 0.062 0.065 10.821 12.684 init_scf_run 11 5.9 0.000 0.001 12.174 12.174 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.174 12.174 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.742 9.837 dbcsr_mm_accdrv_process 20954 16.1 3.824 5.913 7.579 9.453 cp_fm_cholesky_decompose 22 10.9 9.187 9.195 9.187 9.195 grid_integrate_task_list 129 12.3 8.603 8.800 8.603 8.800 wfi_extrapolate 11 7.9 0.001 0.001 8.790 8.791 pw_transfer 1559 11.6 0.091 0.093 8.373 8.382 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.138 8.146 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.911 7.912 fft_wrap_pw1pw2_140 527 13.2 0.543 0.548 7.181 7.194 mp_alltoall_d11v 2429 14.1 7.023 7.174 7.023 7.174 calculate_dm_sparse 129 9.5 0.001 0.001 6.896 7.016 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.538 6.610 grid_collocate_task_list 129 9.7 6.384 6.419 6.384 6.419 fft3d_ps 1301 14.7 2.782 2.797 6.162 6.172 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.092 6.159 density_rs2pw 129 9.7 0.005 0.005 6.037 6.067 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.966000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2770.636364, yerr=169.948801 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 1.261629E+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 32456416. 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 0.087 0.126 85.839 85.841 qs_energies 1 2.0 0.005 0.021 85.143 85.199 ls_scf 1 3.0 0.002 0.016 84.230 84.245 dbcsr_multiply_generic 111 6.7 0.016 0.016 72.937 73.095 multiply_cannon 111 7.7 0.017 0.020 56.047 57.292 multiply_cannon_loop 111 8.7 0.213 0.230 52.663 54.120 ls_scf_main 1 4.0 0.000 0.003 52.519 52.526 density_matrix_trs4 2 5.0 0.003 0.005 47.025 47.114 ls_scf_init_scf 1 4.0 0.001 0.002 28.653 28.669 ls_scf_init_matrix_S 1 5.0 0.000 0.008 27.183 27.242 mp_waitall_1 11316 10.9 22.660 25.326 22.660 25.326 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.084 25.100 multiply_cannon_multrec 2664 9.7 8.210 8.960 15.587 17.261 multiply_cannon_sync_h2d 2664 9.7 13.644 14.964 13.644 14.964 make_m2s 222 7.7 0.009 0.011 13.274 13.711 make_images 222 8.7 0.102 0.111 13.252 13.692 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.779 12.727 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.434 9.301 make_images_data 222 9.7 0.005 0.006 7.777 8.322 hybrid_alltoall_any 227 10.6 0.218 1.856 6.655 7.991 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.609 6.992 7.968 dbcsr_mm_accdrv_process_sort 4760 11.4 6.281 7.167 6.281 7.167 calculate_norms 4752 9.8 5.533 6.307 5.533 6.307 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.097 5.256 mp_sum_l 807 5.4 3.085 4.333 3.085 4.333 make_images_sizes 222 9.7 0.000 0.000 0.730 3.596 mp_alltoall_i44 222 10.7 0.730 3.596 0.730 3.596 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.571 mp_irecv_dv 6231 10.9 2.038 3.545 2.038 3.545 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.267 3.289 arnoldi_extremal 4 6.8 0.000 0.000 3.256 3.286 arnoldi_normal_ev 4 7.8 0.001 0.005 3.256 3.286 build_subspace 16 8.4 0.009 0.012 3.158 3.160 ls_scf_post 1 4.0 0.005 0.028 3.056 3.074 ls_scf_store_result 1 5.0 0.000 0.000 2.871 2.911 dbcsr_special_finalize 555 9.7 0.005 0.006 2.402 2.774 dbcsr_merge_single_wm 555 10.7 0.455 0.561 2.394 2.765 make_images_pack 222 9.7 2.214 2.647 2.216 2.649 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.332 2.584 dbcsr_sort_data 658 11.4 2.198 2.531 2.198 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.478 2.073 2.480 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.196 2.260 buffer_matrices_ensure_size 222 8.7 1.753 2.073 1.753 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.895 1.913 rebuild_ks_matrix 3 7.3 0.000 0.000 1.885 1.903 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.010 1.885 1.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.841000, 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/0a22203fba5dd876101bb33c272b97e0757c22c1_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 2.110591E+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 0.031 0.045 91.467 91.469 qs_energies 1 2.0 0.000 0.000 90.849 90.853 ls_scf 1 3.0 0.000 0.002 89.528 89.532 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.555 75.941 multiply_cannon 111 7.7 0.028 0.041 53.482 56.923 ls_scf_main 1 4.0 0.000 0.001 55.174 55.174 multiply_cannon_loop 111 8.7 0.117 0.124 50.230 53.466 density_matrix_trs4 2 5.0 0.002 0.003 49.499 49.666 mp_waitall_1 9246 10.9 21.411 30.960 21.411 30.960 ls_scf_init_scf 1 4.0 0.001 0.003 30.837 30.839 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.685 29.781 multiply_cannon_multrec 1332 9.7 13.081 17.408 22.366 27.826 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.275 27.291 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.013 21.505 make_m2s 222 7.7 0.008 0.009 15.335 15.910 make_images 222 8.7 1.582 1.921 15.305 15.882 dbcsr_mm_accdrv_process 4041 10.4 0.264 0.451 8.886 10.486 dbcsr_mm_accdrv_process_sort 4041 11.4 8.482 10.034 8.482 10.034 make_images_data 222 9.7 0.004 0.005 8.775 9.652 hybrid_alltoall_any 227 10.6 0.523 2.444 8.224 9.303 mp_sum_l 807 5.4 5.440 8.998 5.440 8.998 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.204 7.837 mp_irecv_dv 3311 11.0 3.184 7.769 3.184 7.769 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.234 7.093 calculate_norms 2376 9.8 6.021 6.821 6.021 6.821 multiply_cannon_sync_h2d 1332 9.7 4.795 6.201 4.795 6.201 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.050 5.253 arnoldi_extremal 4 6.8 0.000 0.000 4.735 4.760 arnoldi_normal_ev 4 7.8 0.001 0.004 4.735 4.760 build_subspace 16 8.4 0.014 0.021 4.464 4.467 ls_scf_post 1 4.0 0.000 0.001 3.517 3.521 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.176 3.414 ls_scf_store_result 1 5.0 0.000 0.000 3.252 3.336 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.223 2.744 3.225 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.251 2.657 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.537 2.615 make_images_pack 222 9.7 2.026 2.432 2.029 2.434 mp_allgather_i34 111 8.7 0.947 2.287 0.947 2.287 dbcsr_sort_data 436 11.2 1.839 2.052 1.839 2.052 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.837 1.839 buffer_matrices_ensure_size 222 8.7 1.010 1.838 1.010 1.838 dbcsr_data_new 4174 10.1 1.614 1.837 1.614 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.469000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 2.701402E+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 0.059 0.114 95.349 95.350 qs_energies 1 2.0 0.001 0.006 94.634 94.653 ls_scf 1 3.0 0.000 0.003 93.237 93.253 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.975 78.291 ls_scf_main 1 4.0 0.000 0.001 58.446 58.452 multiply_cannon 111 7.7 0.038 0.080 53.804 57.632 multiply_cannon_loop 111 8.7 0.100 0.106 50.248 53.914 density_matrix_trs4 2 5.0 0.002 0.004 52.550 52.721 mp_waitall_1 7374 11.0 24.992 34.963 24.992 34.963 ls_scf_init_scf 1 4.0 0.004 0.013 31.172 31.176 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.020 30.109 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.539 27.553 multiply_cannon_multrec 888 9.7 12.626 15.207 21.224 24.408 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.864 23.902 make_m2s 222 7.7 0.007 0.008 17.508 18.736 make_images 222 8.7 1.986 2.300 17.469 18.697 hybrid_alltoall_any 227 10.6 0.620 2.873 9.690 11.193 make_images_data 222 9.7 0.004 0.005 10.079 11.070 dbcsr_mm_accdrv_process 3754 10.4 0.250 0.418 8.114 9.341 dbcsr_mm_accdrv_process_sort 3754 11.4 7.747 8.923 7.747 8.923 mp_sum_l 807 5.4 5.104 8.330 5.104 8.330 multiply_cannon_sync_h2d 888 9.7 6.100 7.564 6.100 7.564 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.665 7.481 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.484 7.253 mp_irecv_dv 2335 11.1 2.469 7.213 2.469 7.213 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.909 6.588 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.960 5.208 arnoldi_extremal 4 6.8 0.000 0.000 5.174 5.194 arnoldi_normal_ev 4 7.8 0.001 0.006 5.174 5.194 build_subspace 16 8.4 0.014 0.020 4.868 4.874 calculate_norms 1584 9.8 4.280 4.609 4.280 4.609 mp_allgather_i34 111 8.7 1.383 3.888 1.383 3.888 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.486 3.806 ls_scf_post 1 4.0 0.006 0.022 3.618 3.636 dbcsr_matrix_vector_mult_local 304 10.0 3.025 3.596 3.027 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.357 3.435 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.737 2.829 make_images_sizes 222 9.7 0.000 0.000 1.067 2.464 mp_alltoall_i44 222 10.7 1.067 2.464 1.067 2.464 make_images_pack 222 9.7 1.824 2.115 1.827 2.118 dbcsr_sort_data 325 11.1 1.894 2.111 1.894 2.111 dbcsr_data_release 9322 10.9 1.341 1.964 1.341 1.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.350000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2187.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 3.380670E+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 0.040 0.052 97.137 97.138 qs_energies 1 2.0 0.000 0.000 96.441 96.448 ls_scf 1 3.0 0.000 0.001 94.803 94.807 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.490 78.825 ls_scf_main 1 4.0 0.000 0.001 58.847 58.848 multiply_cannon 111 7.7 0.072 0.178 51.581 56.065 density_matrix_trs4 2 5.0 0.002 0.003 52.801 52.899 multiply_cannon_loop 111 8.7 0.115 0.125 46.549 49.497 ls_scf_init_scf 1 4.0 0.000 0.002 32.722 32.725 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.561 31.632 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.039 29.051 mp_waitall_1 6438 11.0 22.700 28.574 22.700 28.574 multiply_cannon_multrec 1332 9.7 14.211 17.411 22.173 25.754 make_m2s 222 7.7 0.008 0.009 21.297 22.617 make_images 222 8.7 3.146 3.630 21.246 22.568 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.992 17.065 make_images_data 222 9.7 0.004 0.005 11.919 13.630 hybrid_alltoall_any 227 10.6 0.803 3.828 11.260 13.172 dbcsr_mm_accdrv_process 3641 10.4 0.210 0.403 7.592 9.100 dbcsr_mm_accdrv_process_sort 3641 11.4 7.220 8.682 7.220 8.682 mp_sum_l 807 5.4 4.028 8.112 4.028 8.112 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.151 6.489 multiply_cannon_sync_h2d 1332 9.7 5.501 6.119 5.501 6.119 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.101 6.096 mp_irecv_dv 3229 10.9 2.078 6.020 2.078 6.020 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.625 5.462 arnoldi_extremal 4 6.8 0.000 0.000 5.355 5.369 arnoldi_normal_ev 4 7.8 0.001 0.005 5.355 5.369 build_subspace 16 8.4 0.014 0.021 5.012 5.021 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.577 4.697 calculate_norms 2376 9.8 4.209 4.533 4.209 4.533 mp_allgather_i34 111 8.7 2.136 4.322 2.136 4.322 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.632 3.966 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.712 3.205 3.714 dbcsr_sort_data 658 11.4 3.080 3.506 3.080 3.506 dbcsr_special_finalize 555 9.7 0.006 0.007 2.832 3.253 dbcsr_merge_single_wm 555 10.7 0.537 0.666 2.824 3.245 ls_scf_post 1 4.0 0.000 0.003 3.233 3.240 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.037 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.896 2.964 dbcsr_data_release 10477 10.7 1.590 2.438 1.590 2.438 dbcsr_finalize 304 7.8 0.050 0.062 1.807 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.138000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2673.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 4.629156E+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 0.045 0.060 98.790 98.790 qs_energies 1 2.0 0.000 0.000 97.958 97.963 ls_scf 1 3.0 0.000 0.000 96.040 96.046 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.552 77.784 ls_scf_main 1 4.0 0.000 0.000 61.851 61.852 multiply_cannon 111 7.7 0.089 0.123 55.098 60.726 density_matrix_trs4 2 5.0 0.002 0.003 54.768 54.882 multiply_cannon_loop 111 8.7 0.069 0.076 50.437 52.216 mp_waitall_1 5481 11.0 25.725 30.835 25.725 30.835 ls_scf_init_scf 1 4.0 0.000 0.000 30.572 30.577 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.414 29.444 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.268 27.281 multiply_cannon_multrec 444 9.7 14.164 16.541 21.269 23.413 make_m2s 222 7.7 0.006 0.007 17.606 20.255 make_images 222 8.7 3.732 4.423 17.543 20.193 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.691 15.712 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.059 13.781 make_images_data 222 9.7 0.003 0.004 9.778 12.392 hybrid_alltoall_any 227 10.6 0.793 3.783 9.497 12.091 multiply_cannon_sync_h2d 444 9.7 6.557 7.966 6.557 7.966 dbcsr_mm_accdrv_process 3003 10.4 0.168 0.367 6.801 7.927 dbcsr_mm_accdrv_process_sort 3003 11.4 6.481 7.582 6.481 7.582 mp_allgather_i34 111 8.7 2.870 7.130 2.870 7.130 arnoldi_extremal 4 6.8 0.000 0.000 6.050 6.063 arnoldi_normal_ev 4 7.8 0.002 0.006 6.049 6.063 build_subspace 16 8.4 0.015 0.019 5.653 5.661 mp_sum_l 807 5.4 2.882 4.836 2.882 4.836 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.655 4.836 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.275 4.474 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.650 4.354 mp_irecv_dv 1241 11.2 1.630 4.337 1.630 4.337 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.165 3.708 4.167 calculate_norms 792 9.8 3.560 3.706 3.560 3.706 ls_scf_post 1 4.0 0.000 0.000 3.617 3.621 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.953 3.568 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.411 3.496 ls_scf_store_result 1 5.0 0.000 0.000 3.383 3.441 make_images_sizes 222 9.7 0.000 0.000 1.150 3.331 mp_alltoall_i44 222 10.7 1.150 3.331 1.150 3.331 dbcsr_finalize 304 7.8 0.062 0.078 2.213 2.278 dbcsr_merge_all 275 8.9 0.474 0.527 2.058 2.102 dbcsr_data_release 10123 10.8 1.354 1.986 1.354 1.986 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.980 1.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.790000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3667.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a22203fba5dd876101bb33c272b97e0757c22c1_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 8.787612E+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 0.108 0.140 107.183 107.183 qs_energies 1 2.0 0.000 0.000 105.744 105.756 ls_scf 1 3.0 0.000 0.000 102.880 102.891 dbcsr_multiply_generic 111 6.7 0.023 0.028 76.770 76.912 ls_scf_main 1 4.0 0.000 0.000 65.287 65.288 density_matrix_trs4 2 5.0 0.002 0.003 56.414 56.485 multiply_cannon 111 7.7 0.168 0.236 49.194 50.502 multiply_cannon_loop 111 8.7 0.068 0.070 45.845 46.807 ls_scf_init_scf 1 4.0 0.000 0.000 33.922 33.923 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.531 32.553 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.695 29.711 mp_waitall_1 4569 11.1 21.449 25.223 21.449 25.223 make_m2s 222 7.7 0.007 0.007 23.809 24.660 make_images 222 8.7 4.602 5.001 23.700 24.549 multiply_cannon_multrec 444 9.7 17.893 18.677 22.550 23.162 hybrid_alltoall_any 227 10.6 1.667 3.633 12.757 15.557 make_images_data 222 9.7 0.003 0.004 13.030 15.284 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.844 10.668 multiply_cannon_sync_h2d 444 9.7 8.843 8.896 8.843 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.519 7.539 arnoldi_normal_ev 4 7.8 0.004 0.010 7.519 7.539 build_subspace 16 8.4 0.026 0.036 6.968 6.979 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.590 5.722 dbcsr_matrix_vector_mult_local 304 10.0 5.078 5.365 5.081 5.367 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.044 5.293 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.820 4.912 dbcsr_mm_accdrv_process 1814 10.4 0.269 0.355 4.490 4.615 dbcsr_mm_accdrv_process_sort 1814 11.4 4.175 4.303 4.175 4.303 ls_scf_post 1 4.0 0.000 0.000 3.670 3.681 make_images_sizes 222 9.7 0.000 0.000 1.415 3.530 mp_alltoall_i44 222 10.7 1.415 3.530 1.415 3.530 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.416 calculate_norms 792 9.8 3.243 3.274 3.243 3.274 dbcsr_finalize 304 7.8 0.082 0.090 3.101 3.151 mp_allgather_i34 111 8.7 0.975 2.949 0.975 2.949 dbcsr_complete_redistribute 5 7.6 1.455 1.502 2.789 2.937 dbcsr_merge_all 275 8.9 0.895 0.920 2.884 2.926 dbcsr_data_release 12724 10.6 2.334 2.881 2.334 2.881 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.834 2.834 matrix_ls_to_qs 2 6.0 0.000 0.000 2.417 2.580 dbcsr_sort_data 325 11.1 2.446 2.509 2.446 2.509 dbcsr_new_transposed 4 7.5 0.244 0.253 2.329 2.355 dbcsr_frobenius_norm 74 6.6 2.058 2.131 2.187 2.232 mp_sum_l 807 5.4 1.435 2.231 1.435 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.149 2.220 dbcsr_add_anytype 103 7.2 0.860 0.892 2.149 2.219 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.204 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.183000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7011.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 0a22203fba5dd876101bb33c272b97e0757c22c1 Summary: empty Status: OK